mm \. IERKELEY ‘SCIENCE’ LIBRARY LIBRARY UNIVERSITY OF CALIFORNIA The Hydrothermal System of Long Valley Caldera, California By M. L. Sorey, R. E. Lewis, and F. H. Olmsted GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER lO44—A UNITED STATES GOVERNMENT PRINTING OFFICE,WASHINGTON : 1978 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Catalog—card Number 78-600091 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, DC. 20402 Stock Number 024-001—03107-8 CONTENTS Page . Page Abstract __________________________________________________ A1 Geochemical characteristics of the hydrothermal system— Con. Introduction ________________________________________________ 2 Boron discharge ________________________________________ A24 Interest in geothermal potential of Long Valley __________ 2 Stable isotopes ________________________________________ 26 Previous studies ________________________________________ 2 Thermal characteristics of the hydrothermal system __________ 28 Scope of present study __________________________________ 4 Regional heat flow ______________________________________ 28 Acknowledgments ______________________________________ 4 Thermal regime within the caldera ______________________ 28 Well and spring numbering system ______________________ 5 Natural heat discharge __________________________________ 32 Conversion of units ____________________________________ 6 Convective heat discharge by springflow ______________ 33 Geologic setting ____________________________________________ 6 Near-surface conductive heat discharge ______________ 33 Location and general features __________________________ 6 Advective heat discharge by ground-water outflow to Geologic history ________________________________________ 8 Lake Crowley ____________________________________ 35 Characteristics of caldera ______________________________ 9 Total heat discharge ________________________________ 36 Hydrologic setting __________________________________________ 13 Sources of heat ________________________________________ 37 Climate and precipitation ______________________________ 13 Analysis of conceptual model of hydrothermal system ________ 37. Water budget __________________________________________ 14 Description of conceptual model __________________________ 37 Shallow and deep ground-water subsystems ______________ 18 Numerical Simulation __________________________________ 39 Ground water in the shallow subsystem __________________ 20 Hydraulic characteristics ________________________________ 41 Hydraulic characteristics of caldera fill __________________ 20 Thermal characteristics ________________________________ 46 Geochemical characteristics of the hydrothermal system ______ 22 ConcluSions and implications ________________________________ 56 Chemical patterns ______________________________________ 22 References cited ____________________________________________ 57 Chemical geothermometers ______________________________ 22 ILLUSTRATIONS Page PLATE 1. Generalized geologic map of Long Valley area showing location of wells, springs, and hydrologic of meterologic data stations _______________________________________________________________________________________ In Pocket FIGURE . 1. Diagram of system for numbering wells and springs ____________________________________________________________ A5 2. Index map showing location of Long Valley area ________________________________________________________________ 7 3. Map showing thickness of fill in Long Valley caldera ____________________________________________________________ 9 4. Map showing thickness of fill above densely welded part of Bishop Tuff in Long Valley caldera ______________________ 10 5. Map showing average annual precipitation in the Long Valley basin ______________________________________________ 12 6. Annual precipitation at Lake Mary Store and Long Valley Dam __________________________________________________ 13 7. Relation between precipitation and altitude at stations on the east slope of the central Sierra Nevada ______________ 14 8. Map showing tributary drainage areas and streamflow-measuring stations in the Long Valley basin ________________ 15 9. Relation between precipitation and runoff for streams within Long Valley drainage basin __________________________ 17 10. Map of Long Valley caldera showing configuration of the water table, spring 1973 __________________________________ 21 11. Relations between concentrations of deuterium and 18O in Long Valley waters ____________________________________ 26 12. Map of Long Valley caldera showing location of sampling points for deuterium and 180 analyses ____________________ 27 13. Temperature profiles from core holes and shallow holes in Long Valley caldera ____________________________________ 29 14. Map of Long Valley caldera showing temperature at a depth of 15 m, 1973—74 _____________________________________ 30 15. Map of Long Valley caldera showing near-surface conductive heat flow, 1973—74 __________________________________ 32 16. Relation between temperatures at depths of 15 m and 10 m in Long Valley caldera, October 1974 ____________________ 34 17. Map showing elements used in estimate of thermal ground-water flow into Lake Crowley __________________________ 36 18. Block diagram showing conceptual model of Long Valley hydrothermal system ____________________________________ 38 19. Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with uniform reservoir permeability distribution (case A) ____________________________________________________ 40 20. Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with fault-zone reservoir permeability distribution (case B) __________________________________________________ 42 21. Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with low-permeability-block reservoir permeability distribution (case C) ______________________________________ 43 22. Sketch map of Long Valley caldera showing equivalent hydraulic head in reservoir from 1 to 2 km deep from case A with uniform reservoir permeability of30 millidarcys, hot-spring discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and simulation period of 35,000 years ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 44 III FIGURE TABLE 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 0:01».me q 10. 11. 12. 13. 14. 15. 16. 17. CONTENTS Page Sketch map of Long Valley caldera showing equivalent hydraulic head in reservoir from 1 to 2 km deep from case C with low-permeability-block reservoir west of Casa Diablo Hot Springs, permeability of 45 millidarcys in remainder of western three-fifths of caldera, and permeability of 30 millidarcys in eastern two-fifths of caldera ,,,,,,,,,,,,, A45 Diagrammatic east-west cross section of Long Valley caldera showing steady-state, conduction-only isotherms from model simulations without fluid flow and used as initial conditions in simulations of flow in the hydrothermal system __________________________________________________________________________________________________ 46 Map of Long Valley caldera showing steady-state conductive heat flow above reservoir (from model simulation) without fluid flow ________________________________________________________________________________________________ 47 Diagrammatic east-west cross section of Long Valley caldera showing isotherms from model simulation (case A—uniform reservoir permeability distribution) after 35,000 years with a discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km __________________________________________ 48 Map of Long Valley caldera showing distribution of average reservoir temperature from model simulation (case A—uniform reservoir permeability distribution) after 35,000 years with a discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km __________________________________________ 49 Diagrammatic east-west cross section of Long Valley caldera showing isotherms from model simulation (case C—low permeability-block reservoir west of Casa Diablo) after 35,000 years with discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km __________________________________________ 49 Map of Long Valley caldera showing distribution of average reservoir temperature from model simulation (case C—low permeability-block reservoir west of Casa Diablo) after 35,000 years with discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km __________________________________________ 50 Map of Long Valley caldera showing results from model simulations (case A—uniform reservoir-permeability distribu- tion) in terms of conductive heat flow above a depth of 0.5 km after 35,000 years with discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km ______________________________ 51 Diagrammatic east-west cross section of Long Valley caldera showing isotherms from model simulation (case A—uniform reservoir-permeability distribution) after 350,000 years with hot-spring discharge of 250 kg/s, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km __________________________________________ 52 Transient response since initiation of springflow of average reservoir temperature below Hot Creek gorge for two reservoir depths, from model simulation (case A—uniform reservoir-permeability distribution) with discharge of 300 kg/s in Hot Creek gorge and southeast rim outflow of 110 kg/s ___________________________________________ 53 Diagrammatic east-west cross sections of Long Valley caldera showing isotherms and convection patterns from model simulations (case A—uniform reservoir-permeability distribution) after 35,000 years (part A) and 350,000 years (part B), with cellular convection in reservoir at 1—3-km depth, hot-spring discharge of 250 kg/s, and southeast-rim outflow of 110 kg/s ________________________________________________________________________________________ 54 Effect of nodal spacing in vertical direction on average reservoir temperature under Hot Creek gorge after 35,000 years and 350,000 years of fluid flow in 2-dimensional, vertical-cross-section model simulation of Long Valley hydrother- mal system. Cross section aligned east‘west from Hot Creek gorge to west rim; mass flux through reservoir between 1—2-km depth set at 1/10th of flux in 3-dimensional model, or 25 kg/s ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 55 TABLES Page . Conversion of units ____________________________________________________________________________________________ A6 . Calculations of volume and average thickness of fill in Long Valley caldera ________________________________________ 11 . Water budget for the Long Valley drainage basin for water years 1964—74 ________________________________________ 14 . Estimated runoff from ungaged areas in the Long Valley drainage basin __________________________________________ 16 . Laboratory data for selected cores from test holes in Long Valley caldera _ _ - - , , , , A - _ _ _ , - , A A _ T ______________________ 19 . Selected chemical constituents, temperatures, and isotopic compositions for thermal and nonthermal waters from Long Valley ______________________________________________________________________________________________ 23 . Temperatures of reservoirs feeding hot springs in Long Valley estimated by chemical geothermometers ______________ 23 . Measured concentrations of chemical elements in Long Valley rocks and total amounts discharged from Long Valley into Searles Lake __________________________________________________________________________________________ 25 . Size of leached reservoir required to supply total discharges of B and Cl for various changes in B and Cl concentrations in the reservoir rock ______________________________________________________________________________________ 25 Generalized lithologic descriptions and thermal conductivity data from cores and cuttings at test-hole sites in Long Valley caldera ____________________________________________________________________________________________ 31 Previous estimates of convective heat discharge from Long Valley caldera ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 31 Convective heat discharge by springflow from Long Valley caldera ________________________________________________ 33 Near-surface conductive heat discharge from Long Valley caldera ________________________________________________ 35 Hydraulic and thermal properties for Long Valley hydrothermal model ____________________________________________ 39 Intrinsic-permeability data from Long Valley model and other studies ____________________________________________ 43 Conductive and convective heat discharge for each case of reservoir-permeability distribution after 35,000 years of fluid flow at 250 kg/s in reservoir at 1—2-km depth __________________________________________________________ 51 Average reservoir temperature below Hot Creek gorge from various model runs ____________________________________ 53 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ,THE HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA By M. L. SOREY, R. E. LEWIS, and F. H. OLMSTED ABSTRACT Long Valley caldera, an elliptical depression covering 450 km2 on the eastern front of the Sierra Nevada in east-central California, contains a hot-water convection system with numerous hot springs and measured and estimated aquifer temperatures at depths of 180°C to 280°C. In this study we have synthesized the results of previous geologic, geophysical, geochemical, and hydrologic investigations of the Long Valley area to develop a generalized conceptual and math- ematical model which describes the gross features of heat and fluid flow in the hydrothermal system. Cenozoic volcanism in the Long Valley region began about 3.2 m.y. (million years) ago and has continued intermittently until the pres- ent time. The major event that resulted in the formation of the Long Valley caldera took place about 0.7 m.y. ago with the eruption of 600 km3 or more of Bishop Tufi° of Pleistocene age, a rhyolitic ash flow, and subsequent collapse of the roof of the magma chamber along one or more steeply inclined ring fractures. Subsequent intracaldera vol- canism and uplift of the west-central part of the caldera floor formed a subcircular resurgent dome about 10 km in diameter surrounded by a moat containing rhyolitic, rhyodacitic, and basaltic rocks ranging in age from 0.06 to 0.05 m.y. On the basis of gravity and seismic studies, we estimate an aver- age thickness of fill of 2.4 km above the precaldera granitic and metamorphic basement rocks. A continuous layer of densely welded Bishop Tuf‘f overlies the basement rocks, with an average thickness of 1.4 km; the fill above the welded Bishop Tuff consists of interca- lated volcanic flows and tuffs and fluvial and lacustrine deposits. Assuming the average grain density of the fill is between 2.45 and 2.65 g/cm", we calculate the average bulk porosity of the total fill as from 0.11 to 0.21. Comparison of published values of porosity of the welded Bishop Tuff exposed southeast of the caldera with calculated values indicates average bulk porosity for the welded tufi‘ (including fracture porosity) from 0.05 to 0.10. Because of its continuity and depth and the likelihood of significant fracture permeability in the more competent rocks such as the welded tuff, our model of the hy- drothermal system assumes that the Bishop Tuff provides the princi- pal hot-water reservoir. However, because very little direct informa- tion exists from drill holes below 300 in, this assumption must be considered tentative. Long Valley caldera is drained by the Owens River and several tributaries which flow into Lake Crowley in the southeast end of the caldera. Streamflow and springflow measurements for water years 1964—74 indicate a total inflow to Lake Crowley of about 10,900 US. In contrast, the total discharge of hot water from the hydrothermal reservoir is about 300 US. For modeling purposes, the ground-water system is considered as comprising a shallow subsystem in the fill above the densely welded Bishop Tuff containing relatively cold ground water, and a deep subsystem or hydrothermal reservoir in the welded tuff containing relatively hot ground water. Hydrologic, isotopic, and thermal data indicate that recharge to the hydrother- mal reservoir occurs in the upper Owens River drainage basin along the western periphery of the caldera. Temperature profiles in a 2.11— km-deep test well drilled by private industry in the southeastern part of the caldera suggest that an additional flux of relatively cool ground water recharges the deep subsystem around the northeast rim. Flow in the shallow ground-water subsystem is neglected in the model except in recharge areas and along Hot Creek gorge, where approximately 80 percent of the hot-water discharge from the hydro- thermal reservoir moves upward along faults toward springs in the gorge. Heat-flow data from the Long Valley region indicate that the re- surgent dome overlies a residual magma chamber more circular in plan than the original magma chamber that supplied the Bishop Tuff, and lead to the inference that magma beneath the east part of the caldera was exhausted during eruption of the Bishop Tuff. Seis- mic and teleseismic studies (based on distant earthquakes) also in- dicate that an anomalously hot or partially molten mass persists below 6—8 km under the west part of the caldera. Other evidence, including an estimate of present-day heat discharge of 6.9 X 10'l cal/s, implies that the heat source for the hydrothermal system is related to the main magma chamber rather than to any of the postcaldera eruptive volcanics. Constraints on modeling the natural conditions of heat and fluid flow in the hydrothermal system are provided by applying chemical mixing models to spring discharges and rates of boron discharge into Lake Crowley to yield estimates of 200—300 kg/s of water at tempera- tures between 200°C and 280°C leaving the reservoir under the area of hot-spring discharge. We also estimate 6.9 x 107 cal/s for the total heat discharge at the land surface, based on measurements of spring discharges and temperatures, shallow conductive heat flows, and ad- vective heat losses from warm water discharge into Lake Crowley. Unfortunately, the time over which this heat discharge has persisted is uncertain. Evidence of hydrothermal alteration indicates that hydrothermal activity was present and perhaps more extensive at 0.3 m.y. ago than today, although only relatively recent periods of saline discharge (30,000—40,000 years) from Long Valley can be ac- counted for by analysis of salts in deposits of Searles Lake, downdrainage from the caldera. The total amounts of various hot-spring constituents-such as B, Cl, Li, and As—which have been contributed to Searles Lake by the A1 A2 Long Valley system could have been supplied by leaching of realistic volumes of reservoir rocks. A direct magmatic source would not be required to'supply these elements, even considering that an ad- ditional Searles Lake-size deposit from a previous period of hydro- thermal activity around 0.3 m.y. ago remains undiscovered. Our mathematical model of the Long Valley caldera involves a transient, three-dimensional simulation using numerical techniques to solve the appropriate partial differential equations. The model includes five horizontal layers corresponding to the major rock units identified by seismic—refraction and geologic studies. The simulated hydrothermal reservoir is in fractured Bishop Tuff and precaldera basement rocks at depths from 1 to 3 km. Recharge to the reservoir occurs along the caldera ring fault around the west and northeast rims, and discharge occurs at the surface along Hot Creek gorge and at depth through the southeast rim. Estimates of effective reservoir permeability, assuming an equivalent porous-media flow system, were obtained with the model for several variations of reservoir per- meability distribution. This was done by assigning pressure- boundary conditions in recharge and discharge areas based on water-table altitudes and then adjusting reservoir permeability to yield a throughflow of 250 kg/s. Intrinsic permeability values from 30 ’ to 50 millidarcys (lo—151112) were obtained for a 1-km-thick reservoir covering the entire area of the caldera. A permeability of 350 mil- lidarcys was required for the case of a more areally restricted fault zone reservoir. The values obtained are inversely proportional to the simulated reservoir thickness. Comparisons with laboratory results on cores from Long Valley and the Neveda Test Site (NTS), and well tests in fractured, welded tuff at NTS indicate that permeability values obtained from the model analysis represent an integration of the effects of fracture permeability over the volume of the reservoir rock. Thermal boundary conditions in the model included a constant- temperature distribution at the base, which simulated a magma chamber under the west half of the caldera, and constant tempera- ture at the land surface. Initial temperature conditions were ob— tained from a steady—state solution with no fluid flow in the reservoir. Simulation of heat and fluid flow for a period of 35,000 years (based on the estimated age of Long Valley salts in Searles Lake) indicates that present-day heat discharge could have been sustained for this period by a magma chamber at 6 km with fluid circulation to depths from 1.5 to 2.5 km in a reservoir which is continuous over the area of the caldera. Simulated reservoir temperatures under the Hot Creek gorge area are close to those estimated geothermetrically (200°C— 280°C) after 35,000 years, but are only about 80°C under the south- eastern portion of the caldera as a result of recharge from the north- east rim near Glass Mountain. Cooler temperatures under the east- ern caldera are consistent with results from the 2.11-km-deep test well drilled in that area. To sustain hot-spring discharge with present-day heat flow and reservoir temperatures for periods much greater than 35,000 years, deeper levels of fluid circulation would be required. For a period of 350,000 years, at which time the system would have reached steady- state conditions, fluid circulation to at least 4—5 km would be neces- sary. Consideration of cellular convection in addition to horizontal throughflow in the hydrothermal reservoir does not significantly alter these results. The model simulations and the diverse indications of the age of hot-spring activity are consistent with the concept that hydrother- mal system has functioned intermittently with considerable periods of inactivity—possibly related to climatic variations and chemical self-sealing processes. Additional data from deep drilling in the western part of the caldera are needed to more satisfactorily de- lineate the characteristics of the hydrothermal flow system and the caldera’s geothermal history and to evaluate the adequacy of the simplified hydrothermal model considered in this study. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS INTRODUCTION INTEREST IN GEOTHERMAL POTENTIAL OF LONG VALLEY Geothermal manifestations of Long Valley and adja- cent areas have been recognized since the late nineteenth century (Whiting, 1888; Russell, 1889). No attempts to develop the geothermal resource were made, however, until the late 1950’s and early 1960’s, when several exploratory geothermal wells were drilled at or near Casa Diablo Hot Springs (pl. 1, fig. 2) by Magma Power Company (McNitt, 1963, p. 29; California Division of Oil and Gas, 1972). Although the deepest exploratory well reached only 324 m, temperatures as high as 180°C were recorded (McNitt, 1963, p. 25—29; California Department of Water Resources, 1967). In spite of these favorable in- dications, no further test drilling was done commer- cially until 1976, when an exploratory well was drilled by Republic Geothermal, Inc. to a depth of 2,109 m in the southeastern part of the Long Valley caldera (pl. 1, T3S/R29E—29K, well 66—29). Specific data from the deep well are proprietary but temperatures in the hole were indicated to be too low for conventional genera- tion of electricity and for most nonelectrical uses (W. S. Keys, T. C. Urban, W. H. Diment, and Manuel Mathenson, written commun., 1976). Meanwhile, chemical analyses of hot waters from the shallower test wells and hot springs indicated a hot- water-type reservoir (White and others, 1971, p. 77—80) with minimum temperatures of 180°C on the basis of methods of geothermometry developed by Fournier and Rowe (1966) and Ellis (1970). Calculations by White (1965, table 1) based on relations of boron concentra- tion to heat content of thermal waters indicated a con- vective heat flow of 7 X 107 cal/s. According to a recent estimate by Nathenson and Muffler (1975, table 16), Long Valley is one of the largest hydrothermal convec- tion systems in the western United States in terms of potential for electric-power development. In spite of relatively low temperatures found in the Republic Geothermal well, the Long Valley hydrothermal sys- tem, or perhaps, the heat stored in the rocks below the depths of significant hydrothermal circulation, are likely to be developed commercially, once the environ- mental problems associated with such development can be dealt with satisfactorily. PREVIOUS STUDIES Long Valley and surrounding areas have been the subject of many geologic, hydrologic, geophysical, and geochemical studies since the late nineteenth century. The presence of hot springs, young volcanic rocks, and evidence of earlier hydrothermal activity was recog- nized in the earliest investigations. Among papers de- HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA scribing thermal springs and other hydrothermal man- ifestiations are those of Russell (1889); W. T. Lee (1906); Waring (1915); Stearns, Stearns, and Waring (1937), Blake and Matthes (1938); Waring (1965); and Kaysing (1970). Early regional geologic studies include those of Mayo (1934), who applied the term "Mammoth embayment” to the L-shaped reentrant in the east face of the Sierra Nevada in what is now considered the west half of the Long Valley caldera, and Gilbert (1941), who sum- marized the structure and upper Tertiary volcanic rocks of the Long Valley area. Rhyolites of the so-called "Mammoth embayment” were described by Chelikowsky (1940), whose interpre- tations of the age and stratigraphic sequence have been greatly modified by later workers. Gilbert (1938) gave the name Bishop Tuff to the extensive rhyolite tuff, chiefly of ash-flow origin, that erupted from vents beneath Long Valley at the time the caldera was formed. Putnam (1940, 1949, 1960) showed that the Bishop Tuff erupted shortly after the Sherwin glacial stage (probably middle Pleistocene). Sheridan (1965, 1968, 1971) studied the Bishop Tuff in detail in expo- sures east and southeast of Long Valley, where he con- cluded that two major eruptive pulses were repre- sented. Potassium-argon dates for the Bishop were ob- tained by Dalrymple (1963, 1964) and Dalrymple, Cox, and Doell (1965), which ultimately established an age of about 0.7 m.y. (million years) for the eruptions that formed the Long Valley caldera. Noble, Korringa, Hedge, and Riddle (1972) described the rhyolite at Glass Mountain on the northeast rim of the caldera. Wood (1975) used radiocarbon dating and trace- element correlations to decipher the sequence of late Pleistocene eruptions from the northwest part of the Long Valley caldera northward to the Mono basin. Geologic mapping of most of the caldera and adjacent areas to the east and west, published at a scale of 1:62,500, was done by Rinehart and Ross (1957, 1964), Huber and Rinehart (1965), and Crowder and Sheridan (1972). In spite of all this and other geologic work, the Long Valley area had not been studied systematically as a unit until a coordinated program of geologic, geophysi- cal, geochemical, and hydrologic investigations was undertaken by the US. Geological Survey during 1972—73 (Muffler and Williams, 1976). As a major part of this effort, Bailey, Dalrymple, and Lanphere (1976) presented a synthesis of their work and related studies, which included a tentative model of the caldera system and its evolution. Our present interpretation of the geologic framework of the caldera is based largely on Bailey, Dalrymple, and Lanphere (1976) and on the results of geophysical investigations, which are out- A3 lined below. The earliest published results of geophysical studies of Long Valley consisted of gravity, seismic, and aeromagnetic surveys in the Mono Basin—Long Valley area (Pakiser and others, 1960). The results of the gravity and magnetic surveys in Long Valley were subsequently summarized by Pakiser (1961). Pakiser, Kane, and Jackson (1964) developed a gravity model of Long Valley and discussed the geologic structure of the valley and its relation to the surrounding geologic fea- tures and to the geologic history of the Sierra Nevada and the Basin and Range provinces. Further re- finements of gravity and magnetic interpretations were made by Kane and Mabey (1973), and by Kane, Mabey, and Brace (1976a, 1976b). These studies indi- cated a steep-sided caldera having a porous fill possibly 3 km in thickness. The magnetic results indicated a more magnetic fill in the eastern part of the caldera than in the western part. Seismic-refraction studies (Pakiser, 1968; Hill and others, Pakiser, 1973; Hill, 1976) have strongly supplemented the gravity and magnetic studies in de- lineating caldera geometry and showing the major structural and stratigraphic features of the caldera fill. Gross stratigraphic features inferred from seismic- refraction profiles have been corroborated by total- field-resistivity surveys, direct-current soundings, and electromagnetic soundings, which also were successful in outlining zones of low resistivity that correlate with known hot springs or with zones of hydrothermal al- teration (Jackson and others, 1973; Stanley and others, 1973, 1976). Shallow resistivity anomalies were ex- plored by an audiomagnetotelluric system, which iden- tified two linear zones of low resistivity associated with hot springs and hydrothermally altered rocks (Hoover and others, 1973, 1976). Spontaneous (self) potential surveys indicated streaming potential attributed to upward flow of thermal ground water in the west- central part of the caldera and downward flow in areas to the north and west (Anderson and Johnson, 1973, 197 6). A seismic-noise survey indicated the possible pres- ence of geothermal noise beneath Long Valley (Iyer and Hitchcock, 1976). Microearthquakes were recorded for only a brief period in 1973 (Steeples and Pitt, 1973, 1976). The record showed activity only at the southeast edge of the caldera, at a depth of 10 km, possibly on the Hilton Creek fault. A longer record might have shown activity elsewhere. Studies of teleseismic events (dis- tant earthquakes) indicated delays of P-wave arrivals attributed to presence of low-velocity material beneath the west-central part of the caldera, at depths between 7 and 40 km (Steeples and Iyer, 1976). Heat-flow studies—Lachenbruch (1968, 1970); A4 Lachenbruch, Lewis, and Sass (1973); Sass, Lachen- bruch, and Munroe (1974); Lachenbruch, Sass, Monroe, and Moses (1976); Lachenbruch, Sorey, Lewis, and Sass (1976); Sorey and Lewis (1976)—were based on 11 temperature-gradient test holes located from 0 to 30 km outside the caldera rim and ranging in depth from 1 13 to 271 m, and also on 40 test holes drilled to depths from 5 to 325 m within the caldera for hydrologic as well as heat-flow information. The heat-flow studies, which are discussed in a later section, generally placed constraints on our interpretation of the magmatic his- tory of the caldera and of the effects of ground-water movement on near-surface heat flows in the caldera. The first significant subsurface geologic and geochemical data in the caldera were obtained from nine geothermal test wells drilled at Casa Diablo Hot Springs and one well drilled about 5 km farther east between 1959 and 1964 by Magma Power Company (McNitt, 1963, p. 29). The wells ranged in depth up to 324 m, and the highest temperature recorded was about 180°C (California Dept. of Water Resources, 1967). The first hydrologic investigation for which pub- lished data are available was that of W. T. Lee (1906). Although primarily concerned with the geology and water resources of the Owens Valley, Lee discussed the volcanic features and the thermal activity of the Long Valley area. C. H. Lee (1912) also discussed the water resources of the Owens Valley; although he presented no data for Long Valley, he did include data on precipi- tation, streamflow, evaporation, transpiration, and in- filtration which are valid, with some modification, for Long Valley. Lewis (1974) presented basic data on springs and wells in the Valley. Sorey (1975b) de- scribed the effects of potential geothermal development on the springs at Hot Creek Fish Hatchery. Geochemistry of the thermal waters in the caldera was investigated by Willey, O’Neil, and Rapp (1974); and Mariner and Willey (1976). In addition, estimates of reservoir temperatures and convective heat dis— charge from the caldera on the basis of chemical data were made by White (1965); White, Muffler, and Truesdell (1971); Sorey and Lewis (1976); and Mariner and Willey (1976). Studies of isotopes of hydrogen and oxygen by Mariner and Willey (1976) and Friedman and Smith (1970, 1972) indicated possible sources of meteoric recharge water and have assisted in delinea- tion of the mixing of thermal and nonthermal waters in the caldera. Qualitative results from the Republic Geothermal, Inc. well 66—29 (pl. 1, T3S/R29E—29K), which are gen- erally known within the geothermal industry and are discussed in subsequent sections of this report, include identification of the top of the welded Bishop Tuff at GEOHYDROLOGY OF GEO’I‘HERMAL SYSTEMS 0.85 km and a lack of evidence for basement rocks above the drilled total depth of 2.11 km. Selected sam- ples from the drill cuttings were analyzed for grain density, and one fluid sample was obtained from a drill stem test for analysis of chlorine and stable isotope contents. White (1976, written commun.) suggests, on the basis of the relatively low temperatures measured throughout the drilled interval and previous geologic, geophysical, and geochemical data, that the eastern 40 percent of the caldera is eliminated as an attractive hydrothermal resource, but that the western 60 per- cent continues to have major potential. SCOPE OF PRESENT STUDY The purpose of the present study was to synthesize previous geologic, geophysical, geochemical, and hy- drologic investigations of the Long Valley caldera and to derive one or more generalized conceptual and mathe- matical models of the hydrothermal system. Although fairly abundant areal geologic, surface-geophysical, and geochemical data were available, subsurface drill-hole information, with one exception, was limited to depths less than about 300 m. Thus, any conceptual model must be an oversimplification at this stage. The model analyzed in this study is consistent with con- straints provided by the results of the recent deep test well, previous geologic and geophysical studies, and estimates of heat and fluid discharge from the caldera. Thus, we believe that the gross features of the hydro- thermal system are adequately quantified in the model. However, because data from additional deep drilling could suggest necessary changes and refine- ment in the conceptual model, we present the following analysis more as an illustration of the usefulness of numerical simulation techniques, rather than as a unique representation of the Long Valley hydrother- mal system. The ultimate goal of this modeling effort is to satis- factorily evaluate the potential for energy develop- ment, the optimum methods for that development, and the extent to which development can proceed without unacceptable effects on other resources in the Long Valley area. The modeling study presented in this re- port involves a three-dimensional, transient analysis of the natural thermal and hydrologic conditions within the caldera. As additional detail from deep test drilling becomes available, this basic model can be modified and refined to permit analysis of the energy development problem. ACKNOWLEDGMENTS The writers express their appreciation to R. A. Bailey, A. H. Lachenbruch, D. E. White, and R. O. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA Fournier of the US. Geological Survey for contributing information, insights, and discussions on Long Valley. We are also grateful to R. H. Mariner, R. J. Munroe, T. H. Moses, J. G. Blevins, and N. E. Voegtly for their assistance and cooperation in the collection and analysis of field data. We acknowledge the cooperation of Republic Geothermal, Inc. in providing access to, and information from, their deep test well. We also benefited from access to temperature data obtained by W. S. Keys, T. C. Urban, W. H. Diment, and Manuel Nathenson of the U. S. Geological Survey. In addition, comments by A. F. Moench, Manuel Nathenson, J. W. Mercer, and F. W. Trainer were val- uable in encouraging further analysis and emphasiz- ing that this explanation of the Long Valley Hydro- thermal system is not unique. WELL AND SPRING NUMBERING SYSTEM Wells and springs are assigned numbers according to their location in the rectangular system for the sub- MOUNT DIABLO BASE LINE T.1S. T.2S. MOUNT DIABLO MERIDIAN T.4S. \ \ R.28E. H.29E. \ \ R.1E. R.2E. A5 division of public land. For example, as shown in figure 1, in the number 38/29E—2A1, the part of the number preceding the slash indicates the township (T 3 S.), the part between the slash and the dash indicates the range (R. 29 E.), the number between the dash and the letter indicates the section (sec. 2), and the letter indi- cates the 40-acre subdivision of the section. Within the 40-acre tract wells are numbered se- rially, as indicated by the final digit. .Thus, well 38/ 29E—2A1 is the first well to be listed in the NE1/4: NE1/4 sec. 2, T. 3 S., R. 29 E., Mount Diablo base line and meridian. Springs are numbered similarly except that an S is placed between the 40-acre subdivision letter and the final digit, as shown in the following spring number: 3S/28E—35ESl. The letter Z, substituted for the letter designating _ the 40-acre tract, indicates the well or spring was plot- ted from unverified descriptions; the described loca- tions of such wells or springs were visited, but the data could not be correlated with existing wells or springs. T.3S./R.29E.—2A1 U" A 3 2 1 \ e 7 on (D 10 11 12 15 14 13 \ 18 17 16 19 20 21 22 23 24 \30 29 28 27 26 25 31 32 33 34 35 36 FIGURE 1.—Diagram of system for numbering wells and springs. A6 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 1.——Conversion factors between metric and English units Multiply metric units By To obtain English Units _ , Length millimeters (mm) 3.937 x 10‘2 inches (in.) meters (m) 3.281 feet (ft) kilometers (km) .6214 miles (mi) Area hectares (ha) 2.471 acres (acre) square kilometers (km?) 247.1 acres (acre) 3.861 square miles (miz) Volume cubic centimeters (cm3) 6.10 X 10—2 cubic inches (in3) liters (L) .2642 gallons (gal) 3.531 x 10‘2 cubic feet (ft3) cubic meters (m3) 35.31 cubic feet (its) 8.107 X 10‘4 acre-feet (acre-ft) cubic hectometers (hma) 8.107 x 102 acre-feet (acre-ft) Flow liters per second (US) 15.85 gallons per minute (gal/min) 25.58 acre-feet per year (acre—ft/yr) CIlbiC hectometers per year (hm3/yr) 8.107 X 102 acre-feet per year (acre-ft/yr) Mass grams (g) 3.528 x .10‘2 ounces (oz) kilograms (kg) 2.205 pounds (lb) Temperature: degrees Celsius to degrees Fahrenheit—T = 18°C + 32 Thermal Parameters Multiply "working units" By To obtain SI units Thermal conductivity millicalories per centimeter second - degree (In cal/(cm s~°C)) 0.4187 watts per meter-degree Kelvin (W/(m'°K)) Thermal diffusivity square centimeters per second (cmZ/s) 1.0 x square meters per second 10" (m2/s) Heat flow (Heat—flux density) microcalories per square centimeter'second (cal/(cm2 - 5)) 4.187 X 10‘2 watts per square meter (W/mz) heat-flow unit (HFU) Energy calories (cal) 4.187 joules (J) Heat discharge calories per second (cal/s) 4.187 watts (W) CONVERSION OF UN ITS The metric system is used throughout this report, although some of the original measurements and data were reported in English units. Thermal parameters are given in the more familiar “working units” rather than in the now-standard SI (Systeme Internationale) units. Table 1 lists metric and equivalent English units, and "working units” and equivalent SI units for the thermal parameters. GEOLOGIC SETTING LOCATION AND GENERAL FEATURES The Long Valley caldera is in east-central Califor- nia, in southern Mono County, about 40—70 km north- west of Bishop (fig. 2). The caldera floor is elliptical in plan, 17 by 30 km, with its long axis aligned east-west. The area of the floor is about 450 km2, and the total area of the caldera and its tributary drainage basins is about 830 km2. The caldera margin (pl. 1) is formed by the Sierra \Nevada on the south and west, by a prominent ridge from Bald Mountain to Glass Mountain on the north, and by an unnamed dissected volcanic tableland on the east. Altitudes along the rim range from as low as 2,420 m in the northwest and 2,067 m at Lake Crowley (spillway level) in the southeast to 3,390 m at Glass Mountain in the northeast, 3,823 m at Bloody Mountain in the south, and 3,367 m at Mammoth Mountain in the southwest. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA Long Valley proper occupies the lower, eastern two- fifths of the caldera. The higher, western three-fifths of the caldera includes a complex series of postcaldera rhyolite flows which have been arched and sub- sequently faulted to form a resurgent dome 10—12 km in diameter. The annular moat between the resurgent dome and the caldera rim is occupied in the north, west, and south by still younger rhyolitic to basaltic lavas and in the east chiefly by alluvial and lacustrine deposits. Altitudes within the area of the caldera floor range from 2,067 m at Lake Crowley in the southeast to 2,860 m at the summit of a lava dome within the A7 southwest moat. The resurgent dome rises to maximum altitude of 2,545 m at Lookout Mountain in the northern part, or about 340—460 In above the floor of the adjacent moat. The caldera is drained by the Owens River, which flows eastward across its northern part and then southward into Lake Crowley, and by several trib- utaries around the periphery. Mammoth, Hot, and Convict Creeks are the chief tributaries, flowing gen- erally eastward across the southern part of the caldera into Lake Crowley. Drainage on the resurgent dome has a roughly radial pattern. _- _ "l x I x l \ l \\ O i \ \ '7 \‘s Area of \ ( Index Map \ , \ A \\ ‘ \ 0 Bridgeport \\ \x\ \ \\ K K (>131 ~. (/‘o ‘70 l 833 06> 4 , 4’ \ \ .. ’47 \ ’__J \. —- \ x ' Mono Lake / \ 0 \\ Lee Vining . \ . [/\_/ M ® \ \\ L, Benton\\ 3 _ \\~ .,’/ Station \\ - \m .Bamon \ \ / \ / ' ~ /’ ”Vb—V \ \ a \ éCrow/ey Toms Place / V000 ‘3‘ ’0 MONO co— Q o _ _ _ _ _ __ __ _ _ @‘Siée }- J- INYO co A“ , , \ chic—t'TWILES /\ @\o Bishop 0 1o 20 \ KILOMETERS \..\ \ FIGURE 2.—Index map showing location of Long Valley area. A8 GEOLOGIC HISTORY The events described below are based chiefly on an interpretation of the geologic history of the Long Val- ley caldera by Bailey, Dalrymple, and Lanphere (1976), whose work included geologic mapping, sam- pling, potassium-argon dating on rock samples care- fully selected to provide maximum information about the sequence of eruptive events, petrologic studies, pre- liminary petrographic work, and studies of preexisting petrochemical data. Although some of the conclusions are tentative, the general outline of the history is be- lieved to be reasonably well understood. Cenozoic volcanism in the region now occupied by the Long Valley caldera began about 3.2 m.y. ago with the eruption of basalt and andesite from widely scat- tered centers. The earliest eruptions were followed 3.0 to 2.7 m.y. ago by eruption of rhyodacite (quartz latite) from the Two Teats-San Joaquin Mountain and Bald Mountain areas. All these events probably happened during the latter stages of the last major uplift of the Sierra Nevada. Lindgren (1911), Matthes (1930, 1933, 1939, 1947), Dalrymple (1963, 1964), Bateman and Wahrhaftig (1966), Christensen (1966), and others believed that the Sierra Nevada crest reached approximately its present height by the time of the earliest glaciation of the Pleistocene, and that the formation of the eastern escarpment of the Sierra followed the last major uplift. A different interpretation was reached by Piper, Gale, Thomas, and Robinson (1939), Axelrod and Ting (1961), Axelrod (1962), and Putnam (1962), who in- ferred that the last major uplift was middle Pleistocene or later. The prepondrance of evidence, however, sup- ports the views of Lindgren, Matthes, and later work- ers cited above. In any event, the inference seems reasonable that the earliest volcanic activity in the Sierra crestal region was related to the initiation of major normal faulting that downdropped the area to the east. The next volcanic episode was the eruption of highly silicic rhyolite along an arcuate zone 13 km in length northeast of the present Long Valley caldera in the general vicinity of Glass Mountain. The arcuate zone probably is coincident with an incipient ring fracture related to the earliest stages of the development of the Long Valley magma chamber. The rhyolitic eruptions began at least 1.9 m.y. ago and were episodic to about 0.9 m.y. ago. The major event that resulted in the formation of the Long Valley caldera took place about 0.7 m.y. ago with the eruption of 600 km3 or more of rhyolite from the upper part of a magma chamber beneath Long Valley. The eruptions included two major pulses and occurred GEOHYDROLOGY OF GEOTHERMAL SYSTEMS chiefly as ash flows. Because of the withdrawal of such a large volume of material, the roof of the magma chamber collapsed along one or more steeply inclined ring fractures, forming the present caldera. The amount of subsidence was at least 1 km, and the ellip- tical depression thus formed measured about 29 km in an east-west direction by about 15 km north-south. Rhyolite erupted intermittently from at least 12 vents near the center of the caldera, about 40,000 to 100,000 years after the collapse that formed the cal- dera. During the eruptions the west-central part of the caldera floor was uplifted to form a subcircular resur- gent dome about 10 km in diameter. At the same time, the crest of the dome foundered along north- northwest-trending faults to form a keystone graben 5 km wide. The depression formed by the caldera collapse rapidly filled with water to form Long Valley Lake. The lake reached a high level of more than 2,320 m above sea level, and the early rhyolites of the resurgent dome probably erupted into the lake. Relations of lake-terrace deposits, strand lines, and perlitized or hydrated zones in rhyolites to radiometrically dated volcanic rocks indicate that the lake receded at irregu- lar rates and finally was drained completely by downcutting of its outlet sometime within the last 100,000 years. Recent discovery that basaltic ash in lakebeds in the southeast part of the caldera is similar to that at Black Point in the Mono Basin, which has been dated at 13,500 years before present, suggests the presence of a shallow lake as recently as 10,000—15,000 years ago (Roy Bailey, oral commun., 1976). After the formation of the resurgent dome about 0.7 to 0.6 m.y. ago, porphyritic rhyolite erupted from three groups of centers in the moat between the resurgent dome and the walls of the caldera. The eruptions oc- curred at 0.2 m.y. intervals about 0.5, 0.3, and 0.1 m.y. ago, starting in the north moat and proceding in clockwise succession around the resurgent dome to the southeast and then the west moat. Porphyritic rhyodacite (quartz latite) erupted on the southwest rim of the caldera and near the base of the north-west—north caldera walls beginning about 0.2 m.y. ago and ending about 0.05 m.y. ago. The silica content of these lavas, about 59—74 percent, averages lower than that of the earlier rhyolitic rocks, which suggests derivation from a lower level in the differ- entiated and stratified magma chamber. Mammoth Mountain, on the southwest caldera rim, was the largest center of these eruptions. During the same period as the eruptions of rhyoda- cites of the caldera rim and the latest eruptions of rhyolite in the west moat, basaltic to trachyandesitic HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA flows and cinder cones erupted in the west moat about 0.2 to 0.06 m.y. ago. These mafic to intermediate vol- canic rocks are part of a more extensive chain which decreases in age northward from 0.9 m.y. in the Devil’s Postpile area southwest of the caldera to about 0.013 m.y. at Black Point on the north shore of Mono Lake. Chemically and mineralogically, these rocks are simi- lar to many other Cenozoic rocks east of the Sierra Nevada in the Basin and Range province, and they may not be related directly to the Long Valley magma chamber. The most recent volcanic events in the Long Valley caldera were the eruptions of the rhyolite and rhyoda- cite of the Inyo craters and domes. These features are alined on an apparent north-trending fracture along A9 the east front of the Sierra Nevada from the west moat of the Long Valley caldera northward to the Mono Craters. The youngest of the five lava domes in less than 720 years old (Wood, 1975), and the Inyo Craters, which comprise three phreatic explosion pits on the south side of Deer Mountain, suggest the possible pres- ence of rhyodacitic magma in the Long Valley magma chamber as recently as 450 years ago. CHARACTERISTICS OF CALDERA The present dimensions of the caldera floor result from enlargement of the original caldera by slope re- treat of walls caused by slumping and erosion; the main ring fracture lies within the caldera moat (Bailey and others, 1976, p. 731). (See fig. 3.) 37°45' “‘(H ........ I. \ gnaw. . . \~_ _.." a ."dg EXPLANATION " ‘35 q Line of equal thickness of fill, in kilometers 37035, — ooooooooo Inferred trace of caldera ring fracture ’_-\__-___‘ Boundary of caldera floor a l———~l a’ Location of seismic profile (Hill, 1976, fig. 1) l 0 2 4 6 8 10 J Kl LOMETERS 119°00’ 118°45' FIGURE 3,—Map showing thickness of fill in Long Valley caldera. A10 The caldera probably subsided along a single steeply dipping ring fault (Roy Bailey, oral commun., 1976). We infer a position of the ring fault midway across the steepest part of the Bouguer gravity gradient. The area encompassed within the surface trace of the ring fault is 269 km2 (fig. 4). The fracture is assumed to dip 80° inward. Precaldera basement rocks have a seismic P—wave velocity of 6.0 km/s (Hill, 1976). We infer that the basement surface beneath the caldera is the lower (B) horizon on the seismic-refraction profiles of Hill (1976, figs. 7 and 8). The layer overlying basement in the caldera, which has a P-wave velocity from 4.0 to 4.4 km/s (Hill, 197 6), probably comprises densely welded Bishop Tuff and underlying volcanic and sedimentary(?) rocks that GEOHYDROLOGY OF GEOTHERMAL SYSTEMS predate caldera collapse (rhyolite of Glass Mountain and, possibly, volcanic rocks of Pliocene age). Most of the Bishop Tuff within the caldera is believed to be densely welded (Roy Bailey, oral commun., 1976), as was found to be the case in the Republic Geothermal, Inc. well 66—29 (pl. 1). The layer overlying probable welded Bishop Tuff within the caldera, which has a P-wave velocity rang- ing from 2.6 to 3.4 km/s, is believed to comprise rhyo- lites and rhyodacites of the caldera moat (Hill, 1976, p. 750), and, possibly, also basalt (Deadman Creek area) and tuffs intercalated with the flows. The top seismic layer, which has a P-wave velocity from 1.5 to 1.9 km/s (Hill, 1976), probably comprises unconsolidated epiclastic and volcaniclastic deposits and soft pumiceous tuffs. l 37°45' I " , " a . . ’ K - --_.—" \ EXPLANATION A 1.5 ~—/ Line of equal thickness of fill, in kilometers 37°35' ”— -——~_———-‘ ~ Boundary of caldera floor a l————-l a’ Location of seismic profile (Hill, 1976, fig. 1) I 4 6 8 1O #1 KILOMETERS r—N l 119°oo' ‘ 118°45' FIGURE 4.—Map showing thickness of fill above densely welded part of Bishop Tuff in Long Valley caldera. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA We used seismic sections A’—A’ and B’—B’ of Hill (1976, figs. 7 and 8) for primary control in constructing the isopachous maps of caldera fill (figs. 3 and 4, lines a—a’ and b—b’). Isopachs were extended into adjacent parts of the caldera on the basis of an interpretation of gravity data of Kane, Mabey, and Brace (1976a, b). The two gravity lows in the north-central part of the cal- dera (Kane and others, 1976a, b, fig. 2) probably reflect in part thicker caldera fill and in part material having lower density than that in surrounding areas. We es- timate the maximum thickness of caldera fill to be about 4 km near the eastern gravity low and about 3.5 km near the western low. Thickness of fill above the densely welded Bishop Tuff (fig. 4) was estimated by a procedure similar to that used for estimating total thickness of fill, except that little use could be made of the gravity data of Kane, Mabey, and Brace (1976a, b), and the isopachs were extended from the seismic sec- tions chiefly on the basis of inferences made by Bailey (1974; oral commun., 1976) as to the location and amount of uplift on the resurgent dome. Total volume of caldera fill and volume of fill above the welded Bishop Tuff were computed by planimeter- ing the isopachs shown in figures 3 and 4. Incremental volumes were computed as the product of the thickness increment and the geometric mean of the areas en- closed by the two bounding isopachs. The volumes thus calculated then were added to obtain the total volume of fill, as in table 2. Our estimate of the total volume of fill, 868 km", is somewhat greater than the 810 km3 estimated by Muffler and Williams (1976). The difference between the two estimates is due chiefly to the greater thick- ness of fill along the margin of the caldera inferred from our analysis than that inferred by Muffler and Williams (1976, fig. 2). In their interpretation, Muffler and Williams placed more emphasis on the gravity data; in the present analysis, we placed more emphasis on the seismic results of Hill (1976). In modeling the hydrothermal system, it is useful to know the average bulk density and porosity of the caldera fill. Average bulk density of the fill is computed as pm V_Mo pb =—V~ where pm is the bulk density of the rocks surrounding the caldera (2.67 g/cm3), V is the volume of fill (8.68 x 1017 cm3), and M0 is the mass deficiency calculated by Gauss’ theorem from the residual gravity anomaly (3.33 X 1017 g; Kane and others, 1976, p. 759). Sub- stituting these values we obtain an average wet bulk density of the caldera fill of 2.29 g/cm3. All Average bulk porosity of caldera fill is calculated as p _ ¢ = —g _ p" where d: is porosity (dimensionless), p is grain density, Pb is wet bulk density, and Pw is density of interstitial water. Average grain density of the caldera fill cannot be estimated within narrow limits from available data. Sheridan (1965) found an average grain density of 2.45 g/cm3 on the basis of specific-gravity and porosity meas- urements for 51 samples of ashflow tuff from outcrop areas of the Bishop Tuff southeast of Long Valley. The standard deviation from the mean density of Sheri- dan’s samples was only 0.01 g/cm3, even though bulk density ranged widely, from 0.79 to 2.38 g/cm". How- ever, an average grain density of 2.65 g/cm3 was meas- ured by the Survey’s Hydrologic Laboratory for cut- tings of densely welded Bishop Tuff from the Republic Geothermal test well. Muffler and Williams (1976, p. 723) assumed values of 2.50 and 2.60 g/cm3 on the basis of inferred equivalence values of the Bishop Tuff to those from two ash-flow tuffs from the Creede caldera, Colorado (Ratté and Steven, 1967). TABLE 2.—Calculations of volume and average thickness of fill in Long Valley caldera Area with shallower Depfiéinge ‘3??? 7133" Total volume of fill: 0—1.0 369 363 1.0 —1.5 357 170 1.5 -2.0 324 135 2.0 —2.5 224 94 2.5 —3.0 159 66 3.0 —3.5 111 32 3.5 -4.0 36 7 >4.0 6 1 0 —>4.0 868 Average thickness = 868 km3/369 km2 = 2.35 km Volume of fill above densely welded Bishop Tuff: 0. —0.5 369 174 0.5 —1.0 330 119 1.0 —1.5 172 42 >1.5 42 6 0 —>1.5 341 Average thickness = 341 km3/369 km2 = 0.92 km Volume of densely welded Bishop Tuff and older fill = 868 km3 — 341 km3 Average thickness = 527 km3/369 km2 =, 1.43 km A12 In calculating bulk porosity, we use a lower limit of 2.45 g/cm3 and an upper limit of 2.65 g/cm3 for the average grain density of the caldera fill. Wet bulk den- sity of the fill is 2.29 g/cm3, as calculated above. Den- sity of interstitial water is taken as 0.96 g/cm3 on the basis of an assumed average temperature of 100°C. Bulk porosity, therefore, ranges from 0.11 to 0.21. Both theoretical considerations and empirical evi- dence suggest that porosity is not uniformly distrib- uted throughout the caldera fill. In the following analysis, lateral variations in porosity are not consid- ered. However, the average porosity of the densely welded Bishop Tuff and possible underlying older rocks is believed to be much less than that of the post-Bishop caldera fill. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS In the analysis below, we assume, on the basis of test-drilling and seismic-refraction results, that one- half of the caldera fill above the densely welded Bishop Tuff consists of flows having a porosity of 0.10 and that the other half consists of tuf‘fs and sedimentary rocks having a porosity of 0.45 (see table 5). The average porosity of the post-Bishop caldera fill is, therefore, 0.28. The average porosity of the densely welded Bishop Tuff and older fill is computed as (bl Z 1 _ d) 2 Z 2 4’ _ z1 — 22 where (#1 = bulk porosity of total fill, (112 = bulk porosity of post-Bishop caldera fill, Z1 = average thick- ness of total fill, and Z2 = average thickness of post- 37°45’ — 37°35’ -- EXPLANATION ————— Margin of caldera floor —1000— Line of equal precipitation, in millimeters °|___j__'i) KILOMETERS | 119°00' FIGURE 5.—Map showing average annual precipitation in the Long Valley basin. Data obtained from Rantz (1972) for altitudes above 3,000 m, and from estimates of annual precipitation at 71 locations with altitudes below 3,000 m derived from US. Weather Service climatological records and the work of Spreen (1947). HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA Bishop caldera fill. For a range in total fill porosity of 0.1 1—0.21, we obtain the average porosity of the welded Bishop Tuff in the range from 0.00 to 0.16. These values, being residuals in a series of calcula- tions involving uncertain assumptions, do not help significantly to define the probable average bulk porosity of the densely welded Bishop Tuf‘f within the caldera. More useful information consists of published values for porosity of the welded phase of the Bishop Tuff in exposures southeast of Long Valley, which range from 0.026 to 0.079 (Gilbert, 1938, Bateman, 1965). Because these values are for small volumes of samples collected in the field, they probably do not in- clude significant fracture porosity, which would be present in much larger volumes of rock. The bulk porosity of welded Bishop 'Iiiff in the areas of exposures may, therefore, be somewhat higher than the pub- lished values. However, within the caldera, because of the greater probable thickness of the Bishop ’Ihff than outside the caldera, and because of the overburden, which averages nearly 1 km in thickness, the average bulk porosity of the densely welded Bishop Tuff pro- bably is no greater than the higher published values. In our model, discussed in a later section, we use a porosity of 0. 10 for the upper part of the densely welded 1200 - 1 100 - 1000 - 900 *- 800 ‘ 700 - 600 _ 500 '- 400 - 300 - 200 — 100 - PRECIPITATION, IN MILLIMETERS _, |‘-——Mammoth, California at Lake Mary Store _ OIlllllllllllllllLlllllIll A13 Bishop Tufl‘ (layer 2, table 14) and 0.05 for the lower part (layer 3, table 14), HYDROLOGIC SETTING CLIMATE AND PRECIPITATION The climate in the Long Valley region is influenced greatly by the Sierra Nevada. Precipitation is derived chiefly from storms that originate over the Pacific Ocean and move eastward. About 70—80 percent of the total precipitation in the mountains and a somewhat smaller proportion in the caldera falls as snow from November through April. The remainder falls primar- ily during local warm-season thunderstorms. Because of the orographic effect of the Sierra Nevada, a “shadow” effect is produced east and north of the crest, and precipitation in the caldera is less than near the crest. Average annual precipitation within the Long Valley drainage basin ranges from more than 1,500 mm along a part of the Sierra Nevada crest to less than 300 mm on Lake Crowley, in the southeast part of Long Valley (fig. 5). Variations in annual precipitation from year to year are substantial, as indicated by data for stations at Lake Mary Store and Long Valley Dam (fig. 6). IIllllllJ_LlIlllIllIllIlI II 1921 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 WATER YEAR 700 600 500 400 300 200 1 00 Crooked Creek at PRECIPITATION IN MILLIMETERS . ——>l‘-— Long Valley at the Dam Little Round Valley oIllllllllllllllllllllllIIIjllllllllllllllllllllll|1ll_ 1921 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 WATER YEAR FIGURE 6.—Annual precipitation at Lake Mary Store and Long Valley Dam. Data from written communication, Los Angeles Department of Water and Power, 1976. A14 The lack of adequate precipitation data in the Long Valley basin required the generation of additional data by an indirect method to draw figure 5. Data presented in figure 7 indicate that precipitation is not a function of altitude alone. Spreen (1947) showed that, along with altitude, the amount of precipitation is influenced by other parameters such as slope, orientation, and ex- posure. On the basis of Spreen’s work, we constructed a family of curves relating precipitation to these factors, using available data from along the east slope of the Sierra Nevada. With the aid of these curves, we esti- mated the annual precipitation for 71 data points below 3,000 m altitude; for altitudes above 3,000 m, we used the annual precipitation shown by Rantz (1972). These data were contoured to produce the isohyetal map in figure 5. Little direct information is available concerning air temperatures within the caldera. Indirect evidence from measurements of temperature at shallow depths in test wells suggests that the average annual temper- ature at the land surface ranges from about 6°C at higher altitudes in the west-central part of the caldera 300C:L I I I I I I .Rock Creek 02mm Lake .Bishop Union Carbide .Gom Lake Lake Mary 0 Bishop Creek _ ‘ O Bodio O Lundv Lake In 5 .Long Valley Dam - E 2000 _ . .Mono Lake 2 Bridgeport Darn E 2‘ 9 I- - - < > Lu : . Bishop Airport 2 9 .— _ _ < 1000 l— U) 0 I I l I I I I 0 I 00 200 300 400 500 600 700 800 ANNUAL PRECIPITATION, IN MILLIMETERS FIGURE 7.—Relation between precipitation and altitude at stations on the east slope of the central Sierra Nevada. Data from US. Weather Service climatological records, and written communica- tion, Los Angeles Department of Water and Power, 1976. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS floor to perhaps 10°C at lower altitudes near Lake Crowley, in the southeast. In heat-flow calculations, described in a later section, we use an average value of 8°C. WATER BUDGET A water budget for the Long Valley drainage basin for the water years 1964—74 is presented in table 3. In calculating the budget, we assumed that (1) all the water in the hydrologic system originates as rain or snow and (2) inflow to the basin equals outflow during the budget period except for change in storage in Lake Crowley. The first assumption is not valid if a significant fraction of the hot water in the hydrother- mal reservoir has a magmatic source. However, the results of isotope studies of Long Valley and of similar hydrothermal systems throughout the world (White, 1968) indicates that the proportion of magmatic water in the total inflow or outflow in the basin is very small; it can be neglected in this discussion. The second as- sumption is valid only if the change in ground-water storage in the basin is insignificant in relation to total TABLE 3.—Water budget for the Long Valley drainage basin for water years 1964-74 Inflow: L/s hm3/yr Owens River below East Portal1 ______________ 5,140 162.3 Hot Creek at the gorge2 ______________________ 1,810 57.1 McGee Creek ________________________________ 930 29.4 Convict Creek ______________________________ 805 25.4 Hilton Creek ________________________________ 344 10.9 Rock Creek Diversion3 ______________________ 293 9.3 Laurel Creek4 ______________________________ 168 5.3 Crooked Creek ________ _______________________ 113 3.6 Precipitation on Lake Crowley5 _______________ 168 5.3 Ungaged inflow6 ____________________________ 1,090 34.5 Total inflow (rounded) __________________ 10,900 fl) Outflow: 7 Main venturi at Long Valley Dam ____________ 9,670 305.3 Evaporation of shallow groundwater7 __________ 508 16.0 Evaporation from irrigated grasslandB ‘ ........ 430 13.6 Evaporation from Lake Crowley9 ______________ 469 14.8 Owens River gorge, Main Weir ________________ 121 3.8 Ground-water loss to regional system __________ 110 3.5 Change in reservoir storage, 1964—1974 ________ —90 —2.8 Total outflow (rounded) ________________ 11,200 £470 ‘lncludes water imported from Mono Lake basin: 113.2 hma/hr (3,590 Us) average for the 11-year period of record. 2Includes discharge from the Hatchery springs and hot springs in Hot Creek gorge. “Period of record: 1966—1974 water years. 4Period of record: 7/70—7/73, from Calif. DWR (1973). 5Average 25 mm on 21 km2 of lake surface. f"Includes recoverable water from ungaged drainage and all spring discharge except as described in 2 above. 7From 85 km2 area where water table is less than 2.4 m. ”Same as from a lake—690 mm; 19 km2 irrigated. 9Average 690 mm from 21 km2 lake surface. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA inflow or outflow during the budget period. Although specific data for the budget period are lacking, we be- lieve this condition to be reasonably well fulfilled. On the basis of observed changes in water level in shallow test holes during 1972—74, a change of 1 m in average water level throughout the caldera floor from the be- ginning of 1964 water year to the end of water year 1974 should be an upper limit. Assuming water-table (unconfined) conditions and an average storage coeffi- cient of 0.3, a l-m change in water level over the 450 km2 of the caldera floor would amount to 135 hm3—less than 4 percent of the estimated total inflow or outflow of about 3,800 hm3 to the drainage basin during the 11-year budget period. The inflow items in the budget (table 3) include streamflow measured at eight gaging stations, an es- timate of streamflow from several ungaged areas, and A15 precipitation on Lake Crowley. Except for the precipi- tation directly on Lake Crowley, the annual volume of precipitation in the basin was not estimated as a dis- crete budget item. Instead, the fraction of the precipi- tation that runs off in the stream channels is measured at the eight gaging stations shown in plate 1 and figure 8. A part of the measured flow is direct runoff after storms or from melting snow, but the remainder is base runoff furnished by ground-water discharge into the stream channels above the gaging stations. In this sec- tion of the report we do not estimate either the amount of base runoff or the proportions of the base runoff that are supplied from the deep hydrothermal reservoir (deep subsystem) and from the shallow nonthermal ground-water reservoir (shallow subsystem). Instead, we include estimates of the discharge from the hydro- thermal reservoir in later parts of the report. Deadman, Glass, and Dry Creeks \ Upper Owens River 37°45’ _ 37°35’ — O 5 1O |__L__l KILOMETERS l Hot Creek C Gorge 119°00’ 118°45' FIGURE 8.—Map showing tributary drainage areas and streamflow-measuring stations in the Long Valley basin. A16 Ground-water underfiow into Lake Crowley is not included as an item of inflow because hydrologic infor- mation was insufficient to determine this quantity di- rectly. It is likely, however, that ground-water flow into the lake and into the Owens River below the East Portal gage accounts for a significant part of the com- puted 11 hm3/yr (350 US) difference between total inflow and outflow. In the section, “Thermal charac- teristics of the hydrothermal system,” we estimate a discharge of 4.8 hm3/yr (120 US) of warm ground water (50°C) southeastward across the caldera boundary into southwestern Lake Crowley. Most streamflow originates in the Sierra Nevada along the west and south margins of the basin, where precipitation exceeds potential evapotranspiration and where snowpacks last into the late spring in most years. Seven Sierra streams are measured by the Los Angeles Department of Water and Power. Approximately 30 percent of the measured runoff from the Sierra Nevada originates in the northwest part of the basin drained by Deadman, Dry, and Glass creeks. In the discussion below, this part of the basin is called the upper Owens River area. The flow from these three Owens River tributaries is not measured directly but is calculated from the difference measured at the gages at Owens River below East Portal and at the East Portal of the Mono Tunnel. The average discharge from this drainage area is about 49.1 hm3/yr (1,560 US). Mammoth, Convict, and McGee creeks contribute more than half of the measured runoff to the basin from the Sierra Nevada or about 77.7 hm3/yr (2,460 US). The remainder of the runoff from three small Sierra Nevada tributaries amounts to only about 14 percent of the measured total. Additional runoff, not measured, is generated from about 44 km2 along the lower slope of the Sierra Nevada south of the caldera and from about 127 lain2 between Bald Mountain and Glass Mountain Ridge north of the caldera and around to Long Valley Dam on GEOHYDROLOGY OF GEOTHERMAL SYSTEMS the east. Additional runoff, also ungaged, occurs from 55 km2 of the resurgent dome. To estimate the quantity of water entering the basin as ungaged runoff, the relation between average an- nual precipitation and average annual runoff, shown in figure 9, was used. The area-weighted mean annual precipitation for each of the gaged Sierra Nevada trib- utary basins was obtained from figures 5 and 8, and plotted versus the average annual runoff measured for the corresponding basin. To construct the curve, water losses from runoff above the gaging stations were as- sumed to be negligible and measured flows were used to represent recoverable water in the water-budget computations. The curve is constrained at the left by our assumption, based on the work of Blaney (1933) and others, that runoff is near zero where annual pre- cipitation is less than 250 mm. Estimates of annual runoff (recoverable water) from all the ungaged areas in the Long Valley drainage were computed using figures 5 and 9 and are summarized in table 4. The total estimated runoff from the ungaged areas is 34.5 hm3/yr (1,090 US). On the basis of the curve shown in figure 9, data points for the upper Owens River, Mammoth Creek, and Laurel creek are anomalous. This could be caused by precipitation values which are too high or by runoff which was not measured. Precipitation values for the Mammoth Creek drainage basin are from two precipi- tation gages and five snow courses and afford better control for that basin than for any of the other drainage basins. The relations for the adjacent upper Owens River and Laurel Creek drainage basins also benefit, to a lesser extent, from these data. Thus, if the precipitation-runoff relation in figure 9 is valid, a loss of runoff by downward percolation is indicated for the Mammoth Creek, Laurel Creek, and upper Owens River drainage basins. If the data point for Mammoth Creek is shifted laterally to intersect the curve in figure 9, an increased runoff of about 170 mm or 18.9 TABLE 4.—Estimated runoff from ungaged areas in the Long Valley drainage basin Drainage Bald Mtn-Glass Mountain Ridge ____________________________ East Rim Resurgent Dome ____________________________________________ Sierra Nevada: Ungaged 1 ______________________________________________ Ungaged 2 ______________________________________________ Ungaged 3 ______________________________________________ Ungaged 4 ______________________________________________ Ungaged 5 ______________________________________________ Valley Floor ________________________________________________ Total or average (rounded) ____________________________ Average annual runoff Precip- Area itation (kmzl (mm) (mm) (L/s) IhmJ/yri 81.1 670 190 485 15.30 46.2 410 46 66 2.10 55.2 530 110 188 5.92 3.4 510 94 10 .32 7.3 560 120 28 .89 16.6 560 120 64 2.02 10.1 510 94 30 .95 7.0 510 94 21 .65 215.0 360 31 207 6.54 m W) Ti; 1W) ”374.7, HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA hm3/yr (600 US) is indicated. Similarly, for Laurel Creek the increase in runoff is 66 mm or 0.99 hm3/yr (31 US), and for the upper Owens River it is 56 mm or 10.6 hm3/yr (336 US). Most of the water lost above the gaging stations on Mammoth and Laurel Creeks probably migrates downward through fractures in the shallow Mammoth Lakes basalts described by Bailey (1974) and, along with other shallow ground water and possibly a small amount of hot water (Sorey, 1975b), furnishes most of the 30.4 hm3/yr (964 US) discharge in the springs at the Hot Creek Fish Hatchery. Test drilling shows that basalts in the upper Owens River drainage area are also fractured, and temperature profiles in heat-flow A17 holes indicate downward movement of cold water to depths of at least 200 m around the northwest periph- ery of the caldera, as discussed in a later part of this report. Thus, the indicated 10.7 hma/yr (336 L/s) of water lost in the upper Owens River drainage basin is a possible source of recharge to the deep hydrothermal system, from which a discharge of 200—300 kg/s is es- timated (see section, “Chemical Geothermometers”). All this additional water amounts to about 30.5 hm3/yr (965 L/s) but is not included as a separate item in the hydrologic budget. Instead, the flows from the Hatchery springs and the Hot Creek gorge springs, which include most of the hot spring discharge in the caldera, are included in the measured flow of Hot 1300 I I I T I 1200 — _ 1100 >— 1000 — U) m LIJ t; 900 — E 3 -_J 2 800 '- E 2 9 700 —— p. E g 600 _ / STATIONS — E / 7. Upper Owens River 3' 500 - / / 2. Mammoth Creek — 2 3. Laurel Creek 2 2 400 — / 4. Convict Creek _ E / 5. McGee Creek 2 . 300 _/ 6. Hilton Creek __ 200 — - 100 — .. I I J J l 0 100 200 300 400 500’ 600 MEAN ANNUAL RUNOFF, IN MILLIMETERS FIGURE 9.—Relation between precipitation and runoff for streams within Long Valley drainage basin. A18 Creek at the downstream end of the gorge. On plate 1, the springs in Hot Creek gorge are in T3‘S/R28E—825, along a 1.5-km reach of Hot Creek bounded by the NW-SE trending graben faults. Since 1947, about 69.9 hm3/yr (2,220 US) of water has been imported annually from Mono Lake basin north of Long Valley. During water years 1964—74, im- ported water from the Mono Lake basin averaged 113 hm3/yr (3,590 US). The imported water enters the Long Valley drainage basin through the Mono Craters Tun- nel, flows into the Owens River at East Portal and then to Lake Crowley where it is stored, and subsequently is released at Long Valley Dam to regulate the flow in the Los Angeles Aqueduct. Since 1965, an additional 9.3 hm3/yr (293 US) has been diverted into the Long Valley system from the Rock Creek drainage basin adjacent to the Long Valley basin on the southeast. Most discharge from the Long Valley drainage basin consists of controlled releases from Lake Crowley. The releases are measured with a Venturi meter by the Los Angeles Department of Water and Power. Since the completion of Long Valley Dam in 1941, about 205 hm3/yr (6,490 US) has been discharged from the basin. During water years 1964—74, the discharge averaged about 306 hm3/yr (9,690 US). The increase in outflow for the modern 11—year base period is due to the in- crease in the volume of water imported from the Mono Lake basin. An additional 3.8 hm3/yr (121 US) was re— leased through the main weir at the dam. Natural discharge of ground water occurs as evap- oration from soil where the water table is near the surface and as transpiration from native vegetation. The estimate in table 3 of 16.0 hm3/yr (508 US) of evap- oration of shallow ground water was based on the work of White (1932), Houk (1951), and McDonald and Hughes (1968). Transpiration of ground water by phreatophytes was neglected except for the 13.6 hm3/yr (430 US) attributable to areas of saltgrass (Distichlis spicata) cover along the Owens River and between A1- kali Lakes and Lake Crowley where the water table is less than 1 m below land surface. An investigation along an 11-km reach of the Owens River gorge below Long Valley Dam showed no springs or seeps which might indicate subsurface outflow of water from the Long Valley drainage basin. However, in the modeling study described in the section, “Analysis of Conceptual Model of Hydrothermal Sys- tem,” a ground-water discharge of 3.47 hm3/yr (110 US) from the deep ground-water subsystem (the hydro- thermal reservoir), through the southeast rim of the caldera, is included. This provides an outlet for the flux of relatively cool ground water which is indicated by the temperature profile in the Republic Geothermal, Inc. test well to be moving through the Bishop Tuff GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS under the southeastern part of the caldera. Some natural discharge occurs from the lake surface and from springs, pools, and streams as evaporation. The evaporative loss of water is measured at one sta- tion near the dam by the Los Angeles Department of Water and Power. The long-term average annual pan evaporation losses measured at the dam are about 1,000 mm. Using a pan coefficient of 0.7, this converts to an equivalent lake evaporation of about 700 mm. For a lake the size of Lake Crowley, this amounts to an average evaporation of about 14.8 hma/yr (469 L/s). At the end of water year 1974, 31.2 hm3 more water was stored in Lake Crowley than at the beginning of water year 1964, the start of the 11-year budget period. This amounts to an average annual increase of 2.84 hm3, which is shown in table 3 as a negative outflow item. As shown in table 3, the total average annual out- flow of 354 hm3 exceeds the total average annual inflow of 343 hm3 by only about 3.2 percent, which indicates fairly good agreement. Although the largest items of both inflow and outflow are streamflows measured by standard techniques for which errors of less than 5 per- cent of flows for periods of several years are to be ex- pected, some of the estimated items are sizable frac- tions of the totals. Large errors in these estimates seem to be precluded by the fairly small imbalance in total outflow and inflow. Furthermore, as discussed pre- viously, part or all of the 11 hm3/yr imbalance could be attributed to unmeasured ground-water underflow in Lake Crowley. SHALLOW AND DEEP GROUND-WATER SUBSYSTEMS For purposes of the present analysis, it is useful to consider the ground-water system in the Long Valley caldera as comprising two major parts: (1) a shallow subsystem in which temperatures are not much higher than ambient land-surface temperatures, ground- water flow paths are relatively short and direct from areas of recharge to areas of discharge, and, except where increased locally by evapotranspiration in the areas of shallow water table, the concentrations of dis- solved solids are relatively low; and (2) a deep subsys- tem in which temperatures are commonly much higher than ambient surface temperatures, ground-water flow paths are relatively long and circuitous, and concen— trations of dissolved solids—especially some of the al- kali chlorides, silica, boron, and arsenic—are rela- tively high. The two parts of the system are not everywhere sharply delineated; but, in general, the shallow subsystem is contained in the fill above the densely welded Bishop Tuff. This part of the caldera fill HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA consists chiefly of intercalated lava flows and tuffs in the western part and chiefly of tuffs and lacustrine and fluvial sediments in the eastern part. The lower part of the fill, which contains the deep subsystem, probably consists chiefly of densely welded Bishop Tuff with subordinate amounts of pre-Bishop Tuff volcanic and (or) metasedimentary rocks. As described previously, the average porosity of the materials of the shallow subsystem probably is much greater than that of the underlying densely welded Bishop Tuff of the deep subsystem. Although the shal- low materials include some volcanic flows of low porosity, especially in the west part of the caldera, the intergranular porosity of the tuffs and sediments is high (see table 6). By contrast, the intergranular porosity of the Bishop Tuff in the caldera is very low, and most of the pore space, which probably averages less than 10 percent of the total volume, is in secondary A19 fractures associated with contraction of the rock on cooling or with deformation. Ground water in the shallow subsystem circulates through the relatively coarse-grained, more permeable sediments and, locally, through brecciated or faulted zones in the volcanic flows. Much of the material, how- ever, is clayey, tuffaceous, or altered sedimentary de- posits or dense flows of extremely low permeability (see table 6). Such material forms confining layers which separate aquifers in the shallow subsystem and, more significantly, probably separate the shallow and deep subsystems from each other except locally where high-angle normal faults provide interconnecting channelways for upward- or downward-flowing ground water. Places where flow is assumed to occur vetween the shallow and deep subsystems are described later in the section, “Analysis of Conceptual Model of Hy- drothermal System.” Table 5,—Laboratory data for selected cores from test holes in Long Valley caldera [Sample analyses run by Hydrologic Laboratory, WRD, Denver, Colo., except as noted below] Specific gravit P M. t Tgergvslt “9°3in - . . 1 ermea i l y‘ can no Ni 2 reels 1V1 y N... “13.51.23“ ”a?“ miss; $275,113? (222:3. 35 percent) and thermal conductivities are low, so that the rocks of the shallow subsystem act as a thermal blanket, producing high temperatures in the underlying rocks of the deep subsystem. Vertical per- meabilities measured on the cores are extremely var- ied, ranging from 5.5 X 10’4 to 8.3 X 10+3 millidarcys. The lowest values were obtained for rocks and deposits that are hydrothermally altered to varying degrees. Such altered materials were penetrated in core holes A21 within the present area of hot-spring activity and on the resurgent dome in Little Antelope Valley. The oc- currences of altered deposits suggest that upward movement of hot water from the deep subsystem is restricted to zones where recent fault movements pro- vide conduits through the capping rocks of the shallow system. ‘ Hydrothermal alteration is much less widespread in the west part of the caldera than in the east part. Ac- tive fumaroles and zones of acid alteration occur on Mammoth Mountain and at Casa Diablo Hot Springs, and a zone of hydrothermal alteration parallels the southwest caldera wall near the base of Mammoth Mountain (Bailey and others, 1976). Pyritized rhyolite was found at a depth of 210 m in drill hole DC in the northwest moat. However, no evidence of alteration was found in CH—S (325 m deep) in Smokey Bear Flat T 37°45' — , t \\ \ / I‘V‘i‘i’b \\ \ l\\\ \ ) “3&1“ \\\\\ \\\}l/ l )J Il’ \%,i°\\\\ \\\\\ iii // 212/ r $091 '\ %\°\ \\\\\\\\\ ,l 72,4 ” N \\ // \ meg“ EXPLANATION 02102 0,2389 Well (dot) _or spring (dot with tail) 37035! and altitude of water level, in meters above mean sea level 2440 Water-table contour Altitude, in meters above mean sea level. Dashed where uncertain l 10 J Kl LOMETERS l 119°00' FIGURE 10.—Map of Long Valley caldera showing configuration ofthe water table, spring 1973. 118°45' Data from Lewis. 1974. Contour interval 20 m. A22 and CH—9 (92 m deep) in the upper Dry Creek drainage basin. Permeable zones in the shallow ground-water subsystem in the west appear to be restricted to chan- nels in the flow rocks, resulting from some combination of fractures, cooling joints, and brecciated surfaces. Large lateral flows of cold water in these layers also appear to be responsible for the isothermal nature of the temperature profiles in core holes CH—8, CH—9, and DC, where significant recharge from the surface through more than 35 m of unsaturated materials ap- pears to be unlikely. Secondary permeability caused by recent faulting may be hydraulically effective only in the more compe- tent rocks, such as the welded Bishop Tuff and pre- Tertiary basement rocks, and apparently sediments which have undergone alteration (silicification and zeolitization). Drilling data and lab tests on cores in- dicate that the primary permeability of the postcaldera tuffs and altered sediments and the unfractured flow rocks is relatively low. Therefore, in our conceptual model of the hydrothermal system, the fractured, welded Bishop Tuff and the underlying basement rocks are assumed to provide the hot-water reservoir. If this is correct, these rocks should also be altered, and main- tenance of permeable flow channels may have required either or both continued tectonic activity or successive invasion of unaltered rock by hydrothermal fluids. Ground-water flow in the overlying shallow subsystem is neglected except. in recharge areas around the cal- id’era rim and in the hot-spring discharge area. Second- ary effects, such as mixing of shallow ground water with upflowing hot water and shallow subsurface dis- charge of hot water are discussed in a later section. These effects were not included in numerical simula- tions with the model. GEOCHEMICAL CHARACTERISTICS OF THE HYDROTHERMAL SYSTEM CHEMICAL PATTERNS Surface and ground waters of Long Valley have dis- tinct differences in chemical composition. In large part, the variations in chemical character reflect sharp dif- ferences between deep hot waters and shallow cold waters, and they also indicate the local proportions of these two contrasting types. The California Depart- ment of Water Resources (1967) has identified four principal types of water on the basis of proportions of major dissolved constituents and concentration of dis- solved solids: (1) cold surface and ground waters along the caldera margins in which Ca and H003 are the predominant constituents and the concentration of dis- solved solids is relatively low—usually 30—100 mg/L; GEOHYDROLOGY OF GEOTHERMAL SYSTEMS (2) thermal waters in hot springs in which Na, H003, and often also C1 are predominant and the concentra- tion of dissolved solids is moderate—usually about 1,000—4,000 mg/L; (3) alkali lakes and adjacent ground waters in the east-central caldera, where the water table is near the surface, in which Na and H003 or C03 are predominant and concentration of dissolved solids is relatively high—about 30,000 mg/L; and (4) warm springs and related shallow ground waters that repre- sent mixtures of (1) and (2). Except for type 3 above—a water that results from concentration of other waters by evaporation and transpiration—the types described above are useful is assessing the changes in water chemistry that result as hot geothermal fluid (type 2) mixes with cold, shallow ground water (type 1). In addition, concentrations of B and Cl may be used to estimate the degree of mixing of hot and cold ground waters (Sorey and Lewis, 1976; Mariner and Willey, 1976). Concentrations of B range from less than 0.5 mg/L in waters from springs and wells where tempera- tures are 10°C or less to 14 mg/L in waters from hot springs and geothermal test wells (table 6). Concen- trations of chloride in the same waters range from less than 5 mg/L to 280 mg/L. Molal ratios of Cl/B are 5.8 t 0.2 for most thermal-spring waters, which probably in- dicates a common source at depth. The origin of B and C1 in the hydrothermal waters have been attributed in part to magmatic fluid rising from a magma chamber (California Department of Water Resources, 1967, p. 21). However, we believe that, although a magmatic contribution to circulating meteoric water cannot be ruled out, its proportion must be low because of the meteoric nature of the D and 180 content of the hot- spring waters. Calculations of rock volumes required for leaching of B, presented later, indicate that a mag- matic contribution is not required to supply the ob- served discharge of B. CHEMICAL GEOTHERMOMETERS Results of applying the silica geothermometer of Fournier and Rowe (1966) and the Na-K-Ca geother- mometer of Fournier and Truesdell (1973) to estimate the temperatures of reservoir(s) feeding the hot springs and wells in Long Valley, are given in table 7. Except for the Hot Bubbling Pool, which cools significantly by evaporation, reservoir temperatures based on the cation geothermometer exceed those based on silica. This indicates that the hot—spring waters are mixtures of hot and cold components which have not reequili- brated after mixing because the cation geothermome- ter, being based on ratios of constituents, is less sensi- tive than the silica geothermometer to changes in abso- lute concentration. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA A23 TABLE 6.—Selected chemical constituents, temperatures, and isotopic compositions for thermal and non thermal waters from Long Valley [Data for features a through j from Mariner and Willey (1976); sampled in 1972. Chemical data for features 1, m, and v from Lewis (1974). Features it through t collected by M. Sorey (September 1975), “0 analyzed by N. Nehring, D analyzed by D. White, Cl analyzed by L. Tanner. Features u through 2 collected by B. Evans (June, 1976), D analyzed by D. White. Chemical concentrations are in milligrams per liter] T. surf. Location Locality SiOz Na K Ca 1l‘lCOa Cl B (°C) 8D 5130 a. 3S/28E—13ES3 Little Hot Creek 110 410 30.0 50.0 736 200 10.6 79 — 121.8 — 15.84 b. 3S/28E—32E9 Casa Diablo Well 340 390 45.0 0.9 449 280 214.0 94 — 115.8 —14.16 c. 3S/28E—35ES1 ‘ Hot Bubbling Pool 300 380 25.0 3.3 467 250 13.0 60 — 111.0 — 12.44 d. 3S/29E—21PSl Big Alkali Lake 250 310 37.0 25.0 829 150 7.7 56 —— 123.9 — 16.17 e. 3S/29E—28HSl SE. of Big Alkali Lake 240 400 43.0 22.0 846 170 8.6 49 —123.4 —15.85 f. 3S/29E—31ASI N. of Whitmore Hot Springs 150 310 22.0 15.0 520 170 8.1 58 —121.2 —15.23 g. 3S/29E—34KSl W. of Lake Crowley 205 320 28.0 23.0 696 150 6.2 41 ~124.9 —16.08 h. 38/28E—25H3 Hot Creek Gorge 150 400 24.0 1.6 597 225 10.3 90 — 120.3 — 14.83 i. 2S/27E—25ASl Big Springs 58 23 4.0 5.1 90 6 0.4 11 -115.4 —15.89 j. 38/29E—13C1 Well near N.E. rim 64 28 1.3 5.3 117 3 0.2 10 — 129.5 -17.07 k. 4S/29E—6P S. of Whitmore Hot Springs —- — — — — <10 —- 16 —120.3 —15.26 1. 4S/28E—9FS1 E. of Laurel Canyon — 5 1.0 16.0 46 1 0.0 12 —127.7 —16.68 m. 3S/28E—35KSI Fish Hatche 33 15 3.0 13.0 95 3 0.1 13 —121.0 —15.89 11. 28/28E—14M McLaughlinlCreek — — —— — — <10 —— 8 —130.0 —16.87 0. 2S/27E—27B Deadman Creek at 395 — — — —— — < 10 — 13 — 124.4 — 16.01 p. 3S/27E—7P Deadman Tributary — — — — — <10 — 7 —111.8 —14.71 r. 3S/26E—36P Minaret Summit — — — — — <10 — 5 —110.7 —14.87 s. 28/29E—21D N.W. of O’Harrel Canyon — — — — — <10 — 7 —130.9 — 16.91 t. 4S/27E—6A Mammoth Mountain —- — — — — < 10 — 2 — 109.2 — 14.52 u. 2S/27E—180 Hartley Springs — — — — — —— - 4 — 126.3 — v. 3S/30E—22BSI Watterson Troughs 63 14 3 8 11 0 63 5 11 0 11 — 135.0 — w. 4S/28E—22P Spring above Convict Lake — — — — — — —— 13 —125.3 — x. 3S/29E—2A Well near N.E. rim — —— — — -— — — 14 —129.2 — y. 3S/3OE—5J Wilfred Creek — — — — — — —- 13 — 129.2 — z. 28/29E31 Well E. of Arcularius Ranch — — — —- — —— — 10 —134.0 — 'Total alkalinity calculated as HCOn. 2Average value for 12 samples from Mariner and Willey (1976) and CDWR (1967). 3Corresponds with sample 38/28E—25AS4 in Mariner and Willey (1976, p. 794), which is incorrectly located. Applications of mixing models to estimate mixing ratios and the temperature of the hot-water component at depth have been made by Mariner and Willey (1976), Sorey and Lewis (1976), and Fournier, Sorey, Mariner, and Truesdell (1976). Reservoir temperatures esti- mated from the silica mixing model of Fournier and Truesdell and the silica-chloride mixing model devel- oped by A. H. Truesdell (Sorey and Lewis, 1976) range from 200°C to 225°C and average 210°C. These models suggested that a single reservoir with water similar to that obtained from the Casa Diablo well (that is, 219°C, TABLE 7,—Temperatures of reservoirs feeding hot springs in Long Valley estimated by chemical geothermometers ‘TSiOz . 2TNa'K—Ca Spring or well (°C) (°C) a. Little Hot Creek Spring ______________ 143 172 b. Casa Diablo Well ____________________ 219 3238 c. Hot Bubbling Pool ____________________ 209 4189 d. 38/29E—21PSI ________________________ 196 200 e. 3S/29E—28HSI ________________________ 193 200 f. 3S/29E—31ASI ________________________ 161 176 g. 3S/29E—34KSI ________________________ 182 184 h. 153 192 Hot Creek Gorge Spring ______________ ‘By the method of Fournier and Rowe 11966). 2By the method of Fournier and Truesdell (1973). 3Affected by calcite deposition. "Affected by evaporation. 14 mg/L B, 280 mg/L Cl) underlies or at least feeds all the major hot springs in the caldera. A more recent anyalysis of the geothermometer and isotope data by Fournier, Sorey, and Mariner (1976) using chloride-enthalpy relations suggests a somewhat more complicated circulation system, including a deep aquifer with water at 282°C, 375 mg/L Cl, and 18.8 mg/L B. This water could yield the 219°C Casa Diablo water bs mixing with cold water, and the 201°C Hot Creek gorge water by boiling and steam loss. Oxygen and deuterium isotopic data, as discussed later, are consistent with this interpretation, as are reservoir temperature estimates of 269°C—273°C based on oxygen-isotope composition of dissolved sulfate in thermal waters (McKenzie and Truesdell, 1977; Fournier and others, 1976). Sorey and Lewis (1976) compared the concentrations of B and C1 in each of the major thermal springs with those in the sample from the Casa Diablo well to calcu— late the hot-water component in the measured spring flows. A total hot-water discharge of 250 US was ob- tained, of which approximately 80 percent is contrib- uted by the springs in the Hot Creek gorge. This esti- mate is based on discharge measurements at relatively warm surface temperatures and should be corrected for hot-water density and temperature to yield a mass flux A24 of 250 kg/s and a volumetric flux of 290 US at 210°C. Fournier, Sorey, Mariner, and Truesdell (1976) as- sumed that the same total discharge of B and Cl in the springs was supplied by a hot-water component with 375 mg/L Cl and 18.8 mg/L B and calculated a mass flux of 190 kg/s. The average annual rate of B discharge from the caldera waters in Lake Crowley can also be used to estimate the flux of hot water through the hydrother- mal system if it is assumed that most of the B in Lake Crowley was contributed by the hot-water reservoir. From 1960 to 1973, the average annual discharge of B from sources within the caldera was 136 metric tons per year (Sorey and Lewis, 1976, p. 789). This B flux could be supplied by a discharge of 300 kg/s of water with a B concentration of 14 mg/L (table 6, Casa Diablo well).' Alternatively, the same annual discharge of boron could be supplied by 230 kg/s of water having a boron concentration of 18.8 mg/L, which applies to the 282°C water suggested by Fournier, Sorey, Mariner, and Truesdell (1976). Comparisons of the estimates of mass flux based on boron discharge in individual hot springs with those based on boron discharge into Lake Crowley, indicate that as much as 20 percent of the upflowing hot water discharges in the subsurface and flows laterally into Lake Crowley. The magnitude of this hot-water flux is small compared with the total surface-water and ground-water discharge into Lake Crowley of 11,000 US, but it carries away a significant amount of heat. The rate of heat discharge from the caldera by convec- tion and conduction is discussed in a later section. BORON DISCHARGE Smith (1976) compared the present-day rates of dis- charge of characteristic hot-spring elements such as B, Li, Cl, and As from the Long Valley hydrothermal sys- tem with amounts and ages of evaporite deposits in Searles Lake, downstream in the Owens River system. The salts in Searles Lake are less than 32,000 years old, on the basis of dating of interbedded mud layers, and approximately 70 percent of the B in the salt layers can be accounted for by discharge from Long Valley at present rates for a 32,000-year period. A similar period of spring flow would account for ob- served quantities of K and 804, whereas much of the Li and As from Long Valley was apparently lost in transport to Searles Lake. Smith also found that the amount of Cl now in Searles Lake is two to three times greater than the present Owens River could have supplied from Long Valley over 32,000 years, suggest- GEOHYDROLOGY OF GEOTHERNLAL SYSTEMS ing that Searles Lake received Cl from sources outside Long Valley. In contrast, Bailey, Dalrymple, and Lanphere (1976) found evidence of a wide distribution of fossil gas vents, ancient sinter deposits, and areas of acid alteration, which indicates that surficial hydrothermal activity within the caldera was more extensive as well as more intensive in the past than at present. The extensive development of hydrothermal activity in lacustrine sediments that are about 0.3 my old, based on the 0.28 my age of the interbedded Hot Creek rhyolite flow, suggests that the hydrothermal system may have reached maximum development at that time. Bailey, Dalrymple, Lanphere (1976) suggest a subsequent de- cline and areal restriction in surficial hydrothermal activity, owing to a general reduction in permeability of the intracaldera rocks caused by self-sealing proc- esses of silicification, argillization, and zeolitization. To reconcile the differences between these two esti- mates of the age of hot-spring activity, Smith (1976) suggests that large saline deposits remain undiscov- ered somewhere in the Owens River drainage or that only late states of spring activity contributed elements like B, Li, and K to the Owens River. Smith’s analysis results from intensive search by the US. Geological Survey for borate deposits in the Owens River drain- age, and it seems unlikely that large deposits of B, equivalent in amount to, say, 10 Searles Lake deposits, have been overlooked. A third alternative, suggested by D. E. White (oral commun., 1976), is that hot-spring activity has been intermittent, and that the equivalent of one more Searles Lake deposit (presumably from the 0.3-m.y. activity) may be dispersed in earlier river drainages and sediment-filled basins. The origin of the dissolved material in the hot-spring waters is of interest in this discussion. If magmatic fluids contribute the bulk of elements such as B, Li, Cl, and K, then it is possible that hydrothermal activity was continuous over the past 0.3 m.y. with changes in chemical composition of hot-spring waters occurring in relation to intermittent intrusive activity. Alterna- tively, if these elements are provided chiefly by hy- drothermal leaching of the reservoir rocks, then the apparent variation in chemical discharge would most likely be related to variations in flow in the hydro- thermal system. As discussed by Ellis and Mahon (1964) and White (1968), evidence from deuterium and 15‘0 determina- tions in volcanic hydrothermal systems indicates that only minor (less than 5—10 percent) amounts of mag— matic water are present in the circulating thermal fluids. Isotopic compositions of Long Valley waters, as discussed in the following section, are similar in this HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA regard. However, the contribution of magmatic fluids cannot be completely eliminated, especially if varia- tions in isotopic composition of hydrothermal fluids oc- curred over the 0.7 -m.y. caldera history. In relation to the leaching mechanism, Ellis and Mahon (1964) present data from short-term (20-day) leaching experiments involving volcanic rocks from New Zealand and temperatures from 150°C to 350°C. They found that from 12 to 75 percent of the original chloride and from 3 to 30 percent of the original boron in the rocks could be removed within 20 days, and con- cluded that much of this “easily liberated” material was held on surfaces in the rocks rather than in solid solution in the rock silicates. Over long periods of time, additional releases of rock constituents associated with hydrothermal alteration of silicate structures would be expected. Further evidence from drill cores from New Zealand hydrothermal areas (Steiner, 1955) showed that more than 80—90 percent of the original chloride and boron in rhyolitic rocks from 200°C to 260°C is lost during natural alteration. The reservoir volume required for leaching to supply the bulk of the chemical discharge from the Long Valley hydrothermal system can be calculated as a function of the percentage of original rock constituents which can be removed. The bases for these calculations are the estimates by Smith (1976) of the quantities of B, K, Cl, and Li in Searles Lake which are supplied by the Long Valley system. These totals are listed in table 8, along with measured element concentrations in the caldera rock. These data show that B and Cl discharges place the greatest demand on leached rock volumes. Accord- ingly, in table 9, the required rock volumes are listed as functions of the percentage of leachable B and Cl. Corresponding reservoir thicknesses are calculated by assuming a reservoir area of 220 km2, which covers the western three-fifths of the caldera. TABLE 8.—Measured concentrations of chemical elements in Long Valley rocks and total amounts discharged from Long Valley into Searles Lake Concentration in Total amount rock samples1 discharged’ Element (ppm) (ng109 B ____________________________ 330—53 7 K ____________________________ 416,000—20,000 44 Cl ____________________________ 5300-400 6140 Li ____________________________ 320—49 1 lSpectrographic analysis by C, S. Annell, USGS, Reston, Va. 2From Smith (1976), “Range for 9 samples of Bishop Tuff from Owens River gorge. ‘Sheridan (1965). 5Range for 6 samples of intracaldera early rhyolite and Bishop Tuff from Owens River gorge. 6Calculated from total B, assuming Cl/B : 20. A25 These results indicate that under certain conditions the reservoir volume and thickness required for leach- ing to supply the salts discharged from Long Valley are realistic. One condition is that significant percentages of the original B and Cl in the reservoir rocks be re- moved. Considering a reservoir thickness of 2 km as reasonable, and accounting for an additional period of comparable hydrothermal activity around 0.3 m.y. ago, at least 30 percent of the B and 60 percent of the Cl must have been leached. The New Zealand data dis- cussed above suggest that this is certainly possible. A second condition is that the areal extent of the reser- voir be close to the assumed 220 km2. If it were consid- erably more areally restricted, the required reservoir thickness would become unreasonably large, unless the 80—90 percent B and Cl leaching found in New Zealand has also occurred in Long Valley. If these levels of element removal have not been approached in Long Valley and if the hydrothermal circulation sys- tem is not as areally extensive as assumed here, then magmatic sources must have contributed significantly to the chemical and probably the thermal discharge from the Long Valley caldera. Additional lithologic data from deep test holes in the western caldera are required to resolve these questions. The contrast between the age of the saline deposits of Searles Lake and that of the earliest hydrothermal ac- tivity within the caldera, and the inference of extreme climatic and hydrologic variations during the interven- ing period, lead to the conclusion that the hot springs and associated hydrothermal activity have been in- termittent. This matter is considered further in a later section in terms of the influence of the age of hy- drothermal activity on heat-flow requirements from an underlying magma chamber. TABLE 9,—Si2e of leached reservoir required to supply total discharges of B (md Cl for various changes in B and Cl concentrations in the reservoir rock Change in element leached-rock Leached-rock content of rock‘ volume2 thicknessa Element (percent) (kma) (km) E 75 100 0.45 50 1 50 .68 25 300 1.4 10 750 3.4 Cl 75 214 0 97 50 320 1.5 25 640 2.9 10 1600 7.3 1Assuming original rock B = 40 ppm, Cl : 350 ppm, zVolume = Amount of element discharged/(change in element content of rock >< bulk density of rock). aAssuming reservoir area = 220 km”. A26 STABLE ISOTOPES Comparisons of 18O and deuterium contents of hot and cold Long Valley waters are useful in understand- ing the caldera’s hydrologic system. In comparison with most other isotopically studied hydrothermal convection systems, the Long Valley system appears unusual in terms of the large areal variability in isotopic contents of its meteoric waters and clear evi- dence of the involvement of at least two different meteoric waters in the hot-spring fluids. Figure 11 is a plot of stable isotope date in standard 6 values, parts per mill (0/00) relative to SMOW, for var- ious waters which were sampled at the locations shown in figure 12. Mariner and Willey (1976) used data from GEOHYDROLOGY OF GEOTHERMAL SYSTEMS samples collected in 1972, represented by the dark cir- cles, to suggest that water similar to i (Big Springs) was typical of the recharge to the hydrothermal sys- tem. The bulk of the flow in Big Springs probably originates in the Deadman Creek drainage basin and moves eastward at shallow depth through the exposed basalt flows. As meteoric water percolates deeper un- derground and is heated, the oxygen isotopes react with oxygen in the rocks while hydrogen isotopes re- main almost unchanged. Thus, the isotopic composi- tion of the circulating water would shifi from i to b. Water b from a well at Casa Diablo Hot Springs was interpreted by Mariner and Willey (1976) to be a sam- ple of the hottest water in the system, having a I I I I I I I -110 —. __ 0c g —120 - _ 9.. O '— _. ‘0 430 — ._ l l I l l l I -20 -18 -16 -14 —12 5'“0(°/oo) FIGURE 11.—Relations between concentrations of deuterium and 18O in Long Valley waters. Closed circles represent samples collected in 1972, open triangles represent 1975 samples, open squares represent 1976 samples for which 180 concentrations have not been determined. Sample locations given in figure 12 and table 7. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA geochemically estimated temperature at depth of 219°C. During upflow, if this water mixed in different proportions with cold water similar to j (from an arte- sian well near the northeast rim), the isotopic contents of the various hot and warm springs could be obtained. The lineation of points representing these springs be- tween end points b and j tends to support this mixing model. Additional data from cold springs sample in 1975 and 1976 and annual variations in D/H ratios of snow sam- ples collected at Mammoth Mountain midstation (Friedman and Smith, 1970, 1972; Smith, unpub. data, 1976) allow somewhat more detail in the analysis of recharge and mixing relationships. The more recent data are represented by the open symbols in figure 11. Point t* for the Mammoth Mountain spring was calcu- lated using an average SD of —114 * from snow sam- ples for the years 1970—76 over which a range of —110 to —118 occurred. This point, along with points r, p, and t, which also represent cold waters from the west rim but are based only twice 1975 data, suggests that re- charge water for the hydrothermal system may be A27 slightly heavier than i. If this were so, water b could be a mixed water rather than an end member, consistent with the geothermometer analysis by Fournier, Sorey, Mariner, and Truesdell (1976). Water c from the Hot Bubbling Pool has been strongly fractionated probably by evaporation, similar to many other hot springs of high surface temperatures but without surface dis- charge. The more recent 6D date in figure 11 also indicate differences between cold waters around the north, east, and sourth rims. Water just beyond the east rim (v) is lightest at —135° 0/00. Ground waters from the north and northeast rims (j, n, q, and y) range in 8D from —129 to —132 o/oq. The south rim waters (1 and w) are heaviest at —125 to —128°/oo , and similar to the Hartley Springs water (11) from beyond the northwest rim of the caldera. The 8D content of water z from a windmill on the northeast flank of the resurgent dome is quite light at —l34 0/00. If this water originates as precipitation on the resurgent dome, as suggested by the water table map in figure 10, the dome waters must be relatively light. A contribution of recharge from the I o: f ’l‘ Bald Mtn llGlass Mtn /‘ / ’l‘ //\ A‘ /l //\ n //\\ U s ” ———————————————————————— k owe/rs ‘ ~_ , 0"“ ,_ 4/ ~ \ 37°45 ’- -__, " p ‘ _‘ cs / 9,. \ o\a I? O l \‘ I Q ‘ l/o‘ \ J z \ H’ q \ 9&3 \ y \ 0° “ l 'I p \ I . ) I l \ l \ I \ \ , v ' d / / ll e / I r\\ b I, _ Mammoth g 1 1'2. Mtn ”\t ~ I “(V k / \ a—-\ I \-‘_'a’ \\~’__J\ .......... "” lee Crow/9y o 5 10 W 37°35’ l- L I IKlLOMETERS _ I' 119000, 118945! FIGURE 12.—Map of Long Valley caldera showing location of sampling points for deuterium and 180 analyses. A28 Lookout Mountain area on the dome to Big Springs might account for point i being lighter than the other west rim waters. In view of temporal variations in D/H ratios of snow- fall, a complete set of sample from springs at one or more times would be desirable. However, the data in figure 11, along with our knowledge of ground-water movements in the shallow system, suggest several sources of cold, diluting water in the thermal springs, rather than the single source (i) indicated by Mariner and Willey (1976). For the Casa Diablo Hot Springs area and the warm springs west of the Hot Creek gorge, the south rim and resurgent dome are the most likely sources. For springs in the gorge and Little Hot Creek, the north rim and resurgent dome and possibly the south rim are indicated. Because shallow recharge from the northeast and east rim flows into the Owens River and Lake Crowley, it would not be available for mixing with hot-springs waters further to the west as was previously suggested. The isotope data do suggest, however, that some re- charge from the higher altitudes on the flanks of Glass Mountain may percolate down the rim fractures and flow laterally toward the southwest through permeable zones in the Bishop Tuff. A drill-stem sample of ques- tionable quality from a depth of 1.46 km in Republic Geothermal’s 2.11 km-deep test hole near spring f (fig. 12) yielded a «SD of —126 °/oo, (C1 = 100 mg/L). If this sample was contaminated by drilling fluid, which had a measured SD of —114 °/oo , the relatively cold waters in the Bishop Tufi‘ at this location may have D/H ratios closer to the —129 to —132 ‘70.: range for the northeast- rim waters. THERMAL CHARACTERISTICS OF THE HYDROTHERMAL SYSTEM Measurements of thermal anomalies would seem the most obvious indication of potentially valuable geo- thermal areas. These anomalies may take the form of hot-spring and fumarolic discharge, above-normal ground temperature detectible by infrared imagery or snow calorimetry, and above-normal near-surface heat flow, geothermal gradient, and temperature. As noted by Lachenbruch, Sass, Munroe, and Moses (1976), meas- urements of heat flow, unlike other geophysical meas- urements, may contain direct information on past geologic events because of the long time required for a thermal disturbance to equilibrate by conduction in earth materials. Unfortunately, the time variable and the interaction of convective (or advective) and conductive heat- transfer mechanisms can produce large variations GEOHYDROLOGY OF GEOTHERMAL SYSTEMS from one area to the next and make interpretation of thermal and hydrologic conditions at depth uncertain. With due consideration for these limitations, investi- gations of regional heat flow and the near—surface thermal regime in the Long Valley area were under- taken. These studies are described by Lachenbruch, Sass, Munroe, and Moses (1976) and Lachenbruch, Sorey, Lewis, and Sass (1976); the results which are pertinent to our analysis of the hydrothermal system are discussed below. REGIONAL HEAT FLOW Temperature-gradient and heat-flow measurements were made at 11 stations in the vicinity of (but outside of) Long Valley caldera (Lachenbruch, Sass, Munroe, and Moses, 1976). Evaluation of the thermal anomaly associated with magma beneath the caldera shows no conspicuous effect of the transition between the Sierra Nevada and Basin and Range provinces, possibly a small local heat-flow anomaly beyond the east rim of the caldera (2.2 HFU), and a very substantial anomaly beyond the west rim (3.8 HFU). Lachenbruch, Sass, Munroe, and Moses (1976) con- clude that heat-conduction models for the near-normal heat flow at the east rim suggest that magma beneath the east part of the caldera might have been exhausted during eruption of the Bishop Tuff 0.7 m.y. ago and that the resurgent dome, which subsequently formed in the west-central caldera, overlies a residual magma chamber more circular in plan than the original magma chamber that supplied the Bishop Tuff. High heat flow indicated by the measurement near the west rim is attributed either to a simple shallow magma chamber beneath the western caldera or to recent local magmatism along the Sierra frontal fault system. The analysis of heat—conduction models of cooling magma further suggests that the Long Valley caldera is the surface expression of a deep magmatic system, because an upper-crustal magma chamber could not have furnished molten material throughout the 2-m.y. eruptive history since the early eruptions of the Glass Mountain Rhyolite unless it were resupplied with heat from deep crustal sources. Calculations of heat supply by crustal intrusion of mantle basalt yield minimum intrusion rates of the order of 1 m per century, which implies accelerated crustal spreading. THERMAL REGIME WITHIN THE CALDERA Thermal measurements were made in 29 holes drilled to depths of 30 m or less and in 10 deeper holes (from 52 to 325 m) within the caldera. Well locations are shown in plate 1; LV designates the shallow (30 m) HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA holes; CH designates the deepper holes in which cores were obtained. A few thermal conductivity meas- urements were made to permit rough estimates of conductive heat flows. Selection of drilling sites and depths involved a compromise between areal coverage and vertical definition. As noted by Lachenbruch, Sorey, Lewis, and Sass (1976), the farther beneath the surface we extend our observations the more we can expect to learn about the hydrothermal system associ- ated with the potential resource, but the more costly and time consuming each observation becomes. In ad- dition, adequate information on lateral variability may prove essential to subsequent selection of sites for deep drilling. Temperature profiles from all the test holes are plot- ted in figure 13; identification of individual profiles in the 30-m holes is given by Lachenbrich, Sorey, Lewis, and Sass (1976). Temperatures at 15 m are shown in figure 14 and near-surface conductive heat flows in figure 15. The former differs somewhat from the cor- responding map in Lachenbruch, Sorey, Lewis, and Sass (1976), because it is based on additional data. Methods of drilling and completion to minimize effects of fluid circulation in boreholes are described by Lachenbruch, Sorey, Lewis, and Sass (1976). Thermal conductivity measurements and lithologic characteris- tics of the rocks at each core-hole site are listed in A29 table 10. In contrast to the regional temperature and heat- flow data, thermal data obtained within the caldera are strongly influenced by the hydrologic system. Analysis of the thermal regime has provided useful information on near-surface (0—300 m) hydrology, as discussed be- low. The shallow thermal observations also help to define the nature and extent of the recharge and dis- charge areas for the deeper hydrothermal system and suggest areas where the shallow ground-water subsys- tem is relatively impermeable and shallow tempera- ture gradients might be used to estimate temperatures in the underlying deep subsystem. However, additional information from deep test drilling is needed to evalu— ate satisfactorily the role of shallow thermal observa— tions in geothermal exploration in Long Valley. Temperature profiles in figure 13 show gradients as- sociated with conductive heat flows of 0 to 50 HFU for thermal conductivities (table 11) of about 2 mcal/ (s °C cm). Temperatures at 15 m (and as shallow as 6 m) correlate well with the character of the thermal gra- dients to 30 m, as discussed by Lachenbruch, Sorey, Lewis, and Sass (1976). The grouping of temperature profiles in figure 13 is suggested by the following. Group I regimes, representing heat flows less than 1 HFU, are characteristic of the outer margin of the caldera, suggesting that this is an area of hydrologic 0 ‘* ‘ —:T ' 100 — CH9 150 - CH4 DEPTH, IN METERS 200 - CH8 250 '- | l I I l 1 CH7 CH10 ‘ n”_ CH6 l l l l 1 I 300 i n l L 1 0 10 20 30 40 50 60 70 80 90 1 00 1 1 0 TEMPERATURE, IN DEGREES CELSIUS FIGURE 13.——Temperature profiles from core holes and shallow holes in Long Valley caldera. Data from Lachenbruch, Sorey, Lewis, and Sass (1976), except for CH 8, 9, 10. A30 recharge. Group II regimes, representing conductive heat flows of about 4 to 8 HFU, show little influence of convection down to 300 m in at least one location (CH—1). Group III regimes, representing larger heat flows ranging to more than 50 HFU, show considerable variability with depth and more variability with time than gradients in the other groups (Lachenbruch, Sorey, Lewis, and Sass, 1976). Temperature profiles in group III indicate that they are characteristic of areas of subsurface discharge of hot or warm ground water; the hot springs discharge in a fault zone characterized by near-surface regimes in groups II and III. Tempera- ture reversals in wells CH—3, CH—5, CH—7, and CH—10 and temperature and heat-flow patterns (figs. 14 and 15), indicate that warm ground water moves laterally from Hot Creek gorge toward Lake Crowley. An esti- mate of the heat discharged by this subsurface flow is GEOHYDROLOGY OF GEOTHERMAL SYSTEMS discussed in the next section. Heat flows at the group II sites are comparable to the hydrologically undisturbed value of about 4 HFU meas— ured in granite at Devils Postpile just beyond the west rim of the caldera (Lachenbruch, Sass, Munroe, and Moses, 1976), although the agreement is probably for- tuitous. One explanation for the group II profiles is that they occur in areas where the 200°C reservoir temperatures (from chemical geothermometers) exist at depths of about 1 km or less and heat transfer in overlying rock of low permeability is basically conduc- tive. One core hole in group III, CH—6, was drilled on the resurgent dome in Little Antelope Valley, and encoun- tered the highest temperature (110°C) in any of the test holes. The nearly isothermal part of the temperature profile below 165 m indicates that circulation of heated ///\ I ///\\Ba|d Mtn //Z;Glass Mtn /\ 01 /,\ //\\ I ————————————————————————— 37°45, —- /_\~ J‘JWOWGDS ‘‘‘‘‘ \x . —’ - \ —: s ./ -\_..-.. .—— \ 9/), \\ <3" / \\ \ J’ \‘ a“ ’ \\ k 6‘0 \ 0° \ \ol' ‘ / l l -8.1 g \ l \\ ’0 I' \I \ l / ~1o.5 / I] \l I / I ll \\ / .. 44.8 in Mammoth'lr__ \\ Mtn [IR— EXPLANATION """ Crowley / ,— 10 \ Line of equal temperature at 15 m depth, 37°35. _ in degrees Celsius. Dashed where uncertain ._ - 10.5 Well and temperature at 15 m depth, 0 5 10 in degrees Celsius I I I KILOMETERS I l 1 19°00’ 118°45' FIGURE 14,—Map of Long Valley caldera showing temperature at a depth of 15 m, 1973—74. Data from Lachenbruch, Sorey, Lewis, and Sass (1976) and Lewis (1974). (Dashed-dot outlines early rhyolite outcrops on resurgent dome.) HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA A31 TABLE 10.—Generalized lithologic descriptions and thermal conductivity data from cores and cuttings at test-hole sites in Long Valley caldera Thermal H l D th . . Nooe Location (iii) Lithology (niggfigglswgé ’ ) CH—l 3S/29E—19C 0—160 Tuffaceous sediments, alternately coarse 1.8—2.1 Cashbaugh and fine, some clay-alteration Ranch 160—305 Tuffaceous sediments, hydrothermally 1.7—2.5 altered with permeability reduction CH—3 38/29E—27L 0—52 Porphyritic biotite rhyolite flow 2.1—2.4 West of Lake Crowley CH—4 38/30E—19M 0—90 Ash and sediments 2.1—2.4 East rim 90—187 Medium to coarse sand 2.6 CH—5 4S/29E—5B 0—60 Tuffaceous sediments ' ______ E. of Whit- 60—120 Pebbly sand with obsidian and metamorphlc clasts ______ more Hot Spring 120—200 Biotite rhyolite pumiceous tuff 2.0—2.3 210—265 Hornblende-biotite rhyolite tuff (0.3 my old) 2.1-2.8 CH—6 38/28E-22F 0—150 Clay-altered tuff 2.9—3.5 Little Ante— 150—208 Altered tufl' with sulfide-bearing silica 3.4—5.2 lope Valley veins CH—7 3S/29E—19R 0—70 Sand ______ Near Little 75—165 Altered sediments (silicified, argillized) 3.5—3.8 Alkali Lake ‘ CH—8 38/28E—18D 0—30 Pumice ______ Smokey Bear 30—100 Rhyolite flow ______ Flat 100—325 Aphyrite rhyolite tuf‘f 1.8—2.3 CH—9 3S/27E—20H 0—10 Pumice Upper Dry 10—50 Porphyritic vesicular basalt 3.0—3.6 Creek 50—85 Hornblende-plagioclase andesite (dense 3.8—4.0 to vesicular) 88—92 Olivine-plagioclase basalt 3.3 CH—lO 3S/29E—30E 0—53 Altered Hot Creek rhyolite ______ ‘ Hot Creek Gorge DC 28/27E—34A 0—50 Pumiceous sediments ______ Deadman 50—100 Trachyandesite 3.6—4.4 Creek 100—190 Sand 1.7—2.4 190—193 Pumiceous tufl‘ 1.7 TABLE 11.—Previous estimates of convective heat discharge from Long Reference Valley caldera Reservoir Convective tempera- heat flow Method ture (°C) (X107 cal/s) Sorey and Lewis Measured springflow __ 210 4.3 (1976) Boron discharge ...... 210 5.4 White (1965) Boron discharge ______ 180 7.0 Fournier, Sorey, Measured springflow __ 282 5.3 Mariner, and Boron discharge ______ 282 6.6 Truesdell (1976) ground water is influencing the thermal regime in the upper 200 m. In contrast, the temperature profile in CH—8, drilled in Smokey Bear Flat 5.6 km northwest of CH—6 and within the keystone graben on the resurgent dome, is characteristic of the low-heat-flow regimes of group I. Similar low temperatures and isothermal temperature profiles were found at other core holes drilled farther west of CH—8, suggesting that the circu- lation of relatively cold ground water may be masking evidence of deeper thermal activity in the western third of the caldera. It is clear from the thermal data in figure 13, that deepening any of the holes by 50 to 100 A32 111 could lead to surprises, and that downward extrapo- lation of near-surface temperature gradients is hazardous. In general, our conceptual model of the Long Valley hydrothermal system ignores the details of the shallow thermal and hydrologic regimes. However, the data discussed above were used in calculating the total heat discharge from the caldera, which provides a check on the estimates from chemical mixing models of reser- voir temperatures and hot—water discharge and also provides a necessary constraint on heat-flow simula- tion with the numerical model. NATURAL HEAT DISCHARGE The natural rates of heat and fluid discharge from ‘ GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS the Long Valley caldera provide useful constraints on the simulation of the hydrothermal system. Published estimates of the convective heat discharge are given in table 11. The estimates by Sorey and Lewis (1976) should have been corrected for variable fluid density to yield from 5.1 to 6.4 x 107 cal/s. Calculations based on measured springflow use geochemical mixing models to estimate reservoir temperature and mixing ratios between hot-water and cold-water components. The rate at which heat leaves the hydrothermal reservoir by convection (Q) is computed from Q = M (h, — ho) (1) where M = mass discharge of hot water from the reser- voir, hr = enthalpy of hot water, and ho = enthalpy of I 0/; B ld M l /‘ GI M \ //\ a tn [/7‘ ass In N //1 01 a w 37°45’ — creek _ 55 ’ ‘ 6‘3 ’1?- r ( I, C: n ’j 3 6‘0 9° / ‘\. 0/ l I // ’3 ‘V‘ \ II .j l/ / ‘1‘) 0‘00 \ l l f / / l 9 \ l | \. ' / ’ 3 8 I l \ \. / / (-\,/ . t o I \ ~ I, I p/2I . I I \I l , ~ 2/ 6 o I I . I l \ I, \. l , I l \---k ' l l l l ‘\ I l \ \ ll ‘ V< \ K \2 l — I Mammoth”[7—_ ~ WV 0 Mtn ll? T‘ \ \\~ _—’,-——\\ EXPLANATION /20 ““ ~‘ 37°35! _ Line of equal heat flow, in heat-flow units (HFU) _i Dashed where uncertain . 0 5 10 Well used for Control | I g KILOMETERS I , l 119°00' 118°45’ FIGURE 15.—Map of Long Valley caldera showing near-surface conductive heat flow, 1973—74. Based on temperature data from Lachenbruch, Sorey, Lewis, and Sass (1976) and Lewis (1974) and an assumed thermal conductivity of 2.0 meal/(cm 5 °C). (Dashed-dot line outlines early rhyolite outcoprs on resurgent dome.) HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA recharging meteoric water (:10 cal/gm). The corrected estimate of 5.1 X 107 cal/s noted above is based on a mass discharge of 250 kg/s at 210°C and 214 cal/gm. The corresponding estimate by Fournier, Sorey, Mariner, and Truesdell (1976) of 5.3 X 107 cal/s is es- sentially the same, even though it is based on a reser- voir temperature of 282°C. This is because the same total flux of boron and chloride from the hot-water res- ervoir was assumed, but higher calculated concen- trations of these elements in the hot water are in- volved. The result is a lower mass discharge of 190 kg/s used in equation (1). Convective heat-flow estimates based on boron dis- charge from Lake Crowley involve calculating the mass discharge of hot water from the reservoir needed to supply the average annual tonnage of boron leaving the caldera. Using similar assumptions as to boron and heat contents in the hydrothermal reservoir, the con- vective heat-flow estimates in table 11 exceed the cor- responding estimates based on measured springflow by 25 percent. As discussed previously, these differences are due in part to unmeasured subsurface discharge of hot water from the hot-spring conduits. To determine the total heat discharge by conductive and convective processes and to test the adequacy of the mixing-model results, we have calculated the total heat discharge at the land surface. For the purpose of the following analysis, the heat discharge from the caldera is subdivided into three elements : (1) convec- tive heat discharge by springflow; (2) near-surface con- ductive heat discharge; and (3) advective heat discharge by ground-water flow beyond the caldera margin into Lake Crowley. A33 CONVECTIVE HEAT DISCHARGE BY SPRINGFLOW The estimate of convective heat discharge by springflow (table 12) is based on data of Sorey and Lewis (1976, tables 1 and 2) and Lewis (1974). Dis- charge temperatures are measured values except for the springs discharging into Hot Creek in Hot Creek gorge, where the discharge is assumed to be 93°C (boil- ing temperatures at an altitude of 2,120 m) and the Fish Hatchery Springs, where the weighted-average discharge temperature is 14.4°C (Sorey, 1975b, p. 4). Net enthalpy per gram of discharge from each spring or spring area is calculated as the difference between the discharge temperature and an assumed average an- nual temperature of 8°C at the land surface. The estimate of total convective heat discharge by springflow, about 2.9 X 107 cal/s, is believed to be con- servatively low, because several nonthermal springs in Long Valley that have sizable discharges at more than 8°C were not included. NEAR-SURFACE CONDUCTIVE HEAT DISCHARGE The second element of heat discharge—near-surface conductive heat discharge—was estimated in the fol- lowing manner. As the first step in the calculation, the average annual near-surface geothermal gradient at each test-well site was computed as the difference be- tween the observed temperature at a depth of 15 m and an assumed average annual temperature of 80°C at the land surface, divided by the 15-m depth interval. A depth of 15 m was selected because repeated meas- urements in most of the test wells indicated very small seasonal fluctuations in temperature at that TABLE 12,—Convective heat discharge by springflow from Long Valley caldera [Temperature and discharge data from Sorey and Lewis (1976, tables 172)] Discharge Net enthalpy Heat discharge Temperature Density of water Spring or spring area (Li’s) (”CI (cal/g) (g/cm“) (x106 cal/s) 1. Springs in Hot Creek gorge ____________________________ 247.0 93.0 85.0 0.963 20.20 2. Fish Hatchery Springs ________________________________ 964.0 14.4 6.4 .999 6.16 3. Little Hot Creek1 ____________________________________ 12.0 80.0 72.0 .972 .84 4. Casa Diablo Hot Pool ________________________________ 6.5 60.0 52.0 .983 .33 5. Whitmore Hot Springs ________________________________ 26.0 34.0 26.0 .994 .67 6. Big Alkali Lake2 "1, ____________________________ 8.8 56.0 48.0 .985 .42 7. Little Alkali Lake2 __ ,,,,,,,,,,,,,,,,,,, 3.8 66.0 58.0 .980 .22 8. Chance Spring ______________ 23.0 20.0 12.0 .998 .28 9. TBS/R29E—27, 28 ______________ 3.8 49.0 41.0 .988 .15 10. T3S/R29E—34K _________________________ 1.8 41.0 33.0 .992 .06 11. TSS/R29E—31A ______________________________________ 1.5 58.0 50.0 .984 .07 Total or average __________________________________ 1,298.0 31.3 23.3 0.991 29.40 ‘Based on discharge measurements by L. Willey at E83 and includes corresponding estimates for the other three springs. 2Based on estimates by Lewis to include some unmeasured discharge. A34 depth. Temperature at 15 m at sites where test wells are less than 15 m deep, was estimated using a least- squares linear-regression correlation between temper- atures at 10 m and 15 m. The coefficient of determina- tion for the linear regression was greater than 0.99 for each of three sets of measurements during 1972—74. The data and linear regression for the last set of meas- urements, in October 1972, are shown in figure 16. The pattern of temperature at 15 m depth is shown in figure 14. The area of highest temperature is presuma- bly along the Hot Creek gorge where the hot springs discharge into the creek. Highest measured tempera- tures are in two wells less than 0.6 km southeast of the gorge. The configuration of the isotherms suggests southeastward transport of heat by ground water mov- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ing through aquifers below 15 m depth in the direction of the lateral hydraulic gradient, toward Lake Crow- ley. The pattern of two elongate lobes of high tempera- ture, separated by a zone of much lower temperature shown in figure 14 differs from the pattern at a depth of 10 m inferred by Lachenbruch, Sass, Munore, and Moses (1976, fig. 1). Although the data points are too sparse to choose between the two isotherm configura- tions, we prefer the present interpretation, for two principal reasons. First, the configuration of the isotherms shown in figure 14 is most consistent with the pattern of shallow ground-water flow (fig. 10). Ac- cording to our present interpretation, some of the thermal water that rises near the Hot Creek gorge does not discharge as springs in the gorge but mixes with 38 I I 34- 26—- IN DEGREES CELSIUS 22—- 18..— 14*— TEMPERATURE AT 15 m, Tl 5 , 10..— 6 1 1 T15 =1.29Tm —2.1 — r1 = 0.992 I l l 6 10 14 18 22 26 3O TEMPERATURE AT 10 m, T10, IN DEGREES CELSIUS FIGURE 16.—Relation between temperatures at depths of 15 m and 10 in Long Valley caldera, October 1974. Data from Lachenbruch, Sorey, Lewis, and Sass, 1976. HYDRQTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA shallow ground water and moves laterally toward Lake Crowley. As discussed in a later section, the zone carry- ing most of the hot water lies at depths about 100 In. Second, a temperature survey at 1 m depth in 1972 utilizing additional data points (Lewis, unpub. data, 1973) showed approximately the same pattern as that in figure 14. Although the correlation of temperatures at 1 m with those at greater depths is not nearly as good as that of temperatures at 10 m and 15 In de- scribed above, we believe the pattern inferred from the 1-m temperature data is a fairly reliable indication of the pattern at a depth of 15 In. As the second step in the calculation of near-surface conductive heat discharge, the temperature gradients derived from the measured and extrapolated tempera- tures at 15 m and the assumed land-surface tempera- ture of 8°C were multiplied by an assumed average thermal conductivity of 2.0 mcal/(cm s - °C) (Sass and others, 1974, p. 29 and table 11) for the materials from O to 15 m to give conductive heat flow. Heat-flow values, in ,u cal/(cm2 s) or heat-flow units (HFU), are shown in figure 15. Because they are derived from the tempera- ture at 15 m, the heat flows define an areal pattern similar to that of the temperatures (compare figs. 14 and 15). Much of the area having heat flow less than 2 HFU is affected by recharge of cold ground water. The area of high heat flow, especially above 20 HFU, is charac- terized by convective upflow of thermal water or by lateral advection of thermal water that has moved upward nearby. To calculate the total near-surface conductive heat discharge in the caldera, heat discharge from the area between each pair of heat-flow isograms in figure 15 is computed 'as the product of the area and the geometric mean of the two isogram values. The heat discharges thus obtained are then added to obtain the total as shown in table 13. The total area used is slightly greater than that of the caldera floor, because heat flows within the 10- and 20-HFU isograms southeast of the caldera, at Lake Crowley, are included. The total conductive heat flow, 3.5 X 107 cal/s, represents all the geothermal heat leaving the caldera except that transported to the surface by spring flow or transported beyond the perimeter of the caldera by ground-water flow into southern Lake Crowley. ADVECTIVE HEAT DISCHARGE BY GROUND—WATER ()UTFLOW TO LAKE CROWLEY The final element of heat discharge from the Long Valley caldera is the heat advected by eastward and southeastward ground-water flow from the Hot Creek A35 TABLE 13.—Near-surface conductive heat discharge from Long Valley caldera (from heat-flow data shown‘ in fig. 15) Geometric-mean Heat flow Area heat flow Heat discharge (HFU) (km’) (HFU) (><106 cal/s) > 60 9.5 69.30 6.6 40—60 9.1 49.00 4.4 20—40 28.3 28.30 8.0 10—20 74.4 14.10 10.5 2—10 95.2 4.47 4.3 < 2 232.4 0.50 1.2 Total or average -_' __________ 448.9 7.80 35.0 gorge thermal area into Lake Crowley south of the caldera margin. Warm bulges in the temperature pro- files of four test wells, CH—3, CH—5, CH—7, and CH—10, indicate that most of the thermal ground water moving toward the lake is in an aquifer at depths above 100 m and ranging in thickness from about 23 to 46 m. The heat advected to the lake by this ground-water flow is estimated as follows. First, thermal ground-water outflow is considered as the flow across a vertical section of aquifer near the north shore of the southern part of the lake (line A—A’ in fig. 17). The outflow is calculated as the product of the average hydraulic gradient normal to the transmitting section, the area of the section, and its average hydraulic conductivity at the prevailing tem- perature of the fluid (50°C). The area of the transmit- ting section is the product of its length (3.1 km) and its average thickness (32 m), or 9.9 x 108 cm2. The average hydraulic gradient is calculated as the average of the two gradients of the water table shown in figure 17, or 0.0115. The average hydraulic conductivity at the pre- vailing temperature of 50°C is assumed to be 9.3 m/day. This value is equivalent to a hydraulic conduc- tivity of 4.5 m/day at the standard temperature of 60°F (156°C), which is near the upper limit of values meas- ured for core samples from the test wells (table 6 and Lewis, 1975, table 2). Estimated ground-water outflow is, therefore, Q=AIK=120L/s whereA = area of transmitting section,I = hydraulic gradient, and K = hydraulic conductivity. The heat advected by this outflow is calculated from equation (1) as 5.0 X 106 cal/s, on the basis of an annual average ambient temperature at land surface of 8°C and an average discharge temperature of 50°C. A36 TOTAL HEAT DISCHARGE The total discharge of geothermal heat from the Long Valley caldera is summarized below: Heat discharge (X107 cal/s! , Convective heat discharge by springflow ________ 2.9 Near-surface conductive heat discharge __________ 3.5 Advective heat discharge by ground-water outflow to Lake Crowley ______________________ f 6.9 Total heat discharge ______________________ This estimate of heat discharge exceeds the previous estimates listed in table 11, except for that by White (1965), at least in part because most of the earlier esti- mates do not include all the effects of heat conduction. The close agreement with White’s estimate is partly GEOHYDROLOGY OF GEOTHERMAL SYSTEMS coincidental with offsetting of differences in assump- tions. Sources of error in the present estimate include errors inherent in the simplified assumptions and errors or deficiencies in the data. Previous estimates of convective heat discharge are limited by various as- sumptions in the geochemical models, which are not yet verified by deep-test-hole data, but which appar- ently lead to reasonable heat-flow estimates. Taking into account all the known sources of error, the most likely value of the total natural heat discharge is within the range 5—10 x 107 cal/s. Our estimate of 6.9 X 10'1 cal/s is equivalent to 15 HFU integrated over the 449-km2 area of the caldera floor. In the numerical simulations discussed in the following section, this heat flow represents an important constraint on the required depths of fluid circulation. ~\ \\ ‘\\\‘ \\ ‘~’\ 'I ,' ocmo \ ‘~\ ‘-~\ ‘\ ‘x‘ l I ’z-‘\ \\ ‘\\\ \\ ‘\ ‘-‘_~ \ I \\’,/ \\ \\\ \\ .CH3 \\ \ \\ \ \ \ l \ \ \ l \ l \ -_a \ I I, \ l l \‘\ l W0 I ‘ I i \ l N I I / i \ a W , I \ ('2 / ‘0 / / N / II" / / I \ / / / // \ II I, I, 1/ // \ , / I I / z ‘ I I I ‘w " 1” | l. I I If} \ \_~ I I i I , \ I I \ I ,I '<\ O \ ‘ ‘ I ’ 0.010 / E \ \\ \\ \.CH5 [I / (‘l‘ l. I \ \\ Hydraulic gradi’ents ' i l \ ' l' I I I o‘\ ,/ II I I I I, t; 3 0” I / / ll / 0.01108 / ’.——.—_’ ’,.——__—r ‘ I l \ ‘ ‘ \ l ‘ ‘I ' i I I :5- 4 \x \I I I— J- I Water-tabIe contour / I x I [I \ / \ \\ \ \ \ \ \ \‘~—~ 0 1 2 .— 37°35’ — Transmitting section LAKE CROWLEY \ \ \ \l \ 118°45' FIGURE 17.—Map showing elements used in estimate of thermal ground-water flow into Lake Crowley. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA SOURCES OF HEAT The results of several investigations suggest that a residual magma chamber underlies the west half of Long Valley caldera at the present time. The heat-flow anomaly beyond the west rim and implications of re- supply of magma from deep crustal sources to sustain molten material over the 2-m.y. eruptive history of the area were noted above. In discussing the magmatic his- tory, Bailey, Dalrymple, and Lamphere (1976) explain the concentric zonation of the postcaldera eruptives as a consequence of progressive downward crystallization of a magma chamber that was vertically zoned from rhyolite in its upper part to rhyodacite in its lower part. On the basis of structural analysis of the resur- gent dome, they suggest that at the close of resurgence about 0.6 m.y. ago, the top of the Long Valley magma chamber had risen to at least a 5-km depth and possi- bly to 2 or 3 km. By about 0.2 m.y. ago, lateral en- croachment of basalt dikes on the chamber suggests that it had congealed inward and downward 4 km to a depth of 6 to 9 km. Teleseismic and seismic refraction studies (Steeples and Iyer, 1976; Hill, 1976) indicate that an anoma- lously hot or partly molten mass persists below 6 to 8 km under the western caldera. Gentle gravity gra- dients outside, but sloping toward, the caldera indicate a low-density mass at depths of 8 to 16 km and large enough to constitute an anomaly of about 10 mgal (Kane and others, 1976a, b). Magnetic contrasts within the caldera (lower in the west half) are best explained in major part by extensive hydrothermal alteration of the Bishop tuff in and near the resurgent dome (Williams, 1976). A simplified model which emerges from the above results involves a magma chamber at a depth of 6 to 8 km, which has underlain the west half of the caldera for at least the last 0.3 m.y. A period of 0.3 m.y. is sufficient for equilibration of an overlying conductive regime. This is, of course, an oversimplification: young rhyolitic rocks of the Inyo Domes, which erupted in the west moat as recently as 700 yrs ago, appear to be a mixture of magmas from the residual Long Valley chamber and from the younger Mono Craters magma system farther to the north. This and other recent vol- canic activity suggest that localized areas of relatively shallow hot rock may underly the west moat; their presence may be masked by the flow of cold ground water originating near the west rim. However, if the hydrothermal system has persisted for a period of 0.3 m.y., as suggested by Bailey, Dalrymple, and Lanphere (1976), the paucity of evidence for hydrothermal ac- tivity directly associated with any of the postcaldera eruptive rhyolites and estimates of present-day heat discharge imply that the hydrothermal system is re- A37 lated to the main magma chamber, which is a deeper and larger heat source. ANALYSIS OF CONCEPTUAL MODEL OF HYDROTHERMAL SYSTEM DESCRIPTION OF CONCEPTUAL MODEL On the basis of geologic, hydrologic, geophysical, and geochemical characteristics of the Long Valley caldera discussed previously, we have developed a generalized conceptual model of the hydrothermal system. In many respects the model is oversimplified, and should not be considered a unique representation of the caldera’s hydrothermal system. Although the gross features of the model are consistent with known constraints, ad— ditional data from deep test holes would add consid- erably to the level of detail with which the system could be modeled. Indeed, results obtained from Repub— lic Geothermal’s recent deep test well caused a significant revision of our original model in regard to hydrothermal conditions beneath the eastern part of the caldera. With these limitations in mind, we have analyzed heat and fluid flow in the model, using nu- merical simulation techniques, in order to demonstrate the capabilities of this approach, to provide prelimi- nary quantification of the relationship between heat and fluid flow, and to gain insights as to the hydraulic characteristics of the hydrothermal reservoir. The conceptual model is illustrated in figure 18. The model is three-dimensional, includes the area within the topographic boundary of the caldera floor, and ex- tends to a depth of 6 km. For numerical simulation, the caldera rocks are divided into 5 horizontal layers, cor- responding in general to the major rock units identified by the seismic refraction study (Hill, 1976), the geolog- ical investigations (Bailey, and others, 1976), and the calculations of depths of fill presented previously. The upper layer, 1 km thick, corresponds to the postcaldera sedimentary and volcanic rocks having low P-wave ve— locities (from 1.5 to 3.4 km/s). This layer includes the shallow ground-water subsystem described earlier. Layer 2 corresponds to the densely welded Bishop Tuff and forms a continuous layer over the area of the cal- dera having a seismic P-wave velocity of 4.0 to 4.4 km/s. Layer 3 extends to a depth of 3 km and includes densely welded Bishop Tuff, rhyolite of Glass Mountain, and other precaldera sediments and vol- canics, and, in parts of the caldera, probably some basement rocks. Layers 2 and 3 correspond approxi- mately to the deep ground-water subsystem described earlier. Layers 4 and 5 extending to 6 km are consid- ered as impermeable but thermally conductive granitic and metasedimentary basement rocks. Two layers are used in this depth interval to allow more accurate nu- merical heat-flow simulation. A magma chamber A38 below 7 km under the western half of the caldera, as suggested by the seismic, teleseismic, and heat-flow studies discussed previously, is simulated by a con- stant (with time) but areally variable temperature dis- tribution at the base of the model. The adequacy of this simplification is discussed later in this section. The hydraulic and thermal properties of each layer in the model are listed in table 14. Layer 1 is considered an impermeable cap except along parts of the caldera rim, where recharging ground water moves downward along the ring fault, and in the Hot Creek gorge area, where hot water flows upward along faults and dis- charges in the springs in the gorge. We assume that, although ground-water flow in the shallow sedimen- tary and volcanic rocks may have significant localized —-z—> 395 $ERRA NEVADA o‘°° 003d...» GEOHYDROLOGY OF GEOTHERMAL SYSTEMS effects on the surficial thermal regime, flow in this layer can be considered as separated hydraulically from the hydrothermal reservoir (the deep subsystem), except as noted. Extremely low values of vertical hy- draulic conductivity measured in cores from the upper 300 m tend to support this assumption, although the existence of slow vertical flows through this layer, which may significantly affect the temperature and heat-flow regime at depth, cannot be discounted, par- ticularly in areas of active faulting. Limitations of computing times and the lack of adequate hydrologic data require that the shallow ground-water subsystem be neglected in this analysis, except as a source or re- charge along parts of the caldera rim. The hydrothermal reservoir is assumed to exist in Glass Mtn Gorge Springs LAYER 2 ’ LAYER 3 9M; FIGURE 18.—Block diagram showing conceptual model of Long Valley hydrothermal system. Model consists of 5 horizontal layers having properties listed in table 15. Patternless layers between depts of 1 and 3 km represent the hydrothermal reservoir in fractured, densely welded Bishop Tuff. Recharge to reservoir is by way of caldera ring fault in the west and northeast. Discharge is by way of faults and fractures to springs in Hot Creek gorge. Straight arrows indicate ground-water flow, wavy arrows indicate heat flow. Vertical to horizontal exaggeration approximately 1.6 to 1. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA A39 TABLE 14.———Hydraulic and thermal properties for Long Valley hydrothermal model Thermal conductivity (meal/)cms - °C)) Heat Intrinsic Vertical 050301ch capacity permeability compressibility (cal/)cm" °C)) (m2 X 10‘”) mz/N) Porosity 0.54—0.58 0 110.0 X 10—10 0.35 .54— .58 0.03—0.35 21.0 X 10~1o .10 .54— .58 .03— .35 1.0 X 10‘10 .05 .54— .58 0 1.0 X 10—10 .05 .54— .58 0 1 0 X 10’10 .05 ‘Estimated at 3 times compressibility of sand (Jacob, 1950. p. 334) because of abundance of tuffaceous and lacustrine deposits. 2Estimated at 1/3 compressibility of sand because of induration. the fractured, welded Bishop Tuff in layers 2 and 3 of the model. The possibility that the main aquifer is below the Bishop Tuf‘f in precaldera sediments, tuffs, and lava flows should be recognized. Because the aver- age thickness of the welded Bishop Tuff is about 1.4 km, layer 3 in the model could be considered as includ- ing such an aquifer. The effect of the depth of fluid circulation on temperature and heat-flow distributions was evaluated by simulating flow in layer 2 only, in layer 3 only, and finally in both layers 2 and 3. In each case, ground-water flow is from the higher altitudes along the west and northeast rims toward discharge areas in Hot Creek gorge and along the southeast cal- dera rim. Additional driving force causing flow is pro- vided by density differences between hot and cold parts of the flow system. In some model runs, pressures based on altitudes of the water table in recharge and discharge areas (fig. 10) were specified and reservoir permeability was ad- justed to yield the desired mass flux of water through the reservoir. Similar results were obtained by specify- ing recharge rates and pressures in discharge areas, and adjusting permeability until pressures in recharge areas matched the observed water-table altitudes. Re- charge rates were determined by distributing the total mass flux over the areas of the recharge nodes, with weighting according to water-table altitudes. These procedures yield values for the product of effective res- ervoir permeability times thickness; the ranges in permeability values listed in table 15 represent results from three cases of assumed reservoir permeability dis- tributions, as discussed below. The principal limitation of this procedure is the uncertainty as to pressure losses in the upflow and downflow fault conduits. NUMERICAL SIMULATION To permit numerical simulation of heat and fluid flow, each layer of the model is subdivided into 82 grid blocks or nodes along land-net lines (fig. 19). Finer nodal spacing is used near the discharge areas. Ther- mal water is assumed to discharge only over the sur- face of the node that includes the springs in Hot Creek gorge, and through the southeast rim of the caldera. Only minor differences in computed distributions of pressure and temperature would be expected if a more detailed distribution of hot-water discharge were modeled, because approximately 80 percent of the sur- face discharge from the thermal reservoir is through the springs in the gorge (Sore and Lewis, 1976). The equations and solution procedure used in this study are described in detail by Sorey (1975). The flow equation is V‘[P%(AP—pg)] =0; (2) where p = fluid density, k = intrinsic permeability, ,u = dynamic viscosity, P = fluid pressure,§= gravita- tional acceleration vector, C = fluid-rock com- pressibility, and t = time. Equation 2 is based on conservation of mass and Darcy’s law for nonisothermal fluid flow in porous media. An assumption inherent in this formulation is that fluid flow in the hydrothermal system, although probably controlled locally be permeable zones along faults, can best be described in large scale as a porous medium in which permeability is distributed effec- tively throughout. The energy equation is 0T v - [KmAT1—pc6-AT=(pc)' 6t (3) where Km 2 rock-fluid thermal conductivity, T = rock- fluid temperature? = Darcy velocity vector, c = fluid specific heat at constant volume, and (pc)’ = rock-fluid heat capacity. Equation (3) accounts for conductive and convective A40 transfer of heat under steady-state and transient con- ditions. We assume that thermal equilibrium exists be- tween fluid and solid phases at points of contact and that heat transfer by hydrodynamic dispersion can be neglected in the type of problem considered here (Mercer and others, 1975, p. 2618). Temperature- dependent parameters, p. and c, in equations (2) and (3) were evaluated from tabulated data (Dorsey, 1968) The equation of state relating fluid density to tem- perature is p = pa [1 — ,8(T — To) - Y (T — TOP] (4) where p0 = fluid density at reference temperature To, [3 ; thermal expansivity, and Y = coefficient for second-order fit. Density variations with pressure are neglected. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Simultaneous solutions to the flow and energy equa- tions were obtained by an integrated finite-difference method involving iterative solutions at selected time steps for pressure, temperature, and velocity fields. This numerical procedure offers considerable advan- tages over standard finite-difference methods in terms of reduced computing times and nodal requirements (Narasimhan and Witherspoon, 1976). The time step used to solve the energy equation is continuously in- creased by a factor between 1 and 2, with the limitation that the maximum change in nodal temperatures per time step be less than about 10 percent of the maximum total change expected in the system. Be- cause the response times for pressure changes are much smaller than for temperature changes, the flow system essentially equilibrates to a quasi-steady state within each thermal time step. For a simulation period of 35,000 years, approximately 50 thermal time steps ‘ j Bald Mtn . //‘\Glass Mtn LU LU ///\\ a g ///\\ 0: cc //\ 37°45’ — — Mammot Mtn Lake Crowley o I O 5 10 .— 37 35 — . I QKILOMETERS l l 119°00’ 118°45' FIGURE 19.——Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with uniform reservoir permeability distribution (case A). R denotes recharge node; D denotes discharge node covering springs in Hot Creek gorge. Principal faults shown as solid heavy lines with ball on downthrown side. Arrow denotes ground-water discharge at depth through southeast rim. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA were used; for a 350,000-year simulation, approxi- mately 70 time steps were required. As indicated in figure 19, by the nodes marked R, recharge water enters the upper layer of nodes along the caldera rim. The centers of the rim nodes corres- pond generally to the caldera ring fault which is as- sumed to provide conduits for downflow to the hydro- thermal reservoir. The rationale for simulating re- charge only around the west and northeast rims in- volves several factors. First, isotopic evidence indicates that meteoric water from the west rim is the dominant source of the hot-water component in the thermal springs. Second, water-table altitudes are highest around the west rim (fig. 10), and cold, isothermal tem— perature profiles are observed in core holes drilled in the west moat (figs. 13 and 14). On the basis of data from the 2.11-km-deep test well (pl. 1, well 66—29), water in the deep ground-water subsystem under the eastern part of the caldera is isotopically similar to meteoric waters from the northeast rim. This similar- ity and the nature of the temperature profile observed in the deep well suggest that in addition to the easterly flow of hot water from the west rim, there is a flux of cooler water moving through the welded Bishop Tuff east of Hot Creek, the source of which is presumed to be recharge around the northeast rim. However, in pre- liminiary model simulations which permitted recharge through all the rim nodes by assigning to them a non- zero permeability and connecting them at the land sur- face to nodes held at constant pressure based on corre- sponding water-table altitudes, upflow occurred along the north, south, and east rims. Thus for deep recharge from the northeast rim to occur, artesian heads well above water-table altitudes would have to exist on the flanks of Glass Mountain. Additional water-level data are required to confirm this possibility. The recharge rate from the northeast rim was arbi- trarily set at about one-third of the flux from the west rim; this resulted in reservoir temperatures under the eastern caldera which agree with those observed in the deep test well. Simulation of subsurface discharge through the southeast rim provides the outlet for most of the northeast-rim inflow. The latter would otherwise have to discharge in the springs along Hot Creek gorge. Mixing of these waters would result in consid- erably cooler reservoir temperatures beneath Hot Creek gorge than in the numerical simulations dis- cussed below. The geochemical evidence discussed pre- viously, particularly the analysis by Fournier, Sorey, Mariner, and Truesdell (1976), indicates that such deep mixing does not occur. Permeable channels for the southeast rim outflow, which was modeled as leaving the reservoir through the caldera wall, could be pro- vided by the Hilton Creek fault or possibly by con- A41 cealed faults farther to the east (fig. 19). Additional subsurface information would be required to establish the existence and location of such outflow channels. HYDRAULIC CHARACTERISTICS Ground-water flow in the hydrothermal model be- tween recharge and discharge areas is simulated trhough a horizontal layer of permeable rock from 1 to 2 km thick in layer 2 and/or layer 3. The permeability needed to obtain the observed discharge of hot water can be considered an effective permeability which should be useful in analyzing reservoir response to de- velopment. A similar modeling effort for the hot-water reservoir at Wairakei, New Zealand, yielded tempera- ture and pressure patterns that correlated well with observed response prior to initiation of two-phase con- ditions in the reservoir (Mercer and others, 1975). However, where faults and fractures are primarily re- sponsible for the effective reservoir permeability, an individual well drilled into the reservoir may en- counter considerably larger or smaller permeabilities than those indicated by the model simulation. In figure 19, northwest-trending faults, some of which probably are extensions of the Hilton Creek fault, generally traverse the resurgent dome in the central part of the caldera. In the west moat, a group of north-trending faults are alined with postcaldera ex- trusive rocks (< 0.15 my old) extending from Mam- moth Mountain north toward the Mono Craters. Frac- tures in the welded tuff associated with these major faults, as well as other faults not delineated at the land surface, are considered to provide the major channels for flow in the hydrothermal sytem. Some justification for assuming no hydraulic connec- tion between the upper and lower ground-water sub- systems (except along the ring fault) is that secondary fracture permeability would be less likely to remain effective in the poorly indurated deposits overlying the Bishop Tuff. The occurrence of present-day hot springs along active faults, however, indicates that where poorly indurated deposits are altered and cemented, they can and do have secondary permeability. The ap- parent lack of faulting in the east part of the caldera (with the exception of the ring fault) need not preclude permeable zones in the Bishop Tufl" in that area which could also result from brecciated zones between two major cooling units identified within the Bishop Tuff southeast of the caldera (Sheridan, 1968). Three cases of reservoir permeability distribution are considered. Case A, referred to as the continuous reservoir (fig. 19), postulates a uniform permeability over the area of the caldera. In case B, referred to as the fault-zone reservoir (fig. 20), the northwest- trending fault zone in the central part of the caldera is A42 considered as the only permeable zone within the Bishop Tuff. The volume of permeable rock in the fault—controlled rservoir is assumed to be approxi- mately 40 percent of that in the continuous-reservoir model. Case C, called the low-permeability-block res- ervoir, is similar to case A except for a zone of lower permeability (100 times lower than the remainder of the reservoir) between the west moat and the Casa Diablo Hot Springs area, as shown in figure 21. The rationale for case C involves the consideration that temperatures above 200°C are indicated geochemically in the reservoir underlying the Casa Diablo Hot Springs area, as discussed in the next section. For Cases A and C, numerical computations were made for mass fluxes of 190 kg/s, 250 kg/s, and 300 kg/s GEOHYDROLOGY OF GEOTHERMAL SYSTEMS recharging in the west rim and discharging in Hot Creek gorge, and 110 kg/s recharging in the northeast rim and discharging through the southeast rim. For case B, a flux of 250 kg/s recharging in the north rim and discharging in the gorge was simulated, with no discharge through the southeast rim. In table 15, values of intrinsic permeability obtained in the model simulations are compared with values from laboratory tests on Long Valley cores, with values for ash-flow tuffs at the Nevada Test Site (Keller, 1960; Winograd and others, 1971), and with values for pumice breccia and vitric tuff of the Wairora aquifer in New Zealand (Mercer and others, 1975). The Long Val- ley model results shown are for a mass flux of 250 kg/s discharging in the gorge. l // l l Bald Mtn // lllGIass Mtn / oi 3% 37°45’ ._ We“ —- Gwss I v. aarms" \ \ \ l / I \. \ L 7 | .k x \' l I \Q X / i I I I \ l , \ Casa DI bIOQ Tr \ I, M Mammotr I'll—“ // Mtn y\‘ ‘i , \\\~‘—",v ‘ \U.-~ fl #: _____ Lake “A Crow/0y o 5 10 7‘ l u m KILOMETERS 37°35’ — g - c S I l 119°00' 118°45’ FIGURE 20.-—Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with fault-zone reservoir permeability distribution (case B). Permeability of nodes outside hatch area set to zero. R denotes recharge nodes, D denotes discharge node covering springs in Hot Creek gorge. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA TABLE 15.Jntrinsic-permeability data from Long Valley model and other studies Permeability Reservoir Data source thickness (km) (millidarcysl Continuous reservoir (case A) ______ 1 30 Fault-controlled reservoir (case B) __ 1 350 Low-permeability-block reservoir (case C) ________________ 1 150 Wairakei model2 __________________ 04—085 100 Long Valley cores3 ______________________ 00005—1800 NTS ash-flow tufl's4 ______________________ .04—10.0 NTS welded tufts (fracture) ________ .05—.2 5,000—30,000 ‘This value applies to west three-fifths ofcaldera; permeability equaled 30 md in eastern two-fifths. 2Wairora aquifer consisting of pumice breccia and vitric tuffs as modeled by Mercer, Finder, and Donaldson (1975). "Includes only data from cores of altered rock, flow rocks, and nonwelded tuffs. "Oak Springs Formation (Keller, 1960). sWinograd, Thordarson, and Young (1971). A43 Permeabilities computed from the model are inversely proportional to the assumed reservoir thickness. The range for a reservoir 1 km thick is from 30 millidarcys (md) for the continuous reservoir to 350 md for the fault-zone reservoir. Corresponding values of hydraulic conductivity, which are related to intrinsic permeabil- ity as k =— (5) K M pg vary with temperature within the reservoir. At a tem- perature of 100°C, hydraulic conductivity for the con- tinuous reservoir is 0.09 m/day, compared to 1.0 m/day for the fault-zone reservoir. The reservoir temperature distributions for which the results in table 16 were ob- tained are discussed in the next section. 37°45' _ \ @\ Glass Mtn //\\ 01 l/\ d‘\ _ T 25 , T.3s. W \ f I \ . l r / R l l, / g \I . | \ x ' ‘. 1 ' ‘18 \ or \ I \ \ V I! R \\ V D I Gorg \ \ Case Die lolg \fl _. \ L "r— R \ \< ‘1; MammotilrI—I __ \\ \ 6\ , Mtn IR. \ ' ‘ \\ 1 ’/ \~‘_—” ~\___’_fl_\“ — 77*: ....... O Lnlid ‘9 Crow/0y t“ “,2 0 5 10 L i l Kl LOMETE RS 37°35’ — 1)} _ C S l J 1 19°00’ 118°45' FIGURE 21.—Sketch map of Long Valley caldera showing nodal configuration for numerical simulation of hydrothermal system with low- permeability-block reservoir permeability distribution (case C ). Permeability of hatched reservoir nodes equal to 10—2 times permeabil- ity of other reservoir nodes. R denotes recharge nodes, D denotes discharge node covering springs in Hot Creek gorge. Arrow denotes discharge at depth through Southeast rim. A44 The effective permeability of 100 Ind used in the Wairakei model (Mercer and others, 1975) is within a factor of three of each of the Long Valley model cases. This suggests that the results for the Long Valley model are reasonable, but does not show which of the three cases is most likely. Thermal consideration dis- cussed in the next section of the report indicate that depths of fluid circulation required for the fault-zone model to obtain reservoir temperatures in excess of 200°C are considerably greater than for the other two cases. Permeabilities measured on all cores from the upper 300 m of Long Valley fill range from approximately 10—4 md to 104 md. As shown in table 6, the highest values were obtained on samples of sandy or gravelly tuffaceous sediments, and the lowest values were ob- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS tained from hydrothermally altered tuffs and sedi- ments. In table 16, the range in permeability for cores of altered rock, flow rocks, and nonwelded tuffs are close to or below the range of effective permeability from the model simulations, as are values reported by Keller (1960) for ash-flow tuff at the Nevada Test Site (NTS). In contrast, permeabilities obtained from well tests in fractured welded tuff at NT S (Winograd and others, 1971) are considerably greater than the per- meabilities indicated by the model. This supports the concept that the effective permeability obtained with the numerical model represents an integration of the effects of fracture permeability over the volume of the welded-tuff reservoir. To illustrate circulation patterns in the simulated reservoir, pressures at each node were converted to I /\ l ///\\ Bald Mtn //‘ Glass Mtn //‘ /‘ 4? 01 ’4\ a ’A 37°45’ —‘ creek “ em“ I I / 4r"- 09 a W I I l ‘\ \\ 2%“ l I x I I l \ \ I Mammoth” ‘ Mm [K 1‘ ~ ’ Crowley 0 5 10 l l 4 Kl LOMETE RS 37°35' — ‘1 I 119000! 118045! FIGURE 22.—Sketch map of Long Valley caldera showing equivalent hydraulic head in reservoir from 1 to 2 km deep from case A with uniform reservoir permeability of 30 millidarcys, hot-spring discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and simulation period of 35,000 years. Contours of equivalent hydraulic head in meters above mean sea level. Interval 20 m. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA equivalent or “cold water” heads using the relationship Ho = P/pog + z (6) where p0 = fluid density at reference temperature (10°C) and 2 = altitude of node above sea level. This also permits comparisons of equivalent head distri- butions with the water-table map (fig. 10). Reservoir head distribution in layer 2 at 1.5-km depth for the continuous-reservoir and low-permeability-block- reservoir cases are shown in figures 22 and 23. Both cases show the predominant eastward flow toward the Hot Creek gorge area and the effects of recharge from the Glass Mountain area and discharge through the A45 southeast rim. Comparison of figures 22 and 23 indi- cates the barrier effect of the low-permeability block west of the Casa Diablo area, with the bulk of circula- tion passing north of the barrier. Reservoir permeabil- ity in case C was about 50 percent greater in the western three-fifths of the caldera than in the continu- ous-reservoir case A in order to yield the same flow. The permeability assigned to the low-permeability block was two orders of magnitude lower than elsewhere in the reservoir. Comparisons of reservoir head distribution in figures 22 and 23 with the head distributions in the shallow ground-water system in figure 10 help to clarify some of the hydraulic characteristics of the hydrothermal I 0‘ I \\ €7\Bald Mtn ///‘\Glass Mtn K ’m 0/1 01 ’v 37°45’ — 0'99 LR» C Giass I " ‘ ‘" / ’y./ \ 1’06 I ‘/ I , Diablo , Lake ”K Crowley o 5 10 w / L . J KILOMETERS .\ * q A \_ iv I; ’ . 37°35’ — \ \ LV/y V (a _ \\ \ I \ 119°00' 118°45' FIGURE 23.—Sketch map of Long Valley caldera showing equivalent hydraulic head in reservoir from 1 to 2 km deep from case C with low-permeability-block reservoir west ofCasa Diablo Hot Springs, permeability of 45 millidarcys in remainder of western three-fifths of caldera, and permeability of 30 millidarcys in eastern two-fifths of caldera. Discharge in Hot Creek gorge equals 250 kg/s; southeast rim outflow equals 110 kg/s. Simulation period equals 35,000 years. Contours of equivalent hydraulic head in meters above mean sea level. Interval 20 m. A46 model, in spite of the fact that hydraulic connection between the deep and shallow subsystems was simu- lated only in the recharge and discharge areas. Equiva- lent reservoir heads along the west rim are lower than water-table altitudes because of head losses incurred in downflow to the reservoir. Heads along the northeast rim are above water-table altitudes, which implies that artesian conditions are required in the shallow ground-water subsystem below the uppermost uncon- fined aquifer if recharge to the deep subsystem occurs in that area. Numerous flowing wells and springs on the flanks of Glass Mountain indicate that artesian conditions do exist. As discussed previously, the mag- nitude of reservoir head needed to obtain an assumed inflow of 110 kg/s in the northeast rim is dependent on the value of permeability used. Less head would have resulted if either a higher permeability or less inflow were simulated in this region. However, both quan- tities are constrained to be close to the chosen values; permeability in the eastern caldera is unlikely to ex- ceed that in the western caldera because of the absence of surficial faulting, and inflow from the northeast rim must be sufficient to produce relatively low tempera— tures observed in the Bishop Tuff in the deep test hole. An alternative possibility is that the reservoir in the east part of the caldera is hydraulically separated from the reservoir in the west part, perhaps along the east margin of the intracaldera Hilton Creek fault zone, which could act as a barrier due to sealing effects of hydrothermal alteration. This would have little effect on fluid circulation and temperatures in the model simulations of the western three—fifths of the caldera but might lower the head requirements for recharge from the northeast rim. Antelope Valley Dry Creek Casa Diablo WEST l | l GEOHYDROLOGY OF GEOTHERMAL SYSTEMS In most of the west part of the caldera, equivalent reservoir heads are below water levels in the shallow ground-water subsystem, which indicates a potential for downward flow through the shallow subsystem. Whether such flow actually occurs depends, of course on the vertical permeability distribution. Additional data from drill holes deeper than 300 m in the western part of the caldera are needed to adequately assess the degree of interconnection between the shallow and deep ground-water subsystems. THERMAL CHARACTERISTICS Any model of the hydrothermal system in Long Val- ley is constrained by the estimates of natural heat dis- charge from the caldera and hot-water discharge from the reservoir. Our best estimate of the heat discharge by convection and conduction is 6.9 X 107 cal/s. The ranges in mass flux and maximum reservoir tempera- ture, obtained from geochemical mixing models, are from 190 to 300 kg/s and from 210°C to 282°C, with the highest reservoir temperature corresponding to the lowest mass flux. The magnitude of heat discharge, which requires a heat supply of 15 HFU averaged over the area of the caldera, places severe demands on the underlying heat source, depending upon the length of time it has persisted. In this section we describe the use of the numerical model to evaluate conditions under which an underlying magma chamber could supply the required heat for various periods of time, including the steady-state condition. The important thermal features of the model are dis- cussed with reference to figure 24, which shows steady-state isotherms in an east-west cross section Hot Creek Gorge Owens River EAST 700° O 5 l I 10 4 KILOMETERS FIGURE 24.—Diagrammatic east-west cross section of Long Valley caldera showing steady-state, conduction-only isotherms from model simulations without fluid flow and used as initial conditions in simulations of flow in the hydrothermal system. Dotted pattern indicates simulated reservoir and vertical flow channels through confining layers of shallow ground-water subsystem. Thermal properties of each layer in the model are given in table 15. Lines of equal temperature in degrees Celsius. Interval 100°C. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA passing through Hot Creek gorge. This is the tempera- ture distribution that would exist in the absence of fluid flow, or under conduction-only conditions, with the boundary conditions used in the model. At the bot- tom, a constant temperature distribution is imposed with 800°C under the western half, 300°C under the eastern third, and a linear variation between. As dis- cussed, this boundary condition is intended to simulate magma at 6 km under the western half of the caldera; the 300°C temperature under the eastern part was selected to yield conductive heat flows near 2 HFU at the east rim as constrained by the measured heat flows in that area (Lachenbruch, Sass, Munroe, and Moses, 1976). A constant temperature of 10°C is imposed at the land surface except in the discharge area at Hot Creek gorge. The caldera walls are treated as insulated boundaries. The reservoir depth shown in figure 24 is from 1 to 2 km; other simulations involved reservoir depths from 2 to 3 km and from 1 to 3 km. A47 As discussed previously, the assumption that magma has existed at depths from 6 to 8 km during the last 0.3 m.y. is reasonable. Our transient analysis of the hydro- thermal system also involves a period of approximately 0.3 m.y., which is sufficient for steady-state conditions to be reached. We are primarily interested in thermal conditions after two distinct periods of flow, one being the 30,000—40,000-year period of recent spring flow which has contributed salts to Searles Lake (Smith, 1976); the other being the 0.3-m.y. period since the ear- liest indicated time of extensive hydrothermal activity (Bailey and others, 1976). The initial temperature dis- tribution used in our transient simulation is the steady-state, conduction-only, condition shown in figure 24. For the recent period of spring flow, the use of this initial condition would be valid if the system has had nearly 0.3 my to equilibrate conductively after cessation of the earlier period of hydrothermal activity. This would correspond to the concept of intermittent 37°45’ — “W ///\ T //\ Bald Mm 4‘ Glass Mm /\ 7‘ /\ ’/\ /\ //\ I? l ———————— '7‘ \‘—_-__.—r Mtn \ F0 37°35’ & l, i M th ‘— \~\ ammo I"\ (v Case Diablo 10 IKILOMETERS I 1 19°00’ 118°45' FIGURE 25.—Map of Long Valley caldera showing steady-state conductive heat flow above reservoir (from model simulation) without fluid flow. Lines of equal heat flow in heat-flow units (HFU). Interval 0.5 HFU. A48 hydrothermal activity discussed previously. In con- trast, if spring flow has been continuous over the past 0.3 m.y., our 0.3.-m.y.—peri0d simulation would indi- cate the present-day steady-state thermal condition, even though the assumed initial condition in figure 24 did not exist 0.3 m.y. ago. In figure 25 the areal distribution of heat flow under the initial temperature conditions is shown. The total conductive heat flow is 2.1 X 107 cal/s, and the heat flux varies from 5.7 HFU near the west rim to 2.0 HFU near the east rim. Comparison with the measured total heat flow of 6.9 X 107 cal/s and the measured surficial heat-flow map (fig. 15) indicates the extent to which ground-water circulation affects the thermal regime of the caldera. Simulation of ground-water flow through the hy- drothermal system produces the temperature distri- butions illustrated in figures 26 and 27. These results are for a continuous reservoir (case A) at a depth of 1 to 2 km, with a hot-spring discharge of 250 kg/s persisting for 35,000 years. Temperatures in the east-west cross section in figure 26 indicate the cooling effect of re- charge near the west rim, the increased temperature gradients and heat flows below the reservoir in the west part of the caldera, and the gradual heating of the reservoir fluid as it moves eastward toward the dis- charge area at Hot Creek gorge. The temperature in the middle of the reservoir beneath the gorge is near 200°C, and lateral heat conduction away from the ris- ing column of hot water causes the high temperatures in the adjacent shallow subsystem. The effect of re— charge from the northeast rim is also evident with res- ervoir temperatures of about 80°C in the southeast part of the caldera. These lower temperatures are reason- ably consistent with results from the deep test hole. Antelope Dry Creek Casa Diablo Valley WEST I | I 100° 5 0 L l GEOHYDROLOGY OF GEOTHERMAL SYSTEMS The areal distribution of reservoir temperature shown in figure 27 applies to the midplane of the reser- voir at 1.5-km depth, and could be considered as aver- age reservoir temperatures. This interpretation is, of - course, limited by the coarseness of the vertical nodal spacing used in the model. One observation from the results shown in figure 27 is that a laterally extensive layer of uniform temperature does not exist, although reservoir temperatures below the area of present-day hot-spring discharge around the southeast flank of re- surgent dome are between about 180°C and 240°C. Res- ervoir temperatures below the Casa Diablo area, how- ever, are only about 120°C, whereas measured and chemically estimated temperatures at depth exceed 180°C. The lower temperatures from the model simula- tion result from the proximity of Casa Diablo to the recharge area around the west rim. The low-permeability-block permeability distribu- tion, case C discussed previously, offers an alternative model which produces higher temperatures under Casa Diablo. The temperature distributions at 35,000 years in cross section and plan for this case are shown in figures 28 and 29. A pronounced temperature and heat-flow anomaly west of Casa Diablo is produced by the permeability barrier which diverts the eastwardly ground-water flow to the north. In this case, average reservoir temperatures are near 180°C under Casa Di- ablo as well as under Hot Creek gorge. Existing test- hole data west of Casa Diablo are too sparse and too shallow to confirm the presence of the temperature— heat-flow anomaly shown in figure 28. Additional test drilling might delineate this as a potential area for hot-dry-rock exploration. Results for the fault-zone-reservoir (case B, fig. 20), for a similar period of flow in a more areally restricted Hot Creek Gorge Owens River EAST 10 J Kl LOMETERS FIGURE 26.—Diagrammatic east-west cross section of Long Valley caldera showing isotherms from model simulation (case A—uniform reservoir permeability distribution) after 35,000 years with a discharge of250 kg/s in Hot Creek gorge, southeast- rim outflow of 110 kg/s, and reservoir depth ofl to 2 km. Lines of equal temperature in degrees Celsius. Interval 100°C. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA A49 l //;\ l ///‘\Ba|d Mtn gems Mtn / % ’/,\ s 37°45’ — I I l \ \ Mammo Mtn Lake Crowley 0 5 10 L 4 I Kl LOMETE RS \ 37°35' —- — J 119°00’ 118°45’ FIGURE 27.—Map of Long Valley caldera showing distribution of average reservoir temperature from model simulation (case A—uniform reservoir permeability distribution) after 35,000 years with a discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth ofl to 2 km. Lines of equal temperature in degrees Celsius. Interval 20°C. Antelope Hot Creek Dry Creek Casa Diablo Valley Gorge Owens River WEST l | l | | EAST. " 100° ' O 5 10 1 4L J KILOMETERS FIGURE 28.—Diagrammatic east-west cross section of Long Valley caldera showing isotherms from model simulation (case C—low permeability-block reservoir west ofCasa Diablol after 35,000 years with discharge 0f250 kg/s in Hot Creek gorge, southeast- rim outflow of 110 kg/s, and reservoir depth OH to 2 km. Lines of equal temperature in degrees Celsius. Interval 100°C. A50 reservoir at a 1—2-km depth, yield maximum reservoir temperatures of only about 100°C. These lower temper- atures are caused by the reduced area for heat flow into the reservoir. To obtain higher reservoir temperatures with this model, deeper circulation is required. As discussed in the section “Chemical Geother- mometers,” estimates of actual maximum reservoir temperatures and fluxes of hot water vary. The analysis by Sorey and Lewis (1976) indicates a reser- voir with 210°C water discharging from 250 to 300 kg/s; the analysis by Fournier, Sorey, Mariner, and Trues- dell (1976) indicates a 282°C reservoir discharging from 190 to 230 kg/s. However, convective heat flows for each case, based on equation (1), are not significantly different. This is in agreement with the GEOHYDROLOGY OF GEOTHERMAL SYSTEMS model results which show that, for a given reservoir depth and period of spring flow, maximum reservoir temperatures are inversely related to the mass-flux rate. For example, if in the continuous-reservoir case shown in figures 26 and 27, a mass flux of 190 kg/s instead of 250 kg/s is simulated, reservoir tempera- tures under Hot Creek gorge near 245°C instead of 200°C are obtained. Thus the choice of which geochem- ical model of reservoir temperature and flux to match with these simulations is perhaps not as important at this stage as is the matching of total heat flow from the model with the measured value of 6.9 X 107 cal/s. Heat flows corresponding to the three cases of reser- voir permeability distribution discussed above, are listed in table 16. 37°45’ 37°35’ 10 Q KILOMETERS 119°00’ 118°45' FIGURE 29.——Map of Long Valley— caldera showing distribution of average reservoir temperature from model simulation (case C— low-permeability-block reservoir west of Casa Diablo) after 35,000 years with discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth of 1 to 2 km. Lines of equal temperature in degrees Celsius. Interval 20°C. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA A51 TABLE 16.—Conductive and convective heat discharge for each case of reservoir-permeability distribution after 35,000 years of fluid flow at 250 kg/s in reservoir at 1—2-km depth Reservoir- Conductive Convective permeability heat discharge heat discharge Total case (cal/s) (cal/s) (cal/s) A __________________ 1.6X10" 4.9X107 16.5)(107 B __________________ 1.4)(107 2.2)(107 3.6x 107 C __________________ 1.8x 10" 4.2X107 16.0X107 ‘Does not include 1.4 X 107 calls advected through southeast caldera rim. These results indicate that present-day heat flow could be obtained after a period of 35,000 years with an aver- age depth of hydrothermal circulation of 1.5 km, pro- vided the reservoir is continuous over the western half of the caldera. For a more areally restricted reservoir, correspondingly deeper circulation is required. Conductive heat flows in table 16 were determined from the temperature distributions in layer 1 at a depth of 0.5 km. The areal distribution of surficia] heat flow for the continuous-reservoir case is shown in figure 30. Differences between this map and the meas- ured shallow-heat-flow map (fig. 15) are due in part to the effects of shallow ground-water circulation which were neglected in the numerical model. The most obvi- ous effects are seen in the west moat where apparently shallow recharge cools ground temperatures over a siz- able area, and east of Hot Creek gorge where lateral flow of warm water in the shallow system causes a high-heat-flow bulge to the southeast. II I 395 37°45' C/rgfl ’_\ 4 Case Diablo I _._ / a’ \\ I __________ _ _,- ’ Lake Crow/9V o 5 10 J . KILOMETE RS 37°35' — F l “9000' 118°45' FIGURE 30,—Map of Long Valley caldera showing results from model simulations (case A—uniform reservoir-permeability distribution) in terms of conductive heat flow above a depth of 0.5 km after 35,000 years with discharge of 250 kg/s in Hot Creek gorge, southeast-rim outflow of 110 kg/s, and reservoir depth OH to 2 km. Lines of equal heat flow in heat-flow units. Interval 1.0 HFU. A52 If hot-spring discharge continues for periods greater than 35,000 years, reservoir temperatures cool significantly from the previous results. For the dimen- sions and parameters used in these simulations, tem- peratures would be close to steady state after 350,000 years. The temperature distribution in cross section corresponding to the continuous-reservoir (from 1 to 2 km deep) case after 350,000 years of spring flow is shown in figure 31. Additional heat has been removed from the rocks above and below the reservoir, and res- ervoir temperatures below the gorge have dropped from about 200°C to about 100°C. The total heat flow is now only 2.7 X 107 cal/s, which does not include 0.7 X 107 cal/s advected through the southeast rim of the caldera. The transient response of reservoir temperature under the gorge is plotted in figure 32 for case A and two reservoir depths. In each simulation the early—time increase in reservoir temperature under the discharge area, prior to about 10,000 years, is due to the arrival of hotter water from the west (see fig. 24). After about 10,000 years, discharge temperatures decrease con- tinuously until steady state is reached after approxi- mately 350,000 years. It is possible to calculate the average travel time of water in the reservoir from the time at which the cold temperature wave arrives under the gorge. The effec- tive Darcy, or seepage, velocity Ue, can be computed from equation (3) (neglecting heat conduction) as we =(L/t) [ (WW/p0] (7) with L = the average distance from recharge to dis- charge area andt = time for arrival of the cold temper- ature wave. ForL = 20 km,t = 10,000 years, (pc)’ = 0.6 Antelope Valley Dry Creek Casa Diablo WEST l l I O 5 | I GEOHYDROLOGY OF GEOTHERMAL SYSTEMS cal/ (°C cm3), and pc = 1 cal/(°C cm3), Ve = 0.003 m/d. The average pore velocity, v* = ve/porosity, would then be 0.003/0.1 = 0.03 m/d, indicating that most of the recharge water would show up in the hot springs after about 2,000 years. These calculations apply to a reser- voir 1 km thick. If the Long Valley hydrothermal res- ervoir is less than 1 km thick, pore velocities would be proportionately greater and travel times pro- portionately smaller. The transient responses shown in figure 32 are not expected to represent the caldera’s actual thermal his- tory over the entire period shown. Instead, results for the first 35,000-year period apply to the recent period of boron accumulation in Searles Lake, and would be valid only if the hydrothermal system, after initial ac- tivity, was then inactive for close to the 300,000 years needed for conductive equilibrium to be reestablished. In contrast, results for steady—state conditions (at 350,000 years in fig. 32) would apply to the present day only if hydrothermal activity has been reasonably con- tinuous for the past 350,000 years at close to present rates. Reservoir temperatures under Hot Creek gorge at 35,000 and 350,000 years from all of the numerical simulations are listed in table 18. The effects of both mass—flux rate and reservoir depth on reservoir tem- peratures can be seen. These results indicate that if present-day hydrothermal activity has persisted for only about 35,000 years, estimated reservoir tempera- tures and heat flow could be supplied by a magma chamber at 6 km with fluid circulation to from 1.5 to 2.5 km over the western three—fifths of the caldera. More areally restricted circulation (case B) or greater depths to magma would require correspondingly deeper circulation. Hot Creek Gorge Owens River I I EAST 000° v00 0 10 4 _l KILOMETERS FIGURE 31.—Diagrammatic east»west cross section of Long Valley caldera showing isotherms from model simulation (case A—uniform reservoir-permeability distribution) afier 350,000 years with hot-spring discharge of 250 kg/s, southeast-rim outflow of 110 kg/s, and reservoir depth ofl to 2 km. Lines of equal temperature in degrees Celsius. Interval 100°C. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA TABLE 17.—Average reservoir temperature below Hat Creek gorge for various model runs Reservoir Initial '1‘ after T alter depth T 35,000 yrs 350,000 yrs Run specifications‘ 2 (km) (°C) (°C} °C) 1. Case A, M = 300 __ 1—2 260 179 95 2. Case A, M = 250 __ 1—2 260 203 102 3. Case A, M = 190 -- 1—2 260 245 120 4. Case B, M = 250-_-- 1—2 260 100 60 5. Case C, M = 250_-__ 1—2 260 185 101 6. Case A, M = 300 -_ 2—3 345 262 124 7. Case C, M = 250___- 2—3 345 270 135 8. Case A, M = 250 1—3 300 255 142 ‘Applies to reservoir-permeability distributions in figs. 19—21. 2M = mass discharge in Hot Creek gorge springs in kg/s; additional cooler discharge of 110 kg/s simulated through southeast rim in each run except that for case B. The model results in table 17 can also be used to infer that under steady-state conditions, present-day reser- voir temperature and heat flow could be supplied by conduction from a magma chamber at 6 km only if fluid A53 circulation in the model extended to depths of 4 to 5 km. Alternatively, the effective depth to magma, or the thickness of a conducting, impermeable roof, must be only 1 to 2 km thick. Unfortunately, it is not as yet known whether fluid circulation to such depths does, in fact, occur. From structural considerations, calcula- tions of depths at which the compressive strengths of rocks are exceeded suggest that from 4 to 5 km may be a maximum .(King, 1971). However, recent stable- isotope data showing effects of interactions of volcanic rocks with circulating meteoric ground water indicate that fluid circulation to from 5 to 10 km is possible (J. R. O’Neil, oral commun., 1977, and Taylor, 1976, p. 41). Thermal modeling results show that without deep circulation, hydrothermal discharge near or in excess of present rates could not have been sustained continu- ously over the last 0.3 m.y. by conduction from an un- derlying magma chamber. In this event, an intermit- tent character for the hydrothermal system is implied. 50° I l I I U) D a 400- _ _l I.” U w I“ u] I 8 Q 300 _ E w. I 2 < 5 200 350,000 _ a. z I" p. E o E 100— - Hi In I.“ I o I I I I 102 103 104 105 106 107 TIME, IN YEARS FIGURE 32.—Transient response since initiation of springflow of average reservoir temperature below Hot Creek gorge for two reservoir depths, from model simulation (case A—uniform reservoir-permeability distribution) with dis- charge of 300 kg/s in Hot Creek gorge and southeast-rim outflow of 110 kg/s. A54 The thermal modeling results and the chemical dis- charge records are consistent with the concept of an initial period of hydrothermal activity around 0.3 m.y. ago, followed by nearly 0.25 m.y. of inactivity until the present system was established approximately 35,000 years ago. Under these conditions, hydrothermal circu- lation through the fractured Bishop Tuff at depths of 1 to 3 km in the western three-fifths of the caldera is plausible. The possibility that cellular convection within the reservoir might increase the rate of vertical heat transfer, yielding higher discharge temperatures, and also causing a more uniform reservoir temperature dis- tribution, was also examined. Simulations with uni- form permeability in a reservoir 2 km thick, which included nodal layers 2 and 3, enabled vertical flow components to be determined. The likelihood of cellu- lar convection is related to the dimensionless Rayleigh number (Ra), which for variable fluid properties can be calculated from Sorey (1975) as Antelope Dry Creek Casa Diablo Valley WEST I | l GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Ram = gkL(AT) 02mm Km [1;] I‘m (8) where the subscript m indicates evaluation at the av- erage temperature in the layer in the absence of con- vection, and AT is the temperature difference across the layer of thickness L. The effective thermal expan- sivity, Be, is computed as Be = (p0 - p1)/(p0AT) (9) where p0 and p1 are evaluated at top and bottom, re- spectively, of the permeable layer. For the conduction- only temperature distribution in figure 24, Ram = 1,500 in the west part of the caldera and 500 in the east part. For a horizontal layer of large lateral extent and tak- ing into account temperature-dependent fluid prop- erties, cellular convection should occur for Ram > 70 (Sorey, 1975, p. 33). Thus, if the Long Valley reservoir Hot Creek Gorge Owens River I | EAST 100° J; "R o 1 00 (1 700° 0 \ \ Antelope Dry Creek Casa Diablo Valley WEST I l | 200°’\ @00o\ A Hot Creek Gorge Owens River .__’——- 500° , r0 01 10 J KILOMETERS FIGURE 33.—Diagrammatic east-west cross sections of Long Valley caldera showing isotherms and convection patterns from model simulations (case A—uniform reservoir-permeability distribution) after 35,000 years (A) and 350,000 years (B), with cellular convection in reservoir at 1—3-km depth, hot-spring discharge of 250 kg/s, and southeastmim outflow of 110 kg/s. Lines of equal temperature in degrees Celsius. Interval 100°C. Arrows show inferred direction of fluid flow related to throughflow and cellular convection in reservoir. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA is uniformly permeable and as thick as 2 km, a vigor- ous cellular convection motion would be expected. The results in table 17 for the reservoir from 1 to 3 km deep (run 8), in which convection cells did occur, also include the effects of a 250-kg/s discharge in the hot springs and 110-kg/s discharge through the south- east rim. Comparisons with single-layer reservoir runs with the same throughflows indicate average reservoir temperatures under the gorge which are close to the 2-3-km reservoir results at 35,000 years and 350,000 years. Total heat discharge for run 8 was 10.6 x 107 cal/s at 35,000 years and 4.9 x 107 cal/s at 350,000 years. Isotherms and convective patterns in east-west cross sections for run 8 are shown in figure 33. At 35,000 years a two-convection-cell pattern with axes alined northwest was observed, superimposed on the east- ward- and southward-throughflow regimes. The isotherms form a mushroom configuration similar to that observed in the Wairakei hydrothermal system (Elder, 1965). Comparison with figure 26 for the single-layer reservoir indicates that the convection cells produce more areally uniform reservoir tempera- ture conditions, extending westward from Hot Creek gorge to the Casa Diablo area. At longer times, the cooler throughflow tends to sweep out the cellular m0- A55 tion west and east of the gorge, although some effect is still evident under the gorge. This analysis is limited primarily by the coarseness of the nodal spacing in the vertical direction. It would be anticipated that, if more layers of nodes were added to the simulated reservoir without changing its total thickness, higher rates of heat transfer and higher dis- charge temperatures would have been obtained. At 35,000 years this effect could make circulation in a reservoir from 1 to 2 km deep appear more plausible from a heat-flow standpoint. At longer times, however, the observed tendency for throughflow to sweep out the vertical convective motions indicates that cellular con- vection would be a dominant influence on the thermal regime only if throughflow in the hydrothermal system were more restricted than at present. The effect of the coarseness of the vertical nodal spacing also causes a lowering in vertical heat-transfer rates in the single-layer-reservoir simulations. This ef- fect would be more important at earlier times when the vertical temperature distribution is strongly non- linear, and when conductive heat fluxes from adjacent layers into the reservoir—and hence, reservoir temperatures—are underestimated. Because of the ex- pense involved in using finer vertical node spacing in the 3-dimensional simulations, the magnitude of this m 3 30° I I I U) J Lu 0 m 35,000 YEARS m _______ a: 8 a 200 - L 5 Lu I 3 '2 5 350,000 YEARS n. _._ ------- _ E 100 III-l '— E o > I I“ 3 o I I I I: 0 1O 20 30 40 NODES IN VERTICAL DIRECTION FIGURE 34.—Effect of nodal spacing in vertical direction on average reservoir temperature under Hot Creek gorge after 35,000 years and 350,000 years of fluid flow in 2-dimensional, vertical-cross-section model simulation of Long Valley hydrothermal system. Cross section aligned east-west from Hot Creek gorge to west rim; mass flux through reservoir between 1—2-km depth set at 1/10th of flux in 3-dimensional model, or 25 kg/s. A56 effect was examined for a roughly equivalent, 2-dimensional vertical-cross-section problem. Results, as indicated in figure 34, indicate that the grid spacing in the 3-dimensional model, with 5 nodal layers in the vertical direction, could cause average reservoir tem- peratures to be underestimated by as much as 35 per- cent at 35,000 years and 20 percent at 350,000 years. Obviously, this is a significant effect, which bears di- rectly on the estimates of reservoir depths required for short-term (35,000 years) and steady-state (350,000 years) hydrothermal activity. However, it may be largely offset by the likelihood that the Long Valley magma chamber is actually deeper and less extensive than assumed in this model. Reservoir temperatures obtained from the 3-dimensional model runs beneath the southeastern part of the caldera, near the location of the deep test well (pl. 1, well 66—29), range from 80—90°C at 35,000 years to around 50°C at steady state for the 1— 2-km-deep-reservoir case. Temperatures measured in the test hole were between 50°C and 70°C from below about 200 m to the bottom. Thus the concept of re- charge from the northeast rim and discharge through the southeast rim at rates near 110 kg/s produces thermal conditions consistent with those observed. Other sources for the flux of relatively cool water under the southeastern area cannot be ruled out,,such as re- charge moving down the Hilton Creek fault from the Sierra Nevada southeast of the caldera. Whatever the source, the pattern of ground-water circulation found in the model simulations with discharge through the southeast rim indicates that little of this water dis- charges in the hot springs. However, if no such outlet through the southeast rim exists, this flux of cool water would have to move westward under the intracaldera extensions of the Hilton Creek fault to mix with hot water moving from the west and discharge in the Hot Creek gorge springs. On the basis of geochemical and thermal considerations, we prefer the model analyzed in this study, but more data are needed to resolve this question. It is also possible that within the caldera the east- ernmost extensions of the Hilton Creek fault (pl. 1 and fig. 19) act as a barrier to the westward movement of cool water at depth. The hydraulic characteristics of the reservoir between the regions of hot—water circula- tion under the resurgent dome and cool-water circula- tion east of Hot Creek could be important if future geothermal development involved production of hot fluids from under the dome. If no permeability barrier exists, then movement of the cool water from the east into the production zone may significantly limit the rate or duration of high-enthalpy-fluid production. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS CONCLUSIONS AND IMPLICATIONS Results of previous investigations by the US. Geological Survey and limited information from a re- cently drilled 2.11-km-deep test hole in Long Valley caldera enable a useful conceptual model of the hy- drothermal system to be developed. Because of the lack of sufficient data from deep drill holes and because of generalizations required for numerical analysis of heat and fluid flow, the model is oversimplified at this stage. However, the results obtained from numerical simula- tions with this model permit a useful determination of the general relation between heat flow and the depth and duration of fluid flow, as well as the effective hy- draulic characteristics of the hydrothermal reservoir. The conceptual model is based largely on the follow- ing information from previous studies. Seismic- refraction experiments and geological investigations delineate three major rock units: (1) a near-surface layer of low velocity, largely nonindurated sediments and volcanic rocks, (2) an underlying, continuous layer of densely welded Bishop Tuff, and (3) precaldera granitic and metamorphic basement rocks below depths of 2.5 to 3.0 km. Regional heat-flow, seismic- refraction, and teleseismic analyses, and the recent oc- currence of extrusive volcanism suggest that magma or partially molten rock exists under the western three- fifths of the caldera at depths of 6 to 8 km. The present—day hydrothermal system is dominated by hot water with aquifer temperatures at depth, estimated from the geochemistry of hot-spring waters, of about 210°C to 280°C. If a reservoir with significant areal extent exists, it is most likely to occur in the densely welded Bishop Tuff because of the continuity and depth of the tuff and the likelihood that it retains secondary permeability after fracturing. Hydrologic and geochemical studies, in- cluding determinations of isotopic contents in caldera waters, indicate that recharge to the hydrothermal system occurs around the western rim, supplying a mass flux of 200 to 300 kg/s which discharges in the hot springs and into Lake Crowley in the southeast part of the caldera. Relatively cool temperatures measured in the recently drilled, deep test hole indicate that an ad- ditional flux of water may recharge around the north- east rim and flow southward at depth under the east- ern third of the caldera. Numerical simulations of heat and fluid flow in the model yielded values of effective reservoir permeabil- ity for several alternative cases of reservoir size. For two cases with nearly homogeneous permeability, in a l-km-thick reservoir, permeabilities of 30 to 50 mil- lidarcys were required to produce a hot-spring dis- HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA charge of 250 kg/s. A more areally restricted reservoir case required a permeability of 300 millidarcys. Com- parison of these results with laboratory results on cores from Long Valley and the Nevada Test Site and with data from hydraulic tests in wells in fractured tuff at NTS indicate that equivalent-porous-media per- meabilities determined in the model simulations rep- resent integrations of the effects of fracture permeabil- ity over the volume of the reservoir rock. Estimates of the average residence time of fluid in a 1-km-thick reservoir, based on a 10,000-year arrival time for the cool temperature wave from recharge around the west rim, are close to 2,000 years. For the same mass flux in a thinner reservoir, the estimated residence time would be proprotionately shorter. Thermal contraints on the model results are pro- vided by our estimate of the present-day total heat dis- charge from the caldera of 6.9 x 107 cal/s, and average reservoir temperatures below discharge areas of 210°C to 280°C. Numerical simulations show that these con- ditions could have been sustained for a period of 35,000 years by magma at 6 km under the western three-fifths of the caldera with fluid circulation to depths of 1.5 to 2.5 km in a reservoir covering at least the western three-fifths of the caldera. A more areally restricted reservoir would require correspondingly deeper levels of circulation. Studies of saline deposits in Searles Lake which were derived from Long Valley indicate that the present-day hot-spring discharge has persisted for only 30,000 to 40,000 years. Estimates of reservoir volume and thickness required to leach observed quan- tities of B, Cl, K, and Li are reasonable for a reservoir in the Bishop Tuff, indicating that a magmatic-fluid component providing chemical elements and heat to the hydrothermal system would not be required. In contrast, evidence exists for extensive hydrother- mal alteration about 0.3 m.y. ago which appears to be related to the main magma chamber, rather than to the postcaldera eruptive rocks. Numerical simulations representing a period of 350,000 years, at which time steady-state conditions would be reached, indicate that to obtain present-day heat flow and reservoir tempera- tures above 200°C after this period, fluid circulation to depths of 4 to 5 km would be required. If downward extension of permeable channels to these depths in the basement rocks has not occurred, then the modeling results and the diverse indications of the age of hy- drothermal activity suggest an alternative hypothesis that discharge from the hydrothermal system has been intermittent. It would, in fact, be reasonable to expect that significant periods of inactivity after 0.3 m.y. ago resulted from climatic variations and self-sealing proc- esses which are in evidence today. These possibilities, A57 and the adequacy of the simplified hydrothermal model analyzed in this study, can be evaluated only by deep- test drilling in the western part of the caldera. Under the eastern two-fifths of the caldera, reservoir temperatures measured in the 2.11-km test hole and the model simulations indicate that temperatures are sufficiently low to preclude the possibility of electrical-energy development east of Hot Creek. Tem- perature conditions close to those measured in the test hole were modeled by including a recharge flux of 110 kg/s around the northeast rim and discharge at depth through the southeast rim. To determine if these flow conditions actually exist would also require further deep drilling. Although there presently exists little economic incentive to do more drilling in this area, selection of the east side of the caldera for possible reinjection of residual geothermal fluids could change this situation. Under normal conditions, the model re- sults and geochemical analyses indicate that the cool reservoir water east of Hot Creek does not mix significantly with hot water discharging in Hot Creek gorge. However, should large-scale production of hot fluids occur in the future, the resultant westward mi- gration of this cool water toward the production area would have to be considered. Of importance in this re- gard is the existence of any permeability barrier be- tween the hot- and cold-water areas within the reser- voir, which could have developed from chemical deposi- tion along former fault-plane fluid conduits. REFERENCES CITED Anderson, L. A., and Johnson, G. R., 1973, The application of the self-potential method in the exploration for geothermal energy in Long Valley, California [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1212. 1976, Application of the self-potential method to geothermal exploration in Long Valley, California: Jour. Geophys. Research, v. 81, p. 1527—1532. Axelrod, D. I., 1962, A Pliocene Sequoiadendron forest from western Nevada: California Univ., Pub. Geol. Sci, v. 39, no. 3, p. 195— 268. Axelrod, D. I., and Ting, W. S., 1961, Early Pleistocene floras from the Chagoopa surface, southern Sierra Nevada: California Univ., Pub. Geol. Sci, v. 39, no. 2, p. 119—194. Bailey, R. A., 1974, Preliminary geologic map and cross sections of the Casa Diablo geothermal area, Long Valley caldera, Mono County, California: U.S. Geol. Survey open-file map, 2 p. Bailey, R. A., Dalrymple, G. B., and Lanphere, M. A., 1976, Vol- canism, structure, and geochronology of Long Valley caldera, Mono County, California: Jour. Geophys. Research, v. 81, no. 5, p. 725—744. Bailey, R. A., and Koeppen, R. P., 1977, Preliminary geologic map of Long Valley caldera, Mono County, California: U.S. Geol. Sur- vey, open-file map, 2 p. Bateman, P. C., 1965, Geology and tungsten mineralization of the Bishop district, California: U.S. Geol. Survey Prof. Paper 470, 208 p. A58 Bateman, P. C., and Wahrhaflzig, C., 1966, Geology of the Sierra Nevada in Geology of northern California: California Div. Mines and Geol. Bull. 190, p. 107—172. Blake, A. H., and Matthes, F. E., 1938, The new Casa Diablo “geyser”: Sierra Club Bull., v. 23, no. 2, p. 82—83. Blaney, H. F., 1933, Rainfall penetration, Chap. 6, in Ventura County Investigation: California Dept. Public Works, Div. Water Resources Bull. 46, p. 82—90. California Dept. Water Resources, 1967, Investigation of geothermal waters in the Long Valley area, Mono County; California Dept. Water Resources, Sacramento, California 114 p. California Div. Oil and Gas, 1972, Geothermal map G—51, Casa Di- ablo, 1 in.: 2,000 ft: Sacramento, Calif. Chelikowsky, J. R., 1940, Tectonics of the rhyolite in the Mammoth embayment, California: Jour. Geology, v. 48, p. 421-435. Christensen, M. N., 1966, Late Cenozoic crustal movements in the Sierra Nevada of California: Geol. Soc. America Bull., v. 77, p. 163—182. Crowder, D. F., and Sheridan, M. F., 1972, Geologic map of the White Mountain Peak quadrangle, Mono County, California: US Geol. Survey, Geol. Quad. Map GQ—1012. Dalrymple, G. B., 1963, Potassium-argon dates of some Cenozoic vol- canic rocks of the Sierra Nevada, California: Geol. Soc. America Bull., v. 74, p. 379—390. 1964, Cenozoic chronology of the Sierra Nevada, California: California Univ., Publ. in Geol. Sci., v. 47, 41 p. Dalrymple, G. B., Cox, Allen, and Doell, R. R., 1965, Potassium- argon age and paleomagnetism of the Bishop Tuff, California; Geol. Soc. America Bull., v. 76, no. 6, p. 665—674. Dorsey, N. E., 1968, Properties of ordinary water-substance: New York, Hafner Publishing Co. 673 p. Elder, J. W., 1965, Physical processes in geothermal areas, in Lee, W. H. K., ed., Terrestrial Heat Flow: Am. Geophys. Union Monograph Series, v. 8, p. 211—239. Ellis, A. J ., 1970, Quantitative interpretation of chemical charac- teristics of hydrothermal systems: U.N. Symposium on Devel- opment and Utilization of Geothermal Resources, Pisa, Italy, p. 516—528. Ellis, A. J., and Mahon, W. A. J., 1964, Natural hydrothermal sys- tems and experimental hot water/rock interactions, Geochim. Cosmochim. Acta, v. 28, no. 8, p. 1323—1357. Fournier, R. 0., and Rowe, J. J., 1966, Estimation of underground temperatures from the silica content of water from hot springs and wet steam wells: Am. Jour. Sci., v. 264, p. 685—697. Fournier, R. 0., Sorey, M. L., Mariner, R. H., and Truesdell, A. H., 1976, Geochemical prediction of aquifer temperatures in the geothermal system at Long Valley, California: US. Geol. Survey open-file report 76—469, 34 p. Fournier, R. 0., and Truesdell, A. H., 1973, An empirical Na-K-Ca geothermometer for natural waters: Geochim. et Cosomochim. Acta, v. 37, p. 1255—1275. Friedman, Irving, and Smith, G. I., 1970, Deuterium content of snow cores from Sierra Nevada area: Science, v. 169, p. 467—470. 1972, Deuterium content of snow as an index to winter climate in the Sierra Nevada area: Science, v. 176, p. 790—793. Gilbert, C. M., 1938, Welded tuff in eastern California: Geol. Soc. America Bull., v. 49, no. 12, pt. 1, p. 1829—1862. 1941, Late Tertiary geology southeast of Mono Lake, Califor- nia: Geol. Soc. America Bull., v. 52, p. 781—816. Hill, D. P., 1976, Structure of Long Valley caldera, California, from a seismic refraction experiment: Jour. Geophys. Research, v. 81, no. 5, p. 745—753. Hill, D. P., McHugh, S. L., and Pakiser, L. C., 1973, Structure of the Long Valley caldera from detailed seismic-refraction mea- surements [abs]: Amer. Geophys. Union EOS Trans, v. 54, no. GEOHYDROLOGY OF GEO’I‘HERMAL SYSTEMS 11, p. 1211. Hoover, D. B., Frischknecht, F. C., and Tippens, C. L., 1973, Au- diomagnetotelluric sounding as a reconnaissance exploration technique in Long Valley, California [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1212. 1976, Audiomagnetotelluric sounding as a reconnaissance ex- ploration technique in Long Valley, California: Jour. Geophys. Research, v. 81, no. 5, p. 801—809. Houk, I. E., 1951, Irrigation engineering, v. 1—Agricultural and hyd- rological phases: New York, John Wiley and Sons, 545 p. Huber, N. K., and Rinehart, C. D., 1965, Geologic map of the Devil’s Postpile quadrangle, Sierra Nevada, California: US. Geol. Sur- vey Geol. Quad Map GQ—437, scale 1:62,500. Iyer, H. M., and Hitchcock, T., 1976, Seismic noise survey in Long Valley, California: Jour. Geophys. Research, v. 81, no. 5, p. 821— 840. Jackson, D. B., Stanley, W. D., and Zohdy, A. A. R., 1973, Direct current and electromagnetic soundings in Long Valley, Califor- nia [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1212. Jacob, C. E., 1950, Flow of ground-water, in Rouse, Hunter, ed., En- gineering Hydraulics: New York, John Wiley and Sons, p. 321— 386. Kane, M. F., and Mabey, D. R., 1973, Gravity and magnetic anomalies in Long Valley, California [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1211. Kane, M. F., Mabey, D. R., and Brace, R. L., 1976a, A gravity and magnetic investigation of the Long Valley caldera, Mono County, California [abs]: Am. Geophys. Union EOS Trans, v. 57, no. 2, p. 100—101. 1976b, A gravity and magnetic investigation of the Long Val- ley caldera, Mono County, California: Jour. Geophys. Research, v. 81, no. 5, p. 754—762. Kaysing, W. D., ed., 1970, Thermal springs of the western United States: Santa Barbara, Calif, Paradise Publishers, 36 p. Keller, G. V., 1960, Physical properties of tuffs of the Oak Spring Formation, Nevada in Geological Survey research 1960: US. Geol. Survey Prof. Paper 400—B, p. 396—400. King, G. C. P., 1971, The siting of strainmeters for teleseismic and tidal studies: Royal Soc. New Zealand Bull. 9, p. 239—247. Lachenbruch, A. H., 1968, Preliminary geothermal model of the Sierra Nevada: Jour. Geophys. Research, v. 73, p. 6977-6989. 1970, Crustal temperature and heat production: Implications of the linear heat—flow relation: J our. Geophys. Research, v. 75, p. 3291—3300. Lachenbruch, A. H., Lewis, R. E., and Sass, J. J., 1973, Prospecting for heat in Long Valley [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1211. Lachenbruch, A. H., Sass, J. H., Munroe, R. J., and Moses, T. H., Jr., 1976, Geothermal setting and simple heat-conduction models for the Long Valley caldera: Jour. Geophys. Research, v. 81, no. 5, p. 769—7 84. Lachenbruch, A. H., Sorey, M. L., Lewis, R. E., and Sass, J. H., 1976, The near-surface hydrothermal regime of Long Valley caldera: Jour. Geophys. Research, v. 81, no. 5, p. 763—768. Lee, C. H., 1912, An intensive study of the water resources of a part of Owens Valley, California: US. Geol. Survey Water Supply Paper 294, 135 p. Lee, W. T., 1906, Geology and water resources of Owens Valley, California: US. Geol. Survey Water Supply Paper 181, 28 p. Lewis, R. E., 1974, Data on wells, springs and thermal springs in Long Valley, Mono County, California: US. Geol. Survey open- file report, 52 p. 1975, Data from a 1,000-foot (305-meter) core hole in the Long Valley caldera, Mono County, California: US. Geol. Survey open-file report, 16 p. HYDROTHERMAL SYSTEM OF LONG VALLEY CALDERA, CALIFORNIA Lindgren, W., 1911, The Tertiary gravels of the Sierra Nevada of California: U.S. Geol. Survey Prof. Paper 73, 226 p. McDonald, C. C., and Hughes, G. H., 1968, Studies of consumptive use of water by phreatophytes and hydrophytes near Yuma, Arizona: U.S. Geol. Survey Prof. Paper 486—F, 24 p. McKenzie, W. F., and Truesdell, A. H., 1977, Geothermal reservoir temperatures estimated from the oxygen-isotope compositions of dissolved sulfate in water from hot springs and shallow drill holes, in Proc. Int. Atomic Energy Agency Advisory Group on the Application of Nuclear Techniques to Geothermal Studies, Pisa, Italy: Geothermics, Special Issue (in press). McNitt, J. R., 1963, Exploration and development of geothermal power in California: California Div. Mines and Geol. Spec. Rept. 75, 45 p. Mariner, R. H., and Willey, L. M., 1976, Geochemistry of thermal waters in Long Valley, Mono County, California: Jour. Geophys. Research, v. 81, no. 5, p. 792—800. Matthes, F. E., 1930, Geologic history of the Yosemite Valley: U.S. Geol. Survey Prof. Paper 160, 137 p. 1933, Geography and geology of the Sierra Nevada, in XVI Int. Geol. Cong. Guidebook: Washington, DC, US. Govt. Printing Office, v. 16, part c—l, p. 26—40. 1939, History of faulting movements at the east front of the Sierra Nevada, as indicated by dislocated moraines [abs]: Geol. Soc. America Bull., v. 50, p. 1955. 1947, A geologist’s view, in Peattie, Roderick, ed., The Sierra Nevada: New York, Vanguard Press, p. 166—214. Mayo, E. B., 1934, Geology and mineral resources of Laurel and'Con- vict- basins, southwestern Mono County, California: California Jour. Mines and Geol., v. 30, no. 1, p. 79—88. Mercer, J. W., Pinder, G. F., and Donaldson, I. G., 1975, A Galerkin finite-element analysis of the hydrothermal system at Wairakei, New Zealand: Jour. Geophys. Reserach, v. 80, no. 17, p. 2608— 2621. Mufl'ler, L. J. P., and Williams, D. L., 1976, Geothermal investiga- tions of the U.S. Geological Survey in Long Valley, California, 1972—1973: Jour. Geophys. Research, v. 81, no. 5, p. 721—724. Narasimhan, T. N., and Witherspoon, P. A., 1976, An integrated finite difference method for analyzing fluid flowing porous media: Water Resources Research, v. 12, no. 1, p. 57—64. Nathenson, M., nad Muffler, L. J. P. , 1975, Geothermal resources in hydrothermal convection systems and conduction-dominated areas; in White, D. E., and Williams, D. E., eds., Assessment of geothermal resources of the United States, 1975: U.S. Geol. Sur- vey Cir. 726, p. 104—121. Noble, D. C., Korringa, M.K., Hedge, C.E., and Riddle, G. 0., 1972, Highly differentiated subalkaline rhyolite' from Glass Mountain, Mono County, California: Geol. Soc. America Bull., v. 83, p. 1179—1184. Pakiser, L. C., 1961, Gravity, volcanism, and crustal deformation in Long Valley, California, in Geological Survey research 1961: U.S. Geol. Survey Prof. Paper 424—B, p. 250—253. 1968, Seismic evidence for the thickness of Cenozoic deposits in Mono Basin, California: Geol. Soc. America Bull., v. 79, p. 1833—1838. Pakiser, L. C., Kane, M. F., and Jackson, W. H., 1964, Structural geology and volcanism of Owens Valley region, California, a geophysical study: U.S. Geol. Survey Prof. Paper 438, 68 p. Pakiser, L. 0., Press, F., and Kane, M. F., 1960, Geophysical investi- gation of Mono Basin, California: Geol. Soc. America Bull., v. 71, p. 415—448. Piper, A. M., Gale, H. S., Thomas, H. E., and Robinson, T. W., 1939, Geology and ground-water hydrology of the Mokelumne area, California: U.S. Geol. Survey Water-Supply Paper 780, 230 p. Putman, W. C., 1940, Quaternary geology of the June Lake district, A59 California [abs]: Geol. Soc. America Bull., v. 51, no. 12, pt. 2, p. 1939—1940. 1949, Quaternary geology of the June Lake district, Califor- nia: Geol. Soc. America Bull., v. 60, p. 1281—1302. 1960, Origin of Rock Creek and Owens River gorges, Mono County California: California Univ., Pub. Geol. Sci., v. 34, p. 221—280. 1962, Late Cenozoic geology of McGee Mountain, Mono County, Californiaz‘California Univ., Pub. Geol. Soc., v. 40, no. 3, p. 181—218. Rantz, S. E., 1972, Mean annual precipitation in the California re- gion: U.S. Geol. Survey Basic Data Compilation (2 maps at 1:1,000,000), Menlo Park, Calif. Ratté, J. C., and Steven, T. A., 1967, Ash flows and related volcanic rocks associated with the Creede caldera, San Juan Mountains, Colorado: U.S. Geol. Survey Prof. Paper 524—H, 58 p. Rinehart, C. D., and Ross, D. C., 1957, Geology of the Casa Diablo Mountain quadrangle, California: U.S. Geol. Survey Geol. Quad. Map GQ—99. 1964, Geology and mineral deposits of the Mount Morrison quadrangle, Sierra Nevada, California, with a section on A grav- ity study of Long Valley by L. C. Pakiser: U.S. Geol. Survey Prof. Paper 385, 106 p. Roy, R. F., Blackwell, D. D., and Birch, Francis, 1968, Heat genera- tion of plutonic rocks and continental heat-flow provinces: Earth and Planetary Sci. Letters, v. 5, p. 1'—12. Russell, I. C., 1889, Quaternary history of Mono Valley, California: U.S. Geol. Survey 8th Ann. Rept., p. 261—394. Sass, J. H., Lachenbruch, A. H., and Munroe, R. J., 1974, Thermal data from heat-flow test wells near Long Valley, California: U.S. Geol. Survey open-file report, 43 p. Sheridan, M. F., 1965, The mineralogy and petrology of the Bishop Tuff: Ph.D. thesis, Stanford Univ., Stanford, Calif, 165 p. 1968, Double cooling-unit nature of the Bishop Tuff in Owens Gorge, California [abs]: Geol. Soc. America Spec. Paper 115, p. 351. 1971, Guidebook to the Quaternary geology of the east-central Sierra Nevada: Guidebook, XVI Field Conf., Rocky Mountain Section, Friends of the Pleistocene, Oct. 9—10, 1971. Smith, G. I., 1976, Origin of lithium and other components in the Searles Lake evaporites, California, in Vine, J. D., ed., Lithium resources and requirements by the year 2000: U.S. Geol. Survey Prof. Paper 1005, p. 92—103. Sorey, M. L., 1975a, Numerical molding of liquid geothermal sys- tems: U.S. Geol. Survey open-file report 75—613, 66 p. M. L., 1975b, Potential effects of geothermal development on springs at the Hot Creek Fish Hatchery in Long Valley, Mono County, California: U.S. Geol. Survey open-file report 75—637, 10 p. Sorey, M. L., and Lewis, R. E., 1976, Convective heat flow from hot springs in the Long Valley caldera, Mono County, California: Jour. Geophys. Research, v. 81, no. 5, p. 785—791. Spreen, W. C., 1947, A determination of the effect of topography upon precipitation: Am. Geophys. Union Trans., v. 28, no. 2, p. 285— 290. Stanley, W. D., Jackson, D. B., and Zohdy, A. A. R., 1973, A total- field resistivity map of Long Valley, California [abs]: Am. Geophys. Union EOS Trans, v. 54, no. 11, p. 1212. 1976, Deep electrical investigations in the Long Valley geo- thermal area, California: Jour. Geophys. Research, v. 81, no. 5, p. 810—820. . . Stearns, N. D., Stearns, H. T., and Warning, G. A., 1937, Thermal springs in the United States: U.S. Geol. Survey Water-Supply Paper 679—B, p. 59—206. Steeples, D. W., and Iyer, H. M., 1976, Low velocity zone under Long A60 Valley as determined from teleseismic events: Jour. Geophys. Research, v. 81, no. 5, p. 849—860. Steeples, D. W., and Pitt, A. M., 1973, Microearthquakes in and near Long Valley, California [abs]: Am. Geophys. Union EOS Trans., v. 54, no. 11, p. 1213. 1976, Microearthquakes in and near Long Valley, California: Jour. Geophys. Research, v. 81, no. 5, p. 841—847. Steiner, A., 1955, Hydrothermal rock alteration, in Geothermal steam for power: New Zealand Dept. Scientific and Industrial Research, Bulletin 117, p. 21—26. Taylor, H. P., 1976, Water/rock interactions and the origin of the H20 in granitic batholiths, Thirtieth William Smith Lecture: California Inst. Technology, Div. Geol. and Planetary Sci. Contr. no. 2790, 105 p. Waring, G. A., 1915, Springs of California: US. Geol. Survey Water-Supply Paper 338, 310 p. 1965, Thermal springs of the United States and other coun- tries of the world, a summary: U.S. Geol. Survey Prof. Paper 492, 383 p. White, D. E., 1965, Geothermal energy: U.S. Geol. Survey Circ. 519, 17 p. 1968, Hydrology, activity, and heat flow of the Steamboat Springs thermal system, Washoe, County, Nevada: US. Geol. GPO 789-107 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Survey Prof. Paper, 458-0, 104 p. White, D. E., Muffler, L. J. P., and Truesdell, A. H., 1971, Vapor- dominated hydrothermal systems compared with hot-water sys- tems: Econ. Geology, v. 66, no. 1, p. 75—97. White, W. N., 1932, A method of estimation of ground-water supplies based on discharge by plants and evaporation from soil, results of investigation in Escalante Valley, Utah: US. Geol. Survey Water-Supply Paper 659—A, 105 p. Whiting, H. A., 1888, Mono County: California Mining Bur., 8th Ann. Rept, p. 352—401. Willey, L. M., O’Neil, J. R., and Rapp, J. B., 1974, Chemistry of thermal waters in Long Valley, Mono County, California: US. Geol. Survey open-file report, 19 p. _ Williams, D. L., 1976, Implications of a magnetic model of the Long Valley caldera, California: US. Geol. Survey open-file report 76—439, 7 p. Winograd, I. J., Thordarson, William, and Young, R. A., 1971, Hy- drology of the Nevada Test Site and vicinity: U.S. Geol. Survey open-file report, 429 p. Wood, S. H., 1975, Mono and Inyo Crater eruptions, eastern California—Radiocarbon dating and trace element correlations of late Pleistocene tephra [abs]: Geol. Soc. America Abs. with Programs, Cordilleran Section, v. 7, no. 3, p. 389. UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 1044A PLATE 1 R.27E. GEOLOGICAL SURVEY 1 19°00' DESCRIPTION OF MAP UNITS { . RHYOLITE OF INYO DOMES — Rhyolitic to rhyodacitic lava domes in north- Holocene er western part of caldera. ALLUVIUM ~ Heterogeneous deposits ranging from boulder to clay along peren- Pleistocene { Cal and Holocene nial or intermittent streams or beneath alluvial fans. Includes older deposits now being dissected and near-shore deposits of former Long Valley Lake. Qt GLACIAL TILL — Unsorted morainal debris from Sierra Nevada deposited during Sherwin, Casa Diablo, Mono Basin, Tahoe, Tenaya, and Tioga glaciations. LAKE DEPOSITS — Pebble conglomerate, tuffaceous sandstone, grading eastward into marsh deposits and fine-grained diatomaceous deposits in topographically lower parts of caldera. OI BASALT OF WESTERN MOAT — Flows and cinder cones of basalt, trachybasalt, and trachyandesite in western moat of caldera and farther west, beyond west- ern rim of caldera. Ranges downward in age from 0.9 m.y. near Devil’s Postpile to about 0.06 m.y. RHYODACITE OF CALDERA RIM ~ Hornblende-biotite rhyodacite and rhyolite domes and short, thick flows at Mammoth Mountain, Deadman Creek, and foot of Glass Mountain. Ranges downward in age from 0.18 to 0.05 m.y. RHYOLITE OF MOAT — Post-resurgent coarsely porphyritic hornblende-biotite rhyolite erupted from three groups of vents in caldera moat. Age ranges down- ward from about 0.5 m.y. on northwest flank of resurgent dome to about 0.3 m.y. on southeast flank to about 0.1 m.y. on west flank. Pleistocene EARLY RHYOLITE 7 Aphryic rhyolite, pyroxene rhyolite, and biotite rhyolite erupted after caldera subsidence and exposed in uplifted and tilted blocks of resurgent dome in central part of caldera. Age ranges downward from about Ort 0.7 to about 0.6 m.y. Qer, flows and lava domes; Qrt, tuffs. Oer we BISHOP TUFF — Extensive sheet composed to two cooling units of rhyolite ash- flow tuff erupted from vents now buried within caldera. Consists of crystal—rich nonwelded to welded tuff containing as much as 30 percent phenocrysts of quartz, sanidine, plagioclase, biotite, and iron and titanium oxides. Age about 0.7 m.y. th :1, {E RHYOLITE OF GLASS MOUNTAIN 7 Precaldera hypersilicic rhyolite ranging downward in age from 1.9 to 0.9 m.y. Qgr, chiefly flows; few domes and small intrusive bodies. Qgt, chiefly tuffs, Pelean avalanche deposits, and epiclastic Qgt conglomeratic deposits. Ogr VOLCANIC ROCKS, UNDIFFERENTIATED — Basalt and andesite flows along west and north walls of caldera; less extensive flows on east and south rims; rhyodacite overlying basalt northwest of caldera. Age ranges downward from 3.2 to 2.6 m.y. ;Tv GRANITIC ROCKS — Sierra Nevada batholitic rocks ranging from small masses of gabbro and diorite to large plutons of granodiorite and quartz monzonite. liypabyssal intrusive rocks; minor intercalated sedimentary rocks; all mcta— morphosed to varying degrees. Composition ranges from basalt to dacite. METASEDIMENTARY ROCKS — Chiefly quartz-rich hornfels; subordinate sand» stone, shale, quartzite, metachert, marble, phyllite, and schist. 37°45’ QUATERNARY METAVOLCANIC ROCKS — Include a variety of pyroclastic rocks, flows, and J TERTIARY JURASSIC AND CRETACEOUS TRIASSIC AND JURASSIC ORDOVICIAN TO PERMIAN EX P LA NATION Contact __T___-_.... Fault Dashed were approximately located, dotted where concealed. Bar and ball on downthrown side. Caldera margin Approximate boundary of caldera floor. Concealed caldera ring fracture (not shown) is inward from this line, as shown in figures 3 and 4, Base from US. Geological Survey Mono Craters, 1953, Cowtrack Mtn., 1962, Glass Mtn., 1962, Devil’s Postpile, 1953, Mt. Morrison, 1953, Casa Diablo Mtn. 1953, California 119°00' pm 6370 © Test well 6 Flowing artesian well 0 Stock well A Gaging station v Chemical measurement site in stream <8) Weather station SCALE 1:62 500 1 V2 0 1 2 3 4 5 MILES l—l r—i l—l l—l l—l l——t i—i . g 1 5 0 1 2 3 4 5 6 7 Kl LOMETRES T2 8. . 37°45' ox’~/ l <\_ _...L t”.\ \ 2 5&2 ’ 7,4:_w..._,v,. V ‘1 13 6 ’/6?60-\ / x T.3S. , I?) ,L’K/zs 4 mammoth :K/ , x ( ,v i // . Volcanic vent 1% \ I,“ 9008 631:) Volcanic crater ._. ' T.4S 04 Spring 8/31/1JJ‘1'A Y 578/ - Oar? % Ar 21 . HHHHl—ll—t I—I } it? ’ w T.5S. ,/ //'."‘ \ t tar/(rm 37°30’ 37°30’ a Interior-Geological Survey, Reston, Va.---l977--—W77278 Geology from Bailey and Koeppen, 1977 118°45' GENERALIZED GEOLOGIC MAP OF LONG VALLEY AREA SHOWING LOCATIONS OF WELLS, SPRINGS, AND HYDROLOGIC OR METEOROLOGIC DATA STATIONS This map was printed from negatives prepared by photographic color-separation of author-prepared- colored original. Use of Temperature Surveys at a Depth of 1 Meter In Geothermal Exploration in Nevada By F. H. OLMSTED GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—3 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1977 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Dirertor Library of Congress Cataloging in Publication Data Olmsted, Franklin Howard, 1921— Use of temperature surveys at a depth of 1 meter in geothermal exploration in Nevada. (Geohydrology of geothermal systems) (Geological Survey Professional Paper lO44-B) Bibliography: . 13-25 Supt. of Docs. £10.: 19.16.1044-B 1. Geothermal engineering—Nevada. 2. Geothermal resources-Nevada. 3. Earth temperature—Nevada. I. Title. II. Series. III. Series: Unites States. Geological Survey. Professional Paper 1044-B. TJ280.7.045 551.1’4 77-608303 For sale by the Superintendent of Documents, US. Government Priming Office Washington, DC. 20402 Stock Number 024-001-03028-4 FIGURE \IOSUI 0:000 11. 12. 13. 14. 15. 16. 17. CONTENTS Abstract B1 Introduction 1 Theoretical considerations 1 Equipment and methodology 5 Field application in Soda Lakes area 5 Physical setting 5 8 8 9 Temperature survey in vicinity of steam well Temperature survey in November 1973 Relations of 1-m temperatures to snowmelt patterns Relations of 1-m temperatures to growth of Russian thistle ............................................... 12 Temperature surveys in June and September 1974 12 Annual fluctuation in temperature at three sites 12 Correlation of temperatures at depths of 1 m and 20 m ....................................................... 12 Field application in Upsal Hogback area 16 Geology and topography 16 Depth to water and ground-water movement 16 Temperature survey on June 15, 1975 16 Temperature survey on October 29, 1975 16 Temperature survey on December 2, 1975 16 Temperature at a depth of 30 in, September 1975 21 Correlation of temperatures at depths of 1 m and 30 m ....................................................... 21 Conclusions 21 References cited 25 ILLUSTRATIONS Page . Difference in average annual temperature between land surface and a depth of 1 m in relation to thermal conductivity and geothermal heat flow B2 . Idealized annual temperature fluctuations at land surface and a depth of 1 m for approximate conditions in western Carson Desert area, Nevada 4 . Topographic map of Soda Lakes-Upsal Hogback area showing temperatures at a depth of 30 min shallow test wells, configuration of water table, and direction of shallow ground-water movement in July 1975 .................................. 6 . Map of Soda Lakes area showing character of materials to a depth of 1 m and the depth to the water table in shallow test wells in December 1975 7 . Map showing temperature at a depth of l m near steam well, November 23, 1973 9 . Map of Soda Lakes area showing temperature at a depth of 30 in, fall 1973 10 . Map of Soda Lakes area showing temperature at a depth of 1 m on November 23, 1973 and area of snowmelt in January 1974 11 . Map of Soda Lakes area showing density of Russian thistle on February 5, 1974 13 . Map of Soda Lakes area showing temperature at a depth of 1 m on June 15 and September 20, 1974 .......................... 14 . Fluctuations in temperature at a depth of 1 m at three sites in Soda Lakes area compared with fluctuations in air temperature at Fallon Experiment Station during 1975 15 Correlation of temperatures at a depth of 20 m with temperatures at a depth of 1 m in Soda Lakes area, September 1974 17 Topographic map of Upsal Hogback area showing depth to the water table in test wells, configuration of the water table, and direction of shallow ground-water movement in July 1975 18 Map of Upsal Hogback area showing temperature at a depth of 1 In, June 1975 19 Map of Upsal Hogback area showing temperature at a depth of 1 In, October 29, 1975 ...................................................... 20 Map of Upsal Hogback area showing temperature at a depth of 1 m, December 2, 1975 ........................... 22 Map of Upsal Hogback area showing temperature in test wells at a depth of 30 In, September 1975 ............................ 23 Correlation of temperature at a depth of 30 m with temperature at a depth of 1 m in Upsal Hogback area, December 1975 24 III IV CONTENTS TABLE Page TABLE 1. Characteristics of idealized temperature waves at a depth of 1 m in western Carson Desert ............................................ BB GEOHYDROLOGY OF GEOTHERMAL SYSTEMS USE OF TEMPERATURE SURVEYS AT A DEPTH OF 1 METER IN GEOTHERMAL EXPLORATION IN NEVADA By F. H. OLMSTED ABSTRACT Temperature surveys at a depth of 1 m in the Soda Lakes and Upsal Hogback areas of the western Carson Desert, Nevada, had varying degrees of success in delineating temperature and heat- flow anomalies at greater depths. Optimum results were obtained in a detailed survey of a small area near a fumarole and an aban- doned steam well in the Soda Lakes area. At this location steam occurs above the water table and the unconfined water is at the boiling temperature at depths of only a few meters; also tempera- tures and heat flows are far above background values. Somewhat more equivocal results were obtained in surveys of a larger sur- rounding area in which the temperature range at 1 m was smaller than immediately adjacent to the fumarole and steam well, and in which perturbing effects of geologic, topographic, hydrologic, and other factors not related directly to geothermal heat flow are locally significant. Poorest results were obtained in surveys of the Upsal Hogback area, where temperature of the thermal water leaking into shallow aquifers is much less than that in the Soda Lakes area, the range in temperature at a depth of 1 m is corres- pondingly smaller, and the variability of the geology, topography, depth to the water table, and other perturbing factors is greater than that in the Soda Lakes area. INTRODUCTION In principle, temperature surveys at depths as shallow as 1 or 2 m could be used to delineate deeper thermal anomalies or areas of high geothermal heat flow, provided that temperature differences caused by extraneous factors were not so large as to mask the temperature differences caused by variations in heat flow. Some extraneous factors that affect shallow temperatures are local variations in the character and moisture content of the near-surface materials, topography, depth to the water table, vegetation, land use, and microclimate. The princi- pal advantage of temperature measurements at such shallow depths in comparison with deeper meas- urements is their relatively low cost in both time and money. Earlier investigators have reported both failure and success using shallow temperature surveys in geothermal exploration. In reviewing evidence from a study of shallow temperature gradients in relation to anomalously high heat flow caused by oxidizing ore bodies, Lovering and Goode (1963, p. 39) concluded that temperature measurements at depths of only 1 or 2 m were useless in detecting abnormal geothermal gradients at greater depths. Kappel- meyer (1957, p. 248—254) reached similar conclusions about the use of such measurements to find buried horsts, salt domes, or local concentrations of radio- active material, although he believed it possible in principle to outline areas underlain by fissures carrying thermal water upward into the shallow ground-water system. Kintzinger (1956), in fact, reported the successful use of l-m temperature measurements in outlining an area near Lordsburg, N.M., underlain by hot ground water encountered in wells at depths of only 20—25 m. The present paper (1) briefly reviews the theoretical difficulties and limitations of temperature measure- ments at 1 m depth in determining geothermal heat flow; (2) presents the results of field tests of the method at two geothermal anomalies in west-central Nevada, where fairly abundant comparative tem- perature data are available at depths below the zone affected by fluctuations in solar energy; and (3) evaluates the general applicability of the method in similar geologic, hydrologic, and climatic settings. THEORETICAL CONSIDERATIONS If it were possible to make sufficient measure- ments to define average annual temperature at the land surface and at a depth of 1 m, the temperature at 1 m would be found to exceed that at the land surface, assuming the climate was not changing and heat transfer toward the surface was by conduction. The temperature difference in a horizontal uniform layer would be expressed as _ i M“ ' 10K Bl B2 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS 0 10 20 30 40 50 60 12 12 (I) 2 (D S 10 10 o m I-U ‘9 a f o LLI a z_ 8 8 E ._ O p— O E 3 LL 6 6 LIJ a: 3 '2 0 A- E \L’ ‘5‘ Lu 4 4 [— E 3 «‘5 E / m g .0 ”L: / \k '2 LL 2 2 o / / / 0 0 0 10 20 30 4O 50 60 HEAT FLOW, IN HEAT-FLOW UNITS K=Thermal conductivity, in millicalories per centimeter - degree - second FIGURE 1.—Difference in average annual temperature between land surface and a depth of 1 m in relation to thermal conductivity and geothermal heat flow. where AT0_1 is the temperature difference from land surface to a depth of 1 m, in degrees Celsius; Q is the geothermal heat flow (heat-flux density), in micro- calories per square centimeter per second (u cal-cm'2 oS'l); and K is the thermal conductivity of the layer, in millicalories per centimeter per degree per second (mcal-cm1-°C'1-S'1). In an area of so-called “normal” geothermal heat flow—about 1 pcalocm '2-8'1 or 1 HFU (heat-flow unit)—the temperature difference in material having a thermal conductivity of 2 mcal- cm'1o°C'1-S'1 would be only 0.05°C, which is barely within the range detectable by ordinary temperature- measuring equipment. Temperature differences from land surface to a depth of 1 m for values of heat flow and thermal conductivity representative of conditions in geo- thermal areas in west-central Nevada are shown in figure 1. A K (thermal conductivity) of 0.5 mcal-cm‘1 o°C'1-S‘1 would be typical of dry silt or fine sand; a K of 4 mcal-cm'1o°C‘1oS'1 would be typical of wet coarse sand. Where the thermal conductivity is large, the temperature difference from 0 to 1 m is small, even for a heat flow of 100 times “normal.” Low thermal conductivity, on the other hand, is associated with a large temperature difference (fig. 1). For the simple case just described, if sufficient measurements could be made to define the average annual temperature at 1 m, the only problems in relating temperature at 1 m to geothermal heat flow would be variations in thermal conductivity with TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA B3 TABLE 1.—Characteristics of idealized temperature waves at a depth of 1 m in western Carson Desert Period Amplitude at Amplitude at 1 m depth (°C) Timelag (days) Kind Of wave (days) surface (0C) or = 1.5 x 104011125" Gt = 6.0 x 10'Scmzs“ a = 1.5 x 10‘3cm25‘1 04 = 6.0 x 10' 3cm23" Diurnal .......................................... 1 15.0 2.6 x 10'6 6.2 x 10'3 2.46 1.23 “Weather" .................................... 10 4.0 0.029 0.34 7.8 3.9 Seasonal ........................................ 365 15.0 6.65 9.98 47.4 23.7 both place and time and variations with place in average annual temperature at the land surface. Variations in thermal conductivity are troublesome where the near-surface materials in the area of investigation are not uniform. Variations in average annual temperature are difficult to deal with where topography is irregular and vegetative cover is variable. As shown in figure 1, heat-flow differences of several tens of heat flow units could be masked completely by moderate variations in thermal conductivity or by differences in average annual temperature at the surface of only a few degrees Celsius. In spite of these potential problems, measurement of average annual temperatures at 1 m could be a powerful tool in geothermal prospecting if the area selected for study were fairly level and had uniform soils, vegetation, and land-use patterns. The accurate determination of average annual temperatures at 1 m would require several sets of synoptic measurements during the year, such as at monthly intervals. However, because of this require- ment, much of the advantage of 1-m temperature surveying over surveys at greater depth but at fewer sites is lost. Therefore, it is desirable to consider the advantages and disadvantages in using a single set of synoptic measurements at 1-m depth. The principal advantages of a single set of measurements are the savings in time and expense and the prompt availability of the results for use in interpretation and decisionmaking in further ex- ploration. The chief disadvantage is that, in dealing with a single set of synoptic measurements, another group of variables enters the problem: the areal variations in temperature related to differences in soil-temperature response to periodic and non- periodic fluctuations in solar-energy input. Although diurnal temperature waves are damped to amplitudes too small to be measured at 1 min most climatic and geologic environments, fluctuations having longer periods, including the annual tem- perature wave, have significant amplitude at 1 m depth. In the simple case of a single homogeneous soil or rock layer of semi-infinite extent, the response at the base of the layer to harmonic fluctuations in temperature (sinusoidal waves) at the top surface or land surface can be expressed by two equations: _, 7, Ax=Aoe axP (1) X tx-ro=§ "(f—f (2) where A x is the amplitude of the temperature wave at the base of layer x (°C), A 0 is the amplitude of the temperature wave at the land surface (°C), x is the thickness of layer x (cm), “x is the thermal diffusivity of layer x (cm2 S“), P is the period of the temperature wave (3), and tx — to is the time lag between the temperature waves at the land surface and at the base of layer x (s). As shown by equations ( 1) and (2), the amplitude of a temperature wave decreases exponentially with a linear increase in depth, but the timelag increases linearly with a linear increase in depth. The longer the period of the wave at the surface, the greater will be the amplitude of that wave at any given depth. In general, temperature waves at the land surface, which are caused by fluctuations in solar-energy input, may be grouped in three categories on the basis of their periods: (1) diurnal, (2) “weather,” and (3) annual (or seasonal) waves. The periods of the first and third categories are indicated by their names and are regular. Waves of the second category are irregular in period and are caused by the passage of air masses of different meteorological character- istics. Features typical of each of the three types of waves in the area of study, in west-central Nevada, are presented in table 1. The simple case of idealized harmonic temperature waves being damped con- ductively in a semi-infinite uniform medium is assumed. The two values of thermal diffusivity shown represent the approximate upper and lower limits found in the area of study. The periods and B4 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 2° I I I I I I I I I I I 2° 15 .. . ’ Q) ‘5 2 3 1o - .. 3 10 m I.” a: — G} o 5 5 III-I D 5 o - d 0 Lu 0 E 3 -5 — "" "5 n. E < —10 - - -1o 45 \ I I I I I I I I I J 45 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC ® Temperature at land surface. Wave amplitude=15°C 9 Temperature at 1 m, thermal diffusivity 6.0 x 10'3¢:m1~s‘l . Wave amplitude=10.0°C; timelag=23.7 days Temperature at 1 m, thermal diffusivity 1.5 x10’3cm2-s ‘1 . Wave amplitude= 65°C; timelag =47.4 days FIGURE 2.-—Idealized annual temperature fluctuations at land surface and a depth of 1 m for approximate conditions in western Carson Desert area, Nevada. amplitudes of the waves at the land surface likewise are typical of conditions in west-central Nevada. Stratification of near-surface materials produces effects on the temperature waves which may differ substantially from the effects in the uniform medium described above. Lachenbruch (1959) and Van Wijk and Derksen (1966, p. 177—180) have shown that the amplitude of a temperature wave at a given depth in a multilayered medium can range from several times smaller to somewhat larger than that in a single layer, dependingon the relative thermal properties, especially thermal inertia (contact coefficient), of the layers. As stated earlier, the amplitude of the diurnal waves is too small to be measured in the range of conditions likely to be found in the study area. Although detectable with ordinary temperature- measuring equipment, the amplitude of the “weather” waves is not large enough to be a troublesome factor. The chief problems are associated with the long- period annual waves, where differences in thermal diffusivity can result in differences in phase-lag time of several weeks and differences in amplitude of more than 2°C from place to place. (See table 1 and fig. 2.) Owing to the combined effects of the differences in amplitude and timelag, the variations in temperature at 1 m from place to place exceed 4°C in midwinter and midsummer (fig. 2). Stratification, described above, can cause even larger differences. As shown in figure 2, these differences could be minimized by making the temperature surveys in early April or early October. Although the difficulties described above appear formidable, they are only part of the entire problem. Other significant factors causing variations in temperature at 1 m not related to geothermal heat flow include variations in convective heat transport by water, vapor, and air, and variations resulting from areal variations in the solar-air-earth heat budget, such as the effects of topography, albedo, surface emissivity, wind, and vegetation. These factors could cause perturbations as large as those caused by the areal variations in thermal diffusivity and stratification. Added to these difficulties are the variations in thermal diffusivity with time as the moisture content of the near-surface materials TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA changes in response to precipitation and other weather factors. Clearly, there are serious obstacles to the successful use of synoptic temperature surveys at 1 m depth to determine geothermal heat-flow or temperature anomalies at depths below the range of solar-energy fluctuations. In spite of these obstacles, however, a practical test of the technique was believed warranted. EQUIPMENT AND METHODOLOGY Temperature measurements were made with a two- bead thermistor mounted at the tip of an aluminum - probe 1 m in length, connected by a 3-conductor cable to a solid-state Wheatstone bridge having a digital readout. Temperature was recorded to the nearest 0.01°C. Frequent calibration checks with a mercury- glass thermometer indicated an accuracy of about i0.3°C for individual measurements. In soft coherent deposits, such as damp silt or clay, access holes were driven by hand with a light-weight steel driver. Most of the holes, in sandy deposits, were made with a soil auger 57 mm in diameter. Lithologic character and moisture content of mate- rials were recorded at each site. After insertion of the temperature probe into the material at the bottom of the access hole, readings were made at regular time intervals and equilibrium temperature was computed by a method of extrapola- tion to infinite time described by Parasnis (1971). FIELD APPLICATION IN SODA LAKES AREA The first known thermal anomaly selected in which to field test the effectiveness of l-m temper- ature surveying was northeast of Soda Lakes, in the west-central Carson Desert, about 12 km northwest of Fallon, Nev. (fig. 3). Previous information pertinent to the study consisted of water-level data, chemical analyses of shallow ground water, and temperature-gradient measurements, all from test wells 20—50 m in depth, drilled by the US. Geological Survey and the US. Bureau of Reclamation in 1972 and 1973 (Olmsted and others, 1975, p. 99—118). Additional information included geologic studies by Morrison (1964), gen- eralized gravity data (Wahl, 1965), and generalized aeromagnetic data (US. Geological Survey, 1972). PHYSICAL SETTING The Soda Lakes area is within the Carson Desert, a large basin in the rain shadow of the Sierra Nevada to the west. Precipitation averages less than 100 mm annually (Hardman, 1965), a small part of which is winter snow. Long-term average monthly air temper- B5 ature at Fallon ranges from —0.3°C in January to 22.8°C in July (National Oceanic and Atmospheric Administration, 1975). The severe aridity and summer heat and the generally sandy soils of low water-holding capacity limit the vegetation to scattered perennial shrubs and various annuals that grow during periods of excess soil moisture after heavy rains in the spring or early summer. The Carson Desert is a topographic and structural basin occupied intermittently during the late Pleis- tocene and early Holocene by Lake Lahontan. The Soda Lakes thermal anomaly lies between two loci of late Quaternary basaltic eruptions: Soda Lakes to the southwest and Upsal Hogback to the northeast (fig. 3). The thermal area is underlain by uncon- solidated clay, silt, and sand of Lake Lahontan, intertonguing fluvial, swamp, and marsh deposits, and windblown sand. Present topography, which consists of scattered undrained depressions and intervening sandy flats and dunes, is largely the result of wind scour during the early Holocene (Morrison, 1964). Local relief amounts to as much as 10 m from the bottoms of wind-scoured depressions to the tops of adjacent sand ridges. The land surface slopes generally toward the north-northeast; alti- tudes in the area of the temperature survey range from 1,206 to 1,226 m above mean sea level. Sand and silt, chiefly of eolian origin, predominate within the 0—1 m depth range; less extensive deposits include pebbly sand, probably of stream-channel origin, and clayey silt to silty clay, which represent bottom deposits of Lake Lahontan (fig. 4). The finer grained deposits are exposed in the sides and bottoms of the wind-scoured depressions. Since the early 1900’s, when irrigation began in the area to the south, the water table has risen beneath the surveyed area. The rise may have been as much as 10—15 m on the basis of a reported rise in stage of 18 m at Soda Lakes from 1906 to 1930 (Rush, 1972). Depths to the water table range from about 4 to about 7 m beneath the sandy areas, but are less than 2 m beneath the depressions (fig. 4). As shown by the contours on the water table for July 1975 (fig. 3), shallow ground water beneath the area moves generally toward the northeast. Thermal water from considerable depths, perhaps several kilometers, moves upward along fault- controlled conduits (Olmsted and others, 1975, p. 118). One of these channels steam to the land surface and boiling water to depths of only a few meters in the vicinity of a shallow well about 18 m deep near the southwest corner of sec. 28, T. 20 N., R. 28 E. (fig. 3). The thermal water enters sand aquifers, some of which are 10 m or less below land surface, B6 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 118°50' mfl%fl,rx;}w7j Q? fluids/jJ 24%”. v.1 \ fig/fig T. 20 N. 39° 50' 5'7” 3: I <3; —/ hast?“ V 2')" L4 \ 5'5; Nu T.19 N. Base from USGS 15T topo eries. 4 Kl LOMETERS Soda Lake, Nevada L l I l EXPLANATION 1200 —— Water-level contour Altitude of water-table, in meters above mean sea level. Arrows show inferred direction of ground-water flow --- 20°--- Line of 20°C temperature at a depth of 30 m 18.0 0 Test well and temperature, in degrees Celsius, at a depth of 30 m FIGURE 3.—Soda Lakes-Upsal Hogback area showing temperatures at a depth of 30 min shallow test wells, configuration of water table, and direction of shallow ground—water movement in July 1975. 44 A 39° 34’ . 39° 33' TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA 118°51' 118°50' EXPLANATION O l-meter temperature-measurement site ’-_..-’ Abandoned irrigation ditch Road or trail a," Contact, dashed where approximate Pond or lakebed Test well and depth to water table, in meters Coarse sand, pebbly sand Medium sand Fine sand 600 METERS FIGURE 4.—-Soda Lakes area showing character of materials to a depth of 1 m and the depth to the water table in shallow test wells in December 1975. B8 through which it moves laterally toward the north- east. The temperature anomaly thus produced is strongly asymmetrical toward the northeast (fig 3). TEMPERATURE SURVEY IN VICINITY OF STEAM WELL The first temperature measurements at 1-m depth were made during November 1973 in the vicinity of an abandoned well, reported to have been originally about 18 m deep, near the southwest corner of sec. 28, T. 20 N., R. 28 E. (fig. 3). The site was selected because of suspected high near—surface temperatures. Evi- dence for a shallow thermal anomaly consisted of hydrothermal alteration of the sandy eolian deposits to kaolinite and iron hydroxides and oxides, steam issuing from the old well and from a nearby fumarole, and observed rapid snowmelt in an area coextensive with the exposures of hydrothermally altered deposits. Lateral spacing of the temperature- measurement sites ranges from about 5 min the zone of highest temperature near the fumarole and steam well to more than 100 m on the margins of the strongest thermal anomaly, where temperatures at 1 m were less than 30°C (fig. 5). Temperature at 1 m ranged from less than 30°C to more than 90°C in a small linear band trending northeast and centered on the fumarole (fig. 5). Highest temperatures were associated with the area of most intense hydrothermal alteration and prob- ably indicate the presence of steam near the surface. ‘ The temperature pattern suggests rising thermal water along a northeast-trending elongate or linear conduit, probably fault-controlled. Possible presence of a linear zone approximately perpendicular to the main zone is indicated by the temperature pattern 30- 50 m northeast of the steam well. The high temperatures probably result in large part from heat transport by steam rising from a boiling water table rather than chiefly by heat conduction through the altered clayey sediments above the water table. Depth to the water table within the area shown in figure 5 has not been measured, but measurements in nearby test wells indicate probable depths of 5-7 m outside the hottest part of the thermal anomaly. TEMPERATURE SURVEY IN NOVEMBER 1973 Concurrently with the temperature survey of the small area adjacent to the old steam well, measure- ments were made in a much larger surrounding area. Earlier data from test wells ranging in depth from 9 to 44 m indicated an asymmetrical northeast- trending thermal anomaly centered approximately at the steam well (fig. 6). Temperature at a depth of 30 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS m below land surface ranged from an inferred temperature of more than 120°C near the steam well—boiling temperature at hydrostatic pressure— to less than 20°C on the margins of the anomaly. Lateral temperature gradient toward the southwest was greater than that toward the northeast, and gradients along the northwest margin were greater than those along the southeast margin. With a range in temperature of more than 100°C at 30 m, it was anticipated that the temperature pattern at 1 m would reflect deeper temperatures and that, with much greater density of measurement sites, the temperature pattern at l m might reveal features not apparent from the data at 30 m shown in figure 6. Because of concurrent other work, it was not possible to complete the measurements at about 100 sites within a few days, as would have been desir- able. Instead, the survey required nearly a month, from November 11 to December 7, 1973. Because temperature at 1 m changed considerably during this period, the temperature at each site was adjusted to a midperiod date of November 23 on the basis of repeated measurements at several sites. A linear rate of change of temperature during the period of measurements was assumed, and the same rate of change was used for all sites. Analysis of air- temperature data for the nearby weather station at Fallon suggests that the actual change in tempera- ture in the ground probably was irregular, in response to antecedent “weather” cycles, as dis- cussed previously. Moreover, variations probably existed in the rate of temperature change from place to place, owing to both areal and temporal variations in thermal diffusivity. However, the differences between the interpolated temperatures used and the actual values are believed to be small—on the order of a few tenths of a degree Celsius at most sites. Results of the temperature survey are shown in figure 7. Temperature on the margins of the anomaly, presumably where the geothermal heat flow is normal or nearly normal, was about 12°C; temperature in the warmest parts of the anomaly exceeded 90°C near the steam well and 30°C at another thermal high about 500 m south of the steam well. The general northeasterly alinement of the long axis and the asymmetry of the anomaly are similar to the temperature pattern at 30 m (compare figs. 6 and 7). However, a significant difference is revealed by the greater density of the 1 in data: instead of one thermal maximum centered at the steam well, there were two en echelon elongate maxima, separated by a lower temperature trough, all oriented northeast. The zone of highest temperature northeast of the steam well, where the temperature exceeded 20°C, TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA B9 . 22.3 2 25 .9 24.8 ~---— 50 —— Line of equal temperature at a depth of 1 meter L——l——l—l___l_l O 26.9 . 24.6 >16. EXPLANATION 3 ° Measurement site and temperature at a depth 259 of 1 m, in degreesCelsius Interval, 10 degrees Celsius S: Abandoned steam well F: Fumarole 50 METERS O 17.7 FIGURE 5.—Temperature at a depth of 1 m near steam well, November 23, 1973. extended almost 700 m northeast of the well; this fact suggests that a linear hydrothermal conduit or conduit system may extend at least that far north- east of the steam well. RELATONS OF l-M TEMPERATURE TO SNOWMELT PATTERNS In early January 1974, after a brief snowfall of about 50 mm, oblique aerial photographs of the area near the steam well were taken at low altitude from a US. Navy helicopter. Within 24 hours after precipi- tation ceased, ground temperatures and associated heat transfer from the land surface were sufficiently high to melt the snow completely from three small areas, as shown in figure 7. Two of the snowmelt areas clearly are coextensive B10 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 118°51' 39° 34' EXPLANATION O l-meter temperature-measurement site _,._.I Abandoned irrigation ditch _, _. .— Road or trail Pond or lakebed 25 7. Test well and temperature, in degrees Celsius, ' at a depth of 30 meters. Open circles denote test wells drilled after fall of 1973. S, abandoned steam well — 70 — Line of equal temperature at a depth of 30 meters. Interval, 10 degrees Celsius 39° 33' |'__I—|—_1——I'__T—l I o 300 600 METERS IL- _______.I I l FIGURE 6.—Soda Lakes area showing temperature at a depth of 30 m, fall 1973. TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA 118°51' 118°50' B11 \ I 39° 34' EXPLANATION I l-meter temperature-measurement site ,-—_.—-r Abandoned irrigation ditch _._— a Road or trail Pond or lakebed Interval, 2 degrees Celsius 22 Area of snowmelt 39° 33' r—I—l—f—I—lfi O 300 600 M ETERS l Line of equal temperature at a depth of 1 meter. FIGURE 7.—Soda Lakes area showing temperature at a depth of 1 m on November 23, 1973 and area of snowmelt‘ in January 1974. Sites in northern and eastern parts of area not measured in November 1973. B12 with thermal maxima at 1 m as shown by the November 23, 1973 data, but a third area, on the southeast flank of the principal thermal anomaly, seems not to be related directly to the temperature pattern. However, comparison with the near-surface lithology and the depth to water (fig. 4) suggests that the third snowmelt area indicates a heat sink coincident with a wind-scoured depression. The near- surface deposits in the depression are chiefly clay and silt of low thermal diffusivity in comparison with that of the damp sand of the surrounding area. Moreover, the depth to a warm water table is less than 2 m in the depression, as compared to a depth of more than 5 m in the surrounding higher, sandy area. RELATIONS OF l-M TEMPERATURES TO GROWTH OF RUSSIAN THISTLE Color vertical aerial photographs taken in February 1974 revealed brown color patterns which were approximately coextensive with the areas of maximum temperature at 1 m depth in November 1973 (fig. 8). Later inspection on the ground indicated that the brown patterns were caused by dense stands of dried-up Russian thistle (Salsola Kali L. var. tenufolia Fausch). The thistle appears to thrive in sandy soil, and the density and vigor of the stands correlate directly with the ground temperature (fig. 8). TEMPERATURE SURVEYS IN JUNE AND SEPTEMBER 1974 In order to determine whether the configuration and extent of the l-m anomaly changes with time, a survey was made northeast of the steam well in June 1974, and areal coverage was extended beyond that of the November 1973 survey, mostly toward the east and northeast, in September 1974. The June measurements indicated a pattern similar to that of the previous November, but with minor differences, probably related to areal varia- tions in thermal diffusivity of the shallow materials. (Compare figs. 7 and 9.) The September 1974 measurements showed that the thermal anomaly at l-m depth extended more than 2 km east and northeast of the center of the November 1973 anomaly. A similar pattern at 30 m was established on the basis of extrapolated temperature gradients from measurements in 20-m-deep test wells drilled in October 1974. ANNUAL FLUCTUATION IN TEMPERATURE AT THREE SITES Monthly temperature measurements during 1975 were made at three sites in order to determine differences in the nature of the annual temperature GEOHYDROLOGY OF GEOTHERMAL SYSTEMS wave in three different lithologic and thermal environments. The sites (fig. 9) were selected to represent the approximate range of temperature fluctuations throughout the surveyed area. Site 98, south of the thermal anomaly, is in sand of chiefly eolian origin. Depth to the water table averaged about 4.7 m in 1975. Site 67, also in predominantly sandy deposits but containing slightly more clay and silt than at site 98, is on the northwest flank of the anomaly. Depth to water averaged about 6.7 m in 1975. Site 11, about 380 m southeast of the steam well, is near the north edge of the depression where snowmelt was observed in January 1974, as de- scribed earlier. The deposits at site 11 are lacustrine clay and silt having a moisture content greater than that at the other two sites. As shown in figure 10, temperature at all three sites was always higher than the smoothed-average air temperature at Fallon Experiment Station; the amplitude of the annual temperature fluctuations was less at the three sites than at Fallon; and the temperature waves at the three soil sites lagged the annual air temperature wave at Fallon, as would be expected from theoretical considerations. Differences among the three sites are apparent in figure 10. The temperature at site 11, the hottest location, differs from that at site 98, the coldest location, by amounts ranging from 3°C in late spring to about 11°C in early winter. The temperature difference between sites 67 and 98 ranges from less than 05°C in the spring to about 3°C in the fall. These temperature differences are largely the result of differences in thermal diffusivity at the three sites but are also related to differences in surface conditions and to the effect of a shallow water table at site 11. During most of the year, the thermal diffusivity of the damp clay and silt at site 11 was less than that of the sandier materials at the other two sites; this relation accounts in part for the smaller amplitude of the annual temperature wave at site 11. The smaller amplitude at site 11 probably is also caused by stratification and heat-sink effects of a warm, shallow water table. The range in temperature differences among the three sites during the year illustrates the danger of using synoptic temperature measurements at only one time during the annual cycle in trying to discern patterns related to variations in geothermal heat flow. CORRELATION OF TEMPERATURES AT DEPTHS OF I METER AND 20 METERS The evidence discussed above suggests that, in the Soda Lakes area, a temperature survey at a depth of 1 m might provide a fairly reliable indication of the B13 TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA 118°50' 118°51' EXPLANATION l-meter temperature-measurement site _,._._, Abandoned irrigation ditch ' Road or trail Pond or lakebed Dense stand Very dense stand '////A 600 METERS 300 _ 4 3 O 9 3 39° 33' —' FIGURE 8.-——Soda Lakes area showing density of Russian thistle on February 5, 1974. B14 39° 34' 39° 33' GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 118°50' 118°51' I EXPLANATION l-meter temperature-measurement site Abandoned irrigation ditch Road or trail Pond or lakebed Lines of equal temperature at a depth of 1 meter. Interval, 2 degrees Celsius. Solid lines represent temperature on September 20, 1974; dotted lines on June 15, 1974 Site of periodic temperature measurements at a depth of 1 meter during 1975 r—l—I"_'I—"I—l—_l O 300 600 METE R8 | 98 -——————%L- . , . FIGURE 9.—Soda Lakes area showing temperature at a depth of 1 m on June 15 and September 20, 1974. B15 TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA mp: NFI NF m: ON cN MN NM :ozmm am 252353 in E mcofimgosa 5E» woquEoo No.5 8wa «com 5 mofim 8.23 3 E H mo :33. a 3 353383 E mcofimgoimldfl HEDGE .mpmfi MES“. nofimum unvfiigxm mhmr Own. >02 #00 mum UD< 43—. 23—. >52 mm< «:12 mm". 244. _ _ _ _ _ _ _ _ _ _ .>mZ\ZOJ_n_< QwIFOOS—m ..... . . c \O \ \\\\‘O\\\ I // . . .. \ \ \ \\\ I / O >mZ zen—Jaw". ._.< \\\\\ // // mEDF<.02_>0_>_ > Base from USGS 15' topo series 4 Kl LOMETERS Desert Peak, 1951 ; Carson Sink, 1951; I Soda Lake, 1951; Stillwater, 1950. EXPLANATION 0 Measurement site —18-—- Line of equal temperature at a depth of 1 meter. Dashed where uncertain. Interval, 1 degree Celsius FIGURE 14.—Upsal Hogback area showing temperature at a depth of 1 m, October 29, 1975. TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA the October 29 survey, but there were some differ- ences in detail, owing in part to the greater number of measurement sites for the December survey (fig. 15). The center of the thermal maximum east of the Hogback was about 400 111 west of where it had been in late October, and the thermal minimum between the two maxima had a greater amplitude than it did in October. TEMPERATURE AT A DEPTH OF 30 M, SEPTEMBER 1975 By September 1975, temperatures in the 20 shallow test wells near Upsal Hogback had stabilized sufficiently so that temperatures and temperature gradients below the zone of measurable annual fluctuation could be determined with fair accuracy. As shown in figure 16, the temperature at a depth of 30 m defined a pattern that differed considerably in detail from the patterns indicated by three l-m surveys from June to December 1975. The thermal maximum east of the Hogback, which was noted in the 1-m surveys for October and December, also was the maximum for the 30-m measurements in September. The amplitude of the anomaly at 30 m was about 12°C: temperature ranged from less than 15°C on the northeast edge to more than 26°C at the center, at the test well described earlier. The ampli- tude of the equivalent anomaly at 1 m was about 33° —3.7°C in both the October and December surveys. In contrast to the fairly good agreement between l-m and 30-m temperature anomalies east of the Hog- back, the correlation between the patterns at the two depths appears to be rather poor in the rest of the Upsal Hogback area. The thermal maximum north of the Hogback delineated in both the October and December l-m surveys did not appear as a separate maximum at 30 In: the minimum between the two maxima at 1 m was not present at 30 m. CORRELATION OF TEMPERATURES AT DEPTHS OF 1 M AND 30 M From the differences between the thermal anom- alies at 30 m and 1 m described above, it might be inferred that the correlation of temperatures at the two depths also is poor. As shown in figure 17, this is indeed the case. The least-mean-squares linear regression of temperature at 30-m versus temper- ature at l-m for December 1975 is T30 = 1.02 T1 + 7.55 but r2, the coefficient of determination, is only 0.23; this figure indicates a very poor correlation. Unlike the situation in the Soda Lakes area, where the amplitude of the thermal anomaly at 30 m exceeds 100°C, the amplitude of the anomaly northeast of B21 Upsal Hogback is only 12°C. This difference in large part accounts for the good correlation in the first area and the poor correlation in the second. CONCLUSIONS The results of this study confirm the general principle that the utility of temperature measure- ments at a depth of 1 min geothermal exploration is directly related to the magnitude of the temperature and heat-flow anomalies at greater depths and inversely related to the magnitude of perturbing influences such as local variations in the character and moisture content of the near-surface materials, topography, depth to the water table, vegetation, land use, and microclimate. Temperature measurements at 1 m are most successful in delineating geothermal areas where lateral variations in temperature at greater depths are very large and where near-surface geothermal heat flows are thousands or tens of thousands times larger than background values. In the present study, the detailed survey of the small area surrounding the fumarole and abandoned steam well northeast of the Soda Lakes exemplifies the optimum case. There, the exceedingly high temperatures at 1 m, which exceed 90°C near the fumarole, almost certainly reflect heat transfer by ascending steam rather than by conduc- tion. In such a situation, closely spaced temperature measurements at 1 m may be used to determine the nature and configuration of the near-surface hydrothermal-discharge system. The technique is especially applicable to the study of hot-spring areas where subsurface leakage of thermal water from the conduit system is suspected, or to areas like the one surrounding the steam well, where steam rather than thermal water is present near the surface. Somewhat more equivocal results are obtained with 1-m temperature surveys where convective heat transport by rising steam or hot water does not extend to the land surface or where geologic, topographic, hydrologic, and other factors not related directly to geothermal heat flow are highly variable. The temperature surveys of the Soda Lakes area in November 1973 and in June and September 1974 are examples of this case. The general configuration of the thermal anomaly resulting from hydrothermal discharge into a shallow aquifer system had been established in that area on the basis of temperature surveys in test wells 20—50 m deep. Good correlation between temperature at 20 m and temperatures at 1 m at the test-well sites (fig. 11) suggested that the data for the l-m surveys could be used with some assurance to infer the nature and configuration of the thermal anomaly at greater B22 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS I 118° 50' 118°45 I I ' i i I I I I I I \ I I I 29 M38 I 27 i 26 I . I I I | l \ i I 'i _____ I ______ I II I 4 I u I I \k I i I .. ll I I I F,\ 2 33 1 34 I 35 I I I I E I I I I .I \ . I I \ I : : 3‘” *3”: \ | | 'w Y .43 ________ --(;>——-- p In I ‘ I P :2 (-1.3. ‘I , n ’5 ‘ am i . “a 10 “ ‘ H “2 Ir: id" (3 . upf-L k» EH , r.- 1“" .2 I" .4..- QI’J/I ‘ k — la‘ ’ § 3" fill /’ ‘r I ”y I [/9 l_ Base from USGS l 5' topo series Desert Peak, 1951;Carson Sink, 1951; Soda Lake, 1951 ; Stillwater, 1950 EXPLANATION 0 Measurement site — 13 — Line of equal temperature at a depth of 1 meter. Interval, 1 degree Celsius FIGURE 15.—Upsal Hogback area showing temperature at a depth of 1 111, December 2, 1975. B23 TEMPERATURE SURVEYS IN GEOTHERMAL EXPLORATION IN NEVADA .1 18°45’ “8°50" wxrruz I ::n~ 4 KILOMETERS 21 1| 1 I Base from USGS 15 topo series Desert Peak, 1951 ; Carson Sink, 1951; Soda Lake, 1951, Stillwater, 1950. EXPLANATION Test well — 18 — Line of equal temperature at a depth of 30 meters. 1 degree Celsius Interval , September 1975. FIGURE 16.—Upsal Hogback area showing temperature in test wells at a depth of 30 m B24 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 9 1o 11 12 13 14 15 26 — O - 26 {I} 2 U) _l ”o‘ a O u, 24 — - 24 I! (.7 Lu 0 E O T30=1.02 T1+7.55 8 r2=0.23 t 22 \\ — 22 E O (‘0 LL 0 I p. E 1: 2° 00 — 20 < ’— < ”J c: D E a: 18 O — 18 u] D. E LIJ '— 16 — — 16 O O 14 l 1 1 I 1 14 9 1o 11 12 13 14 15 TEMPERATURE AT A DEPTH OF 1 m (T1), IN DEGREES CELSIUS FIGURE 17.—Correlation of temperature at a depth of 30 m with temperature at a depth of 1 m in Upsal Hogback area, December 1975. depths in considerably greater detail than would be possible from the test-well data alone. Locally, however, particularly at wind-scoured depressions, the effects of topography, geology, and hydrology (especially depth to the water table) on temperature at 1 m are sufficiently large to overwhelm the temperature differences caused by lateral variations in geothermal heat flow. The large differences in the amplitude and phase of the annual temperature waves at a site in clay and silt of a wind-scoured depression in comparison with two other sites on sandy flats illustrate the need to allow for extraneous factors in interpreting the shallow temperature data. The least favorable results are obtained with 1-m temperature surveys where the amplitude of the temperature anomaly at greater depths is relatively small, near-surface heat flows are from a few times to a few tens of times regional background values, and the perturbing effects of the nongeothermal in- fluences are large. The temperature surveys of the Upsal Hogback area represent this case. There, the temperature of the thermal water leaking into shallow aquifers is much less than that in the Soda Lakes area, probably because of greater mixing with REFERENCES CITED B 25 shallow, cool ground water, and the range in depth to the water table, topographic relief, and variability of geologic materials are greater than in the Soda Lakes area. As a result, temperatures at a depth of 1 m have little or no correlation with temperatures at greater depths. The serious risk that arises in attempts to use l-m temperature surveys in places like the Upsal Hogback area is the identification of thermal anomalies that are actually unrelated to true geothermal anomalies at greater depth. This prob- lem is highlighted by the reportedly successful use of shallow temperature surveys for a variety of purposes unrelated to geothermal investigations. Such uses include mapping of buried shallow glacial aquifers (Cartwright, 1966, 1968; Carr and Blakely, 1966; Parsons, 1970); correlation of shallow temper- atures and temperature fluctuations with depth to water or depth to bedrock (Birman, 1969; Cartwright, 1974); analysis of the effects of vegetative cover on the thermal regime of soils and subsoils (Singer and Brown, 1956; Pluhowski and Kantrowitz, 1963; Kane and others, 1975); mapping of buried water-bearing joints, fissures, or caverns in bedrock (Ebaugh and others, 1974); determination of stream reaches receiving ground-water discharge (Hollyday and others, 1972); and determination of the effect of various soil-management procedures in altering the thermal regime of agricultural soils (Van Duin, 1963). All these uses have one principle in common: lateral variations in geothermal heat flow or in temperature below the depth of measurable annual temperature fluctuation are considered a perturbing influence and are treated in the analysis as “noise” in the data. Thus, to the extent that the types of studies listed above are successful, the use of shallow temperature measurements in geothermal studies would be unsuccessful, and vice versa. REFERENCES CITED Birman, J. H., 1969, Geothermal exploration for ground water: Geol. Soc. America Bull., v. 80, p. 617-630. Carr, D. R., and Blakely, R. F., 1966, Temperature variation at a depth of 30 cm in clay till and outwash sand and gravel, in Symposium on remote sensing of environment (ONE and AFCRL) 4th, 1966, Proceedings, University of Michigan, Willow Run Laboratories, Ann Arbor, Michigan, p. 203-214. Cartwright, Keros, 1966, Thermal prospecting for shallow glacial and alluvial aquifers in Illinois [abs.]: Geol. Soc. America Abs. with Program, 1966 Annual Meeting, p. 36. 1968, Temperature prospecting for shallow glacial and alluvial aquifers in Illinois: Illinois State Geological Survey Circ. 433, 41 p. 1974, Tracing shallow groundwater systems by soil temper- atures: Water Resources Research, v. 10, no. 4, p. 847—855. Ebaugh, W. R., Greenfield, R. J., and Parizek, R. R., 1974, Tem- perature patterns in soil above caves and fractures in bedrock [abs.]: Geol. Soc. America Abs. with Programs, Cordilleran GPO 789-090/49 Section 70th Annual Meeting, v. 6, no. 3, p. 171. Hardman, George, 1965, Nevada precipitation map, adapted from map prepared by George Hardman and others, 1936: Univer- sity of Nevada, Agricultural Experimental Station Bull. 183, 57 p. Hollyday, E. F., Scherer, L. R., and Hall, Thomas, 1972, Stream groundwater interface: Temperature gradients and ground- water discharge [abs.]: Geol. Soc. America Abs. with Pro- grams, Annual Meetings, v. 4, no. 12, p. 541-542. Kane, D. L., Luthin, J. N., and Taylor, G. S., 1975, Heat and mass transfer in cold region soils: Alaska University, College, Institute of Water Resources, Publication No. IWR-65, 50 p. Kappelmeyer, 0., 1957, The use of near-surface temperature measurements for discovering anomalies due to causes at depths: Geophysical Prospecting, v. 5, no. 3, p. 239-258. Kintzinger, P. R., 1956, Geothermal survey of hot ground near Lordsburg, New Mexico: Science, v. 124, p. 629. Lachenbruch, A. H., 1959, Periodic heat flow in a stratified medium with application to permafrost problems: U.S. Geo- logical Survey Bull. 1083—A, 36 p. Lovering, T. S., and Goode, H. D., 1963, Measuring geothermal gradients in drill holes less than 60 feet deep, East Tintic district, Utah: US. Geological Survey Bull. 1172, 48 p. Morrison, R. B., 1964, Lake Lahontan: geology of the southern Carson Desert, Nevada: US. Geological Survey Prof. Paper 401, 156 p. National Oceanic and Atmospheric Administration, 1975, Clima- tological Data, Nevada, Annual Summary 1975: US. Depart- ment of Commerce, NOAA (formerly US. Weather Bureau), v. 90, no. 13. Olmsted, F. H., Glancy, P. A., Harrill, J. R., Rush, F. E., and Van Denburgh, A. S., 1975, Preliminary hydrogeologic appraisal of selected hydrothermal systems in northern and central Nevada: US. Geological Survey open-file report 75—56, 267 p. Parasnis, D. S., 1971, Temperature extrapolation in infinite time in geothermal measurements: Geophysical Prospecting, v. 19, p. 612-614. Parsons, M. L., 1970, Ground-water thermal regime in a glacial complex: Water Resources Research, v. 6, no. 6, p. 1701-1720. Pluhowski, E. J ., and Kantrowitz, I. H., 1963, Influence of land— surface conditions on ground-water temperatures in south- western Suffolk County, Long Island, New York, in Geo- logical Survey research 1963: U.S. Geological Survey Prof. Paper 475—B, p. 3186—B188. Rush, F. E., 1972, Hydrologic reconnaissance of Big and Little Soda Lakes, Churchill County, Nevada: Nevada Department of Conservation and Water Resources, Water Resources In- formation Service Report 11. Singer, I. A., and Brown, R. M., 1956, The annual variation of sub-soil temperatures about a 600—foot diameter circle: Am. Geophys. Union Transactions, v. 37, p. 743. US. Geological Survey, 1972, Aeromagnetic map of parts of the Lovelock, Reno, and Millet 1° by 2° quadrangles, Nevada: US. Geological Survey open-file report, scale 1:250,000. Van Duin, R. H. A., 1963, The influence of soil management on the temperature wave near the surface: Instituut voor Cultuur Technick en Waterhuishouding, Wageningen, Tech. Bull. 29, 21 p. Van Wijk, W. R., and Derksen, W. J ., 1966, Sinusoidal temperature variation in a layered soil, ch. 6 in Van Wijk, W. R., ed., Physics of plant environment: Amsterdam, Holland, North Holland Publishing Company, p. 171-209. Wahl, R. R., 1965, An intrepretation of gravity data from the Carson area, Nevada: Stanford University, Department of Geophysics, Student Report, Stanford University, California. ‘ pg 7. DAYS , U. IOWC’ ' The Waters of Hot Springs * National Park, Arkansas-4- Their Nature and Origin * GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—C Prepared in cooperatian with the National Park Service The Waters of Hot Springs National Park, Arkansas— Their Nature and Origin By M. s. BEDINGER, F. J. PEARSON, JR., ]. E. REED, R. T. SNIEGOCKI, and c. G. STONE GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—C Prepared in cooperation with the National Park Service UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1979 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data The Waters of Hot Springs National Park, Arkansas—their nature and origin. (Geohydrology of geothermal systems) (Geological Survey Professional Paper 1044-C) Bibliography: p. C32-C33. Supt. of Docs. no.: I 19.1621044 1. Hot springs-Arkansas—Hot Springs National Park. I. Bedinger, M. S. II. United States. National Park Service. III. Series. IV. Series: United States. Geological Survey. Professional paper ; 1044-C. GB1198.3.A8W37 551.2’3’0976741 77-608354 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, DC. 20402 Stock Number 024-001—03199-0 CONTENTS Page Page Abstract ____________________________________________________ C1 Character of the spring and well waters, eta—Continued Introduction ________________________________________________ 1 . Chemical quality—Continued Purpose of the study ____________________________________ 1 Carbonate geochemistry ___________________________ 018 Metric units ____________________________________________ 2 Isotopic chemistry of carbon ______________________ 20 Acknowledgments ______________________________________ 2 Age of the hot-springs water ____________________ 23 History of spring development and scientific study ________ 2 Radioactivity of the waters in the hot-springs region __ 25 Geologic setting ____________________________________________ 6 The hot-springs flow system __________________________________ 26 Character of the spring and well waters in the hot-springs area 7 Conceptual model ______________________________________ 26 Physical quality ________________________________________ 7 Digital flow models ______________________________________ 28 Flow of the hot springs ______________________________ 7 Description of the hydrothermal model ________________ 28 Temperatures of the hot-springs water ________________ 9 Application of digital models to the hot-springs flow Silica concentration as a temperature indicator ________ 9 system __________________________________________ 29 Temperatures of the cold springs and wells ____________ 11 Discussion of models ________________________________ 32 Chemical quality ________________________________________ 14 Conclusions ________________________________________________ 32 Geochemistry of hydrogen and oxygen isotopes ________ 16 References __________________________________________________ 32 ILLUSTRATIONS Page PLATE 1. Geologic map of Hot Springs National Park and vicinity, Arkansas ____________________________________________ In pocket FIGURE 1. Map showing location of study area ________________________________________________________________________________ C1 2. Map showing locations of the hot springs and hot-water collection lines ______________________________________________ 4 3. Graph showing temperatures and flows of the hot springs ____________________________________________________________ 8 4. Graph showing variation of dissolved-silica concentration with temperature __________________________________________ 10 5. Graph showing maximum measured and calculated temperatures of the hot springs __________________________________ 11 6. Map showing locations of the cold springs and wells ________________________________________________________________ 12 7—11. Graphs showing: 7. Comparison of isotopic composition of the cold and hot springs, Arkansas, waters and of hydrothermal waters elsewhere ________________________________________________________________________________________________ 17 8. Relation between analyzed and calculated partial pressures of CO2 ______________________________________________ 20 9. Relation of alkaline-earth concentration to dissolved-carbonate concentration ____________________________________ 21 10. Relation between calcite saturation and PCO2 in the hot springs ________________________________________________ 24 11. Relation between adjusted carbon-14 age and total carbonate __________________________________________________ 25 12. Diagrammatic model of flow in the Bigfork Chert __________________________________________________________________ 30 13. Diagrammatic model of flow in the Arkansas Novaculite ____________________________________________________________ 31 TABLES Page TABLE 1. Generalized section of sedimentary rocks in the vicinity of the hot springs ____________________________________________ C6 2. Flows of the hot springs in group 1 in 1901 and 1931 ________________________________________________________________ 8 3. Temperatures, in degrees Celsius, of the hot springs ________________________________________________________________ 9 4. Temperatures of the cold springs __________________________________________________________________________________ 11 5. Temperatures of water in the wells ________________________________________________________________________________ 11 6. Chemical analyses of water from the hot springs, cold springs, and wells ____________________________________________ 14 7. Station numbers of wells and springs given in table 6 ______________________________________________________________ 15 8. Hydrogen- and oxygen-isotope concentration of waters in the hot-springs area ________________________________________ 18 9. Carbon-isotope data for waters in the hot-springs area ______________________________________________________________ 19 10. Comparison of the Bigfork Chert modeled and observed data for the hot springs ______________________________________ 31 11. Comparison of the Arkansas Novaculite modeled and observed data for the hot springs ________________________________ 31 III GEOHYDROLOGY OF GEOTHERMAL SYSTEMS THE WATERS OF HOT SPRINGS NATIONAL PARK, ARKANSAS— THEIR NATURE AND ORIGIN By M. S. BEDINGER, F. J. PEARSON, jR.,]. E. REED, R. T. SNIEGOCKI, and C. G. STONE ABSTRACT The 47 hot springs of Hot Springs National Park, Ark., issue from the plunging crestline of a large overturned anticline, along the southern margin of the Ouachita anticlinorium, in the Zigzag Mountains. Rocks in the vicinity of the hot springs range in age from Ordovician to Mississippian. The rocks—cherts, novaculites, sandstones, and shales—are well indurated, folded, faulted, and jointed. The springs emerge from the Hot Springs Sandstone Member of the Stanley Shale near the anticlinal axis, between the traces of two thrust faults that are parallel to the axis of the anticline. The combined flow of the hot springs ranges from 750,000 to 950,000 gallons per day (3.29X 10‘2 to 4.16X 10‘2 cubic meters per second). The flow of the springs is highest in the winter and spring and is lowest in the summer and fall. The temperature of the com- bined hot-springs waters is about 62 degrees Celsius. The radioactivity and chemical composition of the hot-water springs are similar to that of the cold—water springs and wells in the area. The dissolved-solids concentrations of the waters in the area generally range from 175 to 200 milligrams per liter. The main dif- ferences in the quality of the hot water, compared with nearby cold ground waters, are the higher temperatures and the higher silica concentrations of the hot springs. Cold waters in the area generally range from 15.0 to 26.8 degrees Celsius. The silica concentrations of cold ground waters range from 2.6 to 13.0 milligrams per liter, whereas the silica concentration of the hot springs is about 42 milli- J grams per liter. The high silica concentration of the hot springs is due to the increased solubility of silica in hot water. The silica con- centration of the hot springs indicates that the maximum tempera- ture reached by the hot-springs water is no more than a few degrees higher than the temperature at which the springs emerge. The tritium and carbon-14 analyses of the water indicate that the water is a mixture of a very small amount of water less than 20 years old and a preponderance of water about 4,400 years old. The deuterium and oxygen-18 concentrations of the hot-springs waters are not significantly different from those of the cold ground waters. The presence of radium and radon in the hot-springs waters has been established by analyses. Recent (1973) analysis showed the radium concentration to be 2.1 picocuries (10‘12 curies) per liter. Analyses made in 1953 of the radon gas, a radioactive decay product of radium, ranged from 0.14 to 30.5 nanocuries (10‘9 curies) per liter. Mathematical models were employed to test various conceptual models of the hot-springs flow system. The geochemical data, flow measurements, and geologic structure of the region support the con- cept that virtually all the hot-springs water is of local, meteoric origin. Recharge to the hot—springs artesian—flow system is by infil- tration of rainfall in the outcrop areas of the Bigfork Chert and the Arkansas Novaculite. The water moves slowly to depth where it is heated by contact with rocks of high temperature. Highly permeable zones, related to jointing or faulting, collect the heated water in the aquifer and provide avenues for the water to travel rapidly to the surface. INTRODUCTION PURPOSE OF THE STUDY The thermal springs of the Hot Springs National Park, at Hot Springs, Ark. (fig. 1), have been a natural resource of international renown for many years. The springs were known to President Thomas Jefferson, who initiated the first scientific study in 1804. This study, by William Dunbar and George Hunter, marked the beginning of an era of scientific curiosity as to the origin and heat source of the springs. Pine Blufl 100 MILES 0 40 BO 120 160 KILOMETERS FIGURE 1.—Location of study area. C1 C2 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Public interest in the hot springs has been focused primarily on the therapeutic value of the waters, and, in serving such interest, this also has been the focus of Federal management since the area was established as the Hot Springs Reservation in 1832. This emphasis did not change when direct Federal supervision was implemented in 1877, nor when the area was desig- nated as a National Park in 1921. The purpose of the park today is to preserve and protect the hot springs for present and future generations. Long-range planning for park uses takes into consid- eration, however, the prospect of a shifting of emphasis from therapeutic values of the spring waters to the scientific, esthetic, and recreational values of the park as a whole. The existence of the springs and their geologic and hydrologic setting as well as their thermal characteristics play an important role in attracting visitors to the area. The purpose of this report is to describe the hot springs with reference to their flow, temperature, and chemical quality; to present informa- tion on the geologic framework of the hot springs flow system; and to define the nature of the hydrologic and geothermal flow systems as completely as possible with the data available. METRIC UNITS For those readers interested in using the metric sys- tem, metric equivalents of English units are given in parentheses. The English units in this report may be converted to metric units as follows: To convert from— Multiply b -— To obtain~ (English unit) (conversion actor) (metric unit) Inches (in) 25.4 Millimeters (mm) Feet (ft) 3.048X 10—1 Meters (m) Square miles (miz) 2.59X 106 Square meters (m2) Gallons per day (gal/d) 4.38X 10—8 Cubic meters per second (m3/s) Feet per day (ft/d) 3.53X 10—6 Meters per second (m/s) Curie 3.7 X 1010 Becquerel (Bq) Atmosphere 1.013 x 105 Pascal (Pa) Chemical concentrations are given only in metric units—milligrams per liter (mg/ L). For concentrations less than 7,000 mg/L, the numerical value is about the same as for concentrations in the English unit, parts per million. The conversion from temperature in degrees Fahrenheit (°F) to temperature in degrees Celsius (°C) is expressed by: °C=(5/9)(°F—32). Kelvin=degrees Celsius+273.15. ACKNOWLEDGMENTS The authors are grateful for the advice, suggestions, and information given by many associates in Federal and State agencies, and for facts drawn from many published and unpublished reports. We are especially grateful to N. F. Williams, State Geologist, Arkansas Geological Commission, for his advice and interest in the study. Raymond B. Stroud, US. Bureau of Mines, and B. R. Haley, US. Geological Survey, provided ad- vice and assistance on the geologic study. T. W. Car- ney, R. C. Cross, and P. A. Bailey, chemists, Arkansas Geological Commission, provided chemical analyses of rock samples. In the presentation of geothermal models, the authors especially acknowledge the assis- tance of M. L. Sorey of the US. Geological Survey. Mr. Sorey critically examined the flow models presented in the report and provided constructive criticism and technical assistance. The authors are grateful for the cooperation and as- sistance of Garland Moore, Richard Ketcham, and Rock Comstock, of the National Park Service; Bernard Campbell, Superintendent, and Franklyn Hambly, Management Assistant, of the Hot Springs National Park, Ark., and Richard Rasp, Earl Adams, Lewis May, Bernie Braughton, Eugene Stringer, David Car- penter, and Alice Horner, also of the Hot Springs Na- tional Park. We also thank M. D. Bodden, J. C. Chem- rys, C. T. Rightmire, and T. A. Wyerman, who provided l4C, chemical, stable isotope, and tritium analyses of the samples collected in 1972. HISTORY OF SPRING DEVELOPMENT AND SCIENTIFIC STUDY The history of the hot-springs area has been documented in numerous publications, many of which present detailed accounts of some aspect of the springs’ environment and the cultural development of the area. For the purposes of this report, therefore, and to minimize duplication, only those historic events and developmental practices that relate to the technical management of the springs are cited. Early descriptions of the hot springs give different accounts of as many as 72 spring openings, in a belt about one-fourth mile long and a few hundred feet wide, along the southwest slope of Hot Springs Mountain. Excavation and covering of springs, to in- crease and concentrate flows and to protect the springs from contamination, have so altered the natural spring environment that it bears no resemblance to the origi- nal condition. Among the early investigators, Owen (1860) reported 42 springs; Glasgow (1860), 54 springs; Haywood (1902), 46 springs; and Hamilton (1932), 48 springs. In his detailed history of Hot Springs, Scully (1966, p. 139) reported 47 active springs, including 2 exhibition springs. Prior to 1877 some of the springs were walled in and covered by masonry arches to protect them from con- tamination (Scully, 1966, p. 118). By the 1890’s, most WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN C3 of the springs were covered and a complicated piping system had evolved for supplying the bathhouses with hot water. In 1901 the springs were uncovered to give access for sampling, and chemical analyses were made by Haywood (1902). The spring enclosures were opened again in 1931 for cleaning; some of them were deepened, and the present-day (1974) collection system was constructed. The collecting system diverts the flow of 44 springs to a central reservoir, from which the water is redistributed to individual bathhouses. A map of the collection system is shown in figure 2. Since 1948 all the water delivered to the bathhouses has been me- tered. Excess water overflows into Hot Springs Creek when storage reservoirs are full. Through the years (1860 to the present), at least 20 scientific investigations, directly or indirectly involv- ing the hot springs, have been made. Although each study generally had a separate and specific objective, many of the investigators became sufficiently in- terested in the hot springs to try to explain the origin of the water and the source of the heat. The chemical quality of water from the hot springs in Arkansas has been of great interest to man, probably since the hot springs were first discovered. One of the earliest scientific approaches to determine the concen- tration of the hot-springs waters is found in Branner’s (1892) Annual Report for 1891, in which analyses of water samples collected in 1890 are tabulated in grains per gallon. At random times since 1890, analyses have been made for investigations. The purpose of many of these investigations has been to support some therapu- tic claim for the water or to determine whether the chemical concentration of the water has changed. Most earlier investigators concluded that the waters discharged from the hot springs are of meteoric origin, having fallen as precipitation and recharged to the Bigfork Chert in the anticlinal valley lying just north- west, north, and northeast of Hot Springs. Some inves- tigators have attributed some of the recharge to the outcrop of Arkansas Novaculite to the east of the hot springs. Another theory that has been regarded by some as having a degree of scientific validity is that the water may be of juvenile origin, that is, derived from the interior of the Earth and not having previously existed as atmospheric water. Bryan (1922, p. 426) posed the question as to the meteoric, juvenile, or mixed origin of the waters dis- charged from the hot springs. He indicated (p. 447— 448) that the juvenile theory is perhaps more satisfac- tory, although it rests on an insecure foundation in postulating (1) a special igneous mass that is discharg- ing water owing to cooling and recrystallization and (2) a special fault fissure through which the water rises to the land surface. Bryan (p. 444) analyzed the merits of both the juvenile and the meteoric theories, but con- ceded that “a definite conclusion as to theultimate ori- gin of the water in the Hot Springs cannot now be reached.” He pointed out (p. 443—444) that "If the water is juvenile there is presumably a constant sup- ply, diminishing very gradually through the centuries in quantity and temperature * * * If * * * the water has a meteoric origin, it is variable in quantity, fluctuating with the seasons or with groups of years having heavy or light rainfall.” Thus, Bryan recognized the critical value of precise measurements of temperature, dis- charge, and other parameters during a sufficient period of time to provide adequate data on which to base con- clusions as to the water’s origin. Arndt and Stroud (1953) suggested a dual origin for the water. Meteoric water, they believed, entered the spring system through the lower division of the Arkan- sas Novaculite on Hot Springs and North Mountains. They calculated that this source of meteoric water could supply about one-sixth of the flow of the springs. The rest of the water, they considered, could be juvenile water rising from depth. Proponents of the theory of meteoric origin of the spring waters include Weed (1902), Purdue (1910), and Purdue and Miser (1923). Purdue (1910, p. 283) de- scribed the geologic conditions supporting this view and identified the collecting area as the anticlinal val- ley between Sugarloaf and West Mountains. Most of the valley is underlain by the Bigfork Chert, a much- fractured formation of high permeability into which infiltrates water from precipitation. According to Pur- due (p. 284), the occurrence of this formation in anti- clinal valleys, with its highly inclined beds, affords the most favorable condition for the intake of water. He postulates further that the water passes through the Bigfork Chert, beneath the North Mountain syncline, and is forced upward into the Hot Springs anticline to emerge as the hot springs. This suggested movement of water from the recharge area required geologic conditions that account for passage of the water through the Polk Creek Shale, Missouri Mountain Shale, and the Arkansas Novacu- lite, to discharge as it does from the Hot Springs Sandstone Member of the Stanley Shale. Such condi- tions would ordinarily require a fault or faults, with associated jointing and fissuring, that would provide passage through these formations. Several authors have shown such a fault (Arndt and Stroud, 1953; Fel- lows, 1966). Recent mapping by Haley and Stone (pl. 1) confirms the presence of a complex fault system in the area but indicates no conclusive evidence of a large fault at the hot springs. In addition to the faulting, the intensive folding and overturning of formations in the vicinity of Hot Springs are attended by intensive joint- C4 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS y A -(. u 4 1 fl '3 .I0 z ‘1 33"“: ’ .49 E a N A D o M p R \ \3 9 u 09“.“ . who» -°-‘3£’.-~°" EXPLANATION 0 Saving . o VIII max Mcnnolo , ‘ --------- IOOI Nel-vu'u—conociion lino '7 —— I931 Moi-vun—ullocv-on Iinu ecu "OI; unnuhliph“ map 0: :III In “““““ Stem "In an Spr an unusual Poll. I onus (#0! 5pm!” Cull Arch} } . , IOO 50 0 50 I00 FEET e I I . H e I 4 30 IS 0 IS 30 METERS b ) . é Q FIGURE 2.—Locations of the hot p. , Elevations of the hot springs. Hot Springs National Park. Arkansas WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN [From Hamilton and Blood. 193]] Elevation Sprinq (fagsnazgze Point of elevation if other than sprino orifice Remarks level) 682 - — 616 At connection with 4-inch collector. 1 < 680 ----- l 614 At connection with 4-inch collector. 683 l a all At connection with l1-inch collector. " 673 ... ’ 672 609 Overflow elevation of 80.000—callon reservoir. Q 679 609 Overflow elevation of 80.000-qallon reservoir. ------ Dry. not connected to collector line. ‘ 609 Overflow elevation of 80.000-qallon reservoir. I Dry. not connected to collector line. 609 Overflow elevation of 80.000-gallon reservoir. l Dry, not connected to collector line. 1 62l At connection with 4-inch collector. 606 80 feet iron spring. at connection with 8—inch collector. , 654 Not connected to collector line. 645 ’ 635 ----- - - l 655 I ‘ 603 632 597 90 feet from soring's collecting oasin at connection of b .< 6-inch line with lZ-inch collecting min. 597 4 . 597 ‘ “~- ‘ 63l ‘ 6l7 At fornlr point of connection with 3-inch collector. 40 (eet from spring. 6l7 " v 576 Bottom of reservoir. Discharges into central-collectino reservoir. r " 683 681 I i Unnunbered (a)-- 6l3 At connection with l1-inch collector. Unnunbered (b)-- 612 "A Unnunhered (cl __‘ Unnunbered (d)-- 627 At connection with collector line. ' Unnumoeved (e‘ Elevation too low to feed by oravitv into collection lines. Mauric Fordyce well---- 600 /\t connection with 3-inch collector. 20 feet from well. P ‘ , “ spnngs and hot-water-collection lines. C6 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ing and fissuring of the sandstone, chert, slate, shale, and novaculite. It is concluded from the present study that the faults and the associated joints and fissures provide conduits for the water. Prof. D. D. Owen, in his report of 1860 on the hot springs, said the following: “When we reflect on the boundless and never-ceasing flow of thermal waters that must have bathed the sides of Hot Springs Ridge for countless ages * * * and however inexplicable such wonderful phenomena and changes may at first appear, yet, when the chemical principles become properly understood, disclosed by the enlightened and accurate chemical analyses, these obscure geological transformations [and the origin of the water and operation of the springs] can be satisfactorily and clearly explained.” GEOLOGIC SETTING The rocks cropping out in the vicinity of the hot springs are sedimentary rocks, although intruded igneous rocks are exposed in the region (Purdue and Miser, 1923). The sedimentary rocks are relatively old (Paleozoic) and consist of shale, chert, novaculite, and sandstone. The names of the geologic formations, their geologic ages, and positions in the geologic column are given in table 1. Though no significant igneous rocks are exposed in the immediate vicinity of the hot springs, their nearby occurrence has been frequently cited in literature as possible sources for the heat of the springs. The igneous rocks were intruded into the sedimentary rocks during the early Late Cretaceous time (about 90 million years ago). The larger igneous intrusions in the hot-springs region are exposed in two small areas, about 6 miles southeast of the hot springs. Elsewhere in the region, igneous intrusions occur as very small dikes and sills. The sedimentary rocks in the vicinity of the hot springs were originally laid down on a sea bottom of nearly horizontal beds. At present the beds are gen- erally steeply inclined, because of tremendous and complex mountain-building forces in late Paleozoic time. The rocks have been subjected to at least three episodes of structural deformations—two episodes of compression from the south, producing imbricating thrust faults and, third, forces from the north which produced overturning and folding of beds and fault planes and further faulting. The geologic map (pl. 1) shows the edges of the inclined strata where they intersect the land surface. When the formations are crossed from northwest to southeast, they are seen in cross section (pl. 1) to lie in a series of very complexly folded anticlines and synclines, with some associated thrust faults. The hot springs emerge from the plunging crestline of a large overturned anticline along the southern margins of the Ouachita anticlinorium in the Zigzag Mountains. The Zigzag Mountains basically owe their presence to the resistant exposures of the Arkansas Novaculite. The zigzag pattern of the strata is mostly due to tightly compressed folds which plunge south- westward into the Mazarn Basin. The Mazarn Basin is a structural and topographic basin lying south of the Zigzag Mountains. The Stanley Shale, a formation much less resistant to erosion than the Arkansas Novaculite, crops out at the surface of the Mazarn Basin. The structural setting is illustrated by figure 6 in Purdue and Miser (1923). The formations, composed predominantly of shale, TABLE 1.—Generalized section of sedimentary rocks in the vicinity of the hot springs [Modified from Purdue and Miser, 1923] Maximum thickness 53’3“?!“ Formation in Hot Spfgngs area Lithologic description Topography Stanley Shale ridges 8,500 Greenish—black and black shale, Broad valleys with low gra sandstone, and traces and hills. g of t in chert and tuff. a. .9' Hot Springs Sand- 150 Hard, ay quartzitic sandstone, Steep slopes, or narrow, sharp- E cong omerate, and thin inter— crested ridges. g bedded black shale. 2 Arkansas Novaculite 650 Massive- to thin-bedded novacu- High ridges and steep lite, interbedded with black slopes. . clay, siliceous shale, and Devonian tripoli. Siluria Missouri Mountain 195 Green to black shale, a few thin Steep slopes or n Shale, Blaylock, Sand- sandstones, and traces of narrow valleys. stone, and Polk Creek conglomerate. Shale, undifferentiated. 5 Bigfork Chert 700 Thin-bedded chert, highly fractured Steep-sided low ridges IE: and interbedded thin siliceous and round knobs. 1,; shale. E Womble Shale 1,500 Black shale, thin interbedded O lenses of linestone, and very thin sandstones. r affarfifiw—fie r q A WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN C7 include the Womble, Missouri Mountain, Polk Creek, and Stanley Shales. The shales have low permeability, but, locally, limestones in the Womble yield water to springs. Shales generally impede ground-water move- ment, except where open joints and fractures are pres- ent. Wells in shales generally yield meager quantities of water; recharge to shales is also small. The Bigfork Chert typically is highly permeable, ex- hibiting intergranular and fracture permeability. The Bigfork Chert is composed of silt-sized, generally poorly cemented siliceous particles in thin beds 1/2 to 4 in (13 to 130 mm) thick, which have been weathered, leaving a friable material, interbedded with layers of dense chert 4 to 12 in (130 to 450 mm) thick. The dense chert beds were rendered permeable by fracturing, which accompanied the intense folding of the beds, whereas the decalcified silt-sized material has significant intergranular permeability near the ground surface. Wells that yield the largest quantities of water in the region tap the Bigfork Chert. At Belvedere Country Club, northeast of Hot Springs, the Bigfork, tested by Albin (1965), was found to have a transmissivity of 2.67>< 103 ft/d (9.42><10—3 m/s). Many of the springs in the area issue from the Bigfork Chert, and many of the cold-spring emergences are controlled by contact of the Bigfork with adjacent, less permeable formations. This association of cold springs with the Bigfork Chert was noted by Purdue and Miser (1923). The Arkansas Novaculite is composed of three divisions—an upper and a lower division of novaculite, and a middle division of chert. Locally, the upper part of the formation is composed of silt-sized siliceous par- ticles and possesses intergranular permeability. The lower division is generally massive and dense, but is very closely fractured. The middle division is composed mostly of black shale and thin chert beds. The Arkan- sas Novaculite is not as permeable as the Bigfork Chert, but is locally intensely jointed. Some cold springs issue from, and many water wells tap, the Ar- kansas Novaculite. The Hot Springs Sandstone Member of the Stanley Shale is a massive, quartzitic sandstone. Fairly large joints and fractures, as in the novaculite, create some highly permeable conditions, such as at the hot springs. The hot springs emerge from the Hot Springs Sandstone Member near the axis, on the northwest limb, of a southwestward-plunging anticline. The springs emerge between the traces of two thrust faults that are parallel to the axis of the anticline (pl. 1). The locations of the hot springs, shown in figure 2, gen- erally lie along several northeast-trending lines. Ac- cording to Bryan (1924), these lineaments were in- ferred to be the traces of fissures by R. R. Stevens, who first noted their alinement in 1890. Jointing is common in the few exposures of the Hot Springs Sandstone Member in the hot-springs—discharge area. Thus, the hot springs are associated with thrust faults, and with normal faults and joints, on the plunging crestline of the anticline. Upward movement of the hot waters from depth is probably along the permeable fault zones. These fault zones probably carry water to near the surface, where the water follows permeable joints to the spring outlets. Geologic sections in plate 1 show the geologic struc- ture in the vicinity of the hot springs. CHARACTER OF THE SPRING AND WELL WATERS IN THE HOT-SPRINGS AREA PHYSICAL QUALITY FLOW OF THE HOT SPRINGS Flow of the hot springs has been measured at in- frequent intervals by several investigators. The first to attempt measurements was William Dunbar, of the Dunbar and Hunter expedition, in 1803, who measured only the largest springs. In 1860, Glasgow determined the hot-springs flow to be 450,480 gal/d (1.97 x 10‘2m3/s). Probably the first accurate measurement of the total spring flow was made by Weed (1902). Weed, after measuring or estimating the flow of each spring, found the total flow to be 850,000 gal/d (3.72X 10 ‘2m3/ s). The flow of all but a few of the hot springs is piped into a central collecting reservoir. Hamilton, in Hamil- ton and Blood (1931), calculated the maximum rate of filling of the central collecting reservoir to be 960,000 gal/d (4.21 x 10’2m3/s). Park Superintendent Libbey, in 1945, recorded that the central collecting reservoir filled with 15 ft (4.57 m) of water in 7 hours, which represents an average flow of 800,000 gal/d (3.50 X 10‘2m3/s). The volume of the reservoir per unit change in depth is known and provides a means for periodic calculation of spring flow. One large spring emerges at the bottom of the reservoir. The flow of this spring decreases as the depth of water in the reservoir increases. Also, at depths greater than 16 ft (4.88 m), overflow occurs. Flow calculations are thus made during the filling cycle at depths less than 16 ft (4.88 m). In addition, to avoid variable effect of depth in the reservoir on the flow of springs in the reservoir, the flow calculations are adjusted to a depth of 15 ft (4.57 m) by means of a depth-versus-flow rating curve. A hydrograph of the average of calculations of monthly flow since September 1970 is shown in figure 3. Spring discharge has ranged from 750,000 gal/d CS GEOHYDROLOGY OF GEOTHERMAL SYSTEMS (3.29X10‘2m3/s) to about 950,000 gal/d (4.16>< 10‘2m3/s). Fluctuations of spring discharge follow a seasonal cycle during the year; discharge is high in the winter and spring and is low in the summer and fall. One would expect the flow of an individual spring to fluc- tuate seasonally, as does the total spring flow. The flow of the lower display spring (number 32) is shown by hydrograph in figure 3. In addition to seasonal changes in flow, individual springs show long-term changes in flow. Some springs have ceased to flow and new springs have emerged dur- ing the last 170 years. These long-term changes in .. O O O Illlllll vvvvvvvv wrrvv -2 90°_ 4.0 X 10 300— 3.5 x 10-2 HOT SPRINGS FLOW, IN HUNDRED THOUSAND GALLONS PER DAY 700 L10 x 10-2 6.CVllIi[rvr 4—3.5 x 10’4 CUBIC METERS PER SECOND 5.0— — —3.0 x 10-4 SPRING, IN GALLONS PER MINUTE m.....1...........l........ii. —2.5 x 10-4 FLOW OF LOWER DISPLAY 62 — —/_/_/\ _ OUIPIW Spring 61— \ “ 60— — Fordyce Spring 59— — Upp-r 53— display spring 57...l.lllll.i ..... 1......UHlIlHH 55“"'l"" ..... minimunnwnw. TEMPERATURE, IN DEGREES CELSIUS 54— Lower display spring 52 lllllllllllll|llll|l|llllllJL 1970 1971 1972 1973 FIGURE 3.—Temperatures and flows of the hot springs. spring flow have not been systematically documented; many changes have doubtlessly passed, unrecorded. Measurements of Weed (1902), Hamilton (1932), and the present study afford samples for comparisons of long-term variations. Many changes in individual spring flow occurred be- tween 1901 and 1931 because of excavation and con- struction at spring outlets, opening new springs, and drilling the Fordyce well (fig. 2). Hamilton noted that the flow of springs in group 1, which includes those at higher elevations, declined in flow from 168,000 gal/d (736x 10‘3m3/s), in 1901, to 124,000 gal/d (5.43x 10‘3m3/s), in 1931 (table 2). Another group of springs, at lower elevations along the base of a tufa cliff, showed practically no change in flows from 1901 to 1931. These springs (numbers 2, 4, 6, 9, 11, 13, 15, 17, 19, and two unnumbered springs) discharged 313,500 gal/d (1.37>< 10 ‘2m2/s) in 1901 and 315,00 gal/d (1.38X10 ‘2m3/s) in 1931. Thirteen springs that were measured by Weed in 1901 were not included in the 1931 collection system described by Hamilton (1932). Two of these springs are on the Arkansas Rehabilitation Center, one of which, number 39, is used for supplying hot water to the Re- habilitation Center. Four springs measured in 1901 were not flowing in 1931. Locations of six springs were unknown in 1931. Presumably, these six springs were nonexistent or had insignificant flows in 1931. Hamilton lists five springs in his 1931 collection sys- tem that were nonexistent in 1901. Two of these (47 and 48) have declined in flows since 1931, and one (49) is now one of the larger springs in the system. The collection system has not changed since 1931, and no springs have been excavated nor have hot- water wells been drilled, resulting in a relatively sta- TABLE 2.—Flows of the hot springs in group 1 in 1901 and 1931 [From Hamilton, 1932] Flow, in gallons per day Spring number 1901 1931 11 __________________________ 28,800 9,600 3, 5, and 8 ____________________ 39,218 21,800 7 ____________________________ 18,516 1,760 10 __________________________ 18,514 14,400 22 __________________________ 1,723 2,460 23 and 24 ____________________ 10,800 5,000 2 __________________________ 25,847 10,950 27, 28, and 29 ________________ 24,418 (') 47 and 48 ______ Nonexistent 13,500 49 -___do _____ (‘) Maurice Spring - _---do ..... (‘) New ____do ..... 2,400 167,836 ~124,000 'Not measured individually. WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN 09 ble period of spring locations. However, changes in flow rates since 1931 have been noted. Flows of several springs were measured or estimated in 1972 after the springs were uncovered for sampling in January. These measurements showed a general de- cline of flow in the springs located at higher elevations. In 1972 spring number 1 had a very small or no flow—a decline from 28,800 gal/d (1.26X10 ‘3m3/s) in 1901 and 9,600 gal/d (4.21 X 10‘4m 3/s) in 1931. Spring numbers 47 and 48 declined from 13,500 gal/d (5.91X10‘4m3/s) in 1931 to 8,600 gal/d (3.77x 10 ‘4m3/s) in 1972. However, spring number 7 declined from 18,516 gal/d (8.11x10 ‘4m3/s) in 1901 to 1,760 gal/d (7.71X10 ‘5m3/s) in 1931, but rose to 2,800 gal/d (1.23X10 _4m3/S) in 1972. Spring number 49, which was nonexistent in 1901, had a flow of 58,000 gal/d (2.54x10 '3m3/s) in 1972. TEMPERATURES OF THE HOT-SPRINGS WATER Temperature measurements were the first scientific data collected at the hot springs. William Dunbar and George Hunter, in 1804, recorded 67.8°C (Celsius) (154.0°F) for the hottest spring (Weed, 1902). In 1860, the highest temperature measured by Owen (1860) was 644°C (147 .9°F).Glascow (1860) recorded a maximum of 656°C (150.1°F). The maximum temperature mea- sured by the Geological Survey, in 1972, was 61.8°C (143.2°F). Measurements of temperatures of individual springs by several investigators from 1890 to 1953 (table 3) show maximum temperatures of 639°C (147.0°F) in 1901, 644°C (147.9°F) in 1931, and 633°C (145.9°F) in 1952. Thirteen of the same hot springs were measured by Haywood (1902), Hamilton (1932), and Kuroda (1953). The average temperatures of these hot springs when measured in 1901, 1931, and 1952 were 582°C (136.8°F), 57.3°C (135.1°F), and 589°C (138.0°F), re- spectively. These particular data indicate that there has been a slight decline in maximum water temperatures with time (06°F in 10 years). These data are not conclusive evidence of a general decline in temperatures, because of differences in samplings points, variations in tem- peratures with flow rates of individual springs, and temperature fluctuations in spring flows due to air temperature. Temperature fluctuations of individual springs ex- hibit an annual cycle that seems to be in response to the annual cycle in air temperature. Graphs of the temperatures of four springs are shown in figure 3. For some springs, short-term fluctuations corresponding to the annual cycle are evident from detailed temperature records. The short-time fluctuations are due to external environmental factors. The Fordyce Spring (Spring No. 46, fig. 2) temperature, for example, responds to such factors as heat-load changes, caused by opening the air-tight seals on the chamber enclosing the spring, and effects of mixing of the hot-springs water with seepage from nearby rainfall. It would be expected that spring temperature in- creases with increase in spring flow. This relation has been examined by scatter diagrams drawn between flow and temperature for the display springs. A direct relationship is obscured because the effect of ambient air temperature varies seasonally and the high sea- sonal air temperature occurs during the period of low spring flow. SILICA CONCENTRATION AS A TEMPERATURE INDICATOR The solubilities of silica minerals increase with in- crease in temperature. Silica minerals dissolve until the solution is saturated many times faster than they precipitate from an oversaturated solution. Thus, a thermal water will dissolve silica minerals as its tem- perature rises, but as the water cools it will not rapidly lose silica. The silica concentration of water can be TABLE 3,—Temperatures, in degrees Celsius, of the hot springs Date of measurement Spfing 1890 1900 1901 1931 1952 1972 (from (from (from (from (from (from Branner, Haywood, Haywood, Hamilton, Kuroda, (present 1892) 1902) 1902) 1932) 1953) study) m9 m1 M9 $3 an $3 __________ $0 $6 moAmmw a . .NW@@N98P8 pgmouwmmugmww mmmm app» kmmmommm“babbabaemowubhmwmwwmb""' HQNFWwarHSerN°N@.P9<¢.WN@@P< $3 uammmh @WflwNP moummm ,c. mp F??? meow C10 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS used as a measure of the maximum temperature reached by the water (Fournier and Rowe, 1966). The actual silica concentration of water depends on the particular silica mineral to which the water is ex- posed, as well as to the temperature and the rate of dissolution. In the hot-springs system, the main sources of silica are the Bigfork Chert and the Arkan- sas Novaculite. Chert and novaculite are both com- posed of chalcedony, a cryptocrystalline quartz, and microcrystalline quartz. The solubilities of chalcedony and quartz (R. O. Fournier, oral commun., 1972) are shown in figure 4. The analyzed silica concentrations of the samples collected for this study are also shown in figure 4. The cold springs and wells are oversaturated with respect to quartz, but are undersaturated with respect to chal- cedony. Though it is likely that complex silica-bearing clays or other minerals influence the silica concentra- TEMPERATURE, IN DEGREES CELSIUS 70 0 15 30 45 60 80 I I I I I 50—- Hot Springs 40— 30-— 20— Gulpha Gorge Spr. ”(Cluster Spr. Diamond Mineral Spr. McLendon Mineral Spr. _ Whittington Park Whittington Avenuex >(Echo Valley Elizabeth SP" Brown Happv Hollow Spr. Frank Thompsonx Sleepy Valley X Spr. 10 7.... Si02 CONCENTRATION, IN MILLIGRAMS PER LITER X Belvedere Country Iub I I I 3.7 3.5 3.3 3.1 2.9 LOGO/TEMPERATURE, IN KELVINS FIGURE 4.—Variation of dissolved-silica concentration with tem- perature. tion of these waters, it is also possible that, because of their low temperatures, these waters have not had suf- ficient time to reach saturation with the chalcedony. The hot-springs waters have a uniform silica concen- tration (41—42 mg/L). They are oversaturated with re- spect to chalcedony at their temperature when col- lected, but the higher temperature springs approach saturation. The constant silica concentration of the springs, together with the similarities in their other chemical properties, strongly suggests that all the hot springs emanate from a single source having a uniform geothermal environment. From the silica concentra- tion (41.5: 0.5 mg/L), the temperature of the source can be calculated as 63.2: 0.5°C (145.8: 09°F). The differences in temperatures of the individual springs are presumably a result of differing flow paths and rates of approach to the surface from their common source. Silica analyses were made on groups of samples col- lected in 1901, 1952, and for the present study in 1972. These analyses provide a firmer base from which to judge temperature trends with time. The silica analyses of springs in a given group show less varia- tion than overall spring temperatures. As an indicator of maximum temperature, the silica-computed tem- perature is not affected by external factors such as air temperature and flow of the springs. Silica concentrations and maximum temperatures calculated from the silica concentrations (Fournier and Rowe, 1966), as well as maximum temperatures re- corded at various times, are shown in the following table. Maximum Average Number Temperature Year temperature SiOz of calculated measured concentration springs from SiO2 (°C) (mg/ L) sampled concentration 1804 __________ 67.8 -___ __ _-__ 1860 __________ 64.4 -___ __ ____ 1860 _____ 65.6 ____ __ ____ 1901 _____ 63.9 46.6 40 68.5 1931 _____ _ 64.4 _-__ __ ___- 1952 __________ 63.3 42.7 11 64.3 1972 __________ 61.8 41.7 9 63.2 These results are shown graphically in figure 5. Ac- cording to the silica data, the spring-source tempera- ture has been decreasing at an average rate of about 0.077°C (0.14°F) per year since 1901, although data from 1804 to 1931 indicate a lower rate of decline. Also, measured maximum temperatures from 1931 to 1972—a period in which the springs were undisturbed—have decreased at about the same rate (0.07 3°C, or 0.13°F per year) as the spring-source tem- perature decline from 1931 to 1972. This coincidence may be a fortuitous circumstance of sample timing and distribution. In addition to reflecting the change in the ‘77. V 9 9 )e 4' WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN 7G I I I I I I I I I I I I I I I 68—- 66— X X EXPLANATION _ X Maximum measured temperature 64'— 0 Temperature calculated from average Sioz concentration III]. 1830 1850 62— TEMPERATUFIE, IN DEGREES CELSIUS l . I I I I I I 1890 1910 1930 1950 YEAR 1 | 1310 1870 1970 FIGURE 5,—Maximum measured and calculated temperatures of the hot springs. source temperature, variations in measured tempera- tures reflect the sampling distribution, seasonal changes in temperature, and changes in temperature because of changes in flow distribution that are due to natural changes in flow distribution or due to man’s efforts to enclose and control the springs. Thus, al- though the temperature observations may be subject to sampling bias, the silica-concentration trend of the hot springs indicates a decline in temperature. TEMPERATURES OF THE COLD SPRINGS AND WELLS The temperatures of ground water in wells in the vicinity of Hot Springs, other than the hot springs themselves, range from 120°C (53.6°F) to 52.8°C (127.0°F) (tables 4, 5; fig. 6). The warmer of these tem- peratures is the temperature of ground water in the immediate vicinity of the hot springs. The well on the grounds of the Arkansas Rehabilitation Center has a temperature of 52.8°C (127.0°F); the water from a well on the Arlington Hotel lawn, used to supply water to the cooling plant, has a temperature of 322°C (90.0°F; the temperature of the water from the well at the Ar- lington Hotel is reported to be 228°C (73.0°F) ). The high temperatures are associated with the abnormally high geothermal gradient caused by heat conveyed to the surface by the waters of the hot springs. The temperatures of ground waters in and near areas considered as potential for recharge to the hot- springs system (outcrop area of Bigfork Chert and Ar- kansas Novaculite) range from 15.0°C (59.0°F) to 26.8°C (80.2°F). These temperatures, considered in re- lation to the average annual air temperature of 17.7°C (63.9°F), indicate that some of the waters in and near the recharge area are heated by geothermal processes. Ground—water temperatures, in the range of 15.0°C (59.0°F) to 16.7°C (62.1°F), indicate a minimum of geothermal heating and reflect the fact that most re- charge occurs during the cooler seasons of the year. These temperatures indicate that there is a rapid rate C11 of movement of water downward from the surface and proximity to the recharge-source area. Low ground-water temperatures, lower than the av- erage air temperature, occur in the area northeast of the springs and in the outcrop of the Bigfork Chert and the Arkansas Novaculite. Water temperatures from TABLE 4,—Tbmperatures of the cold springs Spring Tempoecrature Date No. Name of Geologic formation (fig. 6) measurement Sl Ozark Lithia ___- 19.6 6—29—72 Bigfork Chert. 82 J. L. Bryant (owner) ________ 15.6 6—29—72 Do. S3 Arbordale ________ 26.8 6— 7—72 Do. S4 Burton Sargo (owner) ________ 17.8 9—12—72 Do. S5 Echo Valley ______ 20.5 1—27—72 Do. S6 Ar-Scenic ___ _-_ 20.7 11—2—71 Do. S7 Cluster __________ 20.8 1—26— 72 Arkansas Novaculite. SS Happy Hollow ___- 17.5 1—24—72 Do. S9 Music Mountain __ 20.9 9—27—72 Stanle Shale. S10 Sleepy Valley _-__ 12.0 1—28—72 Bigfor Chert. Sll McLendon ______ 18.6 9—27—72 Hot Springs Sandstone Member of the Stanley Shale. TABLE 5.—7kmperatures of water in the wells [Thermal gradient = (temp. ('C)-15)/depth of well (ft)] Date Depth Temper- Thermal Well No. of of . . (fig. 6) 3:59 measure- well Egg/lg)“ £690le ment (fl) ormation W1 ___- 15.6 6—29—72 26 0.023 Bigfork Chert. W2 ___- 17.8 9—13—72 30 .093 0. W3 __-_ 16.4 6— 6—72 61 .023 Do. W4 ___- 16.1 7—14—72 170 .065 Do. W5 __-_ 16.9 7—14—72 92 .021 D0. W6 ____ 16.7 7—14—72 30 .057 Do. W7 ___- 15.6 6— 6—72 89 .0067 D0. W8 ___- 16.1 6- 6—72 93 .012 Do. W9 ___- 15.3 6— 7—72 42 .0071 D0. W10 ____ 16.7 6— 7—72 20 .085 Do. W11 __-_ 16.7 6— 6—72 101 .017 Do. W12 ___- 17.4 9—25—72 44 .054 Do. W13 ___- 15.3 9—12—72 67 .0045 Do. W14 ___- 15.6 6— 7—72 126 .0048 Do. W15 ___- 17.6 9—25—72 140 .019 D0. W16 ___- 15.6 6— 2—72 --__ ________ Do. W17 ___- 15.0 6— 2—72 38 .0 Do. W18 ___- 15.8 6— 6—72 78 .010 Do. W19 ___- 16.1 6— 2—72 235 .0047 Do. W20 ___- 16.1 6— 2—72 263 .0042 Arkansas Novaculite. W21 ____ 20.4 9—27—72 120 .045 0. W22 ___- 16.7 11—2—71 90 .019 Bigfork Chert. W23 __-_ 16.1 11—2—71 46 .024 0. W24 __-_ 18.8 1—22—72 300 .063 D0. W25 -___ 18.6 1—28—72 89 .040 Do. W26 ___- 18.0 1—24—72 28 .11 D0. W27 ___- 22.8 3— 8—70 202 .039 Hot Springs Sandstone Member of the Stanley Shale. W28 ___- 52.8 8~ 8—72 336 .11 Do. W29 __-_ 32.2 3— 4—70 200 .086 Do. W30 ___- 18.0 9—22—72 300 .010 Arkansas Novaculite. W31 ___- 16.7 9—22—72 140 .012 Bigfork Chert. C12 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ’ Y” «A r y l y, F r F l EXPLANATION V l & Well numhu Well 0 liLT-mpuouu- ('C) 263 omn L sfifi'ifiq On 237 32223.2.“33’I-u ‘ ‘ W Discharge and a! ho! iprinos ‘ ‘ Oulcroa sun of Ariana-t Nanci-Ii" ). ‘ Oman and o! Billork chon R (alto includes Missouri Mountain, Polk cruk, and Womble Shel") 9,.in..:1.:;:..T........, * 1 l A Gaoino nation 1’ A Partial-record qaqlnq station ‘ 93' 07'30" Base from U.S, Geological Survey Hov Springs Norlh l224,000, I966 Hol Springs Sou"! |124,000, [966 Lonsdale SW uncalled advance pun and Malvern l:62,500, I945 55111145 0” v l,24,000, I972 FIGURE 6.—Locations of _ ‘ WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN 013 RISW RIUW IIINf'I In—-C O I MILES O IKILOIETER the cold springs and wells. C14 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS several wells in the area are higher than the average air temperature. The higher temperatures indicate geothermal heating due to a long residence time of the water in the aquifer. Of the five wells sampled having temperatures higher than the average air tempera- ture, four are flowing artesian wells. The cold-water springs in the area are generally warmer than the well waters. The spring temperatures range from 15.6°C (60.1°F) to 268°C (80.2°F). The de- signation of “cold-water” to these springs is for the purpose of distinguishing them from the hot springs of Hot Springs National Park. The warmer springs in this cold-water group could be correctly classified as ther- mal springs. CHEMICAL QUALITY The ground water of the hot springs contains a vari- ety of chemical species in solution. Knowledge of the chemical types and concentrations is useful in inter- preting patterns of ground-water movement, source of the water, and in determining whether there has been a change in the chemical concentration of the water through the years. Chemical analyses may also hold clues to the longevity of the hot springs. There has been little change in the chemical quality of the hot-springs waters during the period of record from 1890 to 1970. A small change in the silica concen- tration is indicated, and its significance has been dis- cussed. Minor variations in other constituents are con- sistent with normal variations in most ground water. As part of this study, samples for chemical and isotope analyses were collected in 1972 from 15 cold wells and springs in the Hot Springs region and from 9 hot springs. Table 6 gives the results of the chemical analyses made for this study. The concentration of cer- tain dissolved constituents may change between the time of field collection and the time of laboratory analysis owing to loss of gases, temperature changes, and precipitation of solids. To insure that the analyses would represent the natural chemistry of the water as closely as possible, certain analyses were made in the field, and parts of the samples were treated before being sent to the laboratory, to prevent changes before analysis. Temperature was measured at the collection site using thermometers readable to 01°C (0.18°F). Alka- linity, pH, and dissolved oxygen (DO) were also mea- sured as part of the sample collection procedure because they are liable to change by exchange with atmospheric C02 or 02 gas when a water sample is shipped or stored. The precautions mentioned by Barnes (1964) were observed in the pH and alkalinity measurements. The pH values are considered accurate to :0.02 units. The dissolved oxygen analytical proce- dure was that described by Brown, Skougstad, and Fishman (1970, p. 126). A separate bottle of each sam- TABLE 6.—-Chemical analyses of water [Results in milligrams per 9 E z E l- A a 2 8 w E :5 E 8 a E Well or K 3‘ v an o E spring Date of E (ii E 5 E E v .2 number' Name of well or spring collec- 3 v .E g E N E 3 (fig. 6) tion 8 S E ~ ‘3 '5 E. 3 _— = S 3 5: 7.5 a Q --t "' O 0 0-1 3 m < :- E—~ N o 2 Happy Hollow Sprin 17.5 8.4 0.00 0.00 0.00 0.03 0.2 0.3 Belvedere Country ub well ____ 9—25—72 16.8 6.9 .05 .98 .02 .04 .1 .1 Frank Thompson’s house well 9—24—72 16.8 7.5 .04 .02 .00 .02 .2 .1 Sleep Valley Spring 1—28—72 12.0 11 .40 2.1 .13 .10 3.7 10 Bill argo’s wel .V- 9—24—72 18.1 8.1 .20 2.0 . .03 .02 .8 .5 R. B. Yates’ well _____________ 9—25—72 21.2 7.9 .04 .04 .10 .06 5.9 1.5 Music Mountain Spring _____ 9—27—72 20.9 2.6 .08 .04 .00 .00 11 1.4 Cluster Spring _____________ 9426—72 20.8 13 .00 .66 .25 .01 42 2.5 McLendon Mineral Spring . 9-27-72 18.6 11 .00 1.0 .75 .06 40 1.4 Gulpha Gorge Well ,,,,,,, 9—27—72 20.4 13 .00 1.2 13 .02 46 2.8 Eliza th Brown well iiiii 9—24—72 17.4 8.7 .00 2.1 33 .17 55 1.9 Whittington Avenue S ring 1-24—72 18.0 9.4 .0 .79 08 .10 50 2.3 Whittin n Park wel iiiiiii 1~28—72 18.6 11 .0 1.6 11 .12 63 3.4 Echo Va ley Sprin ,,,,,,,,, 1~27—72 20.5 9.7 .0 1.3 15 .08 67 2.9 Diamond Mineral %pring ,,,,, 1-22-72 18.8 12 .0 .37 12 .06 66 3.6 Maurice Hot Spring ,,,,,,,,, 1—20—72 53.3 42 .0 .00 10 .05 45 4.8 Hot Spring No. 17 ,1 iiiiii 1—25—72 55.4 41 .0 .00 00 .05 44 4.6 Hot Spring No. 23 ii 1v26~72 56.2 41 .0 .00 09 .03 44 4.6 Hot Spring No. 33 A 1—26—72 57.6 42 .0 .04 27 .08 45 4.8 Hot Spring No. 46 H 1~18-72 58.3 42 .0 .33 25 .07 45 4.8 Hot Sprin No. 48 H 1—25—72 60.0 42 .0 .02 18 .06 45 4.7 80,000-ga on reservoir ,,,,,,, 1—27—72 61.0 42 .0 .00 20 .04 45 4.8 Hot Spring No. 42 eeeeeeeeeee eve 1—19—72 61.3 42 .0 .01 23 .02 45 4.8 Hot Spring No. 49 eeeeeeeeeeeeeeeeeeeeee 1—21—72 61.8 41 .0 .06 25 .06 44 4.8 'U.S. Geological Survey station numbers of wells and springs are given in table 7. 2Laboratory analysis. aContains trace of hydrogen sulfide (HIS). Nora—Differences in implied accuracy of analyses for aluminum due to differences in types of equipment used. Y "is. i A ‘ S L11: 4 '1...A A him WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN ple was acidified at the time of collection and taken to the laboratory for analyses for aluminum (A1), iron (Fe), manganese (Mn), zinc (Zn), calcium (Ca), mag- nesium (Mg), and strontium (Sr). Analyses for these seven ions, as well as for sodium (Na) and potassium (K), were made by atomic absorption spectroscopy. All determinations were made by using methods standard in US. Geological Survey laboratories (Brown and others, 1970). The analyses of aluminum for samples collected in September 1972 were made from a larger aliquot than was used for the analyses of the samples collected earlier. Thus the precision reported in table 6 C15 chemical characteristics, particularly its silica and chloride concentrations, are unlike those of waters in the hot-springs-flow system. The R. B. Yates’ well, al- though in the geographic and geologic area of interest, is very shallow, and the water has a high concentration of nitrate (NO3=8.3 mg/L) and chloride (Cl=6.4 mg/L). This chemistry and the location of the well suggest that the water is influenced by man’s activity and does TABLE 7.—-Station numbers of wells and springs given in table 6 Station number Name of well or spring ' ' 343110093025301 ________ Happy Hollow Spring IS greater for the samples collectedln September 1972. 343252091002301 ________ Belvedere Country Club well The prec1510n of all analyses is estimated to be plus or 343204093005501 ________ Frank Thompson’s. house wen minus one-half of the last reported digit. That is, a 323333333358; ———————— 1831113 agglgevngmng reported 2.1 mg/L implies a concentration of between 343327093000401 III: R. B: Yates’ well 2.05 and 2.15 mg/L. A reported 0.00 mg/L means that giggg‘éggggggégi ———————— 6413222353321” Spring analysis was made for that constituent but the con- 343105092572001 IIII McLendon Mineral Spring - ‘ - 343130093020801 ________ Gulpha Gorge well stltuent was not found In concentrations greater than 343347092594201 ________ Elizabeth Brown well 0.005 mg / L. 33353093040802 ________ Whittington Avefiiue 1Slpring - - - 5 09 040601 ________ Whitti n Par we The chem1cal.analyses gIven In table 6 are arranged 343231093012801 ________ Echo V ley Sprin (Big Chalybeate) approx1mately in order of increasmg total dlssolved- 343112093033601 ________ Diamond Mineral pring (Lithox) solids concentration. US. Geological Survey station 343051093031101 -------- Mam“. H°t Spring . . . . 343057093031301 ________ Hot Spring No. 17 numbers of the sampling pomts are given in table 7. 343023833831101 ________ got gpring 1:10. 23 U l ' ' ‘ 430 1 0901 ________ ot pring o. 33 ( Pper Disp ay) Two, or poss1bly three, of the analyses g1ven in table 6 343049093031101 ________ Hot Spring No. 46 are not of part1cu1ar use in descr1b1ng the chem1cal 343058093030803 ________ Hot Sprin No. 48 ' _ ' _ ' 343055093031201 ________ 80,000-ga on reservoir h1story of the .hot springs area ground waters. Muslc 343047093031001 ________ Hot Spring No. 42 (Health Services) Mountain Spring 1ssues from Stanley Shale, and its 343035093031001 ________ Hot Spring No. 49 from the hot springs, cold springs, and wells liter, except as indicated] Dissolved solids A g 2 . 8 . 3 1 i a? 1 ‘ 1 1 E 3 1 § §. 5% :2 §8 "—3 5. a .1 x m w A. 2 ,1 H J: I- wv-i e .5 m 5 3‘ v a v a: 3 8 o m 0 g- r: E a E = E 1. ° E z 5 3 3 1 Z I 5 ° “ ° " 8 8 o s g ,2 I .._ a A a: -u "= .2 ° = u. 3 o \O 'E > u E 8 O .82 3 T: 'C 3 g‘ 5 =3 O E 5a.: :3 "‘ E .2 3 E 5% £3 2 g 2 3g an E E e g; g g E E 5 § :2 5 a“ 5 6 a 2 o 5 a“! 5 .5; a. r: 0.00 1.3 0.2 ____ 21 1.4 2.3 0.1 0.0 0.03 11.- 14 11 22 4.58 6.3 .00 1.0 ,2 0.04 5 1.6 1.4 .0 .3 .04 0.08 12 15 15 5.20 .6 .00 .7 .3 .04 20 1.6 .9 .0 .2 .00 .04 12 18 36 4.70 10.2 .01 1.3 ,8 1111 22 14 2.1 ,2 .1 .35 _1_ 36 41 47 24.82 3.3 .00 2.8 1.4 .00 20 7.4 3.6 ,1 1.7 .06 .10 26 36 44 4.62 1.8 .00 5.4 1.4 .02 9 10 6.4 ,1 8.3 .02 .01 51 54 77 5.32 4.8 .01 2.4 1,0 _1_ 229 8.6 4.7 .1 .7 .01 1111 47 60 104 27.12 ___1 .10 4.6 1,7 .08 147 11 2.7 .2 .1 .12 04 146 149 219 6.72 .0 .07 1.7 1.0 .06 145 7,2 1.6 ,2 .1 .39 .01 131 141 232 7.15 .0 .08 2.9 1.5 .02 157 10 2.3 .2 .1 .07 .08 152 165 247 7.10 .0 .24 1.6 .7 .06 183 7.0 2.0 .1 .1 .31 05 164 173 274 6.92 .0 .20 1.5 .9 1111 157 14 2.1 .2 .1 .12 1111 157 164 276 6.69 .0 .26 1.6 1.4 1 _ 227 9.8 1.9 .2 .1 .00 1_ 1 193 200 331 7.6 (3) .11 1.3 .6 1 1 219 7.2 2.0 .2 .1 .06 .1. 196 202 339 7.25 .0 .11 1.9 1.8 1 1 211 12 2.0 .3 .1 .00 _1_ 204 212 354 7.08 .0 .11 4.0 1.5 1 1 156 9.0 1.9 .2 .0 .04 _1_ 189 191 269 7.03 2.0 .11 3.9 1.5 1 2160 7.8 1.8 .2 .0 .00 1-- 184 187 266 27.70 3.6 .11 3.9 1.5 159 8.2 1.9 .2 .2 .09 _1_ 185 188 269 7.52 3.9 .11 4.0 1.5 164 8.2 1.9 .2 ,0 .02 __1 188 193 269 7.13 1.1 .11 4,0 1.5 164 7.8 1.9 .2 ,0 .04 _1 1 187 195 269 7.01 .6 .12 4.0 1.5 165 8.6 1.9 .2 .1 .00 1-11 189 196 276 7.27 2.4 .12 4.0 1.5 165 8.0 1.8 .2 .0 .00 11.- 188 191 275 7.36 3.3 .11 4.0 1.5 159 8.6 1.9 .2 ,0 .04 11 1 188 191 272 6.93 ,0 .11 3.8 1,5 111 155 8,2 1.9 .2 .0 .06 11 1 184 191 268 6.95 .4 C16 not represent the natural conditions of interest in this report. The water from Bill Sargo’s well, too, has slightly high chloride and nitrate concentrations (3.6 and 1.7 mg/ L, respectively) and may not be completely representative of natural conditions. The water from the hot springs is distinctive chiefly by its relatively low mineral concentration (table 6). The mineral concentration of the water probably is low because the rocks associated with the hot springs are made up of only a few minerals, each of which has a low solubility. Most ground water in Arkansas contains from two to to three times more dissolved minerals than the hot-springs water. GEOCHEMISTRY OF HYDROGEN AND OXYGEN ISOTOPES The elements hydrogen and oxygen have several naturally occurring isotopes. Hydrogen has stable isotopes of mass 1, common hydrogen or protium (H or 1H), and of mass 2, deuterium (D or 2H). A radioactive isotope of hydrogen, tritium (3H), also occurs in the environment and is discussed below. Common oxygen has a mass of 16 (160), but oxygen of mass 17 (170) and mass 18 (‘80) also occur. As absolute isotopic abundances or ratios are dif- ficult to analyze with precision, it is customary to mea- sure and express isotopic variations as deviations from an arbitrary standard. These deviations are expressed in delta notation where R sample > 8: ———1 X 1,000. R standard Here, R is the isotopic ratio (D/H; 180/160), and 6 is in parts per thousand, or per mil (0/00) The delta notation is less difficult to use than it seems. The standard to which natural-water isotopic mea- surements are referred is Standard Mean Ocean Water (SMOW) (Craig, 1961a). The concentrations of the var- ious isotopic molecular species in this standard are H2130 g 2,000 mg/L, H2170 g 420 mg/L, and HDIGO g 316 mg/L. H2160 makes up the remainder, and all other species total less than 1 mg/ L. Knowledge of the 170 concentration of a sample provides no more infor- mation than knowledge of the 180 concentration; there- fore, the 170 concentration is generally not reported. The stable isotopic chemistry of water is thus ex- pressed in terms of 8D and 8180. Natural waters are generally depleted in D and 180 relative to SMOW—that is, their 6D and 6180 values are negative. This depletion occurs because the vapor pressure of water molecules containing the heavier isotopes is slightly less than that of common water, H2160. During evaporation and condensation in the hydrologic cycle, molecules containing heavier isotopes are concentrated in the liquid phase. As water evapo- rates from the ocean, the vapor is depleted in D and 18O GEOHYDROLOGY OF GEOTHERMAL SYSTEMS and the amount of depletion becomes greater as the temperature of evaporation decreases. Further isotopic fractionation takes place as water is condensed and reevaporated during atmospheric transport, and the amount of fractionation is inversely proportional to temperature. The D and 180 concentration of meteoric water— that is, water of recent atmospheric origin—varies regularly over the land surface of the Earth. A great number of measurements show that for meteoric water not subject to much evaporation, 6D and 8130 are re- lated by the expression by Craig (1961b): 6D=88180+ 10. (1) The amount of depletion also increases with altitude, latitude, and distance from the ocean. The general pat- tern of isotopic distribution in North America has been mapped (Sheppard and others, 1969). Figure 7 shows the variation of 6D and 8180 from several hydrothermal localities and the trend line for meteoric waters, equation 1. The meteoric waters from the localities shown illustrate the tendency for increas- ing depletion from the oceanic isotopic composition with latitude, altitude, and distance. Figure 7 illustrates that the 180 content of many thermal waters is enriched relative to normal meteoric waters of the same D content. An explanation for this shift might be that the hot waters represent mixing between meteoric water and water emanating from hot or molten rock at great depths within the earth—that is, magmatic water, the isotopic composition of which is not precisely known but which may well be in the range shown in the figure. Were this explanation cor- rect, there should be a shift in 5D, as well as in 8130 toward whatever isotopic composition may be typical of magmatic water. As is particularly well illustrated by the Nevada and Yellowstone Park series, such a deuterium shift does not occur. Because a deuterium shift is absent and because of the existence of another process to explain the 180 shift, magmatic waters are probably not significant contributors to hydrothermal systems. The process bringing about the 180 shift is probably an isotopic-exchange reaction between the water and rock in a system. Such a reaction for a silica mineral can be written as Sil“01"O+H216O-->Si16 02+H2130. Because 180 in most igneous minerals is enriched rela- tive to SMOW, the effect of this reaction is to drive the 8180 in exchanging waters toward positive values. Also, because of the low hydrogen concentration of most of the rock, the SD of waters is relatively un- changed. Isotopic-exchange reactions, as in all chemical reac- tions, proceed faster at high temperatures, and their WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN effects are generally detected only in systems in which the water temperature approaches or is higher than 100°C (212°F). Because of the utility of stable isotopes in pervious interpretative work in such systems, D and 180 measurements were made on several samples from the hot-springs system. Isotopic analytical results are given in table 8 and shown in figure 7. Deuterium in the cold waters has an average 6D: —228.6"/00, whereas the average of the hot springs is insignificantly different at —28.3"/00. These values agree with previous estimates (Sheppard and others, 1969) that deuterium in meteoric water in the central Arkansas region should be within a few per mil of —30. The coincidence of values for the hot- and cold-water sampling points is strong evidence that there is no detectable source of meteoric water contrib- uted to the hot springs from outside the region. The average 6180 of the hot springs is —5.2"/00, and the average of the cold springs is —5.1"/00. The lack of oxygen shift shows that there is no oxygen-isotope ex- change between the hot springs and oxygen-bearing minerals in the aquifer. This lack of oxygen-isotope exchange could be true only if (1) the maximum tem- perature within the hot-springs system is not much higher than the springs’ surface temperatures and (or) (2) the flow rate of the waters through the hottest part C17 of the system is relatively high, so that there is little time for exchange. In summary, the deuterium and 180 data suggest that the concept given elsewhere in this report con- cerning the hydrology of the Arkansas hot springs is reasonable and that (1) virtually all the water in the hot springs is of local, meteoric origin—any theories implying long-distance (that is, many tens or hundreds of miles) water movement, or the presence of juvenile or magmatic water in the springs, must be disregarded—and that (2) the maximum temperature reached by most of the hot-springs water cannot be many tens of degrees higher than the springs’ surface temperature; also (3), the resident time of the hot- springs waters in the heated part of the system is a relatively short time, that is, not more than perhaps a few hundred years. The silica concentration of the hot- springs water, discussed in a previous section, and the 140 concentration, discussed in the following, refine these conclusions further. The isotope of hydrogen is tritium (H3 or T), which has an atomic mass of 3 and is radioactive, with a half- life of about 12% years. Tritium is formed continuously by cosmic rays impinging on the upper atmosphere. This natural tritium is present in precipitation at levels of from 1 to 10 T atoms per 1018 H atoms. I l I I l l 0— 9 Ocean ._ (SMOW) <9 / 4 / “(V 0 Cold and hot springs, Arkansas Q) Larderello, Italy _l 90 The Geysers, California E '50 _ e‘ Hekla, Iceland _ a Q)” LIJ o- Q’Qo / P bl GD 0 ossi e and 3 V‘ . = - 5'30 range of g o O/L‘ Mount Lassen, Californla magmatic waters U) —lO — _. :3 ° 9‘“ i = «Q‘ / Cold Thermal Hasnamboat Springs, Nevada Trend lines of selected / hydrothermal systems /X-\. Upper basin, Yellowstone Park, Wyoming -l50— _ l I l I | l -20 -|5 -IO —5 0 +5 +l0 +l5 18 a Osmow’ PER MIL FIGURE 7 .—Comparison of isotopic composition of waters from the cold and hot springs of Arkansas and of hydrothermal waters elsewhere (after White and others, 1973, fig. 1). C18 Tritium is measured by analyzing its rate of radioac- tive decay in a water sample, and the results are ex- pressed as tritium units (Tu), one of which is equal to a T/H ratio of 1018. Thus, natural tritium is present in the range of from 1 to 10 Tu. Large quantities of tritium are produced by nuclear devices, and the atmospheric testing of such devices— particularly fusion devices (hydrogen bombs)—from the early 1950’s through 1962 raised the level of tritium in precipitation to many times its natural level of from 1 to 10 Tu. Peak tritium levels occurred in the spring of 1963, when precipitation at St. Louis, for example, reached levels of more than 2,500 Tu. Since then, tritium levels have been decreasing at about 30 percent per year. During the 1950’s, tritium levels were in the range of several hundred tritium units, also well above natural levels. The tritium concentrations of samples collected from the hot springs are given in table 8. Results of radioisotope measurements of tritium and also 14C, that are discussed below, are expressed with a statistical-error term corresponding to one standard deviation (10-). There is a probability of two in three that the true value of a quantity is within the 10- range. For Diamond Mineral Spring (table 8), with tritium shown as 1.5:0.4 Tu, there is a 67-percent pro- bability that the true tritium concentration is between TABLE 8.—Hydrogen- and oxygen-isotope concentration of waters in the hot-springs area Well or spring Date of 8D 6”“0 No. ' Name collection SMOW SMOW Tritium (fig. 6) in 1972 0/00 0/00 (TUtla SS __ Happy Hollow Spring __-_ Jan. 24 -30 —5.8 1.1:0.6 W19 Belvedere Country Club Well ________________ Sept. 25 -23 (2) 41.8:20 W20 Frank Thompson’s house well ________________ Sept. 24 -28 (2) 85.3:16 SlO __ Sleepy Valley Spring ______ Sept. 28 -30 -5.5 34.1t2.1 W17 Bill Sargo’s well __________ Sept. 24 -30 (2) 91814.7 W16 R. B. Yates’ well __________ Sept. 25 -28 (2) 27.7:16 SH 2. McLendon Mineral Spring Sept. 27 ~29 (2) 1.5 :08 W21 Gulpha Gorge well ________ Sept. 27 -30 (2) 13:04 S7 4_ Cluster Spring ____________ Sept. 26 -30 (2) .7106 W26 Whittington Avenue Spring Jan. 24 -26 -4.1 28:05 W25 Whittin n Park well ____ Jan. 28 -29 -5.3 12:08 SS __ Echo Va ley Spring ________ Jan. 27 -30 -5.1 3.1 :06 W12 Elizabeth Brown well ______ Sept. 24 -29 (2) .9:0.5 W24 Diamond Mineral Spring __ Jan. 22 -29 -4.9 15:04 Maurice Hot Spring ______ Jan. 20 -27 -3.9 12:03 Hot Spring N0. 17 __,, _ Jan. 25 -30 -5 4 30:05 Hot Spring No. 23 ,1 _ Jan. 26 -28 __ 7:0.8 Hot Spring No. 33 A- A Jan. 26 -28 -5 4 9:0.6 Hot Spring No. 46 ________ Jan. 18 -29 —5.5 31:06 Hot Spring No. 48 ________ Jan. 25 -28 -5.7 .7:0.4 80,000-gallon reservoir ____ Jan. 27 -29 -4.9 1.0i0.7 Hot Spring No. 42 ........ Jan. 19 -28 -5.4 2.7103 Hot Spring No. 49 ........ Jan. 21 -28 -5.6 9:04 ‘U.S, Geological Survey numbers of wells table 7. 2Not analyzed. and springs are given in GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 1.1 and 1.9 Tu, and a 95-percent (20') chance that it is between 0.7 and 2.3 Tu. Of the nine hot springs, four have tritium levels that are so low that they may, within a 95-percent confi- dence limit, contain no tritium; that is, the 20- range of their analyses includes zero. The rest of the hot springs contain tritium, but not more than 3 Tu. A sample that contained 3 Tu could be water that fell as rain in the 1940’s that had an initial tritium concentration of about 10 Tu. This explanation is unlikely, though, be- cause evidence discussed below suggests that the hot- springs waters have ages of thousands of years, and there is another explanation for the 3-Tu level which also accounts for the different tritium concentrations of the several springs. It is possible that some of the hot springs include a small proportion of water of very recent origin which mixes with the main flow of heated water near the springs’ outlets. Such water could contain water from the 1960’s that has tritium levels of several hundred or more tritium units. An admixture of less than 1 per- cent of such water would raise the tritium level to that measured, and such an amount would be too small to measurably affect any other characteristics of the sample. Thus, the tritium level suggests that there is no significant amount of water less than several dec- ades old in any of the hot springs. Of the 14 cold wells and springs, 5 have tritium levels higher than 25 Tu and the rest have levels lower than 3 Tu. Those with little tritium, such as the hot springs, contain no significant amounts of recently re- charged water, whereas those with high tritium con- centrations contain significant quantities of water re- charged within the past 15 years. Their high tritium concentration reinforces the preceding conclusion that the chemistry of Belvedere Country Club well and Frank Thompson’s house well is that of recharge to the system and that Bill Sargo’s well and R. B. Yates’ well could be influenced by man’s activity. The fact that Happy Hollow Spring, although chemically identical with the high-tritium Belvedere Country Club well and Frank Thompson’s house well, contains only 1.1:0.6 Tu is somewhat puzzling. A possible explana- tion is that the water in this spring fell as rain at, say, 5 Tu in the prebomb era (pre-1953) and has no compo- nent of post-1953 water. Sleepy Valley Spring shows only slight effects of mineral—water reactions, and its high tritium concentration is therefore not unexpected. CARBON A'l‘l-Z (}E()(‘.H EM lS'l‘RY The most common inorganic, carbon-bearing chemi- cal species are carbon dioxide (CO2), a gas which read- ily dissolves in water, bicarbonate (HCOg‘), the pre- dominant carbon species in most waters in nature, and carbonate (CO3‘2), found in such minerals as calcite (CaCO3) but also present in low concentrations in solu- WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN 019 tion. The prevalent species in a given solution depend on the hydrogen-ion (H+) concentration of the solution. The pH of a solution is a measure of its H+ activity (pH = —log a H+). A solution having a low pH has a rela- tively high H+ activity, or is acidic; one having a high pH represents an alkaline solution, one having a rela- tively low H+ activity. The interreactions among the three carbonate species are: CO2(gas) + H20 —>H2003, (2) H2003—) H+ + HCO3—y (3) and HCOa‘ —> H+ + cow. (4) Water in contact with a gas phase containing 002 will dissolve C02, according to reaction 2, in an amount proportional to the partial pressure of 002 in the gas (P002). Some of the H2C03 so formed will dissociate, by reaction 3, to bicarbonate and hydrogen ion, which will lower the pH of the solution. If an alkaline substance is present, the H+s° formed will be removed from solution, permitting more H003‘ to form. Carbonate minerals are common alkaline substances and are dissolved by H+ by the reaction Ca003(sond,+H+ —> Ca+2+HC03‘. (5) If a water originally charged with CO2 is brought into contact with a carbonate mineral, the reaction 002+H20+0300390a+2+2HCO3_ (6) will occur. This reaction is the most important in con- trolling the chemical character of waters in the hot- springs area, and together with the solution of silica, discussed previously, also controls the chemistry of the hot springs themselves. In discussing the carbonate chemistry of these wa- ters, the term “total dissolved carbonate” is often used. This term refers to the molar sum of the dissolved car- bonate species H2C03, HCO{, and 003‘2. This is not an analyzed value, but is calculated from the HCO3‘ and pH values given in table 2, and the equilibrium- constant expressions corresponding to reactions 2, 3, and 4. These calculations were made using the com- puter program WATEQ (Truesdell and Jones, 1974). The total dissolved-carbonate concentrations so calcu- lated for the waters sampled are given in table 9. Some of the wells and springs were sampled for analyses of their dissolved-gas concentration. Figure 8 shows the relation between the analyzed partial pres- sure of CO2 (P002), with the PC02 calculated from the values of pH and H003“ given in table 6. Some of the analyzed values are larger than the calculated values, but most agree within 25 percent. Although this may appear a large error, it introduces very little error into the much larger calculated total dissolved-carbonate concentration, which is the parameter of interest. Water entering a ground-water system is exposed to gases from the Earth’s atmosphere and from the gas phase in the soil zone. The soil air is the last gas phase TABLE 9.—Carbon-isotope date for waters in the hot-springs area Total Measured Calculated 6'3C Proportion 6'5C Measured I‘C Adjusted “C Adjusted PDB plant PDB (percentage modern (percentage age (“/oo) carbonate ("/0") t 10-) modern) (years) Well or dissolved spring carbonate No.‘ Name (millimoles (fig. 6) per liter) Happy Hollow Sprin __________ 1.13 Belvedere Country C ub well _- 1.44 Frank Thompson’s house well i. 0.88 Sleepy Valley S ring __________ 1.41 W17 ____________ Bill Sargo’s wel ______________ 21.02 W16 ____________ R. B. Yates’ well ______________ 1.85 S11 ____________ McLendon Mineral Spring ______ 2.78 W21 ____________ Gulpha Gorge well ____________ 3.05 S7 ______________ Cluster Sprin ________________ 3.49 W26 ____________ Whittington venue Spring 2... 3.85 W25 ____________ Whittington Park well ________ 3.94 S5 ______________ Echo Valle Spring ____________ 4.06 W12 ____________ Elizabeth rown well __________ 3.88 W24 ____________ Diamond Mineral Spring ______ 4.15 Maurice Hot Spring ____________ 3.00 Hot Spring No. 17 ____________ 2.70 Hot Spring No. 23 ____________ 2.74 Hot Spring No. 33 ____________ 3.06 Hot Spring No. 46 ____________ 3.18 Hot Sprin No. 48 ____________ 2.97 80,000-gal on reservoir ________ 2.97 Hot Spring No. 42 ____________ 3.18 Hot Spring No. 49 ____________ 3.07 Average of samples indicated by (*) __________________ (2) (2) ~22.5 1.000 —24.0 101.7116 Recharge. —23.7 98.9: 1.5 —22.6 .904 —21.7 (2) ______________ —24.1 ____________ (2) ______________ —23.6 ____________ 3111.82‘:1.3 ______________ —13.8 .620 —14.9 3951-08 63.7 3,730 —14.4 .585 —14.0 35.5:05 60.7 4,130 — 15.8 .670 — 16.1 37.3 $0.4 55.7 4,840 —14.3 .650 —15.6 31.2t0.5 48.0 6,070 —13.8 .565 —13.6 23.7:0.3 41.9 7,180 —14.9 .558 —13.4 22210.3 39.8 7,620 —15.5 .625 —15.0 21210.4 33.9 8,940 ~13.3 .566 —13.6 19.4:03 34.3 8,850 ~14.7 __________________ *—14.1 ____________ 5 ______ *—14.2 __________________ —14.6 __________________ *—14.8 __________________ *—14.3 __________________ *—14.4 __________________ (2) __________________ *—16.1 __________________ —14 6 .608 ______ 4,430 ‘ US. Geological Survey numbers of wells and springs are given in table 7. 2 Not analyzed. 3 Man influenced; not representative. C20 0-050 I l I I I I l I l I I Hot Spring No. 33X Hot Spring No. 12X Hot Spring No. 49x _ Whittington AvenueX _ Hot Spring No. 46X )< Happy Hollow Spring 0.020 — — xEcho Valley Spring 0.010 — xHot Spring No. 23 xWhittington Park 0.005—- — CALCULATED FROM ALKALINITY AND pH PARTIAL PRESSURE OF CARBON DIOXIDE, IN ATMOSPHERES, I l I I I I l L l 003.3003 0.005 0.010 0.020 PARTIAL PRESSURE OF CARBON DIOXIDE. IN ATMOSPHERES, FROM GAS ANALVSIS 0.050 FIGURE 8.—Relation between analyzed and calculated partial pres- sures of 002. to which the water is exposed before it enters the water-saturated zone, so the composition of the soil air controls the type and amount of gases initially dis- solved in ground water. The main difference between the open atmosphere and soil air is that the soil air contains considerably more COZ. In the open atmos- phere, the partial pressure of COZ (PC02) is about 3X10‘4 atm (atmospheres). Water in contact with this PCO2 at 17°C (62.6°F) contains about 1.3)(10’2 mil- limoles per liter (mmol/L) H2C03, or about 0.8 mg/L. In soils, though, much of the CO2 is produced by plant- root respiration and by decay of plant debris, so the P002 in soil air is one or more orders of magnitude higher than that in the open atmosphere. Because of differences in soil and plant types, the soil air PCOz may differ from place to place, but should be reasonably constant within a given area. An analysis of the composition of the gas dissolved in Happy Hollow Spring, one of the waters containing dissolved material which, except for silica, is probably derived mainly from atmospheric precipitation, showed its PCO2 to be 3x10‘2 atm, equivalent to an H2003 concentration of about 1.4 mmol/L, or 87 mg/L. Gas analyses were not made for the other two most dilute samples—from Frank Thompson’s well and the Belvedere Country Club well—but analyses of their HCO;,‘ and H+ (pH) concentrations were made. From these data and from the known equilibrium constants for the reactions be- tween carbonate species (1 and 2 above), the PCOZ of the gas in equilibrium with them can be calculated (Truesdell and Jones, 1974). For the water from Frank GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Thompson’s well, the PCO2 so calculated is about 2x10‘2 atm. The average total dissolved carbonate (H2003+H003‘) in these three samples was 1.15 mmol/L. The chemistry of the remaining cold waters sampled differs from that of atmospheric precipitation, as repre- sented by Happy Hollow Spring, Frank Thompson’s well, and the Belvedere Country Club well, primarily in having higher alkaline-earth (Ca+2, Mg”, and Sr”) and bicarbonate (HC03‘) concentrations. Reactions representing the solution of mineral carbonate are written in terms of an idealized CaC03 mineral. In na- ture, CaCO;, may contain other elements of the alkaline-earth group such as magnesium (Mg+2) and strontium (Sr+2), replacing calcium. These substitu- tions have little effect on the carbonate in the ground water, and so in discussing the measured water chemistry, the total alkaline-earth concentration H (Ca+Mg+Sr) is used, whereas in discussing reactions, only the idealized-Cf2 form is used. Figure 9 is a graph showing the concentrations of alkaline-earth against total dissolved-carbonate con- centrations (HgCO3+HC03‘ + 003—2). On this graph, the point at zero alkaline-earth concentration is at the average total-carbonate concentration of the three most dilute samples (1.15 mmol/L). From this point, lines are drawn showing the trend of compositions which would result from solution of carbonates by reac- tions 5 and 6. Three of the samples—Sleepy Valley Spring, Cluster Spring, and Whittington Avenue Spring—fall near the line of reaction 6. In these sam- ples, the dissolved carbonate in excess of that in the recharge water is half from solution of carbonate min- erals and half from solution of further soil-zone-derived C02. The other samples plot between the two lines, and a greater proportion of their excess carbonate is from solution of carbonate minerals alone. To adjust l"C (radiocarbon) measurements on ground waters so that they can be used to calculate ground- water ages, it is necessary to know the proportion of total dissolved carbonate that has come from solution of soil-zone 002. This factor can be derived from the chemical information in figure 9, but before describing the method, a discussion of the isotopic chemistry of carbon is appropriate. ISOTOPIC CHEMISTRY OF CARBON The element carbon has three naturally occurring isotopes with masses of 12, 13, and 14. Carbon of masses 12 (12C) and 13 (13C) are stable isotopes, whereas carbon of mass 14 (”C) is radioactive. In any chemical reaction such as the solution of 002 to form HCOa‘, or the formation of carbohydrates from C02 by plants during photosynthesis, the light isotope (12C) re- acts faster than the heavier isotope (‘30). The isotope ratios of reactant and product species are different, and WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN C21 2.0 l l ‘1' Line of reaction (eq 5) CaCO3+H+§Ca+HCO— S I .o I Recharge Wu Happy Hollow Spring, Frank Thompson, Belvedere Country Club 0 | ,. I ALKALINE EARTH (CALCIUM+MAGNESIUM+ STRONTIUM) IN MILLIMOLES PER LITER McLendon>< Mineral Spring XSleepy Valley Spring Diamond Mineral Echo Valley Springx xSpring Whittington Park>< Elizabeth Brownx Whittington Avenue G | h u p a Spring GorgeX xCluster Spring Line of reation (eq 6) c02+H20+Cac03,Ca+*+2 Hc03— 0 1 2 3 4 5 6 TOTAL DISSOLVED CARBONATE, IN MILLIMOLES PER LITEH FIGURE 9.—Relation of alkaline-earth concentration to dissolved-carbonate concentration. isotopic fractionation occurs. Carbon-isotope ratios are measured relative to an arbitrary standard, a calcite of marine origin known as PDB and expressed in the 6-notation, as are hydrogen- and oxygen-isotope ratios. 8130 values for samples from a single type of carbon- bearing material are relatively constant, but one material may have quite different 613C values from another. Marine limestones, for instance, have 8130 values close to zero (because a marine carbonate is used as the standard), whereas atmospheric CO2 and plants typical of the hot-springs region are depleted in 13C and have 6130 values of about —7°/00 and —24°/00, respectively. These large differences make 8130 meas- urements valuable in determining the sources of dissolved-carbonate species (Pearson and Hanshaw, 1970). The naturally occurring radioisotope of carbon (”C) has a half—life of about 5,730 years. Natural 1"C is formed by cosmic-ray reactions in the upper atmos- phere and is found at virtually a constant concentra- tion in atmospheric (302. 1"C concentrations are meas- ured and reported as percentages relative to a standard whose l"C concentration is that of an ideal plant, grown in 1950, in the absence of any atmospheric CO2 of in- dustrial origin. 1"C is produced as a byproduct of nu- clear explosions, and in the last 25 years the 1"C level in the atmosphere has risen considerably higher than its natural level (100-percent modern). Plants using atmospheric C02 assimilate 1"C. When a plant dies, the amount of 1"C it contains decreases because of radioactive decay, and the amount remain- ing, relative to that of an ideal living plant, measures the time that has elapsed since its death. This is the principle on which the radiocarbon (”C)-dating method is based. The equation for radioactive decay is sm A std where Asm/Asm is the ratio of sample to standard ac- tivity, A is the decay constant of the isotope (1/8,266 for 1"0), and t is the age in years. (The half-life of an isotope is the value of t when Asm/Astd=0.5; that is, tyz=ln 2/)\=5,730 years for 1"C). For calculation, the decay equation is often written using natural logarithms (In) as Asm Astd ) As discussed, the 002 in the soil air is a product of plant respiration and decay. Because of its plant origin, soil-air 002 has the same carbon-isotopic composition as plants; its 6130 value is about —24°/00, and its l"C concentration is close to 100 percent (in the prebomb era). Soil 002 dissolved by ground water will also in- herit its plant isotopic character, and if soil CO2 were the only source of carbonate dissolved in a ground wa- = exp(—At), t = —8,266 In ( C22 ter, its 1“C concentration would directly reflect the age of the ground water—that is, the time that has elapsed since the water left the soil zone. Ground water generally contains mineral carbonate, which is commonly of marine origin and has a 5130 value near 00/00. Because of its great age relative to the half-life of 1“C, mineral carbonate has no l"C. Solution of 1"C-free carbonate lowers the measured— 1"C concen- tration of carbonate dissolved in ground waters and gives the water a falsely old apparent age. However, if the relative amounts of soil-derived and mineral car- bonate are known, this effect can be corrected. For example, consider a water of an age of zero years, with a total dissolved carbonate of 4.00 mmol/L, of which 3.00 mmol/L is from soil 002 (“0:100 percent) and 1.00 mmol/L from carbonate minerals (”0:0 per- cent). The measured “40 concentration of this water will be ((3.0><1.0:+0(1.0x0.0)) : 75_percent modem, and its apparent age will be t = —8,266 In (75/100) = 2,380 years, instead of its real age, zero years. By knowing the proportions of the sources of carbo- nate to a ground water, it is possible to adjust the measured-”C concentration to its correct value and so obtain a corrected age. The adjustment factor, P, is the ratio of soil-air—derived (=plant) carbonate concentra- tion to total dissolved-carbonate concentrations: P _Cplant 01.0131 (7) For this example, The adjusted l"C concentration is . 1"C (measured) , 1"C (adjusted): —P——- and for this example 75 t 1"C (adjusted) = flu— 0.75 = 100 percent, corresponding to an age of zero, as specified. Chemical information on ground water can often be used to determine the proportion, P, of soil-air-derived carbonate to total carbonate. A graphical representa- tion of changes in water chemistry, such as that given in figure 9, can be useful. Consider Whittington Av- enue Spring. The total carbonate concentration of this GEOHYDROLOGY OF GEOTHERMAL SYSTEMS sample (3.85 mmol/L) is the sum of (1) carbonate dis- solved by atmospheric precipitation in the recharge area, here taken as the average of the three very dilute samples, Happy Hollow Spring, Frank Thompson’s well, and the Belvedere Country Club well shown in figure 9, or 1.15 mmol/L, and (2) additional carbonate from the reactions CaC03+H+-—>Ca+2+H003‘ (5) and 002 +H20 +CaCO3—>Ca+ +2HCO3‘ (6) or, 2.70 mmol/L. Note that in reaction 5 all the bicarbonate is from the solution of mineral carbonate and that an equiva- lent amount of alkaline earth (Ca) also appears in solu- tion. In reaction 6, half of the bicarbonate is of mineral and half of gaseous (plant and soil-air) origin. In reac- tion 6, twice as much bicarbonate as alkaline earth is produced, but the amount of mineral carbonate added still equals the amount of alkaline earth, as in reaction 5. The Whittington Avenue Spring contains 1.34 mmol/L of alkaline earth in solution, and from the pre- ceding paragraph, this will be the amount of mineral carbonate dissolved, as well. That is, of the 2.70 mmol/L total carbonate present in excess of that in the recharge, 1.34 mmol/L is of mineral origin and 1.36 mmol/L (2.70—1.34) of plant origin. To sum these considerations algebraically, (8) where C is the carbonate concentration from the sub- script sources. Crechme and Cgaseous are of plant (=soil- air) origin and Cmineralzcalkaline earth- Equation 8 be- comes Ctotal = Crecharge + Cmineral + Cgaseous 2 Ctotal =Cplant+ Calkaline earth, or (9) The 14C-adjustment factor, equation 7, can now be solved: Cplant Ctotal_Calkaline earth Calkaline earth — _____1_ = 1— ————— (10) Ctotal carbonate Cplantzctotal _ Calkaline earth. Clotal thal For the Whittington Avenue Spring, =1—(1&>=0.65. 3.85 The stable carbon-isotopic composition of the total carbonate dissolved in a ground water sample also re- flects the proportions of the source carbon. As dis- cussed, the 8‘30 value for plants is about *249/00, WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN whereas that for marine mineral carbonate is about 0%». The 8‘30 value of a sample will be in proportion to the amounts and isotopic compositions of its carbonate sources: _ Dlant Cmineral alscsample_613cplant — +813C'mineral C . Ctotal total Because SlsCmmeral = 00/00, the last term is zero and C alacsample=5130plam (fill—t) , (11) which, on recalling equation 7 and substituting an av- erage value for 6‘3Cplam, becomes 613C'sample= (12) —24P . Carbon-isotope data and calculated total carbonate concentrations of samples in the hot-springs area are given in table 9. For each, a value for P, the proportion of plant carbonate, calculated using equation 10, is shown. Also, 8130 values, calculated from equation 12, are given for comparison with the measured 813C val- ues. Agreement between measured and calculated 6130 values is excellent, suggesting that the geochemical processes, described in the preceding paragraphs, adequately model the field situation. The average value of the differences between the calculated and measured 813C values is 0750/00, and the average value of the measured 613C values themselves is 16.70/00. The error associated with the calculated values of P can be estimated as 0.75/16.7 x 100 = 4 percent. This is larger than the l“C-analyzed errors and corresponds to an error band of :500 years for a sample of 4,000 years and of 11,000 years for a sample of 9,000 years. The adjusted ages of the cold wells and springs that were sampled are consistent with inferences previously drawn from their chemical character and their tritium concentrations. The Belvedere Country Club well and Frank Thompson’s well, which represent water re- charging the system, have 1“C concentrations near 100-percent modern. The R. B. Yates’ well, which on chemical criteria is suspected to include recent water showing man’s influence, has both 1“C and tritium levels found only in the last 20 years. The adjusted ages of the remaining cold-water sam- ples range from 3,700 to 8,900 years and increase in age with increasing total carbonate concentration. C23 AGE OF THE HOT-SPRINGS WATER As the analyses in table 6 show, the hot springs have almost identical chemical characteristics, suggesting that they have a common origin. The only detectable differences among them are in their carbonate concen- trations, shown by the different H00; and pH values in table 6, and in the total carbonate concentrations calculated from these values and given in table 9. Calculations have been made to determine whether the waters of the hot-springs system are saturated with respect to the mineral calcite (CaC03)—that is, to de- termine whether the calcium and carbonate concen- trations of the waters are such that they will dissolve more calcite or deposit calcite from solution. The min- eral calcite (CaCO3) dissolves according to the reaction CaCO3—9Ca+2+003‘2. At saturation, the product of the thermodynamic ac- tivities of the dissolved materials ( aCa+2, aco3 -2) equals a constant (KT), the value of which is a function of temperature. That is, at saturation [aCa+2] [acoa— J: KT. It is convenient to express the degree of saturation of a water in terms of the logarithm of the ratio of the ac- tivity product (AP) to the equilibrium constant (KT). The AP of an undersaturated water is less than one, so log (AP/KT) is negative. Conversely, log (AP/KT) of an oversaturated water is positive. In the previous general discussion of carbonate chemistry, equations 2, 3, and 4 were given showing the interreactions among the carbonate species CO2 (gas), H003”, and CO3”. According to these equations, a change in any one of the carbonate species will pro- duce a change in the others. Combining equations 2, 3, and 4, for example, gives 2HCO3*—>CO 2(gas)+C03‘2+H 20, which shows that if 002 gas leaves water, responding to a decreasing (Pc02), the 003‘2 concentration may in- crease. This reaction in turn will increase the AP for calcite and thus the degree of saturation of the water with respect to calcite. Certain of the hot springs visibly evolve gas where they emerge at the Earth’s surface, and because this gas is in part 002, the carbonate chemistry of individ- ual springs should differ as a function of the amount of gas evolved. Measured and analyzed 002 partial pres- sures (Pc02) from some of the springs are shown in figure 8. Comparison of this figure with the total dissolved-carbonate data in table 9 shows that a spring such as number 23 that has a low Pcoz value, which presumably results from evolution of a significant amount of dissolved gas, has a lower total dissolved- C24 carbonate concentration than such springs as numbers 42 and 49 that have higher P002 values. The Few of these waters is as high as 0.035 atm (fig. 8), whereas the Pcozof the Earth’s atmosphere to which they are exposed at the surface is only 0.0003 atm. The hot- springs waters certainly could lose C02 to the atmos- phere as they approach the surface. To follow the influence of CO2 outgassing on the overall carbonate chemistry of the spring water, calcu- lations of log (AP/KT) at various values of PCO2 were made for a relatively cool spring (No. 23, 562°C (133.2°F) ) and for a hot spring (No. 49, 61.8°C (143.2°F) ). The results of these calculations are shown in figure 10. The path of decreasing calcite saturation with increasing Pcoz for spring number 23 is slightly offset from that of number 49 because of the tempera- ture difference between the springs. More important, ”-5 I l I I Hot spring +0.4 — . N0: ‘7 ~ OVERSATURATED +0.3 - - Ho? spring No. 23 : +0.2 - — ¥ 0. < O 3 Collecting ... +0" - reservoir — Z 9 .— < I Hot spring )2 0 0 No, 48 < . U) Lu P: 6|.8°C 3 < 562°C 0 —o.I - V — Ho! spring No. 33 '— —o.2 — « Hot spring, No. 46 Hot sprin -0.3 — No.49 q _ Hot spring No. 42 UNDERSATURATED .0.“ I l l J 0.00 0.0I 0.02 0.03 0.04 0.05 PARTIAL PRESSURE OF CARBON DIOXIDE, IN ATMOSPHERES FIGURE 10.—Relation between calcite saturation and partial pres- sure of carbon dioxide in the hot springs. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS the points corresponding to the other springs sampled lie on or between the two calculated lines. From this comparison we conclude that the differences among the hot springs in pH and HCOa‘ (table 6), in total dis- solved carbonate (table 9), and in P002 (fig. 8) are sup- erficial and due only to different amounts of outgassing in the near-surface distributary system to the springs. Figure 10 shows that certain of the springs are over- saturated with respect to calcite. Table 6, though, shows that the calcium concentration of the springs is identical (44.5 :0.5 mg/L) and that the springs that were sampled had not yet begun to deposit calcite, though they have the potential to do so. If allowed to flow freely on the surface, all the springs would con- tinue to lose 002 and thus would deposit calcite. The tufa associated with the springs shows such deposition does occur. One significant difference between the hot-springs waters and the cold wells and springs in the region is their sodium (Na) concentration. The average sodium concentration of the 12 relevant cold-water samples is 1.8 mg/L (0.08 mmol/L), whereas that of the hot springs is 4.0 mg/L (0.18 mmol/L). There is no difference be- tween the carbonate, sulfate, or chloride concen- trations of the cold and hot springs to balance the ad- ditional sodium as there would be if the sodium were a product of the solution of some mineral. Instead, the sodium must result from a reaction by which it is ex- changed for some other positively charged dissolved ion. Such a reaction can be written Ca+2+Na2 X Zea X 2Na+, (13) in which Ca+2 represents an alkaline earth and Nazx and Ca X represent solids, perhaps clay minerals, in the aquifer system, with X the substrate on which the exchange occurs. This exchange reaction must be considered in cal- culating the proportion of plant to total carbonate to adjust the measured l4C concentrations of the hot springs and to determine their water ages. To calculate this proportion for the cold springs, the expression P: 1 __ Calkaline earth (10) Crotal carbonate was used. In deriving equation 10, it was assumed that only mineral carbonate solution affected the alkaline- earth concentration of the waters, and the validity of this assumption was borne out by the agreement be- tween the calculated and measured stable car— bonisotope concentration (8"‘C) of the samples. In the hot-springs waters, the alkaline-earth concentration is dependent on exchange reaction 13, as well as on the carbonate chemistry, thus equation 10 cannot be used. In discussing the carbonate chemistry of the cold WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN springs, the relationship between the stable carbon- isotope concentration (6‘30) of a sample and the propor- tion of plant to mineral carbonate, C an alacsamplezslscplant <%‘L_> ’ total (11) was presented. This equation can be used directly to find the 1“C adjustment factor, P: C 6130 P: plant = __samnl.e_ . (14) (Canal > alscplant By using an average 6130mm: —24°/00 as before (eq. 12), this becomes P: 613Csample/_ 240/00- Isotopic fractionation takes place between CO3‘2, HCOs‘, and CO2 (gas) as outgassing and 003‘2 produc- tion from H00; occur. However, at the temperature of these springs (g62°C) the fractionation accompanying the reaction HC03'—>H+—>COZ+H20 is nearly equal to and opposite that of the reaction HCO;,‘—>H++CO;,‘2 (Pearson and others, 1972). Thus isotopic fractionation will not interfere with the calculation of P by equation 14. Because the hot springs are chemically and isotopi- cally so similar, their average adjusted l4C concen- trations and ages were calculated (table 9). Only those samples for which both 13C and 14C data are available were used in this average. As table 9 indicates, the proportion of plant carbonate to total carbonate in them, from equation 14, is 0.608, and their adjusted 14C concentration is 58.5-percent modern, corresponding to a water age of 4,430 years. Although the total carbonate concentration of the hot-springs water before near-surface outgassing of CO2 is not known, it certainly cannot be lower than the highest value measured—3.18 mmol/L. This value is plotted with the adjusted age on figure 11 for compari- son with the ages and carbonate concentrations of the cold wells and springs. That the hot springs, cold springs, and wells are part of the same group is clear and is additional evidence that all are part of the same hydrologic system. A hydrologically important inference can be drawn from figure 11. The hot springs are on a trend line defined by both older and younger cold-water samples. Thus, it is probable that the age of the hot-springs water is due largely to its resident time in the cold- water part of the flow system, rather than to an ex- tended period in a zone of heating. A rapid traveltime through the heated part of the system is in keeping with earlier conclusions from oxygen-isotope (‘80) ex- change, suggesting no prolonged period of heating. C25 That the surface temperatures of the hot springs closely approach their probable maiimum tempera- tures, as calculated from their silica concentrations, also implies rapid flow which prevents significant heat loss as the heated water rises to the surface. RADIOACTIVITY OF THE WATERS IN THE HOT-SPRINGS REGION The radioactivity of the waters of the hot springs has been studied by several investigators. The first study was by Haywood (1902), followed by Boltwood (1905), Schlundt (1935), and Kuroda, Damon, and Hyde (1953). Much of the early interest was because of the balneological use of the water. In 1953 studies by Arndt and Damon (1953) were sponsored by the US. Atomic Energy Commission, whose interest was the radon concentration and the source of the radioactiv- ity. The presence of radium in the waters of the hot springs was established by Schlundt (1935) when he determined an average value of 1.38 picograms per liter (1.35 picocuries per liter) of radium for three sam- ples. Recent (July 1973) analyses of the waters by the US. Environmental Protection Agency show a radium concentration of 2.11-0.22 picocuries (10‘12 curie) per liter. Radon, a gas that is a radioactive decay product of radium, was analyzed in the hot springs by Boltwood (1905), Schlundt (1935), and Kuroda, Damon, and 1 0 i l l I Diamond Mineral Spring Elizabeth Brownx 8— _ Echo Valley Spring Whittington Park xWhitfington _ Avenue Cluster Springx Hot Springs Gulpha Gorge McLendon)< Mineral Spring ADJUSTED CARBON—14 AGE, IN THOUSANDS OF YEARS I l 0 1 2 3 4 5 TOTAL DISSOLVED CARBONATE, lN MILLIMOLES PER LITER FIGURE 11.—Relation between adjusted carbon-14 age and total car- bonate. C26 Hyde (1953). Boltwood (1905) reported the radon con- centration of 45 springs, ranging from 0.017 to 9.03 nanocuries (10”9 curie) per liter, with modal value of 0.466 nanocurie per liter. Schlundt found the radon concentration of six hot springs to range from 0.125 to 0.46 nanocurie per liter, with a modal value of 0.32 nanocurie per liter. Kuroda, Damon, and Hyde (1953) reported the radon concentration of 25 hot springs to range from 0.14 to 30.5 nanocuries per liter, with a modal value of 0.82 nanocurie per liter. The variations from spring to spring found by each investigator reflect the combined influence of differ- ences in analytical methods and natural variations. The source of radium and radon in the hot-springs waters is not definitely known, nor have the waters been analyzed for the presence of other radioactive elements. The presence of radium and radon in the region is not peculiar to the waters of the hot springs. Waters from deep wells at Hope, 68 mi (109.4 km) southwest of Hot Springs, and at Prescott, 53 mi (85.3 km) southwest of Hot Springs, ranged in radon concen- tration from 0.05 to 1.88 nanocuries per liter (Kuroda, 1953). Waters from the warm and cold springs near Caddo Gap, 33 mi (53.1 km) west of Hot Springs, ranged in radon concentration from 0.13 to 1.85 nanocuries per liter (Kuroda, 1953). Cold springs in the immediate vicinity of Hot Springs also contained radium and radon. The US. Environmental Protection Agency reported a radon concentration of Happy Hol- low Spring of 1010.15 picocuries per liter and a radon concentration of Whittington Avenue Spring of 0.58 $0.12 picocuries per liter. Kuroda (1953) reported the following radon concentrations, all in nanocuries per liter: Whittington Avenue Spring, 0.36; Whit- tington Park well, 0.03; Happy Hollow Spring, 0.74; and Sleepy Valley Spring, 4.37 and 2.80. In addition, Kuroda (1953) reported the radioactivity of four sam- ples of rainwater ranging from 2.72 to 6.37 nanocuries per liter. Radon concentration of spring waters issuing near the uranium-vanadium-niobium bearing depo- sitis at Potash Sulphur Springs, 6 mi (9.65 km) south- east of Hot Springs, averaged about 15 nanocuries per liter (Arndt and Damon, 1953.) THE HOT-SPRINGS FLOW SYSTEM The geologic, hydrologic, and water-chemistry data and interpretations presented in previous sections of the report provide the bases from which a conceptual model of the hot-springs flow system can be formed. Digital models of the flow system provide a test of the conceptual model and a vehicle for further refinement of the conceptual model. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS CONCEPTUAL MODEL The hot-springs water is meteoric; that is, it is de- rived from precipitation. The water is recharged to formations in the hot-springs region (within a few tens of miles), as opposed to being recharged to formations distant (several tens or hundreds of miles) from the hot springs. The origin and proximity of recharge of the hot-springs water are revealed by the chemical con- stituents and isotopes in the water and the flow varia- tions of the hot springs. The formation or formations that form the recharge area must possess the following general characteris- tics: 1. The outcrop area must be relatively large and permeable, that is, from 3 to 10 mi2 (7.77 to 25.9 kmz), assuming permeabilities that would per- mit average recharge rates of from 2 to 6 in (50.8 to 152.4 mm) per year. 2. The elevations of water levels in the recharge area must be high enough to provide hydraulic head for spring flow. The springs emerge at elevations ranging from 576 to 683 ft (175.6 to 208.2 m) above mean sea level. 3. The recharge formations must be hydraulically connected to the permeable zones that feed the springs. Consideration of the lithology and structure of rocks in the area reveals two formations whose outcrops may serve as recharge areas—the Bigfork Chert and the Arkansas Novaculite. The outcrop area of the Bigfork Chert meets several of the requirements of the hot- springs recharge area—the Bigfork crops out over an area of about 36 mi2 (93.2 kmz) north and northeast of the hot springs, in the area shown in figure 6. The Bigfork Chert generally possesses moderately high fracture and intergranular permeability. Faults and fracture zones are common. Permeable zones extending from the Bigfork Chert to the hot springs could be pro- vided by faults (pl. 1) that connect the Bigfork Chert with permeable zones in the Arkansas Novaculite and the Hot Springs Sandstone Member of the Stanley Shale or that provide a permeable fault zone through the Missouri Mountain and the Polk Creek Shales to the hot springs. The Arkansas Novaculite crops out in the vicinity of the hot springs in a smaller area than does the Bigfork Chert (fig. 6). Within the mapped area in figure 6, the outcrop of the Arkansas Novaculite covers about 13 mi2 (33.7 km2). Permeability of the Arkansas Novaculite varies greatly; in places it is moderately high. The out- crop of Arkansas Novaculite occupies the highest ele- vation in the vicinity of the hot springs. WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN The elevation of head in the recharge area required to provide hot-spring flow depends on such factors as transmissivity of the aquifer, areal extent of the re- charge area, distance of the recharge area from the springs, depth of flow of water en route to the springs, and the temperature of the water in the aquifer. These factors, except for the effect of temperature on hydrau- lic head, are accounted for in the digital models of the aquifer. The effect of the increased temperature of water in the flow system reduces the head required in the recharge area to drive the system. The density of water above 4.0°C (39.2°F) decreases as its tempera- ture increases. Water at 61.7°C (143.1°F) (average hot-springs temperature) is 98 percent as heavy as water at 17.7°C (63.9°F) (average temperature of re- charge water). Therefore, a column of water at 17.7°C (63.9°F), 1,000 ft (304.8 m) in length, will support a 1,020-ft (310.8 In) column of water at 61.7°C (143.1°F). It would be possible for a thermal artesian system to have a lower water-level elevation in the recharge area than in the discharge area. However, under conditions of uniform areal heat flow, it would not be possible for artesian flow to be initiated without sufficient head in the recharge area to drive the system. Water levels in the outcrop area of the Bigfork Chert range from a few feet above land surface (flowing wells) to 30 ft (9.14 m) below land surface. Water levels in the Arkansas Novaculite range from a few feet above land surface to 50 ft (15.2 m) below land surface. As an approximation of the area in which heads in the Big- fork Chert and the Arkansas Novaculite are at a minimum to sustain flow in the artesian system, the outcrop area above 700 ft (213.4 m) mean sea level was chosen. Within the area shown in figure 6, about 8.5 mi2 (22.0><106 m2) of the outcrop of Bigfork Chert is above 700 ft (213.4 m) mean sea level, and almost all the outcrop of Arkansas Novaculite is above 700 ft (213.4 m) mean sea level. However, under about 5 mi2 (13.0x 106 m2) of the Arkansas Novaculite outcrop, the northwest limit of the anticline north of the hot springs, the formation dips northwestward and prob- ably is not in a favorable structural position to supply water to the hot springs. Analysis of streamflow provides information on the magnitude of recharge. Stream discharge and rainfall records have been collected on three small basins in the vicinity of the hot springs (Bedinger and others, 1974). The locations of the gages on East Fork and West Fork of Hot Springs Creek are shown in figure 6. The gage on Glazypeau Creek is located about 9 mi (14.5 km) north-northeast of the hot springs. Each basin is rela- tively small and is underlain principally by Bigfork C27 Chert. The Arkansas Novaculite underlies the higher elevations of East Fork and West Fork of Hot Springs Creek (fig. 6). Base flow is ground-water discharge to the stream and is derived from recharge of precipitation. A rela- tively large part of rainfall on the Bigfork Chert out- crop recharges the subsurface reservoir. Base flows in the East and West Forks of Hot Springs Creek are 61 and 84 percent, respectively, of precipitation (Bedinger and others, 1974). These figures are primarily a mea- sure of the recharge to the Bigfork Chert. Also, the base-flow studies indicate that interbasin transfers of ground water occur. Such interbasin transfers of ground water, probably from several stream basins, supply the water to the hot springs. The large surface area, the great capacity to admit recharge, and the relatively high permeability of the Bigfork suggest that the Bigfork Chert is the principal recharge source to the hot-springs artesian system. The Arkansas Novaculite has a smaller outcrop area of potential recharge, a lower capacity to admit recharge, and a generally lower permeability. However, the Ar- kansas Novaculite is probably a source of part of the recharge to the hot-springs system. Further analysis of the Bigfork Chert and Arkansas Novaculite as the re- charge sources and principal aquifers of the flow sys- tem is made in the section on modeling. The 14C age of the hot-springs water averages 4,430 years. The greater part of the time that the water is in underground circulation the movement is very slow. That is, movement is on the order of a few feet to a few tens of feet per year and is distributed through a large volume of the aquifer. This slow movement of water continues to relatively great depths in the aquifer and probably includes lateral circulation where sufficient heat is absorbed to reach temperatures greater than 60°C. This heat is supplied by conduction through adja- cent rocks of lower permeability. In the absence of fluid circulation, heat is conducted toward the land surface at rates near 1.2x 10—6 calories per second per centi- meter squared, or 1.2 heat-flow units in most of the Eastern United States. In areas of abnormally high crustal radioactivity, the geothermal heat flow may be as high as 2.2 heat-flow units (M. L. Sorey, written comm., 1976). On the basis of an assumed thermal con- ductivity of 0.006 calorie per second per centimeter squared and heat flow of 1.2—2.0 heat-flow units, the normal geothermal temperature gradient in the vicinity of the hot springs should be between 0.006°C/ ft and 0.01°C/ft, although the available subsurface tem- perature information is insufficient to confirm these gradients. With these gradients and a maximum spring temperature at depth of 63°C based on silica C28 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS concentrations, the minimum depth of fluid circulation would range from 4,500 to 7,500 ft. Highly permeable zones, probably related to jointing or thrust faulting, collect the heated water in the aquifer and provide avenues for the water to travel to the surface. Rapid movement of the water from depth to the surface is indicated by the very small decrease in temperature from the maximum temperature attained at depth to the temperature at the surface. DIGITAL FLOW MODELS The purpose of modeling the hot-springs flow system was to test several hypotheses regarding the nature of the flow system. Each hypothesis requires details of aquifer geometry and various combinations of hy- drologic variables, such as area and depth of circula- tion, hydraulic conductivity, and porosity. The model analysis provides a means for estimating values of these variables that can be compared with observed data for the hot springs (such as temperature, dis- charge, 14C concentration, and silica concentration). The equation for the heat-flow model is XD MA1+MA0 H=KN __ X,-X0 __— + X3—X0 YD 2 (MA3+MAO» YD << > *— + — X2_Xo #— + 2 XD 2 (Xrthfli—q—Zfi—Mi» + (XD) (YD) TTOP—X0+DH0 DP0 + SP (X1Q1+X2Q2+X3Q3+X4Q4_X0Q0> The subscripts in the preceding equation refer to node location (a node is the center of a volume element); 0 refers to the nodal location being considered, 1 refers to the node above, 2 refers to the node to the right, 3 refers to the node below, and 4 refers to the node to the left. In the preceding equation, H is the heat-flow residual, in British thermal units per day; KN is the thermal con- ductivity, in British thermal units per day-foot-degree Fahrenheit; X is the temperature, in degrees Fahrenheit; XD is the horizontal node spacing, in feet; YD is the vertical node spacing, in feet; MA is the thickness of the aquifer (equivalent to thickness of the model), in feet; TTOP is the temperature at land sur- face, in degrees Fahrenheit; DP is the depth to the midpoint of the aquifer (average of the depth to the top and the depth to the base), in feet; DH is the tempera- ture gradient at the base of the aquifer, in degrees Fahrenheit per foot; SP is the volumetric specific heat of water, 62.4 British thermal units per degrees Fahrenheit-cubic foot (4.18X 106 joules per degrees Celsius-cubic meter); Q1, 2, 3, 4 is the inflow of water from node 1, 2, 3, and 4, respectively, in cubic feet per day; and Q0 is the outflow of water from the element, in cubic feet per day. The equation for the ground-water-flow model is XD PIMA1+P0MA0 T = — ———————— sl—s0 + YD 2 (PgMA3+P0MA§<105 Silica ____________________________ mg/L__ 40.7 41.7 Carbon-14 age ______________________ yr- _ 5,026 4,430 Maximum head in recharge area above hot springs __________________ ft“ 18 260 The results of the Bigfork Chert model indicate that the known and assumed values of the parameters are within the hydrologic constraints on the system. The model indicates that the available head difference is ample to provide the hydraulic pressure for the spring flow. An additional component of head is produced by the decreased density of the water at temperatures higher than the assumed temperature in the recharge area (18.3°C (64.9°F) ). The ample head available to drive the flow system provides a leeway in several of the estimated values of parameters used in the model, such as thickness and permeability of the aquifer and rate and areal extent of the recharge. For example, a valid model could be designed using lower values for permeability and thickness of the Bigfork Chert, or a valid model could be designed using a lower recharge rate over a larger area or using a higher recharge rate over a smaller area. The critical constraint on the Bigfork Chert model is the high heat flow needed to obtain the observed spring temperature. It was not possible to obtain the 62°C—discharge temperature using a uniform heat inflow of 2.2 hfu, which as discussed previously is the maximum known value of crustal heat flow in the Eastern United States. This heat-inflow deficiency suggests consideration of an alternative model of the Bigfork Chert in which the ground-water flow system supplying the hot springs absorbs heat throughout a larger area. The Arkansas Novaculite was modeled as a nearly vertical two-dimensional plane inclined slightly to the southeast (fig. 13). Flow was modeled to a point in the subsurface near the springs from which flow to the sur- face is probably rapid and through a very permeable zone. Following is a summary of input data for a model of the Arkansas Novaculite: Thickness of aquifer ____________________ ft-_ 500 Porosity ____________________________________ 0.1 Hydraulic conductivity ________________ ft/d“ 1. Thermal conductivity __________ cal/s-cm—“C _ _ 0.0062 Heat flow ____________________________ hfu- _ 2.2 Recharge rate ________________________ ft/yr, _ 0.63 Results of the Arkansas Novaculite model are sum— marized in table 11. One constraint on this model is that the required difference in head between the re- charge area and the springs (1,222 ft, or 372 m) exceeds the available head by a factor of 3. This required head difference is also much larger than the head provided by the water level in the recharge area (420-ft, or 128-m, maximum) and the temperature differential (estimated to be about 160 ft, or 49 m). 1 Known values of hydraulic conductivity of the Arkansas Novaculite indicate that the average value cannot reasonably be expected to exceed 1 ft/d, 0.30 m/d, (at 18.3°C) throughout an area as large as that modeled. lFor g0: = 0.02 and L=8,000 fl, 8H=0.02 (8,000 ft): 160 ft, where 8;) is the change in density, p o is the initial density of the water, L is the depth of circulation, and 81-1 is the head provided by the temperature differential. TABLE 11.—Comparison of the modeled and observed data for the hot springs using the Arkansas Novaculite as the principal aquifer Parameter Modeled Observed Temperature ________________________ °C_ _ 54.4 61.7 Flow ______________________________ its/d- _ 0.85 X 105 1.1 X 105 Silica ____________________________ mg/L__ 36.0 41.7 Carbon-14 age ______________________ yr-- 2,760 4,430 Maximum head in recharge area above hot springs __________________ ft- - 1,222 420 ln————— Recharge area “—4 1 l l l w E // Arkansas Novaculite Hot springs ‘__/ \ \ \ \\\ Heat inflow GENERALIZED CROSS SECTION OF PROTOTYPE FOR ARKANSAS NOVACULITE FLOW MODEL Water inflow (recharge) Heat and water outflow (hot springs) \\~\\ \ \\ Heat inflow DIAGRAMMATIC VlEW 0F CROSS-SECTIONAL DIGITAL MODEL OF FLOW IN ARKANSAS NOVACULITE FIGURE 13.—Model of flow in the Arkansas Novaculite. C32 The Arkansas Novaculite model could also be re- jected on structural considerations. To obtain a dis- charge temperature within 10 percent of the observed spring temperature, the depth of circulation in the novaculite model was 8,000 ft (2,440 m), and heat was added over the lower surface of the slightly inclined plane, 1.52 by 26 mi, or 40 mi2 (2.45 by 41.8 km, or 102.4 km2) (as indicated in fig. 13) at a uniform rate of 2.2 hfu and removed over the upper surface of the plane at rates proportional to the temperature difference be- tween each node and the land surface. Structurally, a continuous aquifer in the Arkansas Novaculite with this lateral extent and depth is of doubtful validity. DISCUSSION OF MODELS Neither the Bigfork Chert model nor the Arkansas Novaculite model, previously described, is entirely satisfactory in describing the hot-springs flow systems. In terms of hydrologic and structural constraints on the model, the Arkansas Novaculite model is the least plausible, whereas the Bigfork Chert model is the most suitable. In terms of thermal considerations, both models require abnormally high heat flow to achieve the observed temperature of the hot springs. An alternative Bigfork Chert model, which could match the observed discharge temperatures with ob- served heat flows in the Eastern United States, would involve lateral circulation of ground water over an area on the order of 50 mi2, which, in turn, would re- quire that the recharge areas be distributed over a larger region than in the Bigfork Chert described pre- viously. Recharge from distances as great as 5— 10 mi would probably be sufficient. Such a flow system may be possible along the plane of the thrust fault(s) that intersects the hot springs and dips to the northwest (pl. 1). The criteria for the recharge areas along the thrust fault(s) would be that recharge occur at distances as great as 5—10 mi north and northeast of the hot springs at elevations above the spring discharge and that there be hydraulic connection with the fault plane at suffi- cient depth to allow water to move laterally toward the springs. Ground water movement would be relatively slow to the point where it intercepts the thrust fault plane. Movement along the fault plane would be rela- tively rapid to the points of spring emergence. Such a model would meet the age requirements indicated by 140 analyses. CONCLUSIONS The hot springs of Hot Springs National Park, Ar- kansas, issue from the Hot Springs Sandstone Member of the Stanley Shale of Mississippian age at the crest of a plunging overturned anticline, along the southern GEOHYDROLOGY OF GEOTHERMAL SYSTEMS margin of the Ouachita anticlinorium. The combined flow of the 47 hot springs ranges from 750,000 to 950,000 gallons per day (3.28>< 10‘2 to 4.16x 10‘2 cubic meters per second). The temperature of the combined hot-springs water is about 62° Celsius. Using silica concentration as an indicator, the maximum tempera- ture of the water at depth is estimated to be not more than a few degrees Celsius higher than the tempera- ture at the surface. Silica concentrations of the water since 1901 and maximum temperature measurements since 1804 indicate a small decline in temperature with time. Tritium and Carbon-14 analyses of the water indi- cate that the water is a mixture of a very small amount of water less than 20 years old and a preponderance of water about 4,400 years old. The radioactivity and chemical concentration of the hot springs are similar to that of cold-water springs in the area. The dissolved solids concentrations range from 175 to 200 milligrams per liter. The main differences in the quality of the hot water, compared with the nearby cold ground water, are the high temperature and higher silica concen- trations of the hot springs. The geochemical data, flow measurements, and geologic structure of the region support the concept that virtually all the hot-springs water is of meteoric origin recharged locally. Recharge to the hot-springs artesian system is by infiltration of rainfall in the out- crop areas of the Big fork Chert and Arkansas Novacu- lite. The water moves slowly to depth where it is heated by contact with rocks of high temperature. Highly permeable zones, related to jointing or faulting, collect the heated water in the aquifer and provide av- enues for the water to travel to the surface. REFERENCES Albin, D. R., 1965, Water-resources reconnaissance of the Ouachita Mountains, Arkansas: US. Geol. Survey Water-Supply Paper 1809—J, 14 p. Arndt, R. H., and Damon, P. E., 1953, Radioactivity of thermal wat- ers and its relationship to the geology and geochemistry of uranium: Arkansas Univ. Inst. Sci. and Technology Ann. Prog. Rept. to US. Atomic Energy Comm., 28 p. Arndt, R. H., and Stroud, R. B., 1953, Thrust faulting near the Hot Springs, Hot Springs, National Park, Arkansas, app. 3, in Amdt, R. H., and Damon, P. E., Radioactivity of thermal waters and its relationship to the geology and geochemistry of uranium: Ar- kansas Univ. Inst. Sci. and Technology Ann. Frog. Rept. to US. Atomic Energy Comm., 28 p. Barnes, Ivan, 1964, Field measurement of alkalinity and pH: U.S. Geol. Survey Water—Supply Paper 1535—H, 17 p. Bedinger, M. 8., Pearson, F. J., Jr., Reed, J. E., Sniegocki, R. T., and Stone, C. G., 1974, The waters of Hot Springs National Park, Arkansas—their origin, nature, and management: U.S. Geol. Survey open-file report, 102 p. WATER OF HOT SPRINGS NATIONAL PARK, ARKANSAS—NATURE AND ORIGIN Bedinger, M. S., Reed, J. E., and Griffin, J. D., 1973, Digital- computer programs for analysis of ground-water flow: U.S. Geol. Survey open-file report, 85 p. Bedinger, M. S., Sniegocki, R. T., and Poole, J. L., 1970, The thermal springs of Hot Springs National Park, Arkansas—Factors af- fecting their environment and management: U.S. Geol. Survey open-file report, 74 p. Boltwood, B. B., 1905, On the radioactive properties of the waters of the springs on the Hot Springs Reservation, Hot Springs, Ar- kansas: Am. Jour. Sci., v. XX, no. 115, p. 128—132. Branner, J. C., 1892, The mineral waters of Arkansas: Arkansas Geol. Survey Ann. Rept. for 1891, v. 1, 144 p. Brown, Eugene, Skougstad, M. W., and F ishman, M. J ., 1970, Methods for collection and analysis of water samples for dis- solved minerals and gases: U.S. Geol. Survey Techniques Water Resources Inv., book 5, chap. A1, 160 p. Bryan, Kirk, 1922, The hot water supply of the Hot Springs, Arkan- sas: Jour. Geology, v. 30, p. 425—449. 1924, The temperatures of hot springs and the sources of their heat and water supply, the Hot Springs of Arkansas: J our. Geol- ogy, v. 32, no. 6, August-September, p. 453—454. Craig, Harmon, 1961a, Standard for reporting concentrations of deuterium and oxygen-18 in natural waters: Science, v. 131, p. 1833— 1834. 1961b, Isotopic variations in meteoric waters: Science, v. 133, p. 1702—1703. Davis, W. E., 1939, Measurement of precipitation above forest canopies: Jour. Forestry, v. 37, no. 4, p. 324—329. Fellows, T. L., 1966, The geology of Hot Springs National Park and vicinity, north-central Arkansas: Dallas, Tex., Southern Methodist Univ., Dept. Geol. Sci., unpub. rept., 27 p. Fisher, D. W., 1968, Annual variations in chemical composition of atmospheric precipitation in eastern North Carolina and south- eastern Virginia: U.S. Geol. Survey Water-Supply Paper 1535—M, 21 p. Foumier, R. 0., and Rowe, J. J., 1966, Estimation of underground temperatures from the silica content of water from hot springs and wetsteam wells: Am. Jour. Sci., v. 264, p. 685-697. Glasgow, William, Jr., 1860, Hot Springs of Arkansas, with their Travertine Formation: St. Louis, Mo., L. Gast Bro. and Co. Hamilton, J. B., 1932, Report of construction of hot water collecting and distributing system, Hot Springs National Park, Hot Springs, Arkansas: Natl. Park Service unpub. rept. Hamilton, J. B., and Blood, R. S., 1931, Plan and profile of hot water collection system, Hot Springs National Park, Arkansas: Unpub. maps in Park Superintendent’s files, Hot Springs Natl. Park, Ark., 3 sheets, dated August 14, 1931. Haywood, J. K., 1902, Report of an analysis of the waters of the Hot Springs on the Hot Springs Reservation: U.S. 57th Cong, 1st sess., Senate Doc. 282, 78 p. Hoover, M. D., 1944, Effect of removal of forest vegetation upon water yields: Am. Geophys. Union Trans, pt. 6, p. 969—975. 033 Horton, R. E., 1923, Transpiration by forest trees: Monthly Weather Rev., v. 51, p. 571—581. Kuroda, P. K., 1953, Radioactivity tables of some natural waters: Arkansas Univ. Inst. Sci. and Technology, 58 p. Kuroda, P. K., Damon, P. E., and Hyde, H. I., 1953, Radioactivity of the spring waters of Hot Springs National Park and vicinity in Arkansas, app. 2, in Arndt, R. H., and Damon, P. E., Radioactiv- ity of thermal waters and its relationship to the geology and geochemistry of uranium: Arkansas Univ. Inst. Sci. and Technology Ann. Prog. Rept. to U.S. Atomic Energy Comm, 28 p. McGuinness, C. L., 1963, The role of ground water in the national water situation: U.S. Geol. Survey Water-Supply Paper 1800, 1121 p. Meinzer, O. E., 1923, Outline of ground-water hydrology with defini- tions: U.S. Geol. Survey Water-Supply Paper 494, 71 p. Minckler, L. S., 1939, 'I‘ranspiration of trees and forests: J our. Fores- try, v. 37, no. 4, p. 336—339. Owen, D. D., 1860, Second report of a geological reconnaissance of the middle southern counties of Arkansas: Philadelphia, Pa., C. Sherman & Sons, p. 18—27. Pearson, F. J ., Jr., Bedinger, M. S., and Jones, B. F., 1972, Carbon-14 ages of water from the Arkansas hot springs, in Intemat. Conf. on Radiocarbon Dating 8th Proc., Royal Soc. New Zealand, Wel- lington: p. 30—341. Pearson, F. J ., Jr., and Fisher, D. W., 1971, Chemical composition of atmospheric precipitation in the Northeastern United States: U.S. Geol. Survey Water-Supply Paper 1535—P, 23 p. Pearson, F. J ., Jr., and Hanshaw, B. B., 1970, Sources of dissolved carbonate in groundwater and their effects on carbon-14 dating: Isotope Hydrology, 1970, Internat. Atomic Energy Agency, Vienna, p. 271—285. Purdue, A. H., 1910, The collecting area of the waters of the Hot Springs, Hot Springs, Arkansas: Jour. Geology, v. 18, p. 279— 285. Purdue, A. H., and Miser, H. D., 1923, Description of the Hot Springs district [Arkansas]: U.S. Geol. Survey Geol. Atlas, Folio 215. Schlundt, Herman, 1935, The radioactivity of the spring waters on the Hot Springs Reservation, Hot Springs, Arkansas: Am. Jour. Sci., v. XXX, no. 175, p. 45—50. Scully, F. J ., 1966, Hot Springs, Arkansas, and Hot Springs National Park: Little Rock, Ark., Pioneer Press, 446 p. Sheppard, S. M. F., Nielsen, R. L., and Taylor, H. P., Jr., 1969, Oxy- gen and hydrogen isotope ratios of clay minerals from porphyry copper deposits: Econ. Geology, v. 64, p. 755—777. Truesdell, A. H., and Jones, B. F., March-April 1974, WATEQ, a computer program for calculating chemical equilibria in natural waters: U.S. Geol. Survey Jour. Research, v. 2, no. 2, p. 233—248. Weed, W. H., 1902, Geological sketch of Hot Springs, Arkansas: U.S. 57th Cong, 1st sess., Senate Doc. 282, p. 79—94. White, D. E., Barnes, Ivan, and O’Neil, J. R., 1973, Thermal and mineral waters of nonmeteoric origin, California Coast Ranges: Geol. Soc. American Bull., v. 84, p. 547—560. G PO 689-03 5 UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 1044—C GEOLOGICAL SURVEY F E ET M ETERS 1000 300 500 150 SEA LEVEL SEA LEVEL F E ET M ET ERS —1000 300 ~ 500 1 50 500 SEA LEVEL SEA LEVEL EXPLANATION Ms Stanley Shale MISSISSIPPIAN Hot Springs Sandstone Member of Stanley Shale ,,,,, MDa DEVONIAN Arkansas Novaculite 0730” _ / SILURIAN Missouri Mountain Shale, Blaylock Sandstone (Silurian), and Polk Creek Shale (Ordovician), un— differentiated “““ . , I 02'30” > ORDOVICIAN Obf 3:??? Bigfork Chert Geology mapped by C. G. Stone in November through May 1973 and C. G. Stone and B. R. Haley in March 1970, as Ow modified from Purdue and Miser (1923) and Arndt and Womble Shale Stroud (1953) (Shown in sections only) (we-v"? vwga gm firm m Faults T indicates overthrust side of thrust faults. Arrows indicate relative directions of movement Contact Boundary of Hot Springs National Park Base from U.S. Geological Survey SCALE 1:24 000 1 1/2 0 Hot Springs North, 1966, and Hot Springs South, 1966 ‘ ' ' ' ' ' r l J z I MILE ‘-I—‘ l .5 O 1 KILOMETER I—l l—l I——-l l-—l I--l CONTOUR INTERVAL 20 FEET NATIONAL GEODETIC VERTICAL DATUM OF 1929 GEOLOGIC MAP OF HOT SPRINGS NATIONAL PARK AND VICINITY, ARKANSAS Numerical Modeling of Liquid Geothermal Systems By M. L. Sorey GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—D UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1978 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Sorey, M. L. ‘ Numerical modeling of liquid geothermal systems. (Geohydrology of geothermal systems) (Geological Survey Professional Paper 1044-D) Bibliography: p. 23-25 1. Geothermal resources—Mathematical models. I. Title. II. Series. III. Series; United States. Geological Survey. Professional Paper 1044-D GBll99.5$67 551.4’9 78-606183 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-03112-4 CONTENTS Page Page Abstract ______________________________ D1 Studies of natural convection _________________ D8 Introduction ____________________________ 1 Circulatory convection ___________________ 9 Previous investigations ___________________ 2 Box models __________________________ 9 Acknowledgments ______________________ 3 General considerations _________________ 9 Numerical model __________________________ 3 Numerical solutions __________________ 10 Partial differential equations _______________ 3 Temperature-dependent parameters ________ 12 Flow equation ______________________ 3 Effect of density variation with pressure _____ 13 Equation of state ____________________ 4 Realistic models _______________________ 14 Energy equation ____________________ 4 Conducting side walls _________________ 14 Numerical solution _____________________ 5 Two- and three-layer models _____________ 15 General procedure ____________________ 5 Hot-spring models _______________________ 16 Numerical formulation of flow equation _______ 6 Summary and conclusions ___________________ 22 Numerical formulation of energy equation ______ 7 References ____________________________ 23 Iterative solution ____________________ 7 Coupling of flow and energy equations __________ 8 FIGURE 9959'!“pr 9° 21. 22. 23. ILLUSTRATIONS Page Interlacing of temperature and pressure steps _________________________________________ D6 Cellular convection in an extensive horizontal layer _______________________________________ 8 Box model for cellular convection with heating from below __________________________________ 9 Convective rolls in box with hx = 2 hz ______________________________________________ 10 Velocity pattern on lower surface of cube for three-dimensional convection ________________________ 10 Influence of grid spacing on steady state Nu, N = number of nodes in x and y directions ________________ 10 Steady-state temperature distribution in box model with aspect ratio = 1.0, Ra = 100, and uniform (symmetric) initial temperature distribution ___________________________________________________ 11 Steady-state temperature distribution in box model with aspect ratio = 1.0, Ra = 100, and nonuniform (asymmetric) initial temperature distribution ________________________________________________ 11 The relationship between Ra and Nu determined numerically for cellular convection in box model ___________ 11 Nusselt number with temperature-dependent fluid properties as a function of cold—side Rayleigh number (eq. 34, dashed lines) and mean Rayleigh number (eq. 35, solid lines) for temperature differences T. —T., between 40°C and 300°C. Critical Rayleigh numbers at Nu = 1.0. Heavy solid line is for properties independent of T (see fig. 9) ______ 12 . Temperature and velocity profiles at midplane (x = hx/2) for convection cell with Ram = 100 and T| -To = 100°C. Temperature curve A is for constant fluid properties; temperature curve B and velocity curve C are for temperature- dependent fluid properties __________________________________________________ 13 . Box model to test effects of density variation with pressure _________________________________ 13 . Physical model for cellular convection with conducting side walls ______________________________ 14 . Temperature distributions in convection model with conducting side walls for w/ L = 2.1 and y/ L‘= 0.5 ________ 14 . Isotherms for 2-layer model with Ra = 100. Isotherms for 3-layer model with Ra = 100 _________________ 15 . Isolated cylindrical conduit hot spring model __________________________________________ 16 Fault plane hot spring model ____________________________________________________ 17 Geometry of hot spring models __________________________________________________ 17 Relationships between dimensionless flow rates (Mopr) and dimensionless temperature drop (1—¢sp) due to conduc- tive heat loss in cylindrical and fault plane hot spring models __________________________ ._ _ _ _ 18 Steady—state temperature distribution in fault plane hot-spring model with discharge = lO‘kg/d, W = 10 m, D = 1 Km, andRa=0(k=0) ______________________________________________________ 19 Steady-state temperature distribution in fault plane hot-spring model with discharge = 10‘kg/d, W = 10 m, D = 1 Km, and Ra = 0 (k = 0), and H = 10 " cal/sec °Ccm2 at land surface ____________________________ 19 Steady-state temperature distribution in fault plane hot-spring model with discharge = IO‘kg/d, W = 10 m, D = 1 Km, and Ra = 20 (solid lines), Ra = 0 (dashed in figure 20) __________________________________ 20 Steady—state temperature distribution in fault plane hot-spring model with discharge = IO‘kg/d, W = 10 m, D = 1 Km, and Ra = 200 _________________________________________________________ 21 IV TABLE 1. 9‘99”?" Symbol bu s 5 II N mom :9. a . S G Wig?" "a: a: para 5 0'1: Ra CONTENTS TABLES Page Temperature—dependent properties of pure water ________________________________________ D12 Results of computer runs with density a function of pressure _________________________________ 14 Heat transfer results for convection in vertical channel with conducting side walls _____________________ 15 Effect of multiple layers on heat transfer in permeable layer with circulatory convection _________________ '15 Spring temperature, TSP, total conductive heat loss, Q, and heat flux at land surface near spring, q, for selected values of coefficient of surface heat transfer H ____________________________________________ 20 NOMENCLATURE Description Dimensions Symbol Description Dimensions Surface area connecting nodes n L2 Rae Critical Rayleigh number for on- — and m set of convection Compressibility coefficient t"/L2 re Radius of cylindrical hot spring L Specific heat at constant volume Lz/tzT conduit 0f flmd 2 2 T Temperature T Specific heat at constant pressure L /t T Tb Temperature at base of hot spring T of fluid models Specific heat 0f SOlid phase Lz/tZT Ts Temperature at land surface in T Length of fault plane conduit L hot spring models Distance between nodes n and m L 5 Time t Distance between node m and in- L l’d Fluid flux vector or Darcian veloc- L/t terface with node n ity Gravitational acceleration vector L/t2 W Half-width of fault plane conduit L Coefficient of surface heat trans- M/tzTL a Vertical compressibility of porous Ltz/M fer medium Thermal conductivity of solid- ML/t3T Thermal expansivity 0f flUid Tfl fluid matrix Coeff1c1ent in second order equa- T'2 Thermal conductivity tensor for ML/taT I tion of state f . . l' . / solid-fluid matrix nterpoa ion actor in imp 1c1t - — ML /t3T exphc1t numerical scheme 1 = —°—” (5) For slightly compressible fluids such as water, we wish to solve the flow equation in terms of pressure rather than of density. Hence, we expand the right-hand side of ( 5 ) as d¢p _ pa¢ 6P app 6T at ' paP at ”99¢“ M” at (6T at’ _ 0P 6P ' ”a— at + dw— at ”Ba—ff (6) where u = fluid compressibility, B = fluid thermal ex- pansivity, and a = vertical compressibility of the po- rous medium. Following Remson, Hornberger, and M012 (1971) we assume that the velocity of solid parti- cles is negligible and deformation of the solid matrix is significant only in the vertical direction. For the prob- lems considered in this study, the thermal expansivity term (pp/3 aT/at can be shown to be small compared with the compressibility terms (pa + ¢pK) 6P/6t owing to the large difference in response times between the thermal and hydraulic systems, that is aT/at << 6P/at. Thus, we will neglect the expansivity term in GEOHYDROLOGY OF GEOTHERMAL SYSTEMS equation 6 and write the final form of our flow equa- tion as k 2_ k - E V (+p7 VP—pg 7)-C 6t (7) where C = p (¢x + a) = compressibility coefficient. EQUATION OF STATE We wish to consider density as a function of tem- perature and pressure. Following Fernandez (1972, p. 54) we can write p = pee-MT-Tok" (P—Po) (8) where p0 = fluid density at some reference state (To Po). Expanding the exponentials in series form and retaining terms of first order yields p = p0[1—/3(T_To)+K(P-Po)]' (9) For geothermal applications where (T-To) may be large, a more accurate form for the temperature de- pendence was suggested by Wooding (1957, p. 274) as follows: p = poll-6(T—Tol—7(T-To)2]' (10) Evaluating (p0,To) at 25°C and setting ,8 = 3.17 X 10" °C'1 and 7 = 2.56 x 10'6 °C'2, yields p—po) to an accu- racy of 2 percent in the range of 25°C to 300°C. Be- twesn 4°C and 25°C a first order expression can be use . In comparing the magnitude of the effects of tem— perature and pressure on the calculation of fluid den- sity, it is seen that, for liquid geothermal systems, the pressure dependence can usually be neglected. For ex- ample, assuming (P—Po) = 100 kg/cm2 (equivalent to about 1,000 meters of head), K = 4.5 X 10'5 cmz/kg gives (p—po)p0 = x (P—Po) = 0.0045. However, for (T—To) = 100°C and B = 5 X 10" °C“, (p—p,,)/p0 = B (T0—-T )= 0.05, an order of magnitude greater. Thus, in the numerical results that follow, equation 10 was utilized to compute fluid density except as discussed in the section on “Box Models”. It should be noted, however, that in solving the flow equation 7 a portion of the density change with time is always derived from the fluid compressibility term ¢pK6P/at. In effect, then, density changes with pressure contribute to transient pressure changes, but calculated values of density are not updated as pressures change, only as temperature changes. ENERGY EQUATION The energy equation in porous media can be written NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS following Mercer (1973, p. 7) as [dmc + (1—¢> pscsl %§—= vim-VT wide-VT (11) where ps = density of solid phase, 03 = specific heat of solid phase, K m = total thermal dispersion tensor, and c = specific heat at constant volume of fluid phase. This relationship is obtained by combining the corre- sponding energy equations for the solid and fluid phases, as developed by Bird, Stewart, Lightfoot and others (1960, p. 314), with terms accounting for energy increases by viscous dissipation and compression ne- glected. To obtain (11) several assumptions are made. First, we assume that the fluid and solid are in ther- mal equilibrium at each point of contact so that a sin- gle temperature, T, can be assigned to the sat- urated medium in a REV. This assumption has been shown to be valid for most porous media problems (Combarnous and Bories, 1973; Holst, 1970; and Green, 1963). However, a single energy equation may not be adequate under certain circumstances, such as forced convection at high flow rates through porous material with relatively low thermal conductivity (Curry, 1970; Combarnous and Bories, 1973). One can also envision difficulties in treating ground-water flow through fractured rock as porous media flow necessi- tating the use of an effective REV which is not small compared with the overall size of t=he system. The flux of heat represented bme-VT is a compos- ite of heat conduction through the solid-fluid matrix and the spreading effects of dispersion in the fluid phase. Bear (1972, p.650) gives the components of the total thermal dispersion tensor as Km =Kc+¢ Kd (12) where RC = therrgal conductivity tensor for solid- fluid matrix and K; = coefficient of thermal disper- sion. Bear states that heat transfer by dispersion is analogous to mass transfer by hydrodynamic disper- sion and results from the distribution of local veloci- ties caused by the presence of grains and molecular diffusion. Mercer (1973, p. 10) develops the functional form of the thermal dispersion tensor as Kdij = fijkl UZUI (13) where fijkl = dispersivity of the medium, a fourth rank tensor, and 0,2,0, = components of velocity in the k and l directions, and v = magnitude of the velocity. Green (1963) has shown analytically and experimen- tally that for relatively uniform porous media, heat transfer due to dispersion is small compared with con- D-5 duction for values of velocity normally encountered in ground water systems. Unfortunately, it is difficult to apply these laboratory results to field situations where considerable variability in permeability may increase the dispersive effect. Indeed, it is possible that where dispersion is important in heterogeneous systems, the assumption of thermal equilibrium between phases may also break down and a more complicated math- ematical description for heat transfer will be required. With this in mind, and considering the absence of in- formation as to the magnitudes of the tensorigl com- ponents of (12), it seems reasonable to treath as a scalar Km, which could represent an enhanced con- ductivity to account for dispersion. For example, Hen- ry and Hilleke (1972, p. 33) used a value of Km which was five times the static thermal conductivity to match the temperature distribution in a coastal salt water-fresh water aquifer. Thus, the energy equation 11 becomes 5% = v-KmVT—padc-VT (14) where E = [¢pc + (1—¢) pscs] = heat capacity of solid-fluid matrix. ' NUMERKMLSOLUUON The flow equation 7 is linear only if we neglect the pressure dependence of p and p; however, both p and u must be treated as functions of temperature, and thus the pressure and velocity fields depend on tempera- ture. The energy equation 14 is nonlinear because of p (T) and c (T), and the temperature field depends on the velocity field. Thus, we must solve a coupled non- linear set of equations. Further, in the geothermal sys- tems we wish to study, properties such as permeability and thermal conductivity will vary spacially and the boundary geometry may be complex. For these rea- sons numerical solutions are required, and a code which can treat multidimensional problems with vari- able geometries would be desirable. A computer pro- gram for slightly compressible heat and fluid flow was therefore developed. GENERALPROCEDURE The basic procedures in the program were derived from the work of Edwards (1969) and Lasseter (1974). Before writing finite difference approximations, we integrate the equations over the volume of a finite ele- ment and apply the divergence theorem to reduce the flux terms to integrals over the area of the element. Thus, for an equation of the form a). a —— = V'flVh at (15) D-6 integration yields, ag—i‘dv=fBVA-dZ- (16) vol area To solve this equation numerically, we discretize the continuum into a finite series of elements, each with a specified volume and a surface area between it and surrounding elements. We must then assume that the volume-integrated factors on the left-hand side are constant over each element, and the surface-integrat- ed terms are reasonably constant over each of the con- necting areas. Consider an element “n” and two adjacent elements m1 and m2 as shown below Writing equation 16 in difference form for this system yields AAn VOln arr—Ki— = An,m1 6n,m1 (vx)n,mi + An,m2 fin,m2 (vx)n,m2 =;An,m Bram (vx)n,m ' (17) To evaluate the flux at each interface, we assume that the average values of [3 and A over the element apply at a particular point or node within the element. The gradient (VA)n,m is approximated as (WOW, = 4m n,m+dm,n (18) In general the distances dmm and dm,n are measured normal to the interface Amm. To preserve continuity of flux and potential across the interface, the coefficient 16”,", must be evaluated as the harmonic mean of the values 13,1, for node m and 6,, for node n. Thus _ Dn,m finfim fln’m _ 3ndm,n+ Bmdn,m where Dmm = dmm + dmfl. Equation 17 can now be written (19) Vol AN; = Ammfinflm (X —>\ ). ”0'" At m 6ndm,n+6mdn,m m " (2°) GEOHYDROLOGY OF GEOTHERMAL SYSTEMS The system of algebraic equations generated by writing equation 20 for each node in the system is solved by iteration. This general procedure puts no in- herent restrictions on the number of nodes connected to node “n” or on the shape of the connecting surfaces. Thus it can be applied, as noted previously, to multi- dimensional problems with simple or complex ele- ment geometries. Narasimhan (1975) refers to this scheme as the IFD (Integrated Finite Difference Method) and gives a detailed conceptual comparison of the IFD method with the finite element method. NUMERICAL FORMULATION OF FLOW EQUATION The flow equation is k — k 6P . -—V _ 2 _ = _. 7 V(+p#Ppgu) Cat () After volume integration, the difference equation can be written AP — kA P —P On VOlnA—tn=;l(pT)n,m -(JDT’ln) (21) +( pzkfng) n,m l where ”n,m is the direction cosine of the angle between the outward normal of “n” towards “m” and the gravi- tational acceleration vect'or g. Evaluating (Ir/Mn”, as the harmonic mean Dmm (‘2‘)" (%)m WWW) (k) 7 ndm,n+ 7 mdn,m and weighting the density by (22) = dm,n 9n. + dn,m Pm (23) n,m pn,m we can rewrite equation 21 as Cn Voln AT? = Z ‘I’n,m (Pm _ Pn) + “n,m (24) m k where \Iln’m = transductance = [— pn’m(7 n,m A k D" m ] and “n,m = [pzn’m (7 )n,m An,m "n,m gl- n,m We wish to solve the flow equation implicitly. Thus, we expand equation 24 into implicit and explicit parts as NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS AP At" = Z {[‘I’n,m(Pm-Pn) +wn,ml m + e[ ‘I’n,m (APm - APn)l} Cn Vol,, (25) where P is the pressure at the old time and AP is the change in P during the time step At. The interpolation factor 9 can be set to 1 for a fully implicit or backward differencing scheme, 1/2 for a Crank-Nicholson implic- it scheme, or 0 for a fully explicit scheme. In solving the flow equation, 9 is set to 1. The explicit part of equation 25 is computed first and then the implicit part is solved by iteration. NUMERICAL FORMULATION OF ENERGY EQUATION The energy equation 14 is —6T [)0 T = V 'KmVT - pUdC‘VT ‘ (14) Before integrating, we rewrite this equation in diver- gence form as E % = V~KmVT - V'pfidcT + T V'pidc - (26) The corresponding difference equation is (Vol gallifti; ; [(KA)”, 3%,? + (pvch) n,m (Tmm _ Tn)] where velocity is defined as positive into node n and Kn)", is the harmonic mean conductivity. Expanding into implicit and explicit terms yields (27) (Von—c)" % = 2m: {[(KA)n,m fififl n,m + (PUdCA)n,m (Tmm ‘ Tn” (28) + 9[(KA)n,m ET_rIn)-_AT_n) n,m + (pUdCA)n,mATn,mll ' Temperature-dependent material properties are eval- uated at some int rmediate time between t and t +At. The procedure b which this point is determined is discussed by Las§eter (1974, App. C). The value of 9 used to solve the nergy equation is adjusted between 0.57 and 1.0 in or er to damp out small oscillations in the solution. The temperature at the interface Tn,m could be evaluated by a linear interpolation D-7 = dm,nTn + dn,me . T nm ’ Dn,m (29) It can be shown, however, and anticipated physically, that the temperature variation in the presence of con- vection is not linear and that the interface tempera- ture is weighted in favor of the upstream node temperature. Numerically, it is also found that insta- bilities develop unless we use this upstream weighting. Thus, we rewrite equation 29 in terms of the adjusted distances d’mm and d’m’n as I / d m,nTn + d n,me (30) Dn,m Tn,m = The energy equation now becomes (Vol a), AA? = 2 [91mm Tm + (:22 — pvch)n,an] m (31) +e[91,,,mATm + 92,,,,,,AT,,] in terms of transductances 91mm = [(%m)+ (pvch)n’m 33mm] n,m and 92mm = [ _ ( KDAn m)+ (PUdCA)n,m-$;—’:l- The instability noted above may develop if 91",", and 92",," have the same sign. To avoid this, we weight the interface temperature in favor of the up- stream node temperature as follows. If (pvdlnm , the mass flow rate from node “m"to “n”, is large and posi- tive, the transductances may both be positive. In this case the weighting d’m , n /D m," on Tm is reduced until 92".," is just negative. If (pud)n’m is large and negative, the transductances will both be negative and the weighting d’n’m/Dnm on Tm is reduced until 91",”, is positive. This computation is done internally by the computer program so that stability is always assured. The accuracy of this weighting technique was verified by Lasseter (1974) by comparing numerical results with analytical solutions for forced convection through porous media. ITERATIVE SOLUTION For each time step, the flow equation 25 is solved in a fully implicit fashion with the interpolation factor 9 set to 1. With the energy equation 31, 6 is adjusted between 0.57 and 1.0 and an implicit-explicit op- timatization method is used. In the latter, a stability limit or time constant Tn is defined for each node as D-8 1 _ (Volp_c)n n- 9* where 9*",m signifies that we sum the conductive components of Q n. m over all connections and take the difference between inflow and outflow of the convec- tive components. 1-,, represents the largest stable time step that could be used to solve equation 31 explicitly for node n. Physically, it is the approximate time re- quired for a node to react significantly to changes in temperature of nodes it is connected to. Thus, for each time step, At, the energy equation is solved explicitly for nodes with 7,, 0, provided the adiabatic temperature gradient 6T/62 = gTfl/cp is exceeded (Landau and Lif- shitz, 1959, p. 9). However, we would expect convec- tive effects on the thermal and hydrologic regimes to be negligible except in geothermal areas with above normal heat flow and temperature gradients. Donald- son (1968) estimates that in geothermal areas the Rayleigh number is in the range of 500 to 5,000; and Home and O’Sullivan (1974) estimate that in the Wairakei, New Zealand geothermal reservoir Ra is near 5,000. Previous investigations of cellular convection in po- rous media have treated either the infinite layer case or box models with insulated, impermeable vertical sides. Results from box model studies have applica- tion to both laterally extensive aquifers where the zone of influence of each of the convection cells forms a box with respect to adjacent cells and to structurally bounded aquifers or channels. The effects of assump- tions of isothermal surfaces above and below the aquifer and insulated side walls are discussed in the section on “Realistic Models.” In the next section, a general description of heat and fluid flow in box model systems will be given and numerical results from the program developed for this study will be compared with analytical, experimental, and numerical results from other authors. BOX MODELS GENERAL CONSIDERATIONS The geometry and boundary conditions for the box model are shown in figure 3. All sides are impermeable and the vertical sides are insulated. Following the analysis by Beck (1972), the problem of the onset of convection is solved by writing steady state forms of equations 7 and 14 for static conditions (no convec- tion) and for a small disturbance upon the static state, under assumptions of constant coefficients and con- stant fluid density except in the gravity term of 7. The two sets of equations are subtracted and the resulting set of equations nondimensionalized. Solutions have been obtained by the linear and energy methods, both yielding the same result for the box model problem D-10 Z fl=v =0 T=T.,vz=0 BY y 1 / l / ‘ 6T_ _ l 532”?" QI=0 hz / —————————— - _ _ _ 3X «— " // h Vx=0 / \ h y / / \\ x X 3§=vy=o T=T1 ,vz=0 Y FIGURE 3.—Box model for cellular convection with heating from below. (Beck, 1972). The critical Rayleigh number, which is the smallest eigenvalue of this eigenvalue problem, depends entirely on the aspect ratios hX/hz and h /hz. The minimum value of Rac is 41r2 for the infinite ayer case, and bounding the fluid tends to make it more stable, that is Rae >412. For values of Ra just above the critical value, several modes of convection are possible. The motion can be either 2- or 3-dimensional with single or multiple cells. In contrast with the corresponding problem of free convection of an enclosed homogeneous fluid, 2-di- mensional motion is possible because a no-slip bound- ary condition at the walls is not required for the porous media system. A roll is defined as a cell with only two non-zero velocity components. It has the ap- pearance of a cylinder with axis parallel to the zero velocity coordinate. Beck’s (1972) solution to the minimum eigenvalue problem shows that rolls are preferred over 3-dimensional cells whenever the height 112 is not the smallest dimension. When rolls do form, they are usually parallel to the shorter side, al- though the overriding rule is for the number of rolls and the direction of their axes to be such that each roll has the closest approximation to a square cross section as possible. For example, for bx = 2.0, by = 0.75, and 112 = 1.0, two rolls form parallel to the y-axis as shown in figure 4. Three-dimensional cells occur when 12x, by, and 112 are nearly the same size, that is, a cube, and when the height 112 is less than both lateral dimen- sions. For a cube, the motion resembles a toroid with vertical axis through the center of the box. The veloc- ity pattern in the (x, y) plane is shown in figure 5. One effect of this convective motion is to increase the heat transferred vertically through the system. The Nusselt number Nu is defined as the ratio of heat flow with convection to heat flow by conduction in the absence of convection. For Ra- 7 _ 6 _ Combarnous and / o: Bories (1973) Lu 5 — m :2) 4 — Z 5 3 - Lu (n (n D Z 2 — <——-THE PROGRAM 1 | J J 1 | 1 10 20 40 100 200 400 1000 2000 RAYLEIGH NUMBER FIGURE 9.—The relationship between Ra and Nu determined numerically for cellular convection in box model. The existence of oscillatory or fluctuating convec- tive states has been investigated by Home and O’Sul- livan (1974) and Caltagirone, Cloupean, and Combarnous (1971) for box models in which the lower boundary is uniformly and nonuniformly heated. Their work includes experimental and numerical ob- servations of regular and irregular variations in the convection cell pattern and the resultant Nusselt number at Rayleigh numbers above about 300. Real time periods for these fluctuations at Ra near 1,000 were approximately equal to the expression (0.006 Lap—c/Km), which yields a range from 200 to 700 years for typical thermal properties in systems with L near 1 sz. Corresponding variations in Nusselt number of i 15 percent were found. Numerical simu- lations using the program for the uniformly heated case indicate similar instabilities at a Rayleigh num— ber of 1,000 for motion which is dominantly unicellu- lar, although the Nusselt number variation was only :6 percent. Thus, from the standpoint of modeling natural convection in geothermal systems, the rather large periodicity and limited quantitative effect on heat flow associated with this oscillatory phenomenon suggest that its practical significance may be limited. TEMPERATURE—DEPENDENT PARAMETERS In previous studies of the critical Rayleigh number and the Ra-N u relationship, the properties it and c were treated as constants and the first-order equation of state (9) was used to relate fluid density to tempera- ture. In the program the relationships M (T) and c (T) can be tabulated or suitable analytical expressions can GEOHYDROLOGY OF GEOTHERMAL SYSTEMS be solved at each thermal time step. Table 1 lists val- ues of p, c, and [.t for pure water at temperatures cover- ing the range found in hydrothermal systems. The viscosity variation with temperature is considerably larger than that for density or specific heat, and the variation suggests that results obtained with a con- stant viscosity assumption may be significantly in er- ror. Actually, thermal expansivity 6, which appears in the equation (33) for Rayleigh number, increases con- siderably with temperature and further complicates the analysis. TABLE 1. Temperature-dependent properties of pure water [From Dorsey (1968)] T P p u 0 (°C) (atm) (g/cma) (g/m~s) (cal/°C g) 10 ————— 1 1.000 1.307 1.000 60 ————— 1 .983 .467 .951 160 ————— 50 .910 .174 .833 260 ————— 50 .788 .109 .763 310 ————— 100 .692 .090 .741 The combined effects of the temperature depen- dence of p, c, M., and 6 on the analysis of cellular con- vection were evaluated for a box model with aspect ratio of 1.0. The N u-Ra relationship was examined for several values of the vertical temperature difference (Tl-T0), using an average temperature in each case of 160°C. The second-order relation (eq. 10) for p(T was used and permeability k was adjusted to keep Ra con- stant as (Tl-T0) varied. For each run, two Rayleigh numbers were calculated. One could be termed the cold-side Rayleigh number Rao, where (34) R00 = gkL(T1_To) [polfofio ] 0 m and the subscript (0) indicates evaluation at T = To. A mean Rayleigh number Ram was also computed as gkL(T1_To) [ pozcmfie] I‘m m Ram = (35) where the subscript (m) indicates evaluation at T = (Tl-T0)/ 2 = 160°C, and Be is an effective expansivity computed as p0_pl po(T1—To) . (36) fie= For each (Tl—T0), eight-point tabulations of #(T) and C(T) were used with linear interpolation between points. The results are shown in figure 10. The heavy solid curve is the same as the curve in figure 9 for the con- stant parameter case. We note that for a given system, computing the Rayleigh number based on M, c, and B NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS 10: 8— / 6_/ I / LU _ E 34‘ Z I— _ _l u.I U) ‘3 Z 2— / / / / 1 I l | I 10 20 4O 60 100 1000 RAYLEIGH NUMBER FIGURE 10. — Nusselt number with temperature-dependent fluid properties as a function of cold-side Rayleigh number (eq. 34, dashed lines) and mean Rayleigh number (eq. 35, solid lines) for temperature differences Tl-T0 between 40°C and 300°C. Critical Rayleigh numbers at Nu = 1.0. Heavy solid line is for properties independent of T (see fig. 9). evaluated at T0 yields a value of Nu which increases with (T,—T0) and is greater than the constant param- eter result for (Tl—T0) greater than about 40°C. In contrast, the Nusselt number corresponding to Ram is less than the constant parameter result and decreases with(T,— To). The position of the Nu-Ram curves may vary with the actual average temperature (T1—T0)/2 due to nonlinearities in the p(T), C(T), p(T) relations, but the choice of (T1—To)/2 = 160°C should be repre- sentative of geothermal reservoirs. The explanation for the lower heat transfer with variable fluid proper- ties (Nu-Ram) may also be related to the nonlinearity in p(T) which causes the centroid value of p for T0 5 T 5 T1 to be greater than p. at (T,-To)/ 2. Because Ra is inversely proportional to [1,, the effective Ra should be less than Ram. A similar trend of decreasing heat transfer with increasing (T,-T,,) was reported by Holst and Aziz (1972a) with n-heptane as the convect- ing fluid. Results in figure 10 also show that the critical Ray— leigh number is influenced by variable fluid proper- ties. In terms of the cold-side Rayleigh number Ra0 the onset of convection occurs near the theoretical val- ue of 41I'2 for (T,-—T,,) < about 40°C but occurs at val- ues of Ra < 41rz for (Tl—T0) > 40°C. For (T,—T0) = 300°C, the critical value of Ra0 is only 1.7. However, the critical value in terms of Ram is always greater than 47r2, reaching a value of 61 for a temperature dif- ference of 300°C. Finally it should be noted that vari- able fluid properties also cause the temperature and D-13 DARCIAN VELOCITY (m/d x10+3l —15 -10 -5 0 +5 +10 +15 I I I I '4: I DEPTH (z/hZ) '01 I 1.0 110 l l l l 1 30 1 50 1 70 1 90 21 0 TEMPERATURE, IN DEGREES CELSIUS FIGURE 11.—Temperature and velocity profiles at midplane (x = [Ix/2) for convection cell with Ram = 100 and T,—T0 = 100°C. Temperature curve A is for constant fluid proper- ties; temperature curve B and velocity curve C are for tem- perature-dependent fluid properties. flow patterns to be asymmetric as shown in figure 11. The asymmetry reflects the fact that the hotter, less viscous fluid near the bottom moves more easily than the colder fluid above. EFFECT OF DENSITY VARIATION WITH PRESSURE The effects of allowing fluid density to vary with pressure according to equation 7 were evaluated for the system shown in figure 12. Two runs were made, one with p (T) and one with p = p (T,P), for(T,-T0) = 1°C and 100°C. For a temperature difference of 100°C across the box, the effect of temperature on density should dominate, as discussed in the section on “Par- tial Differential Equations.” However, for a tempera- ture difference of 1°C, the density increase from top to bottom due to the increase in hydrostatic pressure is 10 times the density decrease due to the tempera- ture difference. , The results of the computer runs are listed in table 2. Although the Nusselt number for run 4 with p = p (T,P) is slightly higher than the value for run 3 with p = p (T), the difference in heat transfer between the various runs were negligible. Similarly, temperature and velocity distributions were essentially the same for each run. Thus, for problems or circulatory con— vection in porous media, density can probably be con- sidered a function of temperature only. D-14 1km To/ Ra,=100 T, — T,=1,100°C 11km B=5><10"°C“ u=5X 1o“ ch/kg T.\ FIGURE 12.—Box model to test effects of density variation with pressure. TABLE 2.—Results of computer runs with density a function of pressure Run Tl-TO (°C) p Nu 1 _____ 100 p (T) 2.60 2 _____ 100 p (T,P) 2.60 3 _____ 1 p (T) 2.60 4 _____ 1 p (T,P) 2.61 REALISTIC MODELS Application of results from box model investiga- tions are somewhat limited because the boundary con- ditions are not realistic. In particular, the assumptions of insulated sides and constant tempera- ture at the top and bottom of the convecting layer may not hold in the field problem. It is possible, however, to use the numerical code to evaluate the effects of GEOHYDROLOGY OF GEOTHERMAL SYSTEMS these assumptions on temperature distributions and heat transfer rates. CONDUCTING SIDE WALLS Consider the model shown in figure 13 which is similar to one proposed by Donaldson (1968, 1970) for simulation of convecting geothermal systems. The permeable channel, labeled B, could represent a frac— tured zone of upflowing liquid, separated from a re- charge area with downflowing water by impermeable blocks, labeled A. A horizontal channel connecting the recharge and discharge areas is not shown at the bot- tom of this model because our interest here is on circu- latory convection in the vertical channel. For this reason we also treat the upper and lower surfaces of Channel B as impermeable. A constant temperature T0 is imposed along the top and sides, and a linear temperature variation from T0 to T1 is imposed along the bottom of the blocks A. The bottom of channel B is maintained at T1, and the vertical sides of the chan- nel are impermeable but thermally conductive. For a particular case in which Tl—T0 = 100°C, w/L = 2.1, and y/L = 0.5, and a Rayleigh number for the channel of 100, the temperature distribution shown in figure 14 would result. The temperature distribution for the case of conduction only (Ra = 0) is also shown and was used as a reference in computing the Nusselt number to account for the influence of the distance w and the temperature To imposed at the sides of the model. Heat transfer rates along various boundaries are listed in table 3. The Nusselt number for the channel is 1.3, com- pared with a value of 2.2 determined for the corre- sponding box model with insulated sides as in figure 7. Vz=0 X To To To A B ’ A Z To Impermeable Impermeable L To / /, / vy=0\ 11/ W Y 6_T=T, —T,, Tl a_T:T, —T, a w vz=0 6 w FIGURE 13.——Physical model for cellular convection with conducting side walls. NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS D-15 FIGURE 14.—Temperature distributions in convectio7 model with conducting side walls for w/L = 2.1 and y L = 0.5. TABLE 3.—Heat transfer results for convection in vertical chan- nel with conducting side walls Location Q for Ra = 0 Q for Ra = 100 Nu (joules/d X 107) (joules/d X 10") Top of channel 4.38 5.76 1.32 Bottom of channel 6.41 7.98 1.24 Top of model _ _ _ _ 24.4 26.3 -1.08 Bottom of model _ _ 29.0 31.0 1.07 A simulation was also performed on the box model in figure 7 with a linear temperature variation imposed along the vertical sides. This simulation also allows heat conduction through the sides and for Ba = 100, Nu = 1.8. These results indicate that analysis of circulatory convection in bounded reservoirs should provide for the retarding effect of heat conduction through the lateral boundaries. The basic physical model in figure 13 can also be used to simulate the combination of convective throughflow from recharge to discharge areas and circulatory or secondary convection in the upflow channel (Donaldson, 1968, 1970). Results pro- duced by using the program are in agreement with the general features of Donaldson’s combined throughflow and secondary convection results; these results indicate that the amount of heat transferred by the secondary flow decreases markedly as the throughflow increases. Numerical analysis of this ba- sic model, including recharge channel(s) at the sides and a reservoir at the base, could utilize more complex boundary conditions and parameter variations to sim- ulate conditions in a specific hydrothermal system. TWO- AND THREE—LAYER MODELS For the case of convection in a laterally extensive reservoir, we could remove the restrictive assumption of constant T along the upper and lower surfaces by considering a multi-layer system with less permeable rock above and below the reservoir. Convective flow in the reservoir would not draw heat evenly from below and heat would not be conducted evenly across the top of the layer. Consequently, isotherms in the less per- meable layers would be bent near the contacts with the permeable layer in contrast to the constant tem- perature contacts in the single-layer model. Donaldson (1962) analyzed the magnitude of this effect for a two-layer system with a permeable layer underlain by an impermeable layer. Both zones had the same finite thickness and were of infinite lateral extent; the upper surface of the permeable layer and the lower surface of the impermeable layer were held at constant temperatures. Donaldson found that the value of Rac was lower in the double-layer case than in the single-layer case, but that the conducting layer tended to retard convection effects for values of Ra above the Rac. Several runs using the program were made to com- pare the results with Donaldson’s results and to extend the models to the three-layer case. The multi- layer-models are illustrated in figure 15. An equiv- alent Rayleigh number was computed using the vertical temperature difference across the permeable layer which would exist with heat flow by conduction only and the thickness L of the permeable layer. In table 4, the steady state Nusselt numbers for each model for Ra = 100 are listed. The retardation effect noted by Donaldson is evident in the decrease in Nu from 2.60 for one layer to 1.19 for three layers; al- though for an aspect ratio of 1.0, the Nusselt number for the three-layer system may have been closer to that for the two-layer case. TABLE 4.—Effect of multiple layers on heat transfer in permeable layer with circulatory convection [S = cell aspect ratio] Model Ra S Nu 1 Layer ———————— 100 1.0 2.60 2 Layers ———————— 100 1.0 1.34 3 Layers ———————— 100 .8 1.19 D-16 110°C Permeable Permeable 1 80° 210°C FIGURE 15.—A, Isotherms for two-layer model with Ra = 100; B, Isotherms for three-layer model with Ta = 100. The corresponding temperature distributions in the multilayered systems as plotted in figure 15 show the distortion of the isotherms in the impermeable layers near the contacts with the convecting layer. Near the constant temperature boundaries, the isotherms flat- ten out but are more closely spaced than if there were no convection. This effect raises the question of how deep to place the constant temperature lower bound- ary to minimize its influence on simulating the con- vective system. We can use an analysis by Birch (1967) of the temperature distribution beneath the sea floor to show that the isotherm distortion decays as e—WZ/xl, where z = distance from the convective boundary and x’ is the distance from peak to trough for a sinusoidal isotherm at z = 0. For 1rz/x’ = 2, the isotherm distortion would be less than 10 percent of the value at z = 0. Hence, distortion will be small at GEOHYDROLOGY OF GEOTHERMAL SYSTEMS the outer boundaries of the multilayer models if x’ 0. An upper limit for H could be ob- tained from experimental data on turbulent air flow over flat plates. For example, Rohsenow and Choi (1961, p. 200) give H = 1.5 x 10'3(pcv)ai, (45) which for air at 15°C yields the following relationship at the land surface between H and mean annual wind velocity v. v H (cm/sec) (cal/sec °C cm’) 10 _________ 4.5 X 10“ 100 _________ 4.5 X 10'5 1000 ————————— 4.5 X 10" Other factors complicate the choice of an effective value of H, such as intermittent air flow, evaporation, and the presence of a dry, low conductivity soil layer. In the latter case, we could relate a particular value of H to a low conductivity layer of thickness 1 = Km/H. For example, for Km = 0.5 X 10'3 cal/sec °C cm, 1 ranges from 0.01 to 1.1 meters for H between 4.5 X 10" and 4.5 X 10‘6 cal/sec °C cmz, respectively. In this study, the effect of using the radiation boundary con- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS dition (44) was examined for a range of H from 10'6 to 10" cal/(sec °C cm2). . Results are tabulated in table 5 for the fault plane case considered previously with spring discharge equal to 1.1 X 105 kg/day. TABLE 5.—Spring temperature, Tsp’ total conductive heat loss, Q, and heat flux at land surface near spring, q, for select- ed values of coefficient of surface heat transfer H H Ts Q q(x = 5m) (cal/sec °C cm‘) 0(5)) (cal/sec) (u cal/sec-cm’) OO ————————— 76.3 1.29 X 105 82 10" ———————— 76.9 1.29 X 105 79 10'5 ———————— 78.9 1.27 X 105 73 10-5 ———————— 91.8 1.13 X 105 43 For values of H greater than about 10", the radi- ation boundary condition yields essentially the same results as the constant temperature boundary condi- tion. However, for H < 10", the spring temperature is significantly greater and the total conductive heat loss and land surface heat flux near the spring less than for the constant temperature case. The temperature dis- tribution for H = 10" cal/sec °C cm2 is shown in figure 21. Land surface temperatures and heat flows vary from 53°C and 43 HFU at a distance of 5 meters to 135°C and a normal conductive heat flux of 3.5 HFU beyond about 1 km from the spring. Thus, if the effec- tive values of heat transfer coefficient in field areas are below 10'5, the radiation boundary condition should be used in simulating the near surface thermal o 1 km 2 km 1 I 25 50 75 100 125 150 175 1 km 180°C FIGURE 21.—Steady state temperature distribution in fault plane hot-spring model with discharge = 105 kg/d, W = 10 m, D = 1 Km, and Ba = 0 (k = 0), and H = 10" cal/sec°C cm2 at land surface. NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS regime. It should also be noted that we are ignoring the details of heat and mass transfer at the surface in the spring. Many factors such as the geometry of the spring orifice, presence or absence of lateral discharge, and evaporative cooling may cause considerable vari- ations in actual temperatures near the surface. It is of interest to consider the time required for the conductive thermal regime to reach equilibrium fol- lowing the development of the spring system. From equation 32, the effective time constant for this 2-di- mensional heat transfer problem could be approxi- mated by the expression L2 fi/4Km, where L2 can be thought of as the area through which heat flows from the spring to the land surface. For example, from fig- ure 20, L2 might represent an area 1 km deep by 1 km wide or 1 kmz. This means that the time constant in this case would be about 9 X 10° days or 25,000 years. The transient numerical simulations for this problem show that temperatures are within 0.1°C of equilibri- um after a period of about 50,000 years, or twice the effective time constant. The thermal diffusivity in this problem was a relatively low 0.0032 cmz/sec; equi- librium times for larger diffusivities would be propor- tionately smaller. If the restriction is removed that the rock adjacent to the spring is impermeable, ground-water circula- tion in this region will affect the thermal regime. Even if the rock permeability is low enough that the critical Rayleigh number of 47r2 is not exceeded, the fluid 0 10°C D-21 would still be unstable because of the high tempera- ture boundary formed by the spring conduit. In mod- eling this problem numerically, the simplification is made that there is no hydraulic connection between the spring conduit and the adjacent ground-water sys- tem. To simplify the problem, a thin impermeable zone is included at the contact between the two re- gions; such a zone may exist in many field situations where chemical deposition has sealed off the conduit from the surrounding rock. Once again, consider the fault plane model analyzed previously, that is, discharge = 1.1 X 105 kg/day, Tb—Ts = 180°C — 10°C (fig. 20), but with a rock per- meability of 1 millidarcy (10'15 m2). In the absence of spring flow, the equivalent Rayleigh number for this case would be 20. With spring flow, the steady state temperature distribution shown in figure 22 would re- sult. The existence of convection cells is not clearly delineated by the isotherm pattern although compari- son with the 125°C and 150°C isotherms from figure 20 show the shifts caused by upflowing fluid near the conduit and downflowing fluid to a distance of about 1.5 km. The actual velocity distribution includes 3 convection cells but with the magnitude of convection decreasing with distance from the spring. As a result of the convection, the temperature drop and total con- ductive heat loss from the spring are about 6 percent less than for the corresponding impermeable rock case. The 25°C isotherms in figure 22 indicate that 2km 175 1km 180°C FIGURE 22.—Steady state temperature distribution in fault plane hot-spring model with discharge = 105 kg/d, W = 10 m, D = 1 Km, Ra = 20 (solid lines), Ra = 0 (dashed in fig. 20). D-22 heat flow near the land surface is not appreciably af- fected by convection in the rock below. Similar results in terms of slight lowering of the conductive heat loss from the conduit and negligible changes in heat flow near the land surface were found for other spring dis- charge rates in systems with Rayleigh numbers below Rac although the distortion of the temperature distri- bution at depth increases somewhat for the higher spring discharge rates. Now, if the rock permeability is increased to 10 mil- lidarcies (10‘14 m2), the Rayleigh number would be 200; and for a discharge of 1.1 X 105 kg/d the resultant tem- perature distribution is plotted in figure 23. In this case the influence of cellular convection set up by the constant temperature boundaries at the top and bot- tom of the model dominates the thermal regime. The effect of the hot spring is evident out to a distance of about 0.5 km over which the 100°C and 125°C isoth- erms show reversals characteristic of cellular convec- tion at large Ra. In this case, convection in the rock causes a more substantial lowering of the conductive heat loss from the spring conduit—0.99 X 105 cal/sec versus 1.29 X 105 for the Ra = O case—and a tempera- ture drop of 84°C compared with 104°C for Ra = 0. The relative effects of the conductive heat loss from the conduit and the enhanced heat transfer caused by convection in the rock are indicated near the land sur- face by the more closely spaced isotherms between 0 and 0.5 km from the spring compared with the spacing at corresponding intervals beyond 1 km. 0 1km GEOHYDROLOGY OF GEOTHERMAL SYSTEMS These results represent preliminary attempts to simulate heat transfer and fluid flow associated with hot spring systems. To model specific field areas, it may be necessary to include nonvertical orientation of the spring conduit, subsurface discharge and re- charge, and boiling near the land surface in the simu- lation. This work is being planned as sufficient data become available from field studies in California and Nevada. As noted previously, energy development in many of these geothermal areas may focus on localized regions of natural discharge. Thus a useful extension of the numerical modeling work described here will be the evaluation of various schemes for energy develop- ment. SUMMARY AND CONCLUSIONS We have discussed the development of a numerical code which can treat problems involving slightly com- pressible fluid and heat transfer in multidimensional porous media. Solutions to the appropriate partial dif- ferential equations are obtained by the integrated fin- ite difference method which is essentially equivalent to balancing mass and energy over finite subregions or elements. The resultant system of finite difference equations is solved by an iterative procedure, and so- lutions to the fluid flow and energy equations are coupled by interlacing in time so that the temperature and velocity fields are interdependent. The useful concepts of fluid and thermal time constants as indi- 1o°c 2 km 775 25/.“ l 1km 180°C FIGURE 23.—Steady state temperature distribution in fault plane hot-spring model with discharge = 105 kg/d, W = 10 m, D = 1 Km, and Ra = 200. NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS cators of nodal response times and numerical stability limits are an inherent part of the numerical scheme. In applying the numerical model to the problems of circulatory convection in saturated porous media, the relevant aspects have been discussed as they pertain to geothermal systems, and we find that results from the program on the relationship between the Rayleigh number and the dimensionless heat transfer coeffi- cient or Nusselt number are in good agreement with numerical and experimental results from other au- thors. Then the numerical model was used to extend these results to include the effects of temperature de- pendent parameters and density variations with pres- sure. Variations in fluid viscosity and thermal expansivity with temperature result in substantial differences in the values of the critical Rayleigh num- ber for the onset of convection and the Rayleigh num— ber-Nusselt number relationship compared with cor- responding constant parameters results. However, consideration of fluid density as a function of pressure produced no noticeable effect on convective motion. Numerical simulations of more realistic models for circulatory convection show that for laterally bounded reservoirs, conduction of heat across the vertical side walls results in significant lowering of the rate of verti- cal heat transfer through the reservoir. For a laterally extensive reservoir, consideration of impermeable or less permeable layers above and below the convecting layer removes the restrictive assumption of constant temperature boundaries on the permeable layer and has the effect of lowering the value of the critical Ray- leigh number Rac while retarding convective heat transfer at values of Ra above Rae. Heat and mass transfer associated with hot spring systems was analyzed to determine the amount of heat lost by conduction to the rocks surrounding the spring conduit. An isolated cylindrical conduit model and a fault plane conduit model were considered, and the temperature drop in the hot spring water between the source reservoir and the surface due to conductive heat loss was determined numerically as a function of flow rate. The steady state temperature distribution for the case where the rock surrounding the spring is impermeable shows that heat loss from the spring dis- torts the normally horizontal position of the isotherms out to distances comparable to the depth of the spring conduit. Conductive heat flux at the land surface is high near the spring but near the normal or back- ground level beyond one conduit depth. The time re- quired for the conductive thermal regime to equilibrate following the development of hot spring activity can be approximated by the expression LZE/ 2Km where L is the depth to the source reservoir. For D-23 unconsolidated sediments with low thermal conduc- tivity, the equilibration time is about 50,000 years for a reservoir at 1 km. The effects of fluid circulation in the rock surround- ing the spring conduit were examined for systems in which the equivalent Rayleigh number (in the absence of hot spring activity) was both above and below the critical value of 412. With Ra <41r2, circulatory convec- tion is set up due to the presence of the hot spring, but it causes only slight effects on the thermal regime in the rock surrounding the spring conduit and on the conductive heat loss and temperature drop associated with the spring. For the case with Ra>47r2, circulatory convection resulting from the temperature difference Tb—Ts between the source reservoir and the land sur- face dominates the thermal and hydrologic regimes and significantly reduces the conductive heat loss and temperature drop for the spring. The results of this investigation demonstrate the usefulness of numerical modeling to describe the nat- ural conditions of heat transfer and fluid flow in geothermal areas. Given preliminary thermal, hydro- logic, geophysical, and geochemical information, it is possible to construct and analyze simplified conceptu- al models of specific hydrothermal systems as a guide to further data collection in undeveloped areas. An application of this technique applied to the Long Val- ley caldera in California is described by Sorey, Lewis, and Olmsted (1978). As sufficient parametric and geo- metric data become available from deep test drilling to refine the model, it can then be used to evaluate rates and magnitudes of energy recovery and effects of development on existing water resources. REFERENCES Bear, J., 1972, Dynamics of fluids in porous media: New York, American Elsevier Pub., 764 p. Beck, J. L., 1972, Convection in a box of porous material saturated with fluid: Physics of Fluids, v. 15, no. 8, p. 1377—1383. Birch, F., 1967, Low values of oceanic heat flow: Jour. Geophys. Research, v. 72, no. 8, p. 2261—2262. Bird, R. B., Stewart, W. E., and Lightfoot, E. N., 1960, Transport Phenomenon: New York, Wiley, 780 p. Bodvarsson, G., 1961, Physical characteristics of natural heat re— sources in Iceland: U.N. Conf. New Sources Energy, Rome, 1961, Proc., v. 2, p. 82—89. Caltagirone, J. P., Cloupean, M., and Combarnous, M. A., 1971, Fluctuating natural convection in a horizontal porous layer [in French]: Acad. Sci., [Paris], Comptes rendus, B, v. 273, no. 20, p. 833—836. Carslaw, H. S., and J aeger, J. C., 1959, Conduction of heat in solids: London, Oxford Press, 510 p. Coats, K. H., George, W. D., Chu, C., and Marcum, B. E., 1973, Three-dimensional simulation of steam flooding: Soc. Petrol. Engr. Jour., v. 15, no. 6, p. 573—592. D-24 Combarnous, M. A., and Bories, S. A., 1973, Thermal convection in saturated porous media: Groupe D’etude Sur Les Milieux Por- eux, Institut Francais Du Petrole, Toulouse, Rapport G. E. 11. Curry, D. M., 1970, Multi-dimensional analysis of heat and mass transfer in porous media: Houston, Texas, University of Hous- ton, unpublished Ph. D. thesis, 197 p. Darcy, H. P. G., 1856, Les Fontaines Publignes de la Villa de Dijon: Paris, Victor Dalmont, 647 p. Donaldson, I. G., 1962, Temperature gradients in the upper layers of the earth’s crust due to convective water flows: Jour. Geophys. Res., v. 67, no. 9, p. 3449—3459. 1968, A possible model for hydrothermal systems and methods of studying such a model: Third Australian Conf. on Hydraulics and Fluid Mechanics, Nov. 25—29. Paper 2580. 1970, The simulation of geothermal systems with a simple convective model: Geothermics, Spec. Issue 2, UN. Sympo- sium on the Development and Utilization of Geothermal Re- sources, Pisa, Italy, v. 2, pt. 1, p. 649—654. Dorsey, N. E., 1968, Properties of ordinary water-substance: New York, Hafner Pub. Co., 673 p. Edwards, A. L., 1969, TRUMP: A computer program for transient and steady-state temperature distributions in multidimen- sional systems: Natl. Tech. Inf. Service, Springfield, Virginia, UCRL 14754, Rev. 2, 22151, 263 p. Einarsson, T., 1942, The nature of springs of Iceland (in German): Rit. Visind. Es., 26, 91 p. Elder, J. W., 1957, Some problems in fluid dynamics: Cambridge, England, Cambridge University, unpublished Ph. D. thesis. 1966, Heat and mass transfer in the earth: Hydrothermal Systems, New Zealand, D.S.I.R. Bull. 169, 115 p. Faust, C. R. and Mercer, J. W., 1975, Mathematical modeling of geothermal systems [abs]: Proc. of Second U.N. Symposium on Development and Use of Geothermal Resources, San Fran- cisco, Calif, May 20—29, 1975, v. 3, p. 1635-1641. Fernandez, R. T., 1972, Natural convection from cylinders buried in porous media: Berkeley, California, University of Califor- nia, unpublished Ph. D._ thesis, 300 p. Fournier, R. 0., White, D. E., and Truesdell, A. H., 1974, Geo- chemical Indicators of subsurface temperature—Pt. 1, Basic assumptions: Jour. Research US. Geol. Survey, v. 2, no. 3, p. 259—262. Green, D. W., 1963, Heat transfer with fluid flowing through po- rous media: Oklahoma, University of Oklahoma, unpublished Ph. D. thesis. Henry, H. R. and Hilleke, J. B., 1972, Exploration of multiphase fluid flow in a saline aquifer system affected by geothermal heating: University of Alabama Bureau of Engr. Research Rept. No. 150—118, 123 p. Holst, P. H., 1970, A theoretical and experimental investigation of natural convection in porous media: University Calgary, Al- berta, Canada, Ph. D. thesis. Holst, P. H., and Aziz, K., 1972a, A theoretical and experimental study of natural convection in a confined porous medium: Canadian Jour. Chem. Eng., v. 50, p 232—241. 1972b, Transient three-dimensional natural convection in confined porous media: Int. Jour. Heat and Mass Transfer, v. 15, p. 73-90. Horne, R. N. and O’Sullivan, M. J., 1974, Oscillatory convection in a porous medium heated from below: J our. Fluid Mech., v. 66, pt. 2, p. 339—352. GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS James, R., 1968, Wairakei and Larderello; geothermal power sys- tems compared: New Zealandeour. Sci. and Technology, v. 11, p. 706—719. Landau, L. D., and Lifshitz, E. M., 1959, Fluid Mechanics: London, Pergamon Press, 536 p. Lantz, R. B., 1971, Quantitative evaluation of numerical diffusion (truncation error): Soc. Petroleum Engineers Trans, v. 251, p. 315—320. Lapwood, E. R., 1948, Convection of a fluid in a porous media: Proc. Cambridge Phil. Soc., v. 44, p. 508—521. Lasseter, T. J ., 1974, Underground storage of liquified natural gas in cavities created by nuclear explosives: Berkeley, California, University of California, unpublished Ph. D. thesis, 273 p. Lasseter, T. J ., Witherspoon, P. A., and Lippman, M. J., 1975, Mul- tiphase multidimensional simulation of geothermal reservoirs: Proc. of Second U.N. Symposium on Development and Use of Geothermal Resources, San Francisco, California, May 20—29, 1975, v. 3, p. 1715—1723. McNitt, J. R., 1963, Exploration and development of geothermal power in California: California Div. Mines and Geology, Spec. Rept. 75. Mercer, J. W., 1973, Finite element approach to the modeling of hydrothermal systems: Urbana-Champaign, Illinois, Universi- ty of Illinois, unpublished Ph. D. thesis, 115 p. Mercer, J. W., Faust, C., and Pinder, G. F., 1974, Geothermal reser- voir simulation: Proc. of National Science Foundation Confer- ence on Research for the development of Geothermal Energy Resources, Pasadena, California, Sept. 23—25, 1974, p. 256— 267. Narasimhan, T. N., 1975, A unified numerical model for saturated- unsaturated groundwater flow: Berkeley, California, Universi- ty of California, unpublished Ph. D. thesis, 244 p. Olmsted, F. H., Glancy, P. A., Harrill, J. R., Rush, F. E., and Van Denburgh, A. S., 1975, Preliminary hydrologic appraisal of se- lected hydrothermal systems in northern and central Nevada: US. Geol. Survey Open-File Rept. 75—56, 267 p. Platzman, G. W., 1965, The spectral dynamics of laminer convec- tion: Jour. Fluid Mech., v. 23, p. 481—510. Raats, P. A. C., and Klute, A., 1968, Development of equations de- scribing transport of mass momentum and energy in soils, in Water in the unsaturated zone, Proc. Wageningen Sympo- sium: Pub. No. 83, International Association of Scientific Hy- drology, p. 535—546. Remson, I., Hornberger, G. M., and Molz, F. J., 1971, Numerical methods in subsurface hydrology: New York, Wiley-Inters- cience, 389 p. Rohsenow, W. N., and Choi, H. Y., 1961, Heat, mass and momen- tum transfer: Englewood Cliffs, New Jersey, Prentice-Hall. Sorey, M. L., and Lewis, R. E., 197 5, Convective heat flow from hot springs in the Long Valley caldera, Mono County, California: Jour. Geophys. Research, v. 81, no. 5, p. 785—791. Sorey, M. L., Lewis, R. E., and Olmsted, F. H., 1978, The Hydro- thermal System of Long Valley caldera, California: US. Geol. Survey Prof. Paper 1044—A, 60 p. Spillette, A. G., and Nielsen, R. L., 1968, Two-dimensional method for predicting hot waterflood recovery behavior: Jour. Petrole- um Technology, v. 20, p. 627—638. Sutton, F. M., 1969, Stability of flow in a model of a hydrothermal system: Victoria University, Wellington, New Zealand, unpub- lished M. Sc. thesis. Weinstein, H. G., Wheeler, J. A., and Woods, E. G., 1974, Numeri- NUMERICAL MODELING OF LIQUID GEOTHERMAL SYSTEMS cal model for steam stimulation in Improved oil recovery sym— posium, Apr. 22-24, 1974: Tulsa, Oklahoma, Soc. Petroleum Eng. preprint 4759, p..253—270. White, D. E., 1957, Thermal waters of volcanic origin: Geol. Soc. America Bull, v. 68, p. 1637—1658. 1961, Preliminary evaluation of geothermal areas by geo- chemistry, geology, and shallow drilling, U.N. Conf. New Sources Energy, Rome, Italy, 1961, Proc. v. 2, p. 402—408. White, D. E., Muffler, L. J. P., and Truesdell, A. H., 1971, Vapor- dominated hydrothermal systems compared with hot—water D-25 systems: Econ. Geology, v. 66, no. 1, p. 75-97. Witherspoon, P. A., Newman, S. P., Sorey, M. L., and Lippman, M. J ., 1975, Modeling geothermal systems: Univ. California, Law- rence Berkeley Lab. Rept. 3263, 68 p. Wooding, R. A., 1957, Steady state free thermal convection of liq- uid in a saturated permeable medium: Jour. Fluid Mech., v. 2, p. 274. 1963, Convection in a saturated porous medium at large Rayleigh number or Peclet number: J our.Fluid Mech., v. 15, p. 527. ~ Umvmsm! m {lMié’QFW-S as. DéPosnonv Hydrology and Geochemistry of Thermal Springs of the Appalachians By W.A. HOBBA, jR., D. W. FISHER, F. ]. PEARSON, jR., and j. C. CHEMERYS GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—E A study of the geologic and hydrologic setting and the physical and chemical character of selected major warm springs of the Appalachians UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1979 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Main entry under title: Hydrology and geochemistry of thermal springs of the Appalachians. (Geohydrology of geothermal systems) (Geological Survey professional paper; 1044-E) Bibliography: p. Supt. of Docs. no.: I 19.16: 1044-E 1. Hot springs—Appalachian Mountains. 2. Geochemistry—Appalachian Mountains. 1. Hobba, W.A. II. Series. III. Series: United States. Geological Survey. Professional paper; 1044-E. GB1198.3.A57H92 551.2’3’0974 79—607987 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-03218—0 CONTENTS Page Page Abstract __________________________________________________ E 1 Geochemistry—Continued Introduction ______________________________________________ 1 Geochemical indicators ________________________________ E21 Purpose and scope ____________________________________ 1 Dissolved gases ____________________________________ 21 Previous studies ______________________________________ 1 Trace elements ____________________________________ 23 Acknowledgments ____________________________________ 3 Isotopes __________________________________________ 23 Historical background __________________________________ 3 Carbonate chemistry and isotopes __________________ 26 Geologic and hydrologic setting, by W. A. Hobba, Jr __________ 4 Chemical geothermometers ________________________ 27 General description of the springs ______________________ 4 Discussion ____________________________________________ 29 Physical setting ______________________________________ 4 Silica-rock spring systems __________________________ 29 Geology and topography ____________________________ 4 Perry County, Pa., area ________________________ 30 Rock structure ____________________________________ 4 Berkeley Springs, W. Va., area ________________ 30 Hydrology ____________________________________________ 11 Minnehaha Springs, W. Va., area ______________ 30 Probable flow systems ______________________________ 11 Warm Springs, Ga., area ______________________ 30 Selected flow systems ______________________________ 13 Carbonate-rock spring systems ______________________ 31 Temperature of thermal waters ________________________ 17 Lebanon Springs, N. Y., area __________________ 31 Fluctuations ______________________________________ 17 Bolar Spring, Warm Springs, Hot Springs, Wells and thermal gradients ________________________ 18 and Falling Spring, Va., areas ________________ 32 Heat discharge ________________________________________ 20 Hot Springs, N. 0., area ________________________ 33 Geochemistry, by D. W. Fisher, F. J. Pearson, Jr., and Conclusions ______________________________________________ 34 J . C, Chemerys ______________________________________ 20 References cited __________________________________________ 35 General principles ____________________________________ 20 ILLUSTRATIONS Page FIGURE 1. Photograph of Homestead Hotel and Hot Sulphur Spring at Hot Springs, Va ________________________________________ E 3 2. Satellite images of thermal spring areas in Virginia, West Virginia, Georgia, and New York ________________________ 5 3. Side-looking radar images showing prominent lineaments and thermal springs and wells in Virginia, West Virginia, and Georgia ________________________________________________________________________________________________ 6 4. Map showing geology, spring, and well locations in Minnehaha Springs area, West Virginia ________________________ 8 5. Cross section parallel to fault passing near Webster Springs and Minnehaha Springs, W. Va., and Warm Springs, Va __________________________________________________________________________________________ 9 6. Map showing lineaments and their relationship to thermal springs in Virginia and West Virginia ____________________ 10 7. Photograph of sandstone quarry looking southwest toward Berkeley Springs, W. Va ________________________________ 11 8. Cross section showing theoretical movement of ground water through a homogeneous isotropic aquifer with an impermeable layer at 1,500 m ____________________________________________________________________________ 11 9. Cross section showing possible movement of ground water through a multilayered folded, faulted, and fractured aquifer such as those in Warm Springs Valley, Va ____________________________________________________________ 11 10. Block diagram showing possible circulation of water to a warm spring located near the crest of an anticline at the intersection of two faults ____________________________________________________________________________________ 12 11. Diagram showing possible movement of ground water through a faulted and fractured anticlinal ridge ________________ 12 12. Diagram showing possible movement of ground water through a faulted and fractured anticlinal ridge bordered by a synclinal ridge ________________________________________________________________________________________ 13 13. Plan view showing section lines for figures 12 and 14 and possible flow lines in the plane of the sandstone formation __________________________________________________________________________________________________ 13 14. Diagram showing possible movement of water along a vertical plane parallel to the strike of an anticlinal ridge ______ 13 15. Block diagram showing possible circulation of water to the warm springs at Berkeley Springs, W. Va ________________ 14 16. Map showing geology and spring and well locations near Berkeley Springs, W. Va __________________________________ 15 17. Graphs showing water temperature of hot spring and stage of French Broad River at Hot Springs, NC ______________ 17 18. Graphs showing temperature and discharge of Bolar Spring at Bolar, Va __________________________________________ 18 19. Logs showing temperature for a cold-water well and temperature and fluid conductivity for a warm-water well ________ 19 20. Temperature log for air-filled "dry” gas well (Pocahontas 21) near Minnehaha Springs, W. Va ________________________ 2O 21. Graph showing isotopes in Appalachian warm springs area waters ________________________________________________ 25 22. Graph showing dissolved silica and temperatures, Appalachian warm spring areas __________________________________ 28 III IV TABLE WPWNH CONTENTS TABLES Page Thermal gradients at wells producing warm or cold water __________________________________________________________ E19 Estimated depths of circulation based on thermal gradients and conduction losses ____________________________________ 20 Flow, temperature, and power output of thermal springs ____________________________________________________________ 21 Temperatures and selected analytical data for waters in Appalachian warm-spring areas ______________________________ 22 Saturation indexes and carbon-isotope ratios for waters in Appalachian warm—spring areas ____________________________ 24 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS By W. A. HOBBA,]R., D. W. FISHER, F. J. PEARSON,JR., and j. C. CHEMERYS ABSTRACT Thermal springs in nine areas in the Appalachians from Georgia to New York were studied in 1975 and 1976 using satellite imagery, local well and spring data, and results of current and early studies by other investigators. All the springs investigated discharge from folded and faulted sandstone or carbonate rocks in valley areas. Where geologic structure is relatively uncomplicated, ground water discharging from thermal springs probably has circulated to great depths roughly parallel to the strike of the bedding and has moved upward rapidly where a fault or faults cross the bedding. Hydrologic and chemical data suggest that most of the water discharging from warm springs in the Devonian Oriskany Sandstone is derived from recharge entering and circulating through that formation. The tem- perature of the thermal springs can be accounted for by “normal” geothermal heat flow if water circulates to minimum depths of 250— 1,600 meters along paths of sufficient length and contact area to absorb enough geothermal heat to raise the temperature of the water to that required to produce the given thermal springs. The discharge at springs where temperature fluctuates very little is primarily water from deep circulation. The discharge at springs where temper- ature fluctuates widely is warm water mixed with variable pro- portions of shallow-circulating cool water. Observed temperatures of the warm springs range from 18° to 41°C; the highest chemical thermometer temperature is 84°C. Agreement among observed, chalcedony, and cation temperatures of the warmest springs suggests reservoir temperatures of 30°—50°C. Dissolved helium, arsenic, potassium, and 8‘80 are considered as geothermal indicators. Tritium analyses are used to calculate frac- tions of old and modern components of mixed waters. Computer cal- culations of carbonate saturation indices show (1) considerable un- dersaturation in silica-rock warm spring waters and (2) carbonate equilibrium in the limestone and dolomite thermal waters. Better values of saturation indices are obtained when analyzed carbon dioxide rather than field pH is used in the computer input data. A method is described for adjusting NC to correct for carbon dioxide outgassing from water samples. INTRODUCTION PURPOSE AND SCOPE The primary purpose of this study was to investigate the geologic and hydrologic setting and the physical and chemical character of several of the major warm springs in the Appalachians. Investigative and analytical techniques in geochemistry, hydrology, and remote sensing have ad- vanced significantly in recent decades, but few detailed studies of the thermal springs of the Appalachians were made during this time. By using new techniques and investigating the larger thermal springs, an at- tempt was made to identify common characteristics that would show the character of the geothermal sys- tems of the region. The first part of this report, pre- pared by Hobba, describes the geologic and hydrologic setting of the springs; the second part, prepared by Fisher, Pearson, and Chemerys, discusses the geochemistry of the springs and inferences of thermal- reservoir temperatures indicated by the chemical data. The present study was a 1-year reconnaissance. Springs and wells were inventoried and water samples collected during the autumn of 197 5 and the spring of 1976. Several cold springs and wells were also sampled to provide comparative data from which properties peculiar to the warm springs could be discerned. Com- plete hydrogeological and chemical data and de- scriptions of the methods of sample collection and analyses are given in Hobba, Chemerys, Fisher, and Pearson (1977). PREVIOUS STUDIES Various aspects of the warm springs in the Appala- chian Mountains have been described in numerous sci- entific papers, some of which date from well back into the nineteenth century. In addition to descriptions of flow, temperature, and chemical character of the springs, many of the papers speculate as to the geologic structural setting and the origin and the depths of cir- culation of the waters. Most of the early inferences to depths of circulation of water were based on simplified assumptions of prevailing gradients and their effect on spring-discharge temperatures. Recent conceptual models, such as those of Lowell (1975), provide more El E2 realistic estimates of depths of circulation. Likewise, many of the earlier interpretations of geologic struc- ture associated with the springs were based on field observations. Now, gross structural features may be revealed by remote sensing from satellites or high- altitude aircraft. Similarly, geochemical techniques were formerly very simple; recently techniques have been vastly improved so that water-rock reactions and chemical equilibria can be modeled and even age dat- ing of water is possible. From this prespective the fol- lowing paragraphs summarize the earlier papers most useful to the present study. McCallie (1908) reported the chemical quality and discharge of Warm Springs, Ga. In a later paper, McCallie (1913, p. 13) inferred a depth of circulation of about 450 m on the basis of the discharge temperature, the mean annual air temperature, and an assumed “normal” geothermal gradient. Because of errors in- herent in the simplified assumption, 450 m probably represents, at best, only a minimum depth of circula— tion. In an extensive study of Warm Springs and the surrounding area, Hewett and Crickmay (1937, p. 34, 35) concluded that the water emerging from the springs enters a bed in the lower part of the Hollis Quartzite on Pine Mountain, moves downdip to a depth of about 1,200 m where the bedding is offset by a fault, moves upward along the fault, and then moves updip along a second permeable bed near the top of the Hollis Quartzite to discharge where the bed crops out. Keith (1904), Stose and Stose (1947), and Oriel (1950) generally agreed that the area surrounding Hot Springs, NC, is faulted and may be described as a structural window. However, Trapp (1970) suggested that the structure is a graben rather than a window and stated (p. 28) that “A fault probably forms part of the artesian system supplying the thermal springs at Hot Springs.” Stose and Stose (1947, p. 644) concluded that the spring water circulates to a depth of about 1,500 m and discharges along a nearly vertical fault. Rogers (1884) and Reeves (1932) describe the warm springs of Virginia. Rogers (1884, p. 577—597) observed that many of the warm springs are on the axes of anti— clines and suggested that water enters the rocks in the anticlinal ridges and sinks along joints and fissures until it reaches a permeable bed along which it then rises to discharge at the outcrop in an adjacent valley. Reeves (1932, p. 28) suggested that water enters a permeable bed at a relatively high altitude on the crest or limb of one anticline and travels along this bed through a syncline to an outcrop at a lower altitude in another anticline. According to Reeves (1932, p. 26), some anticlines are broken by thrust faults, but most of the springs in the region are not on faults, and hence few are fed by water rising along fault planes or associ- GEOHYDROI.()GY ()F GEOTHERMAL SYSTEMS ated fissures. Rogers and Reeves both postulated, be- cause of similarity in chemical composition, that the water from the warm springs is of the same meteoric origin as that from the cold springs. More recently Kulander and Dean (1972) studied gravity and structure across Browns Mountain, Wills Mountain, and Warm Springs anticlines. They re- ported that the rocks beneath Browns Mountain “an- ticlinorium” are cut by at least seven faults, most of which are interpreted as thrust faults dipping to the east. Displacement along a thrust fault cropping out just west of Minnehaha Springs is about 450 m. A simi- lar structure is proposed for the rocks beneath Warm Springs anticline. Heltz and Sinex (1974) studied the chemical equilib- ria in the thermal spring waters of Virginia and re- ported that; (1) neither concentration of dissolved con- stituents nor water temperature has changed in the past 140 years; (2)the waters with temperatures below 25°C are undersaturated with respect to calcite and dolomite, whereas those above 25°C are in equilibrium with calcite and dolomite; and (3) the waters with tem- peratures less than 25°C are probably mixtures of warm and cool ground water. Costain (1976) recently evaluated the geothermal- resource potential in the vicinity of the Warm Springs anticline in Bath and Allegheny Counties, Va. He (1976, p. 40) concluded that meteoric water percolates downward through the ridges, southeast of the anti- cline, to a depth of about 3 km; and that it moves from there along thrust faults that dip gently southeast- ward. The water is assumed to rise relatively rapidly along high-angle cross-faults which intersect one or more of the thrust faults at depth. Hobba, Friel, and Chisholm (1972), in a description of the water resources of the Potomac River basin in West Virginia, showed with a triangular diagram how the chemical composition of the water discharging from Berkeley Springs varies with time; the spring water probably is a mixture of (1) cold water from near-surface sandstone and carbonate rocks and (2) warm Sodium chloride water from deeper aquifers. As- suming steady-state conditions, a geothermal gradient of 21.6°C/km, and sufficient length and area of flow path to intercept all the geothermal heat flow, the water is moving upward from a depth of about 530 in. Lowell (1975, p. 355, 356), using a mathematical model, computed depths of circulation of 1,100 m for Warm Springs, Ga., 1,300 m for Warm Sulphur Springs, Va., and 1,700 m for Hot Springs, N. C. Low- ell’s model consists of two vertical fractures and an intervening horizontal conduit. In the model, he con- siders flow rates and minimum contact areas required to transfer geothermal heat flow (in a region of so- HYDROLOGY AND (LEOCHEMISTRY ()F THERMAL SPRINGS OF THE APPALACHIANS called “normal” geothermal gradient) to the water and also conduction losses as the water ascends to land sur- face. The spring at Lebanon Warm ‘Spring, N. Y., has not been investigated in detail. However, as early as 1843, W. W. Mather observed that the spring was at or near the contact of limestone and slate beds. Mather (1843, p. 106) also observed “ * * * traces of a fault, and of much derangement of the strata in the immediate neighborhood.” Similarly, little has been done on the Perry County Warm Springs in Pennsylvania. Early mention of the springs was made by Claypole, in 1855 (p. 348), who suggested that they may derive their water from sur- face streams. Little scientific study was done until the recent study by Weinman (1976), who reports that the springs are at the intersection of two lineaments he observed on ERTS (Earth Resources Technology Satel- lite) color composite imagery. ACKNOWLEDGMENTS Many people provided valuable assistance during various phases of the investigation. J. B. Foster and R. L. Wait made geophysical logs of wells. T. K. Ellis and G. Goold provided records of wells in Virginia and New York, respectively. R. Rogers, P. Gentry, R. Taylor, and J. D. Henry monitored spring temperatures in Virginia, North Carolina, and West Virginia, re- spectively. S. M. Pickering, Jr., R. P. Lowell, J. K. Cos- tain, and B. Weinman provided helpful discussions and information regarding thermal springs in Georgia, New York, Virginia, and Pennsylvania, respectively. Personnel of the Pennsylvania Geological Survey supplied well information from their files. A. H. Trues- dell helped greatly with advice on the interpretation of the geochemistry of hot-water systems and of various geothermometers. L. N. Plummer provided valuable assistance in devising computer programming tech- niques for the calculated carbonate geochemistry data in this report. F. W. Trainer and F. H. Olmsted fur- nished guidance and support throughout the study; Olmsted assisted in combining two separate manu- scripts on hydrology and geochemistry into the present paper. Finally, the cooperation of all the individuals who permitted access to their property, wells, and springs is particularly appreciated. HISTORICAL BACKGROUND Many of the warm springs included in our investiga- tion have long and interesting histories. While George Washington was surveying for Lord Fairfax in 1748 he E3 bathed in “ye fam’d warm springs” at the present site of Berkeley Springs, W. Va. Later Lord Fairfax granted the springs and 50 acres (20 ha) to what was then the colony of Virginia, with the stipulation that the waters were “to be forever free to the public for the welfare of suffering humanity.” Even before Washing- ton visited these springs the Indians used the springs for bathing, as they did those at Warm Springs, Ga. Later the Georgia springs were to be used by President Franklin D. Roosevelt in treatment of his polio. After discovery of the warm springs by white men, development soon took place. At some springs elabo- rate pools and hotels were built to accommodate .visitors seeking the healing and medicinal powers claimed for the warm waters. Many of these spas flourished in the late 1700’s and early 1800’s when there was no treatment or cure available for ailments such as rheumatism, arthritis, polio, diabetes, gout, stomach disorders, or kidney stones. The hotels and spas built at many of the springs continue to flourish as resorts today, as does the Homestead Hotel at Hot Springs, Va. (fig. 1). Some like the Warm Springs Hos- pital at Warm Springs, Ga., continue to flourish as hospitals. Many of the buildings at some spas fell into disrepair, were destroyed, or were converted to other uses, but some warm springs, not included in this study, still sustain prosperous resorts such as those at White Sulphur Springs and Capon Springs, W. Va.. FIGURE 1.—Homestead Hotel and Hot Sulphur Spring (foreground) at Hot Springs, Va. E4 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGIC AND HYDROLOGIC SETTING By W. A. Hobba,jr. GENERAL DESCRIPTION OF THE SPRINGS The part of the eastern United States covered by this report extends from western Georgia to eastern New York. This area lies primarily in the Valley and Ridge province, which is characterized by a series of northeast-trending valleys and ridges. The maximum altitudes in the central part of this area (West Virginia, Virginia, and North Carolina) range from 1,200 to 1,400 m above sea level. The valley floors, where the warm springs commonly emerge, are at ele- vations of 670— 700 In in the Virginias and at about 400 In in North Carolina. Toward the north and south the maximum altitudes decrease; in New York the maximum altitude is about 700 m and the valleys are at about 210 m; in Georgia the maximum altitude is about 430 m and the valleys are at about 240 m. The thermal springs are generally at the valley margins or a short distance up the valley walls. One spring in North Carolina discharges upward through about 4.5 m of alluvium on the valley flat. Most of the others discharge directly from bedrock or through wea- thered residuum. Most of the thermal springs are in or near small rural communities surrounded by forested hills and farmed valleys. Growths of green algae and water cress are common in the warm springs and their runoff channels. Temperatures of the spring waters range from 17.7° to 41.7°C. On cool days vapor clouds commonly form over the warmer springs and their runoff channels. Re- cent discharges measured at the springs range from 4 to 293 US; these are probably minimum discharges be- cause at some sites leakage occurs around the measur- ing site through the alluvium or other unconsolidated material. PHYSICAL SETTING GEOLOGY AND TOPOGRAPHY Most of the warm springs in the eastern United States are in the folded and faulted Valley and Ridge province of the Appalachian Mountains. Ten of these spring groups having relatively large yields and tem- peratures above 15°C were selected for this study. These spring groups comprise: Lebanon Springs, N.Y.; Perry County Warm Springs, Pa.; Berkeley Springs and Minnehaha Springs, W. Va.; Bolar Spring, Hot Springs, Warm Springs, and Falling Spring, Va.; Hot Springs, NC; and Warm Springs, Ga. Although many of the springs are widely separated geographically, similar geology and topography characterize each of the warm-spring areas. All the springs discharge from sandstone or limestone; and, with two exceptions, every spring is in a valley on the crest or flank of an anticline. (The exceptions are Lebanon Spring in New York and Hot Springs in North Carolina. They dis- charge from faulted limestones of complex structure, Lebanon Spring on a hillside and Hot Springs in a synclinal valley that has been described by some writ- ers as a graben and by others as a horst.) As early as 1840, W. B. Rogers observedthat 46 of 56 thermal springs he studied in the Appalachians of Virginia and West Virignia are on or near anticlinal axes. Regard- ing the thermal springs of Virginia, he (Rogers, 1843, p. 331) concluded that, in general, the springs “issue from the lines of anticlinal axes, or from points very near such lines.” Topography, lithology, and structure are important factors in the occurrence and character of thermal springs in the Appalachian Mountains. Topography controls potentiometric head relations that make pos- sible deep circulation of the water and its return to the surface. The large yields of some of the thermal springs, and the fact that the water retains appreciable heat until it reaches the surface, indicate the presence of effective flow conduits. These conduits probably are related to both lithology and structure. Both the sandstone and the limestone of this region can provide the permeability required to permit the rapid upward migration of warm water—sandstone through frac- tures and intergranular openings, limestone through solution—widened fractures. In the more “plastic” or less competent shale, the weight of the overlying rock may lead to decrease in permeability with depth; under this condition the upward movement of warm water is slow, and the water has an opportunity to cool as it moves upward just as it warmed when it moved downward. The depth of circulation in shale is not known; the scarcity of thermal springs in shale may reflect either lack of deep conduits in this type of rock or a tendency for the water to cool after rising from depth, as already suggested. The effect of structure on occurrence of the thermal springs is considered in the next section of this report. ROCK STRUCTURE The occurrence of warm springs in the eastern United States suggests a relationship between the structure of the rocks and the deep flow systems that develop and produce the warm springs. As noted ear- lier, the warm springs are typically on the crests or limbs of anticlines. Several structural features in these settings favor vertical permeability: openings along bedding planes, open tension fractures, and open faults and other fractures common to folded rocks. Interest- HYDROI.OGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS E5 ingly, thermal springs are unknown along faults in the Appalachian Plateaus province where the rocks are relatively flat lying and faulted and the relief is high. Similarly, at many localities in the Valley and Ridge province the rocks are folded, faulted, and fractured, topographic relief is great, and the same geologic for- mations from which warm springs issue at other 10- calities are present, but all the springs are cold. Appar- ently deep circulation paths have not been developed at these sites. The warm springs are found where deeper circulation paths have been developed preferentially to the shallower circulation paths. It is relatively easy to identify the probable dip of bedding planes and the general orientation of tension fractures, but it is more difficult to identify faults. However, satellite imagery is helpful in identifying linear features, some of which represent faults or frac- ture zones. Some faults, especially those that cross the strike of the bedding of the rocks, stand out well on ERTS imagery. Figure 2 shows a NW—SE trending fault passing through Webster Springs, Huntersville, and Minnehaha Springs, W. Va., and through Warm Springs, Va. This fault gives rise to a warm spring at Huntersville and perhaps to two other warm springs just northwest of Lexington, Va. In fact, a warm spring occurs each place the fault cuts the sandstone- limestone aquifer formed by the Oriskany Sandstone and Helderberg Group. The apparent northwest end of the fault lies near Webster Springs, where it may yield the salty water observed in springs and wells. (One salty cold spring discharging in the bed of the Elk River was reported by Price and others (1936, p. 101) to contain about 4,000 mg/L chloride. During this inves- tigation an attempt was made to sample the same spring but it could not be found. However, a well in Webster Springs, sampled during the present study, FIGURE 2.—-Satellite images of thermal spring areas in A, Virginia H and West Virginia, B, Georgia, and C, New York. (Lettered arrows ‘ indicate lineaments.) E6 (LEOl-IYDROLOGY ()F GEOTHERMAI. SYSTEMS contained 3,240 mg/L chloride.) What appears to be this same fault is also apparent on the SLAR (side- looking airborne radar) image (fig. 3). Although the SLAR image is much sharper and at a larger scale than the ERTS image, faults are more readily apparent on the ERTS image. However, the offset of ridges by faults is more apparent on the SLAR image. In the Minnehaha Springs area, hydrologic data can be used to aid in the interpretation of fault data ob- tained from the imagery. Figure 4 shows the geology of the area around Minnehaha Springs. Well 404, 14 m deep, penetrated only sandstone. The water had a specific conductance of 140 ,umhos/cm and a pH of 5.8; both these parameters are typical of water from the sandstone. According to the geologic map, well 405 should be in limestone, but the well is 26 m deep and ‘ ‘ il‘u .‘ ‘\ ‘ . 1:1,: 9' "3, FIGURE 3.A—Side-looking radar images showing prominent lineaments and thermal springs and wells in A, Virginia and West Virginia, and B, Georgia. (Lettered arrows indicate lineaments.) HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS yields water with a specific conductance of 80 ,umhos/ cm and a pH of 4.9; the water quality indicates that this well also taps the sandstone formation. If both wells tap sandstone, the northwest-trending fault passes south of well 405 and the northern block is downdropped relative to the southern block, thus dis- placing the outcrop of the sandstone on the northern block westward. Figure 5 represents an interpretation by Kulander and Dean (1972) of the geology in a vertical section parallel to this fault and about one mile north of it (see fig. 6). Their section is based on seismic, well, gravity, and magnetic data. The warm springs at Minnehaha Springs and at Huntersville both occur at points in the outcrop of the Oriskany Sandstone where the fault parallel to the line of the section intersects northerly trending faults. The northerly trending fault at Hun- tersville was mapped by Kulander and Dean (1972). Another fault passing directly through Minnehaha Springs is visible on the ERTS image and was mapped by the authors (fig. 2). It trends about N. 3° W. and extends about 74 km. A second fault just east of Min- nehaha Springs strikes about N. 3° E. It appears to be about 105 km long, reaching 68 km south of Natural Well, Va. The latter fault was found in the field at Ruckman Spring (a large cold spring, 408 fig. 4) prior to observing it on the ERTS imagery. These observa- tions of the occurrence of springs along faults or linea- ments are in agreement with those of Weinman (1976), who pointed out that the warm spring in Perry County, Pa., lies at the intersection of two lineaments observed on color-composite ERTS imagery. He also found that the warm water at the spring is detectable on low— E7 altitude thermal imagery. Werner and Medville (1975) pointed out that in West Virginia both warm and cold springs occur on lineaments or at the intersections of lineaments as observed on ERTS imagery. The lithology and structure at Minnehaha Springs, Berkeley Springs, and Perry County Warm Springs are very similar (fig. 7): each group of springs is in a valley and discharges from the Oriskany Sandstone on the east flank of an anticline. Weinman (1976) mapped two lineaments in the vicinity of Perry County Warm Springs which intersect at the springs just as similar lineaments do at Minnehaha Springs. At Berkeley Springs no intersecting pair of lineaments was ob- served on the satellite imagery; however, a single sub- tle but extensive lineament was observed that extends from Berkeley Springs northwestward across Pennsyl- vania along much of the course of the Allegheny River to the southeastern shore of Lake Erie. The southeast- ern extent of this lineament is not well defined, but it may follow part of the lower course of the Potomac River. The lithology and structure at Warm Springs, Ga., are somewhat similar to that at warm springs in the Oriskany Sandstone in West Virginia and Pennsyl- vania. The warm springs in Georgia, in a valley on the northwest flank of an anticline formed by the Hollis Quartzite, discharge at the contact of the quartzite (which forms Pine Mountain) and the overlying Man- chester Schist. Both schist and quartzite are cut by faults and lineaments (see arrows in fig. 2). One linea- ment passes through the large cold spring southeast of Warm Springs; another one passes northeast through a 73-m well which produces warm water. What may be a FIGURE 3.—Continued. E8 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ‘ 38°15' EXPLANATION PENN SYL VAN IAN Pottsvillc Group Tonoloway Formation Mm w I Mauch Chunk Formation Wills Creek Formation MISSISSIPPIAN Greenbrier Limestone McKenzie Formation SILURIAN I 0 [3| 0 Chemu D g Formation Clinton Formation E E ._. (I) 3‘ Bra lier ale Tuscarora Sandstone i3 3' 0RDO VICIAN Hamilton Formation Juniata Formation DEVON IAN Marcellus Shale ‘0 ._ ‘ _ - i WV?“ 5. ’. .. pring Mfiflrfiflwf " “’9‘" Oriskany Sandstone /4 ,fi’ 7"?“ 0 "ii iii". - f / ./-//, ' m Well ;//' "/[W ‘03 HelderbergGroup ///,- , " r/ V/ «’7 ’D‘o “/\ > y, ./. 1 y, 0 1MILE '//// /\ , / l i I l l I I 1 I71 l | ’ 334 ‘7/7/ \ 1 V: 0 1K|LOMETER ,, , > 3-“ l 7/, LLLLL—U—U—I——#—' Base from U.S. Geological Survey Warm Springs 1 262,500,1948 CONTOUR INTERVAL 40 FEET FIGURE 4.—Geology, spring, and well locations in Minnehaha Springs area, West Virginia. third lineament strikes eastward and passes close to the warm spring. Many lineaments have been mapped in the warm springs area of Virginia by the authors. Most of the known warm springs in this area are on lineaments, observed on the satellite imagery (fig. 6). Dennison and Johnson (1971, p. 506) suggested that the thermal springs here may be related to Eocene volcanism or that they represent a still younger deep source of heat (magmatic). Regardless of whether intrusion has oc- curred, the combination of deep fracturing, the local geology, high relief and structure favors deep circula- tion of ground water here. The lineaments mapped near Lebanon Springs, N. Y., stand out well on imagery with snowcover on the ground (fig. 20). One prominent northwest-trending lineament passes through the spring, as does a lobate northeast-trending feature (not indicated) that may be structural. The satellite imagery offers good evidence that the thermal springs occur in valleys where lineaments or faults cut steeply dipping beds of sandstone or lime- stone. Although many thermal springs occur on the crest or flanks of anticlines, the anticlinal structure itself is not the only factor contributing to their occur- rence. The tension fractures that parallel the anticlinal axis and the dipping rocks are conducive to deep circu- lation parallel to the structure. However, vertical FIGURE 5.—Cross section parallel to fault passing near Webster Springs and Minnehaha Springs, W.Va. and Warm Springs, Va. (Modified after Kulander and Dean, 1972. Approximate line of cross section is along lineament trending southeasterly from Hun- tersville to Warm Springs in fig. 6.) HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS E9 NW SE A A' Minnehaha Springs 1000 _ \\ Huntersville 64> Dc °° °° mm - 2°°° XI \\ 0 —— \\ " o _I-§l -, ‘ 4000 U) -1000 — I “J L] E u.| LL 2 -2000 —- Db Dm ,u..““““\\w«~““v — -8000 '3000 — 40,000 . ‘ Cambrian . 1242319“, " \ __ 42,000 4000 — 0:: 44,000 NW SE Browns Mountain ills Mountain Warm Springs Penmyivmign A Anticline I Anticlina Antictin- LSSQL 0 —— Mississippian —_-_: _ 0 U) Devonian ____;"\ _ 2000 “W E Silurian _ 10,000 E i- 4000 — “ / g Ordovician _ LL Cambrian 20,000 . I T 30,000 8 16 24 32 KILOMETERS EXPLANATION , r E] Chemung Tonoloway Formation Formafion Brallier Shale Wills Creek 2 Formation < z [I a [El 5 Harrell Shale D < McKen21e E < j Formation (I) a El D Marcellus Clinton Shale Formation V s \ E Oriskany Tuscarora Sandstone \ Sandstone Helderberg Group ORDOVICIAN Juniata Formation Martinsburg Shale Trenton Group (Limestone) Black River Group Rocks of Chazy age Beekmantown Group E10 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ' a \Www .m «M‘f - ‘\ F ’, Q \ J' ' 3 ‘ ow/ , r '\ ' V °"‘ M’ ., , -p ,7 / 38°OO’ .1 80°OO’ FIGURE 6.——Map showing lineaments and their relationship to thermal springs in Virginia and West Virginia. HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS Oriskany ' _ Sandstone , Marcellus Shale FIGURE 7.—Sandstone quarry looking southwest toward Berkeley Springs, W.Va. The spring is about 5 km from this site and dis- charges from the sandstone near the shale contact. faults or lineaments, rock type, dip of the rock, and subsurface thrust faults may be the most important factors contributing to the occurrence of the warm springs. For example, a warm spring may be just as likely to develop on a monoclinal fold as it would on an anticlinal fold having similar lithology faults, topog- raphy, and hydraulic head. Table 3, in a later section, summarizes the structural features at each of the thermal springs. HYDROLOGY PROBABLE FLOW SYSTEMS All the springs investigated discharge in valleys whose adjacent ridges are underlain by fairly steeply dipping rocks. The warm springs in Georgia, West Virginia, and Pennsylvania are on the flanks of anti- clines; in Virginia they are commonly near the crests of anticlines. It is of interest to consider likely hydrologic conditions where a warm spring is (1) in a valley near the crest of an anticline, (2) on the flank of a ridge formed by an anticline, or (3) on the flank of a ridge formed by a limb of an anticline. The Virginia springs are in valleys at or near the crests of anticlines, where any bedding planes open to depth would be expected to permit flowlines of upward moving water to converge at the crest of the anticline. Figure 8 shows the theoretical flow lines of ground water in a homogenous, isotropic sand aquifer. Figure 9 shows inferred paths of ground-water movement Ell «92 ’8" $6 a: LU }_ Em 1500 — ' ’I'I"“ \\ L: W LL or am <§ $3 $2 <3 ‘ So -0 E: in 33 051000— ‘ 33> \ / / ,l\\\\ \ / r' o I I la. \ l l / II 8< \ 8 l r‘° 1 | l 9% I ‘1 n. 1 l l l1: 1 i ll l l :1 ’9 0 1 l l llxl I . I l ll.\ 0 1000 2000 METERS FIGURE 8,—Theoretical movement of ground water through a homogeneous isotropic aquifer with an impermeable layer at 1,500 m. (After Toth, 1963.) crumpled beds crumpW Cold spring Hot sprung Cold spring ’fi FIGURE 9.—Possible movement of ground water through a mul- tilayered folded, faulted, and fractured aquifer such as those in Warm Springs Valley, Va. Much of the flow (not indicated by ar- rows) is into the figure and upward beneath the hot spring. through the consolidated folded rocks in Virginia, in a plane normal to the strike of the rocks and at a place E12 where the anticline crest forms a topographic low. To complete the picture, add low angle thrust faults at depth, a vertical fault or fracture at the spring, and possibly a fault or fracture in the plane of the cross section. Any such factors would contribute to the de- velopment of a system of conduits that permits ground-water circulation to great depths Where head is sufficient to drive the system. The overall flow system may look something like that in figure 10, but with much of the flow being into the plane of the figure along open tension fractures, bedding, or fault planes and upward beneath the spring. The occurrence of springs along the crests of anti- clines in valleys or in water gaps in the Valley and Ridge Province of Virginia and West Virginia has been noted by Rogers (1884), Price, Haskins, and McCue (1936), Hobba, Chemerys, Fisher, and Pearson (1977), and Clark, Chisholm, and Frye (1976). Clark, Chisholm, and Frye (1976) made statistical analyses of the yields of wells in various geologic and topographic situations. Working mainly with wells in limestone, they determined that yields along the axes of anti- clines are significantly better than those along axes of synclines. Statistical tests by Siddiqui and Parizek (1972) showed that variation in structural control is not a significant factor in the productivity of wells not on fracture traces. The relatively high well yields re- ported by Clark, Chisholm, and Frye (1976) may reflect the presence of fractures that parallel the anticlinal axes. These fractures, along with deep vertical faults and perhaps deep horizontal thrust faults, could permit a deep flow system such as that in figure 9 to develop. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Predominantly deep circulation favors warm springs; predominantly shallow circulation favors cold springs. The second type of flow system, which may develop on an anticline forming a topographic high, is repre- sented by figure 11, which shows the flow pattern in a vertical plane normal to the strike of the rocks. Water moves down through the open fractures (joints and bedding-plane openings) to as great a depth as per- meability and head allow, and then it returns to the land surface quickly along a fault or through other fractures. It seems unlikely, because of the generally short distances between the crest of the ridge and the warm spring, that a system like this would be capable of developing deep enough circulation to produce a warm spring (fig. 11). However, theoretical computa- tions by Toth (1963) for a homogenous isotropic aquifer with an impermeable boundary at 1,500 m depth sug- gest that circulation to a depth of about 700 m is possi- ble (fig. 8). His model roughly approximates the situa- tion at some of the warm springs. Although circulation to this depth is sufficient to produce some of the moderately warm springs, it is insufficient to produce the warmest springs. Figure 12 depicts the third type of geologic setting common at some of the warm springs. Here a ridge is formed by the flank of an anticline. The warm springs discharge from a sandstone unit underlain by carbon- FIGURE 10.—Possible circulation of water to a warm spring located near the crest of an anticline at the intersection of two faults. Much of the flow is into the plane of the figure along open tension fractures, bedding, or fault planes and upward beneath the spring. FIGURE 11.~—Possible movement of ground water through a faulted and fractured anticlinal ridge bordered by a synclinal ridge. Aquifer has impermeable boundary at 1,500 m depth. Flow pattern is in a vertical plane normal to the strike of the rocks. HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS Limestone Sandstone Siltstone \‘_.__ FIGURE 12.—Possible movement of ground water through a faulted and fractured anticlinal ridge bordered by a synclinal ridge. A ridge is formed by a flank of the anticline so that the water may flow by way of two routes—along a vertical fracture normal to A-A’ and along the aquifer shown by the stippled pattern, in the plane of the cross section. ate rocks and overlain by shale and siltstone. Water enters the ground and may flow to the spring by way of two routes or flow systems: The one indicated by the arrows in figures 12 and 13 shows the flow to be almost entirely within the plane of the sandstone formation. Figure 14 shows recharge and movement to the spring along a vertical fracture approximately normal to the line of A—A’ (The lines ofthe sections in figures 12 and 14 are shown in figure 13.) This type of setting exists at the Warm Spring in Georgia and in West Virginia and Pennsylvania. A significant amount of hydrologic data is available in the Vicinity of the warm spring at Berkeley Springs, W.Va. The situation is depicted in figure 15, and the data suggest that the flow system shown in figures 12 and 13 could be operating there with a small amount of water being added from the formations above or below the sandstone. Similar flow systems may be operating at the Perry County Warm Springs and at Minnehaha Springs. Sl-Zl.l-Z(I'I'lil) FLOW SYSTEMS Figures 15 and 16 show the warm spring at Berkeley Springs and the wells along the contact of the shale and sandstone along the ridge that trends southwest from the spring. The water table in figure 15 is based on reported and observed water—level data from the FIGURE 13.—Plan view showing section lines for figures 12 and 14 and possible flow lines in the plane of the sandstone formation. FIGURE 14,—Diagram showing possible movement of water along a vertical plane parallel to the strike of an anticlinal ridge. wells in the diagram. Wells 302, 304, and 308 are drilled, at least in part, into the Oriskany Sandstone. Wells 304 and 308 were reportedly 46 In deep and are either dry or produced only wet clayey sand near the El4 FIGURE 15.—Block diagram showing possible circulation of water to the warm springs at Berkeley Springs, W.Va. Water quality data suggests that most of the water discharged at the spring has circu- lated only through the sandstone. bottom. Well 302 is the only one in the sandstone that has not been destroyed. It is 121 m deep; in the upper 116 m it penetrates the Marcellus Shale, and in the bottom 4.5—6.0 m it penetrates the Oriskany Sandstone. During drilling some water entered the well at less than 18 m, but it was cased off with 19 m of casing. When this well was first visited on August 19, 1975, the water level was 31 m below land surface. At this time the well had filled with loose sand to a depth of less than 107 m. In order to get a water‘sample from the Oriskany Sandstone, the well was cleaned out on October 16, 1975. At this time water was observed squirting into the well above 49 m, and the water level recovered to about 49 In. On May 19, when the well was sampled again, the water level was 50.1 m, and some of the water, still squirting into the well, was entering between 29 and 32 m. The higher head in the shale favors movement of water from the shale into the sandstone. Prior to removal ofthe sand, the water level was higher because fine sand had partially plugged the well bore, impeding flow down the hole and into the Oriskany Sandstone. The water on January 14, 1976, was at about 189 m altitude—the same as the warm spring at Berkeley Springs. This level was about 37 m lower than the altitude of the main stream channels draining to the northeast on either side of the Warm Springs Ridge. Wells 304 and 308 reportedly were drilled to depths of about 46 m and were either dry or had wet clayey sand near the bottom. If these reports are accurate, then the water table at these wells may GEOHYDROLOGY OF GEOTHERMAL SYSTEMS be lower than the altitude of the warm spring. Despite a difference in water-level, this area could still be the recharge area for the warm spring, because the column of warm water rising at the spring could be driven by a column oflesser head ofcooler, more dense water in the recharge area. For example, a 600.0-m column ofwater at 22° C at the springs could be balanced by a 599.0-m column of water at 13° C in the recharge area. If ground water from the area of wells 302, 304, and 308 does not discharge to the warm spring, it must discharge to Sir Johns Run, or to the Potomac River, north or west of the warm spring. Subsurface flow to- ward the north or west is possible, but permeability across the strike of the bedding probably is not great enough to permit a high flow rate. Also, if such flow were happening, the water that now discharges at the spring probably would be diverted underground to Sir Johns Run or the Potomac River. Assuming recharge of about 880,000 L/day per km2 into the sandstone-limestone aquifer, the recharge area necessary to maintain the flow of the warm spring is about 10.6 kmz. If the actual area of recharge is assumed to be all the ridge to a distance of 6.1 km south of the spring and parts of the valleys on either side of the ridge (which is 1.3 km wide), the recharge area is about 9.8 kmz, or approximately the inferred area of recharge. However, if a large amount of re- charge were derived from the streams on either side of Warm Spring Ridge, the actual recharge area would be indeterminate. The chemical content of the water discharging from Berkeley Springs suggests that nearly all the water is derived from the Oriskany Sandstone. However, the watershifts from a calcium magnesium bicarbonate type at high flow toward a sodium chloride type at low flow. This shift indicates that a part of the water is derived from either a deeper formation or the overlying shale. Rough computations can be made to estimate the likelihood that all the spring water is derived from recharge entering the outcrop of the Oriskany Sandstone. First, recharge is assumed to be 880,000 L/day/km2 or 320 mm/yr in the 9.8-km2 outcrop area. This high rate is believed reasonable because of the permeable nature of the Oriskany Sandstone in this area. Grimsley (1916 p. 224) described an outcrop at Rock Gap Run eight miles south of Berkeley Springs as forming "an apparently solid wall sloping at a high angle to the east valley, but when struck with a ham- mer it is found to be soft and crumbly, it being almost impossible to secure a solid hand specimen ofthe rock.” Grimsley (1916, p. 224) described another outcrop, about four miles north of Berkeley Springs, as a con- glomerate "composed of small or large white pebbles imbedded in a loose white sand matrix which crumbles HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS E15 EXPLANATION ‘ El ‘ Chemung Formation Tonoloway Formation E z Bralher Shale Wills Creek Formation >5 3 ’ S if Hamilton Formation > g McKenzie Formation E :‘ ° E , _ v :- ‘g 9R3 Marcellus Shale Clinton Formation Oriskany Sandstone Geologic contact 0 Helderberg Group J Well (Geology after Grimsley, 1916) 1 1/2 0 1 MILE l 1 l I 1 1/2 0 1KlLOMETER |_____._1.___1___—1 CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL Base from US. Geological Survey Hancock 1:24,000,1951 (Pr 1971), Great Coupon 1:24.000, 1958 (Pr 1974) and Stotlers Cross roads 1:24.000, 1958 FIGURE 16.——-Map showing geology and spring and well locations near Berkeley Springs, W. Va. spring enters a BOO-m wide outcrop of the Oriskany Sandstone from 18 km south of the spring and 4.8 km north of the spring, only two-thirds of the measured to a gravel.” Thus, it is reasonable to infer a large amount of recharge from precipitation on poorly cemented rocks such as these. If all the recharge for the E16 flow of the spring can be accounted for. However, ad- ditional recharge is known to come from the overlying Marcellus Shale, and perhaps some also comes from the underlying limestones. If the spring derives water from as far as 18 km to the south, additional recharge could also be derived from Rock Gap Run, which flows across the sandstone outcrop (where Grimsley de— scribed the rocks). Thus, at least two-thirds of the dis- charge of the spring is probably derived from direct recharge on the Oriskany. R. T. Hurr (Lohman, 1972, p. 53) reports an average hydraulic conductivity value of9 m/day for medium to fine alluvial sand. This value converts to a transmissivity value (T) of about 900 m'Z/day for the Oriskany Sandstone, which ranges in thickness from 90 to 120 m and averages about 100 m. Using a slope (I) of 2.33 X 10‘“ for the water table from the spring to Rock Gap Run, the length of aquifer (L) required to supply the actual discharge of the spring (Q), 110 liters per second, was computed by the relationship Q : TIL and found to be 4,500 m. Thus, the water discharging at Berkeley Springs could be derived from recharge entering 4,500 m of the Oriskany Sandstone outcrop if sufficient recharge were available. However, all or most of the recharge comes from precipitation, and if our earlier assumption of 880,000 L/day/km2 for re- charge is reasonable, then recharge must be entering the rocks over as much as 18—23 km of the outcrop. It should be mentioned here that other nearby areas apparently duplicate the geology at Berkeley Springs (Hobba and others, 1972). The hydrology is similar too in that the water levels in the sandstone hills (formed by the Oriskany Sandstone), are lower than water levels in the adjacent shale valleys, much the same as the area south of Berkeley Springs. However, in these other areas, large cold springs are usually located some distance downgradient and along the strike of the rocks. At Berkeley Springs the only known large spring downgradient and along strike is the Warm Spring. Apparently the only difference between this warm spring and the cold ones is the depth of water circulation. The hydrologic situation near Baker, W. Va. (96 km southwest of Berkeley Springs) suggests that deep circulation may have occurred there in the past. The geologic conditions at this site are essentially the same as those at Berkeley Springs, except that the rocks dip to the west and a large cold spring discharges in the valley from a cavern in the Tonoloway Forma- tion, which lies stratigraphically about 107 m below the Oriskany Sandstone. A dye-tracer study showed that the spring is hydraulically connected to a large sinkhole in the same limestone in an upland valley about 3 km northeast of the spring. The large cavity at the spring is below the intersection of a thrust fault GEOHYDROLOGY OF GEOTH ERMAI. SYS'I‘EMS and a vertical fracture. At some time in the geologic past, when this large conduit was smaller and buried beneath hundreds of meters of rock, it is conceivable that it supplied water to a thermal spring. The age of the cavern is unknown, but it may be very old, since it is supplied water from such a large sinkhole which drains an area of about 2.5 kmz. Regardless of whether this system ever supplied water to a warm spring or not, cavernous circulation systems do develop parallel to the anticlinal ridges along certain formations and perhaps along faults. If developed to a great enough depth, such a system may give rise to a warm spring. The hydraulic system at Hot Springs, Va. may be similar to the one near Baker, W. Va. One well (607) at Hot Springs obtains warm water from a 0.6-m-high cavern at a depth of 230 In. No significant amount of water was encountered in the well above that depth, and well yield exceeded 30 US for a 53 hour pumping period. This high sustained yield suggests a water- filled cavern system at depth. Thermal springs having various temperatures and yields are found at both Hot Springs and Healing Springs, Va. which are 5.6 km apart along the crest of the anticline in Warm Springs Valley (fig. 10). A topographic high separates the two thermal-spring areas in the Valley. A well drilled in this topographic high (near well 611) was 91 m deep; this well penetrated caverns and was dry. If the water table is at or below 91 m, its altitude is near that of the thermal springs at Hot Springs but well above that of the springs at Healing Springs. On the basis of the yield of Healing Springs and Hot Springs and an as- sumed rate of recharge of 730,000 L/day/km2 (270 mm/yr) in the area between the springs, this recharge would supply half the total water discharged at the two spring areas. The other half of the water required is available in the part of the valley north of Hot Springs and that part south of Healing Springs plus any contri- bution from the ridges on either side of the valley. Warm Springs in Georgia discharges from a sandstone formation which dips 25°-30° beneath the Manchester Schist. Hewett and Crickmay (1937), in- ferred that the recharge for Warm Springs is derived from precipitation on Pine Mountain (fig. 3B). They concluded (1937, p. 33) that there are lower and upper permeable zones in the Hollis Quartzite, separated by a nearly impermeable middle unit. Water enters the lower permeable bed on Pine Mountain and flows downdip to where the rocks are broken by Towaliga Fault. At this point the water moves upward into the upper permeable beds and then back updip to dis- charge at the warm spring. Not enough new data are available to draw definite conclusions as to the pattern of circulation at this warm spring. If Warm Springs derives its water from the area of Pine Mountain south HYDROLOGY AND GEOCHEMIS'I'RY ()F THERMAL SPRINGS OF THE APPALACHIANS and southwest of the springs, and if the water flows to depth and if it is heated by a normal geothermal heat flow, then the springs may be part of a flow system similar to the type Hewett and Crickmay (1937) have described. However, a flow system similar to the one described for Berkeley Springs is also possible. That is, recharge entering the Hollis Quartzite on Pine Mountain may circulate to a depth parallel to schistos- ity along open fractures, joints, and faults and rise along a highly permeable zone to discharge at the springs. Figure 2 indicates that a lineament may pass through the warm springs. In either case, the water at the springs would have been derived principally from recharge on exposures of the Hollis Quartzite. TEMPERATURE OF THERMAL WATERS I’I.L'(l'l'l'A'I‘IONS The temperatures of the thermal springs investi- gated range from 17.7°C at Perry County Warm Springs in Pennsylvania to 417°C at Hot Springs, NC. Continuous recording thermographs were installed at Warm Springs, Va., and at Minnehaha Springs, W. Va., in the fall of 1975. The thermographs showed tem- perature fluctuations of about 03°C at Warm Springs over an 8-month period and less than 1°C at Min- nehaha Springs over a 3-month period. Recording instruments were also used by Hewett and Crickmay (1937) at Warm Springs, Ga., and by Costain (1976) at Hot Springs, Va. The temperature at a central discharge point at Warm Springs varied from 30.9°—31.2°C in a year and a half. However, the tem- perature at opposite ends of the 150-m-long discharge belt were 261°C and 283°C. Thus, there is a difference of about 5 degrees between the warmer and cooler wa- E17 ters discharged by the springs. Our measurements of the temperature of the combined flow from all the sources varied only 0.4°C when sampled in December 1975 and May 1976. Costain (oral commun., 1976) found that the temperature of Boiler Spring at Hot Springs, Va., was essentially constant during a 1-month period during which a thermograph was in- stalled. Our measurements at the times of sampling showed a temperature variation of only 0.1°C at Boiler Spring. However, the temperatures of other nearby warm springs fluctuated as much as 122°C during the same period. Weekly measurements were made by observers at Berkeley Springs, W. Va., and at Hot Springs, N. C., using maximum-temperature thermometers. The tem- perature at the discharge point observed at Berkeley Springs varied from 220°C to 222°C; that at Hot Springs varied from 368° to 41.7°C (fig. 17). (To date, 417°C is the highest known temperature measured at any of the thermal springs in the eastern United States.) The major temperature fluctuations here are caused by cool river water moving into the spring house during times of high flow in the river. Figure 17 shows that when the river stage rises to a level higher than about 1.5 m, water may move into the overflow pipe or through the alluvium and reduce the tempera- ture of the spring. From 1949 to 1955 flow and temperature meas- urements were made at Bolar Spring by the Virginia District of the U. S. Geological Survey. Figure 18 shows that there is generally an inverse relationship between discharge and temperature. This indicates that the spring water probably is a mixture of deep circulating warm water and shallow circulating cool water. December 1975 January 1976 February March 1 April May June I July 51015202530 510152025 30 510152025 510152025 30 5 10152025 30 510152025 30 51015202530 510152025 lllllllllllllllllllll f. 1 SPRING TEMPERATURE, IN DEGREES CELSIUS 8 ‘3 g 8 8 I F | | Spring averflow, submerged in river / Approximate level of — overflow pipe RIVER STAGE. IN METERS M IllTIlllll lll‘llllllllll FIGURE 17 .—Water temperature of hot spring and stage of the French Broad River at Hot Springs, NC. E18 GEOHYDROI.OGY OF G EOTHERMAL SYSTEMS DISCHARGE , IN _. 0'1 3 DISCHARGE, IN CUBIC FEET PER SECOND 0| TEMPERATURE, IN llllllllllllllllllllllllllllllllllll||||lllllTlllllllllllllllllllllllllHlll[llllll 500*— O z _ 8 400— UJ U) E Lu 0. 300—- , _ O: LLI '— 3 200 — 100 - 24 _, (I) 2 U) _l uJ 20 — _ O U) LU LU 0: 0 Lu 15 — - D 13 r— —— 1949 1950 1951 1952 1953 1954 1955 FIGURE 18.—Temperature and disc Weinman (1976) measured the temperature of Perry County Warm Springs periodically; he found that it fluctuated from 17.5° to 19.5°C. He also found that even though this spring is not very warm, the temperature difference is great enough to appear on low altitude thermal imagery. Our measurements at the times of sampling varied only from 17.8°C in the fall to 177°C in the spring. The temperatures at Lebanon Springs, N. Y., and Falling Spring, Va., were measured only at times of sampling. Lebanon Springs varied from 219°C in the fall to 18.7°C in the spring. Falling Spring varied from 250°C in the fall to 208°C in the spring. These lower temperatures during the spring suggest that the dis- charge is a mixture of deep-circulating warm water and cool shallow water. WELLS AND THERMAL (LRADIENTS During this investigation three wells were inven- toried that produced warm water. Two are in Virginia and one is near Warm Springs, Ga. Temperature and other geophysical logs were obtained on one of the open wells (607) at Hot Springs, Va. The temperature and fluid conductivity logs are shown in figure 19. This hole is reported by the driller to be cased to 47.2 m with 150—mm casing. The only water reported is warm water harge of Bolar Spring at Bolar, Va. the fluid conductivity log shows two types of water in the well: one of lower conductivity above 109.7 m and one of higher conductivity below 109.7 m. Although the log shows that conductivity increases from 750 to about 1,100 umhos/cm below 122 In, much of this increase is probably caused by the increase of temperature with depth because the conductivity of point samples col- lected at 122 and 227 m were 700 and 750 umhos/cm, respectively, when cooled to the same temperature at land surface. The yield of the well is substantial; (it was test pumped at 30 Us for 53 hours). When sampled with a pump that would pump only 0.41L/s, the tem- perature and specific conductance of the discharge in- creased from 150°C and 380 micromhos to 202°C and 670 micromhos, after 21/2 hours of pumping. During the period, the water level rose 70 mm in the well. The rise in temperature, conductance, and water level with pumping is attributed to the warm water of higher conductivity entering the bottom of the well and mov— ing upward. The expansion of this warm water was enough that the water level in the well increased al- though pressure head actually decreased. Figure 19 shows the temperature log for well 302 near Berkeley Springs, W. Va., and that of the warm well at Hot Springs, Va. At well 302 the cool water is flowing into the well from the Marcellus Shale, down from a cavern between 229.8 and 230.4 m. However, the well bore, and out into the Oriskany Sandstone, the HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS El9 40 — 60 — .q I“ S 'B rt! 80 — i .1 c N rn 100 — ~360,umhos Cold-water well, Berkeley Springs, West Virginia ~750 ,umhos 120 — m Warm-water well,— 0: Hot Springs. E Virginia Lu 2 E :E 140 — '— CL uJ D 160 — g Q 11 8 O o z D c O E 180 — z o "I 200 — 220 — ~ 1100 ‘umhos 23° I I I I I 11.7°12.2° 15.6” 211° 263° 322° TEMPERATURE, IN DEGREES CELSIUS FIGURE 19.—Logs showing temperature for a cold-water well and temperature and fluid conductivity for a warm-water well. to of which lies at about 1 16 m. This well is thought to be located in the recharge area for the warm spring of Berkeley Springs. The warm water in the well at Hot Springs is flowing in at the bottom of the well, up the well bore, and out into the overlying formations. This well is thought to be located in the same discharge area as the thermal springs at Hot Springs, Va. Increases in temperature and conductance with time of pumping were common to all the wells that produced warm water. The temperature and conductance of well 702 near Falling Spring, Va. steadily increased to 272°C and 1,050 micromhos after more than one hour’s pumping. With continuous pumping the temperature of water from this well is reported by the tenant to reach 35°C. After one and one-half hours of pumping, the temperature and conductance of well 907 near Warm Springs, Ga., had increased to 33.8°C and 100 micromhos. This is 26°C higher than the highest tem- peratures observed at the Warm Springs (Hewett and Crickmay 1937, p. 5). Possibly this well is tapping the conduit that supplies warm water to the spring, or perhaps the water from the well is mixed with a smaller amount of cool water than the water from the spring. Thermal gradients were computed using bottom hole temperatures or maximum temperatures of water pumped from wells (table 1). It is apparent from table 1 that some of the shallow hot wells have thermal gra- dients about 100 times that computed for well 302 near Berkeley Springs. The small gradient at well 302 can be explained by the downward flow of cold water in the well. The large gradient in the warm wells can be ex- plained by the upward movement of warm water from depth. The thermal gradients for test well 703 in Virginia and gas well Pocahontas #21 in West Virginia are considered about normal for the eastern United States. Figure 20 is the temperature log for gas well Pocahontas #21. It shows a nearly linear increase in temperature with depth. This well is 3,111 m deep and TABLE 1.—Thermal gradients at wells producing warm or cold water Thermal Bottom-hole Wells gradient' E13 temperature (°C/Km) (m) (“Cl Vir inia: omestead hot well (607) __________ 94.2 230.4 32.8 Webb hot well (702) ______________ 470 50.9 35 West Virginia: Wachter domestic well (302) ______ 6.6 121.3 11.9 Gas well Pocahontas #21 __________ 21.4 3,111.4 78.2 Georgia: Phillips hot well (907) ............ 234 73.1 33.8 ‘Thermal gradient computed from bottom hole temperature and local shallow ground-water temperature. For water wells. highest measured or reported water temperature was as- sumed to be bottom hole temperature. E20 1500 2000 DEPTH, IN METERS 2500 3000 3111 | | | I l I - 5o 55 60 65 7o 75 TEMPERATURE, IN DEGREES CELSIUS FIGURE 20.—Temperature log for air-filled "dry” gas well (Pocahon— tas 21) near Minnehaha Springs, W. Va. is 3 km southwest of Minnehaha Springs, W. Va. It penetrates the Cambrian and Ordovician rocks that underlie the warm spring areas of West Virginia, Virginia, and Pennsylvania. The thermal gradient here is 21.4°C/km. On the basis of this thermal gra- dient, the depth of circulation at Minnehaha Springs is estimated to be more than 450 m. Table 2 shows minimum depths of circulation at these warm spring areas inferred on the basis of known geothermal gra- GEOHYI)R()I.OGY ()F GEOTHERMAL SYSTEMS TABLE 2.—Estimated depths of circulation based on thermal gradients and conduction losses Estimated Minimum Thermal conduction depth of Springs gradient 1055 Circulation 1°C/Km) 1%) m (m) Lebanon Springs, NY _____________ ‘ 26.6 5 500 Perry Co. Warm Springs, Pa ________ ‘ 26.6 3 250 Berkeley Springs, WV ______________ ‘ 23.0 3 500 Minnehaha Springs, WV . ,,,,,,,,,, ‘-’ 21.4 2 450 Bolar Spring, Va ,,,,,,,,,,,,,,,,,,, ‘ 3 20.0 2 600 Warm Springs, Va ,,,,,,,,,,,,,,,,,, ‘ “ 20.0 6 1,300 Hot Springs, Va ____________________ ‘ 3 20.0 8 41,600 Falling Spring, Va ,,,,,,,,,,,,,,,,,, ‘ ” 20.0 2 51,450 Hot Springs, NC ____________________ " 24.8 3 1,300 Warm Springs, Ga ,,,,,,,,,,,,,,,,,, 7 21.2 5 750 ‘Am, Assoc. Petroleum Geologists and U. S. Geol, Survey, 1976. 2Measured in nearby gas well Pocahontas #21. “Lowell, 1975, p. 356. 'For the hottest spring, 3Costain (1976), p, 31) reports a thermal gradient of about 93° C/km in a 305 m test hole near Falling Spring, Va. If this is the representative gradient for the Virig‘nia area, minimum depths ofcirculation may be more than twice that shown for the Virginia springs. “Stose and Stose, 1947, p. 644. THewett and Crickmay, 1937, p. 35. dients, assumed temperature equilibrium at depth, and estimated conduction losses. Conduction losses were estimated from measured spring flows, tempera- ture, and the following assumed values (Lowell, 1975, p. 353—356): thermal conductivity of the rock (2.5 W m’1 °C“‘), specific heat of water (4.2 x 10"‘j kg’l °C"), thickness of the resistance ('heatltransfer) layer equals depth of circulation, and fracture lengths equal 1 km for small springs and 10 km for large springs. HEAT DISCHARGE An impressive amount of heat is carried to the earth’s surface and released by the thermal springs. Computations based on flow and temperature data show that the heat released at larger but cooler springs such as Falling Spring and Bolar Spring, Va, is much greater than that released at the smaller but warmer springs (table 3). Since the discharge at these cooler springs is a mixture of warm and cold water, the com- ponents of hot water must be either larger or hotter than that emerging at the smaller warmer springs. Thus, it seems that areas of the springs with the greatest heat output (as power equivalent) would be the best areas to investigate for geothermal potential. GEOCHEMISTRY By D. W. Fisher, F. J. Pearson, Jr., and J. C. Chemerys GENERAL PRINCIPLES Several combinations of processes can lead to the ap- pearance of thermal water1 at the earth’s surface. The chemical and isotopic measurements discussed here ‘Thermal water is considered herein to be water having a temperature above the mean annual temperature near the earth‘s surface. HYDROLOGY AND GEOCHEMIS'I‘RY ()F THERMAL SPRINGS OF THE APPALACHIANS E21 TABLE 3.—F low, temperature, and power output of thermal springs Spring Spring owner or name Date No. Power output Average annual Structural ( 1975) Flow Temperature at land surface air temperature features (US) (”Cl (Wattsx 10") 1°C! at spring 101 Lebanon Springs, NY ,,,,,,,,,,,,,,,,,, 10— 4 6.6 21.9 .369 8.3 F 201 Perry County Warm Springs, Pa ........ 10— 6 8.83 17.8 .287 10.0 A, L 301 Berkeley Springs, W. Va ______________ 10— 2 103 21.8 4.55 11.1 A, F 401 Minnehaha Springs, W. Va 111111111111 , 11—18 13.88 20.4 .580 10.6 A, L 501 Bolar Spring, Va ______________________ 12—16 130 22.2 6.35 10.6 A, F Hot Springs, Va ________________________ ___- __,, 1-“ ___- __,_ A,F 601 Octagon Spring __________________ 11—21 ___ 36 5 w“ ___- __s- 602 Boiler Spring ____________________ 11—21 3 78 39.9 __1- 1-“ 1___ 603 Hot Sulphur Spring .............. 11—21 __,, 36.5 __,_ ____ ____ 604 Magnesia Sprin 1111111111111111 11—21 .03 16.2 ___- ,e__ ____ 605 Cold Magnesia pring ____________ 11—21 8.5 16.3 ____ _-,_ ,___ 606 Soda Sprin ____________________ 11—21 13.4 ,___ ____ ____ Total ,,,,,,,,,,,,,,,,,, 11—26 63 20.8 2.79 10.6 ____ Warm Springs, Va 11111111111111111111 “-2 -1" __-- ____ A-" A, F 631 Children’s Pool __________________ 11—25 11.6 35.4 A-“ ____ ____ 632 Men’s Pool ...................... 11—25 28.4 35.0 _,-_ ____ ____ 633 Ladies’ Pool ____________________ 11—25 11.7 34.8 ____ 1-1- ___- Total __________________ 11—25 51.7 34.8 5.29 10.6 ,___ 701 Falling Spring, Va 11111111111111111111 12—15 204 24.9 12.3 10.6 A, F 801 Hot Springs, NC ______________________ 12— 6 .5 39.3 .09 10.0 F 901 Warm Springs, Ga .................... 12—10 55.8 30.9 3.37 16.7 A, F A Crest or flank of anticline. F Fault or lineament. L Intersection of lineaments or faults. were intended to identify which of several possible combinations of processes gave rise to the observed sur- face temperatures of the thermal springs and wells sampled. The processes can be envisioned as follows: Ground water is in some way heated—for example, by deep circulation in a region of so-called "normal geothermal heat flow.” The water presumably attains temperature equilibrium with the aquifer rock at depth. This water may then flow to the surface rapidly enough so that it neither loses appreciable heat by conduction to near- surface cooler rocks nor mixes with cooler, shallow- circulating ground water. The surface temperature of such a water is that of the region in which it was origi- nally heated. On the other hand, the heated water may ascend slowly, losing heat by conduction, in which case its surface temperature will be below its maximum temperature at depth, but it will show no evidence of having been mixed with shallow, cooler water. Finally, if the heated water mixes with cool, shallow water be- fore issuing forth at the surface, its surface tempera- ture will also be below its maximum temperature at depth, but it should have characteristics which show that it is a mixture. The measurements made on the samples from the systems described here were intended to permit the estimation of the maximum temperatures attained by the heated component of each of the systems. The meas— urements and their use are described in detail below. Briefly, they include tritium measurements to esti- mate the amount of cool shallow water present in each sample, measurements of dissolved constituents which are useful as geothermometers or as qualitative indi- cators of heat-generating reactions, and consideration of the carbonate geochemistry of the samples as indica- tors of the aquifer rock type through which the water has passed. GEOCHEMICAL INDICATORS Cation and anion distributions and total mineraliza- tion are useful for delineating differences between waters from various sources. Values of dissolved—solids concentrations, a measure of total mineralization, cal- culated as the simple sums of analyzed cations, anions, and nonvolatile neutral constituents, are listed in table 4. Variations in cation and anion distributions are dis- cussed under the individual warm-spring areas. DISSOLVl-ZD GASES Dissolved gas concentrations can also be helpful in understanding the geochemistry of waters. High dis- solved oxygen is typical of a water that has had little contact with organic matter or sulfide minerals during its subsurface travel. On the other hand, the presence of methane and (or) hydrogen sulfide in a water signifies highly reducing conditions in the aquifer through which the water flows. Dissolved oxygen con- centrations of the sampled Appalachian warm springs exceed >3 mg/L except for those in the Hot Springs- Warm Springs area of Virginia, where levels range from 0.1 to 2 mg/L. A hydrogen sulfide concentration of 0.75 mg/L in the Warm Springs sample (631) suggests that sulfide minerals may be consuming the dissolved oxygen of the recharge to these warm springs. Roberts, Friedman, Donovan, and Denton (1975) have proposed the use of helium concentrations in soil E22 GEOHYDROLOGY OF GEOTHERMAI. SYSTEMS TABLE 4.—-Temperatures and selected analytical data for waters in Appalachian warm-spring areas [Locations and hydrologic data for these wells and springs are given in Hobba, Friel, and Chisholm (1977)] Temperature (°C) Tritium Dis- Po- Dis- solved Cal- tas— Sam- solved silica cium Sodium sium Helium ple Owner or name Date Ob- Chal- solids (SiOz) (Ca) (Na) (K) (He) 5‘50 No. served cedony Quartz Cation (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (Mg/L) TU +1cr °/.,., New York—Lebanon Springs area 101 Lebanon Spring __________________________ 10—75 21.9 10.7 45.6 6.6 271 12 35 6.9 1.2 Tr<2 11.6 $10 —9.9 101 5—76 18.7 1.1 36.2 ~.4 266 9.0 40 5.5 1.0 1.0 19.9 $1.3 —10.7 109 John Koepp ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 10—75 14.0 7.8 42.7 10.0 271 11 31 7.0 1.3 <2 1.0 $1.4 — 10.0 109 5476 12.2 7.8 42.7 10.3 253 11 31 7.3 1.3 1.3 —.4 $.5/—9.8 114 B. Schell _______________________ 10—75 10.9 —11.3 23.9 —9.5 698 6.0 133 28 .8 <2 .3 $.9 — 116 F. Amlaw ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 1&75 10.4 16.1 50.8 —23.3 352 14 49 2.8 .4 <2 .6 $ 1.0 —9.4 Pennsylvania—Perry County Warm Springs area 201 Warm Springs __________________________ 10—75 17.8 1.1 36.2 ‘199 197 9.0 38 1.6 .5 <2 6.0 $1.0 —7.7 201 5476 17.7 —5.0 30.1 —19.8 193 7.4 40 1.8 .5 <.5 8.3 $.6 —8.2 202 Lloyd Hetrick ........................... 10~75 12 —21.4 13.9 —17.6 117 4.2 25 .6 .6 <2 55.6 $3.0 — 216 Stambaugh Spring ,. 10—75 12.8 —5,4 29.7 ~30 96 7.3 17 1.4 .6 <2 15.5 $1.4 —8.8 218 M. Loy Spring ,,,,,,,,,,,,,,,,,,,,,,,,,, 10,75 12.0 —11.8 23,4 —12.1 308 5.9 65 3.1 .9 <2 78.1 $4.1 —7. West Virginia—Berkeley Springs area 301 Berkeley Springs ........................ 10—75 22.2 2.9 37.9 —3.8 258 9.5 45 4.1 1.0 <2 2.7 $.8 —8.7 301 5—76 21.9 1.1 36.2 —6.1 249 9.0 47 4.3 .9 <5 2.6 $.6 —8.6 302 F. Wachter ______________________________ 10—75 13.0 29.5 63.7 A176 430 20 80 6.2 .6 — 36.2 :24 -7.9 302 5—76 12.8 23.3 57.7 ~17.5 418 17 83 6.8 .6 <.5 40.2 $2.1 ~‘82 303 F. Wachter _______________________________ 9—75 13.9 25.4 59.8 —14.4 405 18 55 17 .4 <2 45.3 $2.6 — 307 H. Hovermale _____________ .. 10—75 16 —3.4 31.7 —20.9 596 7.8 106 3.8 .7 <2 71.8 $3.7 -7.8 West Virginia—Minnehaha Springs area 401 Minnehaha Springs ______________________ 11—75 20.1 —1.1 34.0 —10.1 212 8.4 39 2.2 ,8 Tr<.1 2.5 $.9 —8.5 401 6—76 20.3 —2.2 32.9 “9.8 213 8.1 38 2.2 .8 <.1 1.3 $.6 -8.6 402 Camp Minnehaha ________________________ 11—75 11.6 16.1 50.8 —23.5 364 14 61 4.2 .4 <3 41.0 $2.4 —8.1 402 6—76 13.5 13.5 48.3 -24.5 363 13 64 3.9 .4 <3 — — 403 A. T. White ............................ 11—75 13.3 —7.6 27.6 «12.8 160 6.8 36 1.2 .8 — — —- — 406 F. Davis _______________________ 11—75 10.3 —16.6 18.7 -19.2 164 5.0 36 .4 .8 <1 62.9 $3.6 -7.2 407 R. Shinaberry ____________ ,_ 11—75 11.8 —7.6 27.6 17.5 80 6.8 14 2.6 1.5 <.1 87,6 $4.5 -8.0 410 T. Rose _______________________________ 6—76 13.8 —5.9 29.3 96.9 5650 7.2 129 1920 12 .6 —.3 $.6 —7.6 Virginia—Bohr Spring area 501 Bolar Spring ____________________________ 12—75 22.2 7.8 42.7 3.0 323 11 58 1.6 2.3 Tr<.2 24.4 2.0 -8.0 501 6—76 20.2 —2.2 32.9 1.2 288 8.1 54 1.6 2.0 <.3 27.2 $16 —8.1 Virginia—Hot Springs—Warm Springs area 602 Boiler Spring ............................ 11—75 39.9 31.4 65.5 41.4 802 21 132 7.0 13 0.7 ~3.1 $1.8 —7.4 602 6—76 40.0 29.5 63.7 44,1 788 20 126 6.9 14 .3 .7 $.5 48.3 605 Cold Magnesia Spring ,,,,,,,,,,,,,,,,,,,, 11—75 15.2 —1.8 33.3 7.7 302 8.2 60 1.7 2.9 <.1 41.0 $2.4 -7.9 605 6—76 14.0 —6.7 28.5 5.5 271 7.0 53 1.5 2.5 — 43.5 $2.2 —9.1 607A Homestead Hot Well __ 11—75 19.6 18.6 53.2 17.4 653 15 113 2.9 5.8 1.0 2.4 $2.3 —7.9 6078 2—76 17.9 —5.9 29.3 13.2 317 7.2 53 2.6 3.0 — — — 607C 2776 19.5 16.1 50,8 35.9 697 14 119 4.9 11 — — — — 607D 2—76 32.2 27.5 61.8 40.7 735 19 123 5.6 13 — — — — 619 M. Dunn Spring ,,,,,,,,,,,,,,,,,,,,,,,,,, 11v75 15.8 .8 35.9 3.0 394 8.9 80 2.1 2.6 Tr<.2 36.0 $2.8 v7.7 619 6—76 14.4 $59 29.3 —1.0 342 7.2 73 1.8 2.1 <.3 34.2 $2.0 —8.3 625 Valley View Cottages ____________________ 11—75 12.0 —11.3 23.9 —10.3 349 6.0 60 1.5 1.2 <.1 4.9 11.8 —7.9 627 T. H. Bonds Well _______________________ 11—75 10.5 —14.4 20.9 $201 446 5.4 96 5.7 ,6 <.1 64.6 $3.5 — 631 Warm Springs (Children's Pool) ___________ 11—75 35.4 31.4 65.5 25.0 607 21 112 3.7 7.4 .1 .9 $9 -7.9 631 6—76 35.4 29.5 63.7 22.4 601 20 119 3.7 6.9 <.3 1.6 $.6 —8.5 633 Ingall's Airport __________________________ 12—75 9.1 724.1 11.2 8.5 9 3.8 .5 .4 .2 <.1 68.3 $3.4 — Virginia—Falling Spring area 701 Falling Spring .......................... 12—75 25.0 25,4 59.8 38.4 848 18 158 3.8 16 <.3 24.4 2.2 —7.7 701 20.8 10.7 45.6 30.0 610 12 117 3.1 10 Tr<.3 32.6 $1.9 —7.8 702 H. D. Webb Well 27.2 36.8 70.7 74.3 1260 24 234 6.6 2.5 2.0 10.9 +1.0 '79 North Carolina—Hot Springs area 801 Hot Springs ______________________________ 12—75 39.3 47.7 80,9 37.3 719 31 135 10 10 .7 —.2 :2.0 —6.3 801 5—76 41.4 50,5 83.6 35.4 719 33 144 9.6 9.8 .4 —.5 $7 -6.4 803 City Well ................................ 12—75 15.2 38.5 72.3 54.0 202 25 13 6.6 4.5 <.1 40.6 $2.4 -6.2 803 5—76 15.2 41.7 75,3 56.3 194 27 11 6.9 4.3 <.1 33.3 $1.9 —7.0 804 Fairview Water .......................... 12—75 12.5 ‘50 30.1 25.4 54 7.4 4.9 .9 1.5 <.1 52.2 $3.0 ‘59 805 Bubbling Springs ________________________ 12—75 13.2 7.8 42.7 5.9 155 11 19 ,3 2.2 <.1 6.0 21.0 —6.2 810 P. Lovin __________________________________ 5—76 15.0 715.4 19.8 4.8 210 5.2 23 1,8 1.3 — 48.9 $2.4 —6.8 Georgia—Warm Springs area 901 Warm Springs ,,,,,,,,,,,,,,,,,,,,,,,,,, 12—75 30.9 29.5 63.7 26.0 192 20 22 1.2 3.8 <1 .8 _.9 —4.5 901 5—76 31.6 25.4 59.8 26.8 181 18 20 1.2 3.7 — .8 $.5 ‘50 905 South Spring ............................ 12775 17.7 —4.6 30.5 49.4 14 7.5 .3 .9 .6 <2 34.6 $2.4 —4.4 905 5—76 17.8 ~98 26.3 50.8 12 6.5 .2 .8 .5 <.1 30,3 :17 —5.2 907 J. Phillips .............................. 12—75 33.8 31.4 65.5 35.6 119 21 12 1.0 4.0 <.1 -1.0 $2.2 -4.3 909 A. Baxley ................................ 12—75 17.0 47.7 80.9 44.0 170 31 19 10 3.5 <.2 15.1 +1.4 -4.2 HYDROLOGY AND GEOCHEMIS'I‘RY OF THERMAL SPRINGS OF THE APPALACHIANS gas as a geothermal prospecting tool. They observed close correlation between helium levels and proximity to hot-water seeps at several locations in the western states; the effect was ascribed to escape of helium from the hot water into the local soil gas. Heat and helium are generated during the radioactive decay of uranium and thorium minerals. Decay of 238U to stable me and 4He evolves =35 X 1010 calories of heat per gram of helium created (from data in Birch, 1954). Calculations based on these data show that the heat produced by the number of decay events needed to add l,u.gHe/L is about 35 X 103 calories. If all this heat were absorbed and retained by the water, its temperature would rise about 35°C. Of course, in a ground-water system, much of this heat is dissipated by conduction in the aquifer material itself, so that the observed temperature in- crease will be smaller than the theoretical 35°C/ ugmHe. However, an elevated dissolved-He concentra- tion is still an indication of a subsurface heat source consisting of a concentration of radioactive minerals in or near the aquifer, and independent of the depth of circulation of the water. Dissolved helium cannot be used as a quantitative geothermal indicator largely because of the nature of helium gas. It is the least soluble and among the most mobile of all gases and may escape from solution before the water reaches the sampled spring or well, produc- ing the soil gas effect noted by Roberts and at the same time reducing the dissolved helium level. Helium con- centrations of the Appalachian water samples are listed in table 4. Water at equilibrium with 5 X 10‘6 atmospheres of He (sea-level value for the normal at- mosphere) will contain less than 0.01 ug/L of the dis- solved gas. Thus, all positive findings of He noted in the table (i.e., Trace, >0.1p.g/L and greater) indicate substantial supersaturation with respect to the atmos- phere. Dissolved carbon dioxide analyses provide an inde- pendent check on the field pH and aklalinity determi- nations. In this study, the gas analysis CO2 values help to provide better carbonate system data. Accurate val- ues of carbonate ion activities are obviously essential for geochemical studies of water-carbonate mineral reactions. Ordinarily, field measurements of pH and bicarbonate alkalinity, together with calculations of activity coefficients and ion pairing of the major sol- utes, are used to determine the carbonate ion ac- tivities of a water. An alternative method, involving calculation of a pH from the analyzed CO2 pressures and field alkalinities, was used in determing carbonate species activities for several of these Appalachian wa- ters.The calculations are described by Pearson, Fisher, and Plummer (in preparation). Carbonate mineral sat- uration data derived by both methods are listed in table 5. E23 Carbonate ion activities are very sensitive to pH. The analyzed CO2-calculated pH method may be ad- vantageous for calculating carbonate ion activities of high Poo, waters, which may evolve COZ, thereby in- creasing pH during pH determinations. We consider the errors in analyzed P002 values to be <5 percent for PCO2 >5 X 10‘3 atm. Calculation of pH from the P002 depends on the relationship co. + H20 2 H+ + Hco; with an equilibrium constant aH+aH(fU:; KP : __ C02 Poo.) al—IaO Noting that pH = —log am, or 10m = 1 + y 3H and rearranging, we obtain 10',“ = 1— “HCO; Pcog ango KPCO2 From (A), an error in PCO2 of 5 percent would corre- spond to a change of about 5 percent in 10"”, or a pH difference of 0.02 units. In contrast, direct pH meas- urements in ground waters are normally no more pre- cise than :0.05 units (Langmuir, 1971; Pearson and Rettman, 1976). - Recent studies of the C02 system in the oceans have made increasing use of measured PCO2 rather than pH as one of the parameters employed for determinations of carbonate ion activity (for example, Li and others, 1969). A discussion of the relative merits of various approaches to these determinations was presented by Park (1969). TRACE ELEM ENTS Dissolved trace elements are sometimes useful qual- itative indicators of geothermal waters. Truesdell (1976) has reviewed recent geothermal studies in which trace element concentrations help to define geothermal reservoirs. Arsenic was found in 8 of the 10 warm springs sampled, indicating a possible correla- tion between this element and the thermal history of the waters. However, the levels of arsenic determined (0.5—0.2 Mg/L, the lower detection limit) do not permit an assessment of the role of arsenic as a thermal indi- cator in these waters. ISOTOPES Ratios of stable-oxygen isotopes and stable-carbon isotopes, and the concentration of the radioactive isotope of hydrogen, tritium, were measured in many of the samples. Oxygen isotopes are qualitative indica- tors of the geography of the area of recharge of a E24 Gl‘ZOHYl)R()L()GY 0F GEO'I‘HERMAI. SYSTEMS TABLE 5,—Saturation indexes and carbon-isotope ratios for waters in Appalachian warm-spring areas Carbonate chemistry based on measured pH Carbonate chemistry based on analyzed C02 Calcu- Total Saturation index Meas- Ana- Total Saturation index Ad- Sam- Owner or name Date Meas— lated car» ______7 ured Calcu- 1 zed car- justed ple ured Pm 2 bonate Cal- D010> 8‘3Cpm, lated A.) 2 bonate Cal- Dolo— 5'3Cpm, No. pH (atmi) (mM/L) cite mite °/.,.. pH (atm) (mM/L) cite mite °/m New York—Lebanon Springs area 101 Lebanon Spring ______________ 10—75 7.60 0.0038 2.86 +0.11 +0.33 — 11.9 7.60 0.0039 2.86 +0.11 +0.32 —11.9 101 5+76 7.72 .0029 2.89 +03 —.19 +119 7.55 .0043 2.95 —.14 —.52 —12.1 109 John Koepp ,,,,,,,,,,,,,,,,,, 10—75 7.60 .0034 2.74 —.30 +.71 + 13.1 7.87 .0018 2.66 —.03 —.18 — 12.8 109 5-76 8.00 .0013 2.63 +07 —.05 —11.9 7.86 .0018 2.66 —.06 —.32 —12.0 114 B Schell ____________________ 10—75 7.20 .0093 3.53 —.20 ~~1.32 — — _ _ _ _ _ 116 F Amlaw 77777777777777777777 10—75 7.55 .0055 4.23 —.04 —.33 —11_5 — — _ _ _. _ Pennsylvania—Perry County Warm Springs area 201 Warm Springs ______________ 10—75 7.20 .0077 2.53 —.59 +2.05 +13.4 7.43 0045 2.39 +36 +1 58 +13 0 201 5—76 7.62 .0029 2.29 +.16 +1.28 —12.6 7.48 0040 2.34 —.30 +1 56 +12 8 202 Lloyd Hetrick ________________ 10—75 6.85 .0088 1.62 +1.45 +4.01 -— — —— — — — — 216 Stambaugh Spring ,,,,,,,,,, 10—75 6.50 .0160 1.73 +2.04 +4.84 —12.9 —— —— — — — —- 218 M. Loy Spring .............. 10—75 7.20 .0092 3.34 +37 —.95 —12.3 — — —— — -— — West Virginia—Berkeley Springs area 301 Berkeley Springs ____________ 10—75 6.99 .0163 3.35 —.59 — 1.85 — 15.2 6.98 0166 3.36 —.60 — 1.86 — 15 2 301 5—76 7.46 .0054 2.89 —.12 —.98 —13.9 6.94 0180 3.36 —.64 +2 02 +14 9 302 F. Wachter ,,,,,,,,,,,,,,,,,, 10—75 6.80 .033 5.78 —.34 —1.80 —15_5 —— — — —— — _ 302 5—76 7.07 .0174 4.88 +27 +1.33 +15.5 — — — — — 303 F. Wachter ,,,,,,,,,,,,,,,,,,,, 9—75 7.30 .0095 4.14 +24 +.75 — — —— — — — — 307 H. Hovermale ________________ 10—75 6.97 .037 8.35 —.02 —.50 —15.2 _ _ _ _ _ _ West Virginia—Minnehaha Springs area 401 Minnehaha Springs __________ 11—75 7.88 .0014 1.94 +04 —.51 +114 7.82 .0016 1.95 —.02 —.62 —11.4 401 6—76 7.90 .0013 1.90 +04 —.54 —11.4 7.72 .0020 1.93 +.13 + 90 +115 402 Camp Minnehaha ____________ 11—75 7.30 .0095 4.24 —.21 —.89 —14.4 — — — — — — 402 6—76 7.30 .0096 4.23 —.16 +.86 — — — — -— — — 403 A. T. White __________________ 11—75 7.20 .0045 1.59 —.87 +3.30 — — — — —- — — 406 F. Davis ____________________ 11—75 7.63 .0022 1.96 +36 +2.06 —13.1 — — — — — — 407 R. Shinaberry ________________ 11—75 5.50 .055 3.11 +3.59 +8.10 —22.9 — — — — — — 410 T. Rose ______________________ 6—76 7.68 .0039 4.61 +28 +.17 +2.1 7.95 .0021 4.50 +.58 + 71 +2 3 Virginia—Bolar Spring area 501 Bolar Spring ________________ 12—75 7.48 .0063 3.58 +.06 +.25 +9.2 7.30 0097 3 71 -.12 — 61 +9 5 501 6—76 7.42 .0065 3.28 +09 —.68 —10.1 7.40 0067 3 29 +.10 + 72 +10 1 Virginia—Hot Springs—Wann Springs area 602 Boiler Spring ________________ 11—75 6.65 0.117 10.23 +0.07 +0.05 +5.6 6.61 0.127 10.48 +0.03 +0.12 +5 7 602 6—76 6.55 .147 10.92 —.05 —.32 +5.8 6.69 .107 9 96 +09 —.04 +5 4 605 Cold Magnesia Spring ________ 11—75 7.20 .0095 3.28 —.37 +1.35 +8.8 — — — — — — 605 6—76 7.15 .0099 3.13 —.51 +1.70 +9.8 — —— — — —— — 607A Homestead Hot Well __________ 11—75 7.15 .022 6.85 +.14 —.01 +7.5 6.68 066 8 58 —.32 — 94 +8 9 619 M. Dunn Spring ______________ 11~75 7.32 .0080 3.56 —.11 —.77 —10.4 — — — — — — 619 6-76 7.42 .0058 3.19 +09 —.90 +7.4 — — —— — — — 625 Valley View Cottages ________ 11—75 7:30 .0103 4.57 +.17 —.77 +7.7 — — ——— — — — 627 T. H. Bond’s well ............ 11—75 7.48 .0080 5.32 +26 —.50 -— 7.26 .0134 5.61 +03 +95 — 631 Warm Springs 11—75 7.32 .0101 3.46 +24 +21 +8.2 7.08 .0175 3.66 .00 +26 +8.5 631 (Children’s Pool) ,,,,,,,,,,,, 6—76 7.20 .0132 3.50 +.14 —.06 +8.8 7.10 .0165 3.59 .04 —.26 +9.0 633 IrLgall’s Airport ______________ 12—75 4.50 .055 3.07 +7.0 —14.3 — — — — — — — Virginia—Falling Spring area 701 Falling Spring ______________ 12—75 7.32 .0139 5.63 +.41 +.38 +6.8 7.13 .0217 5.91 +22 .00 +7.1 701 6—76 7.32 .0109 4.49 +.17 +.16 +8.4 7.10 .0177 4.77 —.O5 —.60 +8.8 702 H. D. Webb well ____________ 11—75 7.80 .0275 7.80 +.49 +.66 +8.0 6.60 .098 10.09 —.05 +.43 +9.5 North Carolina—Hot Springs area 801 Hot Springs __________________ 12—75 7.50 .0045 2.15 +31 +30 +9.5 7.28 0074 2 23 +.10 +.12 +9.7 801 5—76 7.50 .0043 2.03 +35 +27 +9.2 __ _ ._ _ _ _ 803 City well ____________________ 12—75 6.75 .0179 2.68 +1.61 +3.16 +193 — — — — — — 803 5—76 6.88 .0127 2.36 +1.57 +3.06 — 19.2 —- —- — — — — 804 Fairview Water ______________ 12—75 7.10 .0020 5.75 +2.26 +4.53 + 12.0 — — — —- — — 805 Bubbling Springs ____________ 12—75 7.75 .0016 1.84 + .49 — 1.07 — 13.4 — — — — — — 810 P. Lovin ______________________ 5—76 8.10 .0009 2.25 +04 +05 — — — — — — —— Georgia—Warm Springs area 901 Warm Springs ______________ 12475 7.40 .0053 2.15 —.48 —.97 — 13.7 7.20 0083 2 24 +.67 +1.28 +140 901 5—76 6.95 .0146 2.35 —.97 +1.89 +145 7.19 (0083) 2 17 —.73 +1.40 —14.0 905 South Spring ________________ 12~75 4.00 .288 1.22 +7.4 —14.8 —30.1 — —— — —— 905 5—76 4.90 .024 1.05 +6.9 — 13.8 +24.3 —— — — — — +- 907 J. Phillips .................. 12—75 6.42 .028 1.82 +1.93 +3.69 —16.4 — — — — — — 909 A. Baxley ____________________ 12—75 7.02 .0081 1.89 +1.22 +2.98 —20.0 — — — — —— HYDROI.()(}Y AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS ground water and of whether a ground water has reacted with rock at temperatures greater than about 150°C. Carbon isotopes indicate the types and extent of chemical reactions between ground water and carbon- ate minerals. Tritium has a relatively short half life (12.3 years), and nearly all tritium now in the hydro- sphere was injected within the last 20 years. The tritium content reflects whether a ground water was recharged more than about 20 years ago or more re- cently, and frequently it is possible to use intermediate tritium values to estimate the proportions of young and old water in a mixed sample. Stable isotope ratios are reported in 6 notation. 8 represents the deviation in permil (parts per thousand) of the isotope ratio in a sample from that of a standard. The 18O/lSO ratio of the oxygen standard (SMOW) is similar to that of ocean water; the 13C/‘gC ratio of the carbon standard (PDB) is similar to that of marine limestone. Tritium concentrations are reported in tritium units (TU). One TU equals 3.2 picocuries per liter (pCi/L). Waters on the land surface are generally depleted in 180 relative to SMOW—that is, their 8‘80 values are negative. This depletion occurs because the vapor pres- sure of water molecules containing the heavier isotopes is slightly less than that of common water, H2‘60. Dur- ing evaporation and condensation in the hydrologic cy- cle, molecules containing heavier isotopes are concen- trated in the liquid phase. As water evaporates from the ocean, the vapor is depleted in 18O and the amount EXPLANATION 0 Warm springs 20 r— X Wells and cold springs Spring Autumn )0 Perry County area, Pa. Other Va. and W. Va. spring areas Berkeley Springs area, W. Va. MEAN ANNUAL TEMPERATURE, IN DEGREES CELSIUS l 18 6 oSMOW IN PERMIL A \Warm Springs area, Ga. Hot Springs area, N.C.\ E25 of depletion becomes greater as the temperature of evaporation decreases. Further isotopic fractionation takes place as water is condensed and reevaporated during atmospheric transport. The amount of fraction- ation is inversely proportional to temperature, and also varies with altitude, latitude, and distance from the ocean. Figure 21A shows the variation of 8‘80 values of samples from this study with mean annual tempera- ture. The lines on the figure have a slope of 0.5 permil/ °C, within the range of 0.4 to 0.7 permil/°C found in more extensive studies elsewhere (Dansgaard, 1964). The two lines on the figure represent the seasonal range in 8‘80 values to be expected. Samples collected in the spring are characterized by more negative 6‘80 values which reflect the lower temperature winter pre- cipitation; the less negative 8180 values of the autumn samples reflect higher temperature summer precipita- tion. Oxygen isotope ratios of ground water can be changed by chemical and isotope exchange reactions between the water and the aquifer rock. Because the 5‘“O value of rock-forming minerals is strongly posi- tive, reactions between water and rock will make the 18O values of the water more positive. Such reactions occur only very slowly at temperatures below 150~ 200°C. Thus, anomalously positive 6180 values may in- dicate that a ground water has been exposed to high temperatures (White and others, 1973). As figure 21A shows, the 6180 values of the thermal Autumn, 1975 " 3° Lu 0 o D '2 '— o < _| I ’— D: 0 Z (If LLI UJ CE 0 LLI _ O — 46 I l | I l l l | | l 0 20 40 60 80 100 TRITIUM UNITS B FIGURE 21.—-Isotopes in Appalachian warm springs area waters. A, Mean annual temperature and variation ofvS18 O values of samples. B, Latitude and variation of tritium concentration of samples. E26 spring samples are within the ranges of values of the cold wells and springs in the same areas, allowing for the seasonal variations. This is evidence that the areas of recharge to the thermal spring systems are within kilometers or at most tens of kilometers from their dis- charge points. Further, it provides no evidence that would suggest that the thermal spring waters have been as hot as ISO—200°C. Figure 21B shows the variation of the tritium con- centration of samples from this study with latitude. The line on the figure is drawn through the highest tritium values of samples collected in the autumn of 1975. The increase in tritium with latitude, by a factor of roughly 3 from 32°—42°N, and the tritium values themselves are what one would expect to find in ground waters which are a mixture of waters recharged since the period of major tritium injection began in the early 1960’s (Stewart and Farnsworth, 1968; Stewart and Wyerman, 1970). The line may therefore be taken to represent the approximate tritium content of shallow, recently recharged ground waters at these latitudes. Samples from the thermal springs at Warm Springs, Ga., Hot Springs, N.C., and Hot Springs and Warm Springs, Va., contain no tritium. Water from these springs is thus at least 20 years old. Tritium in the thermal springs at Bolar and Falling Spring, Va., was about 25 TU in 1975. From the line in figure 21B, the maximum tritium content of shallow ground water at the latitude of these springs can be estimated to be 64 TU. If the tritium content of the thermal component of these springs is zero, the proportion of thermal (T) to shallow, cool (C = 1 ~T) water is: 25 = T(0) + 64(1 — T) T = (64 — 25) /64’= 60%. The remaining thermal springs contained from 2 to 12 TU in 1975, a level corresponding to a thermal-water fraction of from 97 to 81 percent. (ZARBONATE CHEMISTRY AND ISO’I‘OPES The carbonate chemistry of ground waters—whether the waters are in equilibrium with carbonate aquifer minerals—and the stable carbon isotope ratios of dis- solved carbonate, give qualitative indications of water residence times and flow paths. Enough information is available on the warm and cold-spring samples for this study to permit a detailed interpretation of the carbon- ate chemistry of most of the spring systems. Table 5 shows both the analyzed PCO2 and PCO2 values calculated from the pH and other chemical analyses. In some samples, such as the 1975 samples 101, 301, 401, there is good agreement between the calculated and GEOI—IYDROLOGY OF GEOTHERMAL SYSTEMS analyzed POO2 values. In others, such as 1975 samples 631 and 702 and 1976 samples 301, 701, and 602, the calculated PC02 values are lower than the analyzed values. Saturation indices (SI) for the minerals calcite and dolomite were also calculated and are given in table 2. Both PCO2 and SI values were calculated using WATEQF (Plummer, and others, 1976). SI values cal— culated from measured pH and alkalinity values range from great undersaturation (Slcalcm,<—3) in high- tritium, high-P002 recharge samples to apparent over- saturation (SIcalcite > + 0.2) in such warm-water sam- ples as 631, 701, 702, and 801. In these apparently oversaturated samples, however, P002 calculated from pH and alkalinity is lower than that analyzed. No matter how carefully a field pH is measured, there is always a possibility that some C02 outgassing will occur during collection and the measurement pro- cedure, causing an erroneously high pH to be read. To test this possibility, the saturation index calculations were repeated assuming that the measured P002 values were correct (Pearson, and others, in prep.) The results of these calculations are also shown in table 5. Consider the samples of 631 as an example. On the basis of measured pH, both samples are oversaturated with calcite and have lower calculated than analyzed Pc02_ When recalculated from the analyzed PCOz’s, how- ever, both samples are just saturated with calcite and have lower calculated than analyzed pH’s. The un- changing measured temperature and the identical (within measurement precision) PCO2 and tritium val- ues suggest that the chemistry of water from 631 should also be constant. The recalculated SIC,“cite and pH values better reflect a constant chemistry than do the measured pH values and the SI values calculated from them. Thus the SI’s based on analyzed PCO2 prob- ably best represent the true state of the water chemis- try. As water recharging a ground-water system passes through the soil zone, it dissolves CO; from the soil atmosphere. Soil CO2 is a product of plant respiration and decay, and it has 513C values near —25°/00. If this water then passes into a marine limestone 0r sandstone with carbonate cement of 6‘3Cz0°/00, the limestone or cement will dissolve in the soil C02- charged water and produce bicarbonate with 8'3C val- ues as positive as —10 t0 ~12°/00. In waters in rock where marine carbonate is plentiful, further reactions between rock and water may make the 813C value even more positive. Thus the 813C content of a ground water is a measure of the extent of reaction between the water and carbonate rock. The measured 613C values are given in table 2, and in Hobba, Chemerys, Fisher, and Pearson (1977) where HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS the method of collection and analysis is also described. If C02 gas has been lost from a sample before or during the collection of the 513C sample, the analyzed values will not be a true measure of the 813C value of the source water before outgassing because of the isotopic- fractionation factor of several permil between C02 gas and dissolved carbonate species. If the PCO2 of the source water is known, as from a dissolved gas analysis, the correct 813C can be calculated from the value measured on an outgassed sample (Pearson and others, in prep.) The 613C values of those samples on which the activities of carbonate species were calcu- lated from analyzed P002 values were also adjusted for the effects of outgassing. The adjusted 6‘30 values are given in table 2. The Shame values, based on analyzed PCO2 where ap— propriate, and the 813C values reflect the types of rock through which the various waters have traveled. Some ofthe cold spring samples, including numbers 407, 633, and 905,.have the chemical characteristics of soil-zone waters. They have high Pco2 values, pH values lower than 6.0 and 813C values about -25°/00. Their tritium contents show that they were recently recharged. As such waters move deeper into a ground-water system, they encounter and dissolve limestone and dolomite. This process lowers the Pcoz, raises the pH, and makes the 813C values become more positive. Water long in contact with carbonate rock reaches saturation with calcite; however, water in contact with carbonate rock for only a short time, or in rock containing little car- bonate, may be undersaturated. The samples reflect this pattern. Thermal waters is- suing from sandstone or quartzite in Perry County, Pa., Berkeley Springs, W. Va., and Warm Springs, Ga., are all undersaturated with calcite, though their low tritium contents suggest a sufficiently long residence time for saturation with calcite to have been reached if they had been exposed to carbonate rock. Samples that are associated with carbonate rock (such as 109, 602, 631, and 801) and that contain little or no tritium are saturated with calcite. Those which have significant tritium contents and which represent either relatively recent waters or mixtures of waters of short or long residence times may or may not be saturated with calcite. (2H EMICAL (Ll-IOTHERMOMl‘L'l'ERS The measured temperature of a thermal spring or a hot-water well provides a direct indication of a subsur- face heat source. However, thermal reservoir and discharge-water temperatures may not be equal; heat loss by conduction can lower the temperature of the thermal water considerably below that of the heat source. A number of geothermometers have been de- E27 vised for estimating reservoir temperatures directly from the chemistry of thermal water samples (Trues- dell, 1976). Successful application of geothermometers depends on the attainment of equilibrium in a temper- ature dependent water-rock reaction in the thermal re- servoir, and no subsequent change in thermal water composition. The assumptions inherent in the use of geothermometers are discussed in Fournier, White, and Truesdell (1974). Mixing of hot water with shallow, cold water often occurs in thermal spring systems. The effects of the admixture of cold water must be taken into account before estimates of reservoir temperatures can be made. In general, the temperature of the mixed water will be determined by the enthalpies of the hot and cold components (Fournier and Truesdell, 1974). At tem— peratures below boiling, the enthalpy of water (in calories per gram) is numerically equal to its tempera- ture (Tobsmed) of the mixed—water spring will be equal to the mean of the fractions of cold (F cum) and hot (Fm) water times the respective temperatures, or: Tcold ' Fcold ' + Thot ' Fhm = Tobserved (1) where Fl'Ul(l + thl 2 1 (2) The mixed water will also contain an intermediate concentration (C) of the geochemical indicator, so that: Ccold ' Fcold + Chot ' Fhot = Canalyzed (3) The indicator concentration and the temperature of the cold water can be estimated for favorable cases. The geothermometer relationship between temperature and indicator concentration of the hot water Thu! 2 $(Chm) (4) then provides a unique solution ofthe equations for the mixing fractions and the temperature of the thermal reservoir in systems for which conductive heat loss is negligible. Alternatively, if the concentration of any conservative constituent ofthe hot water is known, and if this concentration differs significantly from that of the cold water component, then the mixing fractions can be determined and a hot water temperature may be calculated directly from the first three equations, and a reservoir temperature may be estimated from avail- able geothermometers. An example of this alternative scheme for calculating mixing ratios is the use of the assumption that tritium in the hot component should E28 be negligible. Then the second term in equation 3 is zero, and the cold fraction is readily calculated. The oldest and among the most generally useful of the chemical geothermometers make use of the silica content of the sample. Reactions between water and common alumino-silicate minerals are frequently of the form: cation-Al-silicate + H20 —> Si02((.iss,,l,vk.d, + cation + Al-Silicate(s) where the reactant cation-Al-silicate may be a sodium, potassium, or calcium feldspar or clay mineral and the product Al-silicate is another clay mineral. The result of such reactions is to increase the dissolved silica con- tent of the water until a silica mineral such as quartz (at relatively high temperatures) or chalcedony (at 80 — EXPLANATION - Spring 70 __ >< Well, <25% modern water A Well, > 25% modern water SiOz, IN MILLIGRAMS PER LITER GEOHYDROLOGY OF GEOTHERMAL SYSTEMS lower temperatures) precipitates. The degree to which chemical equilibrium between dissolved silica and the precipitating mineral is approached depends on the temperature and the time available for the reaction to proceed. The silica geothermometer assumes attain- ment of equilibrium between dissolved silica and a par- ticular silica mineral at the reservoir temperature, with no precipitation of dissolved silica after the water has left the heated reservoir. Determination of the silica content of the water sample will then indicate the thermal reservoir temperature, provided the geothermometer has been properly calibrated for the appropriate form of silica involved in the reaction. Fig- ure 22 is a plot of dissolved silica concentrations versus observed temperatures for the Appalachian spring and well samples. The curves represent concentrations of dissolved silica in equilibrium with quartz and with chalcedony at the indicated temperatures. They are de- rived from silica solubility data determined by a 0 I I I I 0 10 20 3’0 40 TEMPERATURE, IN DEGREES CELSIUS I I I I I I so 60 7o 80 90 100 FIGURE 22.—Dissolved silica content and temperatures of Appalachian spring and well samples. HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS number of workers and reported by Fournier (1973). From the plotted points of figure 22, all the Appala- chian spring samples are seen to be supersaturated with quartz; 2 of them (the North Carolina warm spring samples) are supersaturated with chalcedony. Well water data illustrated in figure 22 were divided into 2 age groups according to sample tritium content relative to the expected tritium level of fresh recharge in the particular geographical area (from fig. 21B.). Dissolved silica concentrations of 8 of the high-tritium well waters are above chalcedony saturation, indicat- ing rapid low-temperature dissolution, by fresh re- charge, of silica from shallow earth materials. Pre- sumably little or no re-precipitation occurs because of slow equilibration rates for the more stable silica pha- ses. On the other hand, the dissolved silica- temperature relationship for the older well waters is similar to that for the springs; a general increase in dissolved silica with increasing sample temperature is evident from the figure. Indicated quartz and chal- cedony temperatures of the samples are included in table 4. Another geothermometer that has been applied with fair success in thermal systems is the Na-K—Ca or cat- ion thermometer (Fournier and Truesdell, 1973). Un- like the silica geothermometers, which are based on laboratory measurements of silica solubilities, the cat- ion geothermometer is empirical and the particular mineral-water reactions that make it work are not known. The formulation of this geothermometer for waters of less than 100°C and with a molar ratio VCa/Na greater than 1 is: t(°C) = 1647 log(Na/K) + <4/3) log(m/Na) + 2.24 — 273.15 where Na, K, and Ca are the concentrations of sodium, potassium, and calcium, respectively, in moles per li- ter. Cation temperatures are also listed in table 4. Tem- peratures calculated on samples with low Ca, Na, and K concentrations have large errors associated with them because of the imprecision of analysis at low con- centrations. Temperatures calculated for some samples with higher Ca, Na, and K concentrations are nega— tive, showing that the waters have not reached the minimum temperature necessary for the reactions that give rise to the geothermometer effect to become ac- tive. Because the cation geothermometer is empirical, this minimum temperature cannot be given un— equivocally. Several of the warmest samples, including E29 those from Warm Springs, Ga. (no. 901), Hot Springs, NC. (no. 801), and Warm Springs and Hot Springs, Va. (no’s. 631, 607, and 602), have cation thermometer temperatures in reasonable agreement with both measured and chalcedony temperatures. The coolest of these are samples 607 A and B, which are about 20°C. Other samples, however, such as 301 and 702, with measured temperatures above 20°C have negative ca- tion temperatures. Because all samples above 30°C do have reasonable cation temperatures, this might be a minimum temperature for reliability of this geother- mometer. At relatively low temperatures (<50°C), geother- mometer estimates will not be particularly accurate because the basic premise behind their operation may not be valid, that is, equilibrium of the temperature dependent reaction may not be attained. However, fair success of the chalcedony and cation thermometers is indicated by the data for the warmest of the Appala- chian spring systems. This agreement between chal- cedony, rather than quartz, with measured and cation temperatures, together with the relatively low (<50°C) maximum temperatures suggests that chalcedony rather than quartz is controlling the silica concen- trations in these systems. DISCUSSION SI LICA-ROC K SPRING SYSTEMS The Perry County, Pa., Warm Springs (201) and Berkeley Springs (301) and Minnehaha Springs (401), W. Va., issue from the Oriskany Sandstone; and the Georgia Warm Springs (901), from the Hollis Quartzite. Certain chemical characteristics of these springs are similar because of their similar aquifer rock type, and it is convenient to discuss them to- gether. All have relatively low concentrations of dis- solved solids—181 to 258 mg/L—as compared to the values of 266 to 848 mg/L in waters from carbonate rocks. A trace of dissolved helium was found in only one sample from Minnehaha Springs (401), and none was found in any of the other three silica-rock warm springs. Generally higher helium levels were found in samples from the carbonate warm springs and hot wells. The four silica-rock thermal springs contain little (201) to no (901) tritium, and all but Minnehaha Springs (401) are significantly undersaturated with calcite and have 513C values of —13°/00 or more nega- tive. These undersaturated waters could not have been exposed for any extended period to aquifer rock con- taining carbonate minerals, and so they must have flowed from recharge to discharge through the low- carbonate sandstones from which they issue. Although Minnehaha Springs (401) issues from the Oriskany E30 Sandstone, the water is saturated with calcite. This spring is in the vicinity of the Virginia thermal springs, all of which are associated with limestone. The fact that Minnehaha Springs is saturated with calcite suggests that, unlike the other sandstone springs, its water has been exposed to carbonate rock. PERRY COUNTY, PA., AREA Waters from warm spring 201 and samples 202 and 216 are derived from the Devonian Oriskany Sandstone. Spring 218 issues from the Silurian To- noloway Limestone. The sample from spring 218, with a tritium content of 78 TU (fig. 22 and table 4), is virtually entirely modern water, and the concen- trations of chloride, nitrate, and sulfate suggest the possibility of human influence on the water composi- tion. The Oriskany Sandstone samples are all dilute calcium bicarbonate waters. The only pronounced compositional differences between the warm spring and the other two waters are the slightly elevated silica content and the generally higher mineralization of the warm spring, suggesting a longer period of flow in the aquifer for this sample. The quartz geother- mometer temperature for the 1975 sample from spring 201 is probably a conservative estimate of maximum temperatures at depth in the aquifer. BERKELEY SPRINGS, W. VA., AREA The warm springs issue from the Oriskany Sandstone. Well 302 taps both the Oriskany and the overlying Marcellus Shale, whereas well 303 is en- tirely in the shale. Sample 307 is from a well completed in the Tonoloway Formation. The waters are of calcium bicarbonate and calcium-magnesium bicarbonate types, with relatively high sulfate in waters from the shale. The thermal-spring water is the most dilute water of those sampled in the area; sample 307, which from its tritium content of 72 TU appears to be all modern water, is the most highly mineralized sample. Silica temperatures are highest for wells 302 and 303 in the Marcellus Shale. These waters are super- saturated with chalcedony at the observed tempera- tures (fig. 22). However, recharge to the shale in this area is local and the measured tritium levels suggest that these samples contain about 50 percent recent re- charge. Measured temperatures do not indicate significant local heating of the rock; therefore, the high silica contents in these samples very likely result from silica from weathering reactions having had insuffi- cient time to equilibrate with the more stable chal- cedony phase. Indicated cation thermometer temperatures are all much lower than observed temperatures; however, GEOHYDROLOGY OF GEOTHERMAL SYSTEMS predicted temperatures for the warm springs are the highest in this group of samples. The relatively high potassium content of the spring water appears to be the only significant chemical indicator that may be associ- ated with the source of heat. There is little likelihood that temperatures at depth in this aquifer system would exceed the indicated temperature of about 38°C for the warm spring. MINNEHAHA SPRINGS, W. VA., AREA The warm springs in this area flow from the Oris- kany Sandstone. The calcium bicarbonate compo- sitions of the waters are generally similar to those in the Perry County and Berkeley Springs areas, except that sulfate constitutes almost 30 percent of the anion equivalents in the Minnehaha warm spring samples 401. Well 402 is completed in the Hamilton Formation, stratigraphically above the Oriskany. Samples 403 and 407 from the Oriskany Sandstone contain high (>20 mg/L) nitrate concentrations, and well 407, with a tritium concentration of 88 TU, is modern water. Com— positions of these well waters are very likely affected by human activities. Spring 406 issues from the To- noloway Formation. Its tritium level of 63 TU indicates that this calcium bicarbonate water is also recent. Highest silica concentrations in the Minnehaha Springs area samples are found in well 402. The water is near equilibrium with chalcedony at the observed water temperature, whereas warm spring water 401 is considerably below chalcedony saturation. As is the case in the Berkeley Springs area, the stratigraphi- cally and hydraulically higher formation contains young (high tritium) water, which according to the silica geothermometers, has been heated more than the thermal spring water in the area. Cation temperatures for spring 401 and well 402, although much lower than observed temperatures, do correspond with the ob- served trend in temperature. Levels of potassium in the warm-spring and well waters are similar to those noted at Berkeley Springs. Also the observed trace of dissolved helium in the 1975 warm spring sample sug- gests that the spring water has undergone some heat- mg. High temperatures at depth in the aquifer do not appear likely; the quartz thermometer temperature indicated for the warm spring water (about 34°C) is probably a conservative estimate of maximum temper- ature for water in the Oriskany Sandstone in this area. WARM SPRINGS, GA., AREA The warm spring (901) and cold spring (905) flow from the lower Paleozoic Hollis Quartzite. Hot well 907 penetrates the younger Manchester Schist, but it may HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS also draw water from the underlying Hollis Quartzite. Well 909 is completed in the Carolina(?) Gneiss of former usage, which is older than the Hollis Quartzite. Spring 905 yields extremely dilute, modern water with a composition unlike that of the warm spring. Chemically, the warm-spring and hot-well waters are similar. Both are waters of calcium-magnesium bicar— bonate type, and relatively high in dissolved oxygen. The concentration of dissolved solids in the spring water is slightly higher than that in the hot well water (table 4). Well 909 differs from the warm spring and hot well in that sodium amounts to 25 percent of its cation equivalents, and dissolved oxygen is very low. Tritium analyses indicate that spring 905 is almost entirely modern water, and that water from well 909 is approximately half young water. Warm spring 901 and hot well 907 do not contain significant components of modern water. Chalcedony- and cation-thermometer temperatures are in fair agreement with each other and with the observed temperatures for the warm spring and hot well. The hot well probably draws water from the same source that supplies the spring. The agreement be- tween geothermometer and observed temperatures for these waters indicates that the reservoir temperature in the warm spring system is approximately equal to the surface temperature of the water; little heat has been lost by conduction, and no evidence of mixing is apparent from the data. Sample 909 geothermometer temperatures are also in good agreement with each other, but they are con- siderably higher than the observed temperature. The agreement between the chalcedony and cation geo— thermometers may be fortuitous; high dissolved silica may result from low-temperature surface weathering reactions, and mixing of two cold waters can result in mixed-water compositions from which high cation temperatures would be predicted. If well 909 is indeed a mixture of hot and cold water which has been cooled conductively, the hot reservoir temperature can be cal- culated as follows. From the tritium content of this sample (15 TU) and that estimated for shallow ground water in the area (35 TU, fig. 21B), the sample appears to contain about 45 percent cool, shallow water. If the silica content of this water is 7.0 mg/L (from samples 905 representing high-tritium, local, shallow ground water), the silica content of the hot water can be calcu- lated to- be about 50 mg/L, which corresponds to a chal- cedony temperature of 70°C for the hot component. A similar calculation for the cation temperature of the hot component of sample 909 gives 53°C, which is below the chalcedony temperature. This suggests that deep waters hotter than 60—70°C probably do not exist in this area. E31 (IARBONATE-ROCK SPRING SYSTEMS LEBANON SPRINGS, N.Y., AREA The spring and wells sampled in the Lebanon Springs area all tap the Stockbridge Limestone of Early Cambrian to Early Ordovician age. Well 114 also draws water from the overlying Walloomsac Slate. Samples 101, 109, and 116 are calcium-magnesium bicarbonate waters; sample 114 is of the calcium sul- fate type. The close similarity in chemical composition of waters from the thermal spring (101) and artesian well 109 suggest that they may be derived from the same source. Seasonal chemical variability in spring 101 probably results from changes in proportions of modern water feeding the spring, as indicated by the tritium analyses (table 4). This temporal variability has diminished, tritium has decayed, and heat has been lost by conduction by the time the water reaches well 109. Wells 114 and 116 were sampled to provide data for comparison with the spring analyses. These waters were thought to be typical of shallow wells in the Stockbridge Limestone. However, sample 114 is a cal- cium sulfate water, unlike that of the spring. Sample 116 is chemically similar to the warm spring water, although relatively higher in total dissolved minerals and depleted in dissolved oxygen and in tritium; the chemical composition indicates that this water has been in the aquifer longer than the spring water. On the other hand, spring 101 and well 109 samples are higher in potassium, chloride, bromide, and strontium than is the well 116 water, and the samples contain detectable amounts of helium. These differences may be associated with the path of circulation and the source of heat for the spring. The amount of dissolved helium in the warm spring, 3 lug/L, indicates that the water may have been heated by as much as 35°C. Neither the silica nor the cation geothermometer in- dicates a high reservoir temperature for any of the samples from the Lebanon Springs area. If the spring water is assumed to consist of a mixture of a heated, quartz-saturated old component with 13 percent of a cold, modern water with negligible silica, then the silica content of the undiluted thermal water would be 1 + (1—0.13)X 12:14 mg/L SiOz, and the corresponding quartz temperature would be 51°C. (The estimate of 13 percent is based on the measured tritium content of 11.6 TU, as compared with the content of 90 TU esti- mated for fresh recharge; see fig. 218.) This value of temperature is probably a reasonable upper limit on aquifer temperatures at depth in the area. The carbonate chemical parameters are in keeping with the source rock of these samples. Stockbridge Limestone samples 109 and 116, with no tritium and hence an age greater than 20 years, are saturated with E32 calcite and have 8'3C values suggesting that their dis- solved carbonate is about an equal mixture of soil—air C02 (813C20250/00) and marine limestone (8130:0700). Lebanon Springs itself (101) and sample 114 may be slightly undersaturated with calcite. Lebanon Springs contains tritium, and it is thus an admixture of water too young, perhaps, to have reached calcite saturation. Sample 114 draws water, in part, from the Walloomsac Slate, which may contain too little calcite to permit calcite saturation to be attained. BOLAR SPRINGS, WARM SPRINGS, HOT SPRINGS, AND FALLING SPRING, VA., AREAS The warm springs sampled in these areas flow from the Ordovician Lowville Limestone; sample 633 is from the Silurian rocks of Clinton age; the other cold springs and cold wells sampled draw water from the Lowville Limestone, or from the Trenton Limestone, and Mar- tinsburg Shale or rocks of Black River age, which in— clude the Lowville Limestone. Hot wells 607 and 702, completed in the Lower Ordovician Beekmantown Group, draw water from elevations below those of nearby springs in the Lowville Limestone. Sample 607A represents a mixed water consisting of compo— nents from two or more producing zones. The February 1976 samples from well 607, on the other hand, were obtained from isolated intervals; sample 607B repre- sents the upper aquifer water and sample 607D is from the deepest water—bearing zone. Except for the extremely dilute sample 633, dis- solved solids of the waters sampled in the four areas ranged from 288 to 1,260 mg/L (table 4). Calcium and magnesium account for nearly all the dissolved cat- ions; bicarbonate and sulfate in varying proportions are the major anions. Small amounts of dissolved helium were found in one or more of the samples taken from the warm springs and hot wells in the four areas and in one sample from spring 619. The concentration of helium in hot well 702, about 2 ug/L, suggests that this water may have been heated by as much as 70°C. Tritium in warm springs 501 and 701 indicates that these samples con- tain one-third to one-half modern water; presumably the temperature of the older component may be sub- stantially higher than the observed temperatures of the flowing springs. The carbonate chemistry of these thermal springs is in keeping with their origin from a carbonate rock. All are saturated with calcite, and their 8‘3C values are more positive than —10°/00. Calcite saturation and 813C values in this range are typical of waters of moderate to long residence times in limestone. Boiler Spring (602), which is saturated with dolomite as well as with calcite, has a 8‘3C value more positive GEOHYDROLOGY OF GEOTHERMAL SYSTEMS than —6°/00. 6‘3C values in this range, with calcite and dolomite saturation, suggest either that the Boiler Spring water had a longer residence time than that of the other springs or that the other springs are mixtures of a Boiler Springlike water with a younger water. Comparisons of observed temperatures with silica thermometer temperatures for the warmest springs (602 and 631) and for hot well sample 607D indicate slight undersaturation of the waters with chalcedony. Cation thermometer temperatures for these samples are in fair agreement with observed temperatures, and here, as in other warm spring areas, the higher cation temperatures are associated with relatively elevated potassium levels in the waters. Relatively high cation temperatures are indicated for spring 701; nearby hot well 702, with an observed temperature higher than that of the spring, has a predicted cation temperature of <0°C. On the other hand, indicated silica tempera- tures for these samples are in the reverse order. The modern water component of the hot-well sample, indi- cated by its tritium content, apparently influences the mixed-water chemistry enough to invalidate the cation thermometer results. Samples 701 and 702 can be interpreted as mixed, conductively cooled waters, and by making certain as- sumptions, the maximum probable temperature of their hot component can be calculated. Thus, if it is assumed that the tritium content of the thermal com- ponent is 0 TU and that of the local cold ground water in the autumn of 1975 was 64 TU (fig. 218), the tritium contents of samples 701 and 702 suggest that the sam- ples contain about 60 and 85 percent ofa thermal com- ponent, respectively. If it is further assumed that the silica content of the cold ground water is 7 mg/L, the silica content of the hot component can be calculated using the expression SiO . _ ‘ SiO 2mIx 7 X proportion cold 2hot — proportion hot The silica content of the hot component of both samples calculated this way is about 26 mg/L; this figure cor- responds to a chalcedony equilibrium temperature of 40°C. It is not possible to calculate a cation tempera- ture of the undiluted thermal component because val- ues for the calcium, sodium, and potassium contents of the cold component are not known. However, the cation geothermometer expression (above) includes the cation concentrations as ratios and so mixing with a presum- ably dilute water should not significantly change the calculated temperature. Thus, for samples 701, at least, the cation temperature supports the conclusion from the calculated chalcedony temperature that the HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS maximum temperature of the thermal component is about 40°C. The occurrence of large warm springs issuing from the Lowville Limestone over a distance of more than 15 miles, and the good yields of warm water from the two wells in the Beekmantown nearly 10 miles apart sug- gest the existence of an extensive warm reservoir in this part of the Appalachians. Boiler Spring (602) and the warm component of the mixtures 701 and 702 have concordant chalcedony, cation, and measured tempera- tures that suggest this reservoir is at about 40°C. HOT SPRINGS, N.C., AREA Waters sampled in the area are from rocks of Cam- brian age. Warm spring 801 and well 810 are in the Shady Dolomite. Spring 805 flows from the overlying Rome Formation, whereas wells 803 and 804 are com- pleted in the older Murray Shale and Hesse Quartzite, respectively. Although the warm spring is saturated with dolomite, its gross chemistry is predominantly of the calcium sulfate type, unlike the water from Shady Dolomite well 810, which contains equivalent amounts of calcium and magnesium as major cations and bicar- bonate as the predominant anion. Spring 805 is similar in chemical type to the water from well 810. The warm spring water, in addition to being of a different chemi- cal type, is also more highly mineralized than are the other samples from the area (table 4). Well 803 water is unusual in that iron accounts for nearly 18 equivalent percent of the cations. Highly re— ducing conditions in the aquifer are indicated by the presence of measurable concentrations of dissolved methane in the water. On the other hand, measurable oxygen was present at the time of sampling, as shown by determinations in the field of dissolved oxygen, and the high tritium concentration indicates a modern- water component of about 80 percent. Apparently, when this well is pumped, a large amount of shallow, modern, oxygenated water is drawn up with the highly reducing older water in the shale, resulting in an un- stable solution containing ferrous iron and dissolved oxygen. The oxygen is rapidly consumed, producing a voluminous precipitate of ferric hydroxide within a few minutes after sampling. Tritium levels indicate that samples 804 and 810 are nearly entirely recent water. Spring 805 contains a les- ser component of young water, while the warm spring (801) contains only relatively 01d water. Small quan- tities of dissolved helium were found in the warm- spring water but not in any other samples from the area. The higher helium concentration (from the sam- ple collected in December 1975) may indicate heating of the water by radioactive decay processes to 24°C above normal ground-water temperatures. E33 The young waters from wells 804 and 810 are under- saturated with chalcedony; older water from cold spring 805 is near saturation. Warm spring 801 and well 803 water, on the other hand, are supersaturated with respect to chalcedony. The relatively low flow (~0.5 US) of the warm spring, as well as the observed seasonal variability in its temperature (table 4 and fig. 17), are consistent with the possibility of appreciable conductive cooling taking place between the heat source and the site at which the spring flows. If such conductive cooling occurs, then the water temperature near the heat source may be considerably greater than the measured outlet temperature, quite possibly as high as 50°C (equilibration with chalcedony). Cation thermometer temperatures suggest that the chalced- ony temperature is the probable reservoir temperature. If mixing rather than conductive cooling accounts for the observed chalcedony supersaturation of warm spring 801, then a hot reservoir temperature can be calculated from the equations described above, using appropriate values of temperature and dissolved silica for the cold component and assuming chalcedony sat- uration of the hot water component. Solution of the equations requires an analytical expression for the chalcedony-temperature curve. An equivalent graphi- cal solution can be obtained, as illustrated in figure 22. Point A represents the December, 1975, warm spring sample. The cold component, at equilibrium with chal- cedony at a temperature of 142°C, is indicated as point B. Tie-line BA which extends to the high-temperature intersection with the chalcedony saturation curve at point C, gives the temperature and dissolved silica con- tent of the hot water component. This procedure yields a reservoir temperature of 77°C and silica content of 57 mg/L for the hot water. Interpretation of the thermal history of the water from well 803 is difficult. Quite possibly the high dis- solved silica level derives from a lack of time for the silica from weathering reactions in the high-tritium cool water to reach equilibrium with chalcedony as seems to have occurred at the Berkeley Springs wells in the Marcellus Shale. The cation geothermometer could not be expected to predict accurate temperatures for the formation water because cation concentrations and ratios will be strongly influenced by the modern component of the mixed-water sample. The tritium content of water from well 803 can be interpreted as indicating that the 1975 sample contains about 20 per— cent hot and 80 percent cold water, and that the cold component contains 8 mg/L Si02. If this were the case, the undiluted hot component would have a silica con- tent of about 90 mg/L, corresponding to a chalcedony temperature of 105°C. These temperatures exceed the E34 concordant chalcedony and cation temperatures of sample 801 too much to be entirely credible, and it is thus likely that the high silica content of sample 803 is due to supersaturation with chalcedony at low temper- ature rather than to equilibrium at high temperature. Thus, in the Hot Springs, NC, area, reservoir temper- ature predicted by the chalcedony geothermometer is 77°C if the warm spring is a mixture of hot and cold waters, or about 50°C if the spring is conductively cooled. Cation geothermometer results suggest that the lower temperature is the more likely. CONCLUSIONS All evidence accumulated thus far indicates that the heat at the thermal springs investigated is derived from rocks at depth in a region of “normal” geothermal heat flow. The spring waters are mixtures of warm, deep ground water and shallow, cool ground water in variable proportions. In general, the warmest springs discharge the highest proportion of deep-circulating warm water. Many thermal springs are in valleys on the crests or flanks of anticlines where steeply dipping beds of sandstone or limestone intersect faults or lineaments. Some thermal springs are in areas of complex geology and are associated with faulted limestone or sandstone. Deep circulation of most of the ground water dis- charging in the thermal springs is in the general direc- tion of the strike of the rock strata, along open tension fractures, bedding planes, and faults. Some deep circu- lation possibly occurs also along fault planes crossing the strike of the rocks. Further work would be required to confirm hydrolo- gic interpretations at the sites studied. The best sites for further study might be Berkeley Springs, W. Va., and Hot Springs, Va., because of the availability of much hydrologic information for those areas. The in- terpretations made from studies of those sites probably would be applicable to many other thermal springs in the East. Further work would be required to determine the geothermal potential of a particular site. Probably the best sites to consider for this type of study would be Falling Spring or Bolar Spring, Va. or Hot Springs, N. C.; these Virginia springs have the highest energy out- put of the ones included in this investigation. Although the Hot Springs in North Carolina have a visibly small discharge, there are reliable reports of hot water dis- charging beneath the French Broad River near the Hot Springs, and in the alluvium near the mouth of Spring Creek. Thus, the area probably has a much higher en- ergy output than indicated for the spring in table 3. This study produced no evidence of very high reser- voir temperatures in the aquifers. Oxygen—isotope GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ratios of the waters fall within ranges expected for re- charge at the latitudes of the springs; no positive shift in 6180 is observed. Thus, temperatures at depth in the spring systems are almost certainly less than 150°C. Geothermometers cannot be expected to provide very accurate results in systems with low reservoir temper- atures. Observed high dissolved-silica levels of some cold-water samples (such as those from wells 302, 303, and 402, for example) which presumably have never been heated will obviously invalidate the estimates of temperature from the silica geothermometer and will also raise doubts about the indicated temperatures of nearby warm-spring waters. Nevertheless, a trend of increasing dissolved silica with increasing tempera- tures is observed in the Appalachian warm springs. Spring and hot-well samples at temperatures >25°C have indicated chalcedony temperatures within about 10°C of the observed values, so that equilibration of dissolved silica with this mineral is clearly an impor- tant reaction associated with heating of the water. Cation-geothermometer temperatures of the warmest spring waters and of two of the three sampled hot-well waters are also in fair agreement with observed tem- peratures. The concordance of observed chalcedony, and cation temperatures for these warmest waters in- dicates that the chemical processes underlying the suc- cess of the cation geothermometer in other systems are also operative in the Appalachian thermal springs, and further, that reservoir temperatures in these systems probably do not exceed the chalcedony values of 30°— 50°C. Refinements can be made in geothermometer esti— mates of hot reservoir temperatures if the hot and cold components of a mixed warm water can be distin— guished. Chemical analyses of cold-spring and well waters in the thermal spring areas proved inadequate for estimating the extent of mixing; the composition of the hot component could not be deduced from compo— sitions of an assumed cold recharge water and the sampled spring water. However, tritium mea- surements distinguish clearly between modern and relatively old waters. Tritium determinations of frac- tions of cold (young) and hot (old) water, together with estimates ofthe silica content of the cold component, do provide improved silica geothermometer estimates of reservoir temperatures in some of the mixed spring systems. Potassium and helium seem to be most significant among the qualitative geothermal indicators consid- ered. Relatively high potassium levels (which are partly responsible for high cation geothermometer temperatures) are found in the warmer springs and in two of the three hot wells. Dissolved helium is present, in amounts from traces to several tenths of a lug/L, in HYDROLOGY AND GEOCHEMISTRY OF THERMAL SPRINGS OF THE APPALACHIANS most of the warm springs issuing from carbonate rocks. Only one sample from the silica-rock warm springs contained a detectable amount of helium. Arsenic may also be a useful geothermal indicator for application in the Appalachian warm spring systems if analytical sensitivity for arsenic can be increased so that a few hundredths of a ,ug/L can be determined with confi— dence. Several observations regarding the geochemistry of the spring waters are apparent from the tabulated data. The tritium analyses show that the warm spring waters are older than those of nearby cold springs. Also, in several of the warm springs the values for 813C are substantially higher than those in nearby cold springs; the high values indicate a longer or slower travel path in the formation of the warm water. Carbonate saturation indices for the thermal waters are higher than those calculated for nearby cold springs in the same formation. Three of the silica-rock warm springs are considerably undersaturated in cal- cite; the carbonate thermal waters and Minnehaha Springs, on the other hand, are near or above calcite saturation. The technique of using analyzed carbon dioxide con- centrations rather than field pH in calculations of car— bonate equilibrium data helps considerably to reduce discrepancies between different samples from a spring of supposed constant composition and also to reduce or eliminate apparent carbonate supersaturation of a number of spring waters. The proposed method for de- termination of an adjusted 613C, using the gas analysis value for C02, also leads to more nearly constant com— positions for a number of the Appalachian springs. The resulting improvements in the geochemical data lend support to the idea that outgassing of carbon dioxide during the preparation of a 13C sample, or in the course of measuring pH, may lead to erroneous conclusions about the carbonate chemistry of a water sample. The use of analyzed carbon dioxide concentrations may be generally helpful in geochemical studies of ground waters. REFERENCES CITED American Assoc. Petroleum Geologists and US. Geological Survey, 1976, Geothermal gradient map of North America: US. Geol. Survey, 2 maps, 1:5,000,000. Birch, F., 1954, Heat from radioactivity, Ch. 5 in H. Faul, Ed., Nu- clear Geology: 414 p. and xvii, John Wiley & Sons, New York. Clark, W. E., Chisholm, J. L., and Frye, P. M., 1976, Water resources of the upper New River basin, West Virginia: West Virginia Geol. and Econ. Survey and West Virginia Dept. Nat. Resources, Div. Water Resources, River Basin Bull. 4. Claypole, E. W., 1885, Second Geological Survey of Pennsylvania, Report F2, a preliminary report on the paleontology of Perry County: Board of Commissioners for the Second Geological Sur- vey, Harrisburg, Pennsylvania, 437 p. Costain, J. K., 1976, Geology and geophysical study of the origin of E35 the Warm Springs in Bath County, Virginia: Virginia Polytechnic Inst. and State Univ., Dept. of Geol. Sci., 44 p. Dansgaard, W., 1964, Stable isotopes in precipitation: Tellus, v. 16, p. 436—468. Dennison, J. M. and Johnson, R. W., Jr., 1971, Tertiary intrusions and associated phenomena near the Thirty Eighth parallel frac- ture zone in Virginia and West Virginia: Geol. Soc. Am. Bu11., v. 82, pp. 501—508. Fournier, R. 0., 1973, Silica in thermal waters: laboratory and field investigations: Proc. of the International Symp. on Hydrogeo— chemistry and Biogeochemistry, Tokyo, 1970, v. l—Hydrogeochemistry, Clark, Washington, D. C., p. 122—129. Fournier, R. 0., and Truesdell, A. H., 1973, An empirical Na-K—Ca geothermometer for natural waters: Geochim. et Cosmochim. Acta, v. 37, p. 1255—1275. 1974, Geochemical indicators of subsurface temperature— Part 2, Estimation of temperature and fraction of hot water mixed with cold water: U.S. Geol. Survey Jour. Research, v. 2, no. 3, p. 263—270. Fournier, R. 0., White, D. E., and Truesdell, A. H., 1974, Geochemi- cal indicators of subsurface temperature—Part 1, Basic assump- tions: U.S. Geol. Survey Jour. Research, v. 2, no. 3, p. 259—262. Grimsley, G. P., 1916, Jefferson, Berkeley, and Morgan Counties: West Virginia Geol. and Econ. Survey County Geologic Rept. Series, 644 p., 3 pl. Heltz, G. R. and Sinex, S. A., 1974, Chemical equilibria in thermal spring waters of Virginia: Geochim. et Cosmochim. Acta, v. 38, pp. 1807—1820. Hewett, D. F. and Crickmay, G. W., 1937, The Warm Springs of Georgia, their geologic relations and origin: U.S. Geol. Survey Water-Supply Paper 819, 40 p. Hobba, W. A., Jr., Chemerys, J. C., Fisher, D. W., and Pearson, F. J., J r., 1977, Geochemical and hydrologic data for wells and springs in thermal-spring areas of the Appalachians: U.S. Geol. Survey Water-Resources Investigations Report 77—25, 36 p. Hobba, W. A., Jr., Friel, E. A., and Chisholm, J. L., 1972, The water resources of the Potomac River basin, West Virginia: West Virginia Geol. and Econ. Survey and the West Virginia Dept. Nat. Resources, Div. Water Resources, River Basin Bull. 3. Keith, Arthur, 1904, Description of the Asheville quadrangle (North Carolina-Tennessee): U.S. Geol. Survey Geol. Atlas, Folio 116. Kulander, B. R., and Dean, S., 1972, Gravity and structures across Browns Mountain, Wills Mountain, and Warm Springs anticlines—gravity study of the folded plateau, West Virginia, Virginia, and Maryland, in Lessing, Peter, ed., Appalachian structures: origin, evolution, and possible potential for new ex- ploration frontiers—a seminar: West Virginia University and West Virginia Geol. and Econ. Survey, p. 141—180. Langmuir, Donald, 1971, The geochemistry of some carbonate ground waters in central Pennsylvania: Geochim. et Cos- mochim. Acta, v. 34, p. 1023—1045. Li, Yuan-Hui, Takahashi, T. and Broecker, W. S., 1969, Degree of saturation of CaCO3 in the oceans: Jour. Geophys. Research, v. 74, no. 23, p. 5507—5525. Lohman, S. W., 1972, Ground-water hydraulics: U.S. Geol. Survey Prof. Paper 708, 70 p. Lowell, Robert F., 1975, Circulation in fractures, hot springs, and convective heat transport on mid-ocean ridge crests: Geophys. Jour. Research Astr. Soc. 40, p. 351-365. Mather, W. W., 1842—43, Geology of New York—Part 1, Geology of the first geological district: Albany, New York, Carroll and Cook. McCallie, S. W., 1908, A preliminary report on the underground waters of Georgia: Geol. Survey of Georgia, Bull. 15, 370 p. 1913, A preliminary report on the mineral springs of Georgia: E36 Geological Survey of Georgia Bull. 20, 190 p. Oriel, S. S., 1950, Geology and mineral resources of the Hot Springs Window, Madison County, North Carolina: North Carolina Dept. Conserv. Devel., Div. Mineral Res, Bull. 60, 70 p. Park, P. Kilho, 1969, Oceanic 002 system: An evaluation of ten methods of investigation: Limnol. and Oceanog, v. 14, no. 2, p. 179— 186. Pearson, F. J., Jr., Fisher, D. W., and Plummer, L. N., in preparation, Correction of ground-water chemistry and isotopic composition for the effects of C02 outgassing: Geochim. et Cosmochim. Acta. Pearson, F. J., Jr., and Rettman, R. L., 1976, Geochemical and isotopic analyses of waters associated with the Edwards Lime- stone Aquifer, Central Texas: Edwards Underground Water Dis- trict Report, San Antonio, Texas, 35 p. Plummer, L. Niel, Jones, Blair F., and Truesdell, Alfred H., 1976, WATEQF—A Fortran IV version of WATEQ, a computer pro- gram for calculating chemical equilibrium of natural waters: U.S. Geol. Survey Water—Resources Investigations Report 76—13, 61 p. Price, P. H., Hoskins, H. A., and McCue, J. B., 1936, Springs of West Virginia: West Virginia Geol. Survey, 146 p. Reeves, Frank, 1932, Thermal springs of Virginia: Virginia Geol. Survey, Bull. 36, 56 p. Roberts, A. A., Friedman, 1., Donovan, T. J., and Denton, E. H., 1975, Helium survey, a possible technique for locating geothermal reservoirs: Geophys. Research Letters, v. 2, no. 6, p. 209—210. Rogers, W. B., 1843, On the connection of thermal springs in Virginia with anticlinal axes and faults: Assoc. Am. Geologists and Naturalist, Report of lst, 2nd, and 3rd meetings 1840—1842, p. 323—347. 1884, A reprint of the annual reports and other papers on the geology of the Virginias, by the late William Borton Rogers: New York, D. Appleton and Co., 832 p., 8 pls, map, and 8 charts. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Siddiqui, S. H., and Parizek, R. R., 1972, Application ofnonparamet- ric statistical tests in hydrogeology: Ground Water, v. 10, no. 2, p. 26—31. Stewart, G. L., and Farnsworth, R. K., 1968, United States tritium rainout and its hydrologic implications: Water Resources Re- search, v. 4, p. 273—289. Stewart, G. L. and Wyerman, T. A., 1970, Tritium rainout in the United States during 1966, 1967, and 1968: Water Resources Research, v. 6, p. 77—87. Stose, G. W. and Stose, A. J., 1947, Origin of the hot springs at Hot Springs, N. G: Am. Jour. Sci., v. 245, p. 624—644. Stumm, Werner, and Morgan, James J., 1970, Aquatic Chemistry: New York, Wiley-Interscience, 583 p. Toth, J ., 1963, A theoretical analysis of groundwater flow in small drainage basins: Jour. of Geophys. Research, v. 68, no. 16, pp. 4795—4812. Trapp, Henry, J r., 1970, Geology and ground-water resources of the Asheville area, North Carolina: North Carolina Dept. of Water and Air Resources, Div. of Ground Water, Ground-Water Bull. 16. Truesdell, A. H., 1976, Geochemical techniques in exploration: Proc. 2nd UN Symposium on the development and use of geothermal resources, p. liii—lxxix. Weinman, B., 1976, Geophysical, geochemical, and remote sensing studies of Pennsylvania’s thermal springs: Pennsylvania State University, unpublished MS Thesis. Werner, E., and Medville, D., 1975, Relation of lineaments and hy- drology in a carbonate terrain of West Virginia: EOS, v. 56, no. 6, p. 358. White, D. E., Barnes, Ivan, and O’Neil, J. R., 1973, Thermal and mineral waters ofnon-meteoric origin, California Coast Ranges: Geol. Soc. America Bull., v. 84, p. 547—560. W. Va. Geol. and Econ. Survey, 1970, Oil and gas fields of West Virginia: 1:250,000 map. Chemical Studies of Selected Trace Elements in Hot-Spring Drainages of Yellowstone National Park By R. E. S'I‘AL’FFER, E. A. jENNE, and]. w. BALL GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1(l44—F UNITED STATES GOVERNMENT PRINTING OFFICE. WASHINGTON : 1980 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Stauffer, R. E. Chemical studies of selected trace elements in hot—spring drainages of Yellowstone National Park. (Geohydrology of geothermal systems) (Geological Survey professional paper ; 1044-13) Bibliography: p. FIB-F20. Supt. of Docs. no.2 I 19.1621044-17 1. Trace elements. 2. Geochemistry~Yellowstone National Park. 3. Hot springskYellowstone National Park. 4. Yellowstone National Park. I. Jenne, Everett A., 1930- joint author. I]. Ball, J. W., joint author. III. Title. IV. Series. V. Series: United States. Geological Survey. Professional paper 1044-F. QESI6.T85$76 551.2’3’0978752 79-607187 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, DC. 20402 Stock Number 02h-001-0329h-5 CONTENTS Page Page Abstract __________________________________________________ F1 Results—Continued Introduction ________________________________________________ 1 Statistical definitions __________________________________ F7 Acknowledgments ______________________________________ 2 Octopus and Azure Springs ______________________________ 7 Sample sources ____________________________________________ 2 Oxidation and diffusion processes ____________________ 7 Lower Geyser Basin _____________________________ ‘_ ’_ _____ 2 Precipitation reaction ______________________________ 10 Yellowstone Canyon ____________________________________ 2 Unnamed sulfide-bearing springs in Yellowstone Canyon__ 11 Madison River system __________________________________ 3 Madison River system __________________________________ 14 Sampling methods __________________________________________ 4 Sorption controls on arsenic in geothermal waters ________ 15 Analytical methods ________________________________________ 4 Arsenic flux of the Madison River ________________________ 17 Results ____________________________________________________ 5 Potential ecological significance of arsenic ________________ 17 Data reliability ________________________________________ 5 Conclusions ________________________________________________ 18 Pure analytical error ________________________________ 5 References cited ____________________________________________ 18 Sampling errors ____________________________________ 6 ILLUSTRATIONS Page FIGURE 1. Sketch map showing rivers and major hydrothermal areas sampled in Yellowstone National Park, Wyo ________________ F3 2. Schematic diagram of Madison River and Yellowstone Canyon sulfide-bearing mixed spring systems and sampling sites ________________________________________________________________________________________________ 4 3. Physical and chemical parameters of two hot springs and their drainages ____________________________________________ 7 4. Concentration factors of several parameters for two hot springs and their drainages __________________________________ 8 TABLES Page TABLE 1. Analytical precision of solute data ____________________________________________________ ,_ ........................... F5 2. Chemical and physical parameters of Octopus Spring and its drainage ________________________________________________ 6 3. Chemical and physical parameters of Azure Spring and its drainage ________________________________________________ 7 4. Selected physical and chemical properties of additional hot waters __________________________________________________ 11 5. Chloride, aluminum, iron, and manganese in selected geysers and hot springs ________________________________________ 11 6. Chemical and physical parameters of Calcite and two other unnamed sulfide-bearing springs in Yellowstone Canyon with the drainage of the mixed spring system __________________________________________________________________ 12 7. Estimated solute losses following mixing in Yellowstone Canyon mixed spring and Madison River systems ____________ 13 8. Chemical and physical parameters of the Madison River drainage ____________________________________________________ 15 III GEOHYDROLOGY OF GEOTI—IERMAL SYSTEMS CHEMICAL STUDIES OF SELECTED TRACE ELEMENTS IN HOT-SPRING DRAINAGES OF YELLOWSTONE NATIONAL PARK By R. E. STAUFFER,‘ E. A. jENNE, and j. W. BALL ABSTRACT Intensive chemical studies were made of S(—II), 02, Al, Fe, Mn, P, As(III), As(V), and Li in waters from two high-Cl, low Ca-Mg hot- spring drainages in the Lower Geyser Basin, a warm spring system rich in Ca and Mg in the Yellowstone Canyon area, and the Madison River system above Hebgen Lake. Analyses were also made of other representative thermal waters from the Park. Soluble Al concentrations were generally low (< 100 ug/L and fre- quently <50 pig/L) except in Azure Spring in the River Group of Lower Basin (~500 ug/L). Approximately 90 percent of the Al was found to precipitate in the upper drainage channel and nonflowing satellite pool of Azure Spring and was accompanied by an 8 percent (250 ug/L) loss of the Li flux. Solute Fe and Mn concentrations were typically very low (<10 ug/L) in the alkaline high-Cl thermal waters. Soluble reactive P and total P concentrations were low (<2 ug/L as P) in all of the high-Cl, HCO§-buffered hot springs. The Firehole, Gibbon, and Madison Riv- ers contained 4 to 6 ug/L. Total solute As acted conservatively in the alkaline drainages studied, including the Firehole-Madison River. However, the As(III)/As(V) ratio was strongly bimodally distributed (~50 or <0.1) according to whether dissolved S(—II) or 0.2 was dominant in the thermal water. Oxidation of As(III) proceeded rapidly at the elevated temperatures (30° to 90°C) in the drainages, but only following the oxidation of dissolved S(—II). In contrast to the Firehole River, ap- proximately 60 percent of the thermally related As flux in the Gib- bon River drainage basin is precipitated prior to the Gibbon‘s conflu— ence with the Firehole at Madison Junction. The As/P atomic ratio is typically ~500 for the alkaline hot-spring waters, and ~15 during base-flow conditions on the Madison River. The differential toxicities of the As species, the variable As(III)/As(V) ratios, and the very large As/P ratios all suggest that As may be ecologically important in the Park’s thermal waters. INTRODUCTION The trace-element chemistry of hot springs and geothermal drill-hole waters has attracted interest in the past because of the hypothesized relationships be- tween hydrothermal solutions and base metal ore genesis (Barnes, 1967). Trace-element chemistry is also valuable in assessing both subsurface physical- chemical conditions and potential environmental im- pacts of geothermal commercial exploitation. The principal solutes present in hot-water- dominated geothermal fluids, SiOz, Na, K, Li, Ca, Mg, Cl, H003, 804, F, and B, have been investigated in diverse geothermal settings throughout the world (El- lis, 1970). The concentrations of the prominent noncon- servative solute species (Si02, Ca, Mg, F, K) have been rationalized using mineral equilibria models (Ellis, 1967, 1970, 1973). The rare alkalis, Cs and Li, and the metalloid, As, are regarded by Ellis (1970) as weakly conservative because they are sometimes partially pre- cipitated in the epithermal zone. The elements Cl, Br, and B are considered highly conservative, once dis- solved in the hot circulating water. Ellis (1969, 1973) interprets the low levels of Fe, Mn and Al in HCOa-buffered hot waters as convincing evi- dence that solubility controls are acting on these geochemically abundant elements Sulfides of Fe, Cu, Pb, and Zn are being deposited at depths of 250 to 800 m at both Waiotapu and Broadlands, New Zealand (Browne, 1969), from hot water containing up to 1.7, 10, and 2.2 ug/L (micrograms per liter) of the latter three elements, respectively (Ritchie, 1973). Heavy- metal mineralization at depth is largely absent at Wairakei, although some pyrrhotite (FeS) is found. The abundant mineralization at Broadlands is attrib- uted to the high concentration (120 mg/L) of solute sulfide species, S(-II), in the effluent. At Wairakei, S(——II) is 12 mg/L, or about one order of magnitude larger than most of the values reported for Yellowstone (Thompson and others, 1975) or Steamboat Springs, Nevada (White, 1967). Spectrographic analyses of evaporated residues of typical alkaline Yellowstone thermal waters (Rowe and others, 1973) indicate levels of transition series elements comparable to New Zea- land thermal waters (Ellis, 1969). Greater than milligram-per-liter concentrations of Fe, Mn and Al commonly occur in “acid-sulfate-type” geothermal waters as a result of attack on the country 'Present address: Water Chemistry Laboratory, University of Wisconsin, Madison. Wis- consin 53706. F1 F2 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS rock by sulfuric acid (White and others, 1971). The sul- furic acid is thought to be derived from oxidation of S(—II) in the near-suface zone. White (1967) noted that the semimetallic elements As, Sb and Hg are especially mobile in geothermal sys- tems and are present at relatively high concentrations. Sb and Hg sulfides are epithermally'deposited along with sinter at Steamboat Springs (White, 1967) and at Broadlands (Weissberg, 1969). Neither of the two prominent arsenic sulfides (orpiment, A5283; realgar, ASS) was observed as a discrete mineral, either in the sinter or at depth in drill cuttings. However, As co- precipitated with stibnite (Sngg) as ~5 percent impur- ity (Weissberg, 1969). Weissberg noted that Tl, Ag, Au, and Hg, in addition to As, were highly enriched in stibnite in the near surface at the Broadlands. Arsen- opyrite in minor amounts has been ejected from a Wairakei drillhole. Weissberg (1969) estimated that less than 10‘4 percent of the As flux from a Broadlands drill-hole discharge was precipitated with the amor- phous silica, in contrast to 0.5 to 5 percent of the Sb. Research on the thermally related fluxes of Hg and As into the Waikato River (Axtmann, 1975) was motivated by geothermal power development in the Taupo Basin, New Zealand. Elevated levels of both As and Hg were found in the sediments of Lake Aratiatia below the discharge of theWairakei effluents. The ma- crophytic dominant in Lake Aratiatia, Lagarosipon major, featured an As concentration factor (over the Waikato River concentration of 39 ug/L) of 5,300 (Reay, 1973). However, because of the very large As flux in the Waikato (estimated at 158 metric tons/yr, Axtmann, 1974), less than 4 percent of the As entering the river system is removed by vegetation before dis- charge into the sea. The attenuation of geothermally related trace ele- ments in hot-spring drainages and receiving waters has apparently not been previously studied in the United States. Boylen and Brock (1973) and Zeikus and Brock (1972) reported chemical analyses of the Firehole River (Yellowstone Park, Wyo.) at stations selected to show qualitative influences of hot-spring influents on the river biota. Among the elements which can properly be thought of as geothermal trace ele- ments, only P04 data were reported and were seriously biased by an As interference (Stauffer, 1980a, b). The present paper reports chemical studies of Oz, S(—II), Al, Fe, Mn, As, P, and Li in two hot-spring drainages in the Lower Geyser Basin, two warm-spring drainages in the Yellowstone Canyon, and the Madison River system above Hebgen Dam. The differential ef- fects on the solutes of temperature, pH, and redox changes accompanying adiabatic cooling and source water mixing are examined using the conservation of mass principle and employing statistical contrast with Cl as the prominent conservative geothermal tracer. Our use of ratio estimation involving C1 as a tracer follows the early initiatives of Ellis and Wilson (1960). The intensively studied thermal drainageways are also compared chemically with representative springs sam- pled in the Upper, Norris, and Mammoth thermal basins. ACKNOWLEDGMENTS The cooperation of the US. National Park Service in facilitating the sampling program for this study is gratefully acknowledged. The authors thank A. H. Truesdell for his help in collecting a set of samples .in 1973 and for many fruitful discussions with him; and J. M. Burchard and DR. Nordstrom for assistance in sample collection and onsite'analyses. Fluoride data were provided by D. K. Nordstrom. The authors ac- knowledge J. M. Thompson for analytical assistance and for his generous contribution of river samples for As comparisons, and John Hem and R0. Fournier for helpful discussions. SAMPLE SOURCES LOWER GEYSER BASIN The most intensive geochemical studies on hot- spring drainages were on 'Octopus Spring (Marler, 1973, p. 435), in the lower White Creek area, and Azure Spring (Marler, 1973, p. 574), in the River Group (Fig. 1). Octopus Spring, with its long (~50 mi) distinct drain- age channel, relatively constant flow regime, easy ac- cessibility, and luxuriant microbiological flora, is one of several sites in the lower White Creek area where the ecology of thermophilic blue-green algae and asso- ciated bacteria has been intensively studied by T. D. Brock and coworkers (Brock, 1967, 1969; Brock and Brock, 1966, 1967, 1968a, b, 1969a, b; Walter and others, 1972). Azure Spring has a well-developed drainage channel ~100 m in length discharging westward into the Firehole River. The hotter satellite pool on the south- west edge of Azure Spring’s main pool was also sam- pled. Azure is one of a large number of springs in the River Group on which major chemical constituents have been determined (Thompson and others, 1975). Sampling stations along the two drainageways were selected to correspond to ~10°C intervals in water temperature. YELLOWSTONE CANYON An unnamed sulfide-bearing spring was studied, which is 20 to ‘30 m vertically above the west bank of TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK 111°OO' F3 110°50' : l Cable Car n 48 40 .5;— J’ADISON 6 West Entrance Yellowstone National Park 48°30' YELLOWSTONE NATI 0 NA PARK WYOMING Continental Divide PA RK COU NTY TETON COUNTY GIBBO RIVER Gibbon Falls 2 01 Firehole Falls 0° .' ,, o AzureSprmg, °, °°">’ Basin _" .. . River Group \In J17 , '0 C 0‘0 Octopus Spring ’17 ' e Creek l SCALE 6 8 KILOMETERS 2 4 MILES FIGURE 1.—Sketch map showing rivers and major hydrothermal areas sampled in Yellowstone National Park, Wyo. the Yellowstone River, about 0.5 km south of Tower Falls. About 5 m below its source in a hillside fissure, the spring water (46°C) mixes with water from a cooler (14°C) spring (fig. 2). Below the confluence the mixed water runs steeply downslope until it encounters a rel- atively flat, poorly drained area, 5 to 10 In above the Yellowstone River. Samples were taken in both tribu- tary spring drainages about 1 m above the confluence, immediately below the confluence, at the bottom of a small riffle, at the base of the steeply dropping com- bined'drainage section, and at a point intermediate between the confluence and the base. Below station 5 the drainage into the Yellowstone River was indistinct. MADISON RIVER SYSTEM Stations 1 and 2 are on the Firehole and Gibbon Riv- ers, respectively, each about 1 km above their conflu- ence forming the Madison River. Station 2 is F4 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS downstream from the entry of Terrace Spring drain- age. Stations 3 and 4 are located on the Madison River, about 1.8 and 6.0 km, respectively, by road below the confluence (fig. 2). Station 5 is at the West Yellowstone Madison River Gaging Station. Station 6 is at the spillway of Hebgen Dam. All of the spring and river samples used in the drainage studies were collected during daylight hours, September 18 to 27, 1974. The entire period was pre- cipitation free on the Yellowstone plateau. Sampling sites (numbers) Distances (horizontal) Yellowstone Canyon 1. 40°C source water a. ~l m 2. 14°C source water b. ~1 m 3. Station below confluence c. ~0.5 m 4. Down drainage station d. ~10 m 5. At base of steeply sloping drainage e. ~15 m Madison River 1. Firehole River a. 1.0 km 2. Gibbon River b. 1.0 km 3. Madison below confluence c. 1.8 km 4. Madison Canyon d. 6.0 km 5. Madison at West Yellowstone e. 12.0 km gaging station (inactive) 6. Spillway of Hebgen Lake 1'. ~30 km FIGURE 2.—Schematic diagram of Madison River and Yel- lowstone Canyon sulfide-bearing mixed spring systems and sampling sites. SAMPLING METHODS Water samples from hot springs and their drainages were obtained and processed as described by Ball, Jenne, and Burchard (1976). Briefly, the samples were obtained with a Masterflex2 portable electric pump equipped with silicone tubing. Filtered samples were obtained by attaching a barrel-less pressure filter as- sembly (142 mm diameter, 0.1 p.m pore size) to the silicone outflow tube of the pump (Kennedy and others, 1976). This arrangement allowed sample collection from boiling hot springs without air contacting the sample, and with a maximum temperature drop of 2°C. 1Mention of brand names is for the convenience of the reader and does not constitute endorsement by the US. Geological Survey. A segmented aluminum pole was used to suspend the inlet tube over, and into, the large hot springs. Sam- ples for S(-II) were fixed in the field by adding zinc acetate (50 mg/250 mL of sample) immediately after filling the polyethylene bottle. A 1 L trace-metal sam- ple was collected and acidified to a pH of 1.0 to 1.3 using concentrated redistilled HN03. Another sub- sample was acidified with concentrated HCl (1 mL/250 mL sample) for determination of As (III), As (V), P, and some major cations. Filtered-unacidified samples were also used for major ion analyses. Most samples were acidified in the mobile laboratory within an hour of collection; samples from the Yel- lowstone Canyon area, Mammoth hot springs, and Hebgen Dam were acidifed within 10 hr of collection. No HCl-acidified samples were obtained from the latter areas; hence filtered HNO3-acidified samples were used for the As(III), As(V), and P analyses. ANALYTICAL METHODS Water temperature, pH, Eh, and dissolved 02 were determined onsite. Spring temperatures were meas- ured with a calibrated mercury thermometer on sam- ples rapidly withdrawn in a Thermos-type bottle clamped to the sampling pole. Temperatures in the dis- charge channels were measured by direct immersion of the thermometer. Eh and pH were measured onsite using a Sargent Model PBX specific ion meter and a flow-through cell (Ball and others, 1976). Dissolved oxygen was measured using the azide modification of the Winkler procedure (American Public Health Asso— ciation, 1971); the fixed samples were titrated within 3 hr of collection. Total alkalinity and F were determined in a mobile field laboratory on the day of sample collection. Alka— linity was determined by titration with standard H2S04 to a pH 4.5 end point. Both uncomplexed and total F activities were measured using an Orion specific ion electrode and procedures specified by the manufacturer. Li, Na, K, Mg, and Ca were determined by flame atomic absorption using a Perkin Elmer Model 303 or 306 spectrophotometer, following procedures issued by the manufacturer. Si02 was analyzed by direct current plasma spectroscopy using a Spectrospan III operating in single element mode. Cl and 804 were determined by cation exchange using a refinement (StaufTer, 1980b) of the method of Mackereth (1963). The refinement was necessary to cope with the high F levels in Yellowstone Park ther- mal waters. P, As(III), and As(V) were determined fol- lowing an adaptation (Stauffer, 1980b) of the molybdate procedure of Johnson and Pilson (1972a). TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK F5 Fe, Mn, and Al were determined using flame atomic absorption spectrometry, following 8-quinolinol MIBK extractions. Al was also extracted from filtered, un- acidified, cooled water samples on the day of sampling following the procedure of Kennedy, Zellweger, and Jones (1974). Field-extracted Al (back extracted from MIBK into 1.0 molar HN03) was subsequently analyzed by flame atomic absorption. Total dissolved Sb was analyzed using a modification (Stauffer, 1977) of the Yanagisawa, Takeuchi, and Suzuku (1973) atomic absorption procedure. Total dissolved S(—II) was determined on fixed sam- ples using the potentiometric procedure of Baumann (1974). RESULTS DATA RELIABILITY PU RE ANALYTICAL ERROR Identification of chemical processes acting in the hot-spring drainages depends in part on the magnitude of the analytical errors. Estimates of the analytical coefficients of variation (c.v.) for pertinent solute levels and the most important solute species are contained in table 1. The errors (table 1) apply to individual var- iates; the variance of the mean X of n independent identically distributed (statistically) X values is given by equation 1 (Hogg and Craig, 1970): VarX = Var Km. (1) The mean square errors (squared bias plus variance) for K and Li are the smallest. The low levels of Ca and particularly Mg in the “typical” alkaline Upper or Lower Basin hot springs imply large analytical coeffi- cients of variation. In general, the low levels of 804 and high concen- trations of F in the Upper and Lower Basins result (table 2) in significant potential biases in 804 esti- mates obtained using the cation-exhange method. The problem is ameliorated considerably in the river wa- ters, and in hot-spring waters from north of Madison Junction, because of large associated increases in the SOi/F molar ratio (Rowe and others, 1973; Thompson and others, 1975). Because of the dominance of Cl in most of the waters being considered here, the Cl data has a c.v. of 1 percent, about equal to that for data published elsewhere on these waters (Rowe and others, 1973; Thompson and others, 1975). Errors associated with the As and P species are dis- cussed in detail elsewhere (Stauffer, 1980a, b). The P levels in the typical alkaline hot-spring waters are TABLE 1.-——Analytical precision of solute data Coefficient Analytical Level of variation“ Solute method‘ (mg/L) (c.v.) (percent) Remarks Li __________ Flame AA .......... 0.50 0.3—0.5 C.v. insensitive to Li ____________ do ________________ 5.00 .& .5 level. Na ________ Flame AA, 100 .5 C.v. becomes progres- 330 nm line sively worse for levels Na rrrrrrrrrr do ________________ 500 .2— .3 under 100 mg/L .......... Flame AA__________ 10.0 .2— .3 Probably most sensi- tive method. .1 3.6 1.0 2.0 1.0 2.5 Sensitive to SO. and do 5.0 1.0 O; interferences. Cl __________ Conductivity, 50 1.0 C.v. increases approxi- cation exchange. mately as the term 1 + (SO. + F)/Cl] when F, 80., and Cl are xressed in milliequiv- aents. SO. __________ do ____________________________ 1.0 Cva increases with non- zero F, SiO, levels. F __________ Specific ion 30 2.0 electrode. Alka- Titration __________ 200 1.0 Noncarbonate alkalinity linity may be 225 percent of total. As ________ Spectre uhommetnc .20 2.0 C.v. increases with de- Mo— creasin level, sensi- As __________ do ________________ 2.00 .3 tive to 4 at low level. Al ________ Flame AA, solvent .10 8.3 extracted in field. Al __________ do ________________ .025 12.0 Al ________ extracted in lab ._ .200 7.0 Sb ________ Flame AA, lab .25 4.0 Recovery dependent upon solvent extracted. volume solvent extrac- Sb __________ do ________________ .5 2.0 ted. Mn ________ Flame AA, lab .001 5 solvent extracted. Mn .......... do ________________ .010 2 Fe ........ Flame AA, lab .003 1.5 solvent extracted. Fe .......... do ................ .020 2.5 02(aq) ...... Winkler ____________ 1.0 5.0 S" ........ Specific Ion 8.0 1.0 electrode. 8’2 __________ do ________________ 3.0 3.0 *AA= Atomic absorption. “Coefficient of variation (0. v.) = 100 s,\/ at when 3, is the estimated standard pure analytical error for an individual replicate near the detection limit of ~2 ug/L. At these low levels the P estimates are also positively biased by the very large concentrations of Si02 invariably present in geothermal waters. The relative root-mean-square errors and c.v. for total As are among the smallest for any of the solutes analyzed in these hot-spring waters. Both the accuracy and precision of the total As compo- nents, As(III) and As(V), suffer, however, from uncer- tainties in the preservation of As(III) at the time of sampling of the boiling hot-spring waters, as well as methodological constraints associated with molybdenum-blue (Stauffer, 1980a, b). The precision of the total F data decreases at the low concentrations found in high-Ca waters in Yellowstone Canyon. The c.v. for S(—II) is likely to be small compared with ZnS oxidation biases which may have resulted from 3 months of storage of the samples prior to analysis. Pos- sible oxidation effects (resulting in negative biases) were not evaluated. However, Brock, Brock, Bott, and Edwards (1971), citing thesis work of Pachmayr, state that the ZnS precipitates are stable against oxidation for weeks. The sulfide data reported here are likely to be minimum values. F6 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 2.—Chemical and physicial parameters of Octopus Spring and its drainage ['l‘emperature, Eh, pH, and 02 are field determinations; other analyses performed after return to Menlo Park Laboratory, except as noted] Concentration Parameters‘ Station No. of spring and its drainage Difference2 Ratio8 1‘ 2 3 4 5 6 7* (7-1) (7/1) Sample 74WA __________________________________________________ 112 114 115 116 117 118 113 .-.. .... Temperature (°C)‘ .. ______________ 82—85 63—75 53—61 50 40 30 23—26 A59 5n.a. p ____________________________________ 7.78 .--. 8. 8.6 8.8 9.0 8165 +87 n.a. Eh (mV) ___________________________ +18 +242 +288 +305 +289 +332 +314 n.a. 02 (a ) (mg/L) __________________________ . 3.93 5. 5.90 6.65 6. +59 n.a. Alkalinity (mg/L) .-.- __________________ 340 . ._ _-_ .._ ._ ... 367 +27 1.08 Na (m /L) ______________________ 307 317 322 326 336 335 332 +25 1.08 K (mg ) ... 15.2 15.4 15.5 16.1 16.5 16.4 16.4 +1.25 1.08 Li (mg/L) . 3.42 3.53 3.57 3.64 3.72 3.75 3.72 +.30 1.09 Mg (mg/L) . .009 .018 .007 .008 .010 .009 .010 .--. .-.. Ca (mg/L) . .54 1.27 .64 64 .79 59 .73 +.19 1.35 SO. (m 21 _ -_ ... ____ _. __ 230 +2 1.14 Cl (mg ) _____ 262 269 269 262 275 286 284 +22 1.08 F (mg/L) _______ 22. . __ __ ... _- _- 25.0 +3.0 1.14 SiOz (mg/L) ....... 270 256 257 258 261 216 219 —51 .815 :(li‘f/L) ........... <15 .... .-- ..- ...- ...- <15 ...- na Al( )(ug/L) 41 .... .... .... ._ ... 47 +6 n.a. Al (Ff) (Mg/L) 47 . .. ._ .. ... 16 —31 n.a. A1 (U1) (pg/L) _ 51 54 64 26 9 33 47 —4 n.a. Fe (Fl) (pg/L) _ 2.8 .... .... .... .- ... 3 +.2 n.a. Fe (Ul) (pg/L) . 1.4 .... . .. ... .... ._ ..- .... n.a. Mn (Fl) (pg/L) . ...... 3.0 . .. . .. .... . .7 72.3 n.a. Mn (Ul) (Mi/L) ... ........................ 3.4 3.7 3.4 3.1 5.0 1.1 .8 —2.6 n.a. As (V) (pg ) ...................................... 1,380 ... .... ... .... .... 1,610 +230 us As (V) + AS (111) (pg/L) ............................... 1,455 1,525 1,510 1,545 1,575 1,630 1,620 +160 1111 Sb (pg/L) ......................... 58 ...- .... ._ . .... ...- ... n.3, n.a. P‘(p.g/L) __.__.-.' .............................................. 2 _ . .. . . .9 v1.1 n.a. *Filtered samples, except as noted. Stations 2—6 were unfiltered. 'Ff = filtered and field extracted; Fl = filtered and lab extracted; U] = unfiltered and lab extracted. 2Concentration at station 7 minus that at station 1. “Concentration ratio of station 7 to station 1. ‘Temperature variations in response to discharge periodicity was measured at selected stations only. 5Not applicable. “Solub e reactive phosphorus as P (Strickland and Parsons, 1968;. The alkalinity data were adjusted for partial proto- nation of F (Ellis, 1963) and incomplete protonation of HC03 at the pH = 4.5 alkalinity end point. Corrections for the protonation of dissolved borates and silicates were not made. However, the protonation of dissolved silicates may introduce an error in some instances (D. V. Vivit and others, unpub. data, 1978). Among the solvent-extracted elements quantified by atomic absorption (Al, Mn, Fe, and Sb), Al and Sb suf- fer from the largest analytical variances. Low levels of A] tax the sensitivity constraints on Al analysis by flame; flameless atomic absorption for Al is inherently imprecise. Furthermore, the field extractions of Al have error components resulting fom the following fac- tors: (1) Field constraints prevented extracting the samples immediately after sampling; in fact, the time lag varied from 1 to 24 hr; (2) the solution temperature during the extraction procedure varied from about 6° to 50°C, with possible effects on solvent-extraction effi- ciency. The Sb procedure is influenced by SiOz levels; the high and variable levels of SiOz in the hot-spring waters and the very complex polymerization chemistry of SiO2 contribute the major analytical component of variance to the Sb analyses (Stauffer, 1977). SAMPLING ERRORS Although physical-chemical conditions in the springs are relatively invariant, the estimated concen- trations of solutes at fixed points along the down drainages vary (time scale = minutes) with spring dis- charge. The channel-water temperatures are positively correlated with discharge, and evaporative cooling re- sults in a concentration of solutes in the remaining liquid phase. From the Bowen equation (Bowen, 1926) it can be shown that 95 percent of the heat dissipation of 85°C (spring orifice temperatures) and 88 percent of the heat dissipation at 50°C is evaporative (latent heat transfer). A latent heat transfer model implies a 0.18 to 0.19 percent increase in conservative solute concentra- tion for each 1°C of evaporative cooling. The sun’s radiant heat input may approximate 15 cal/g water. A purely latent heat transfer model approximates total water losses through evaporation in the channel. Thus, the small amount of sensible heat exchange is approx- imately balanced by dissipation of the added radiant energy. Drainage stations could not be synoptically sampled; thus variable spring discharge introduces a within- station and a between-station component of sampling variance to the estimated solute concentrations. Based on the coefficient 0.19, and the observed short-term temperature variations in the Lower Basin drainages studied (tables 2, 3), the “sampling” component (c.v.) of solute concentration errors is ~1.0 percent, a value larger than the analytical “pure error” for the major alkali ions but $ analytical c.v.’s for the major anions (table 1). The analytical and sampling errors should be TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK TABLE 3.—Chemical and physical parameters of Azure Spring and its drainage [Temperature Eh, pH, and 02 are field determinations; other analyses performed after return to Menlo Park laboratory, except as noted] F7 Concentration Station No. of spring and its drainage Difl‘erenoe‘ Ratio’ Parameter' 1 W 1* 2 3" 4 5“ (5— 1) 5/1 1/1I 1 74WA _ 136 132 135 133 137 134 imgeiature cc 39 86 64—70 45—63 4e50 3341 — 49 1.09 ‘n.a. H _________ 7.49 8.62 9.18 9.25 9.20 9.42 +30 n.a. ' n.a Eh (mv) _____ -76 —185 —36 +75 +351 —342 +527 n.a. n.a )(mg/L) ,_,, _________________ .0 . .38 1.78 .4.78 +4.78 n.a. n.a gllfa‘linity (mg/L) ,,,,,,,, 198 _ __ 204 ,__ 209 +11 1.06 -_._ Nak (m /L) ________________ 322 328 332 337 353 +31 1.10 1.04 K (mg 1 ,,,,,,,, 13.3 14.5 15.0 14.6 14.6 +.30 1.06 1.03 1.1 (mg/L) ,,,,,,,, 3.23 3.10 3.16 3.16 3.30 +07 1.02 1.10 Mg (mg/L). ........ .023 .054 .013 .027 .075 +05 2.68 .354 C8 (mb/L) __. ........ 1.20 1.14 S1,21 1.49 5{1,25 $.05 1% 1371 ,,,,,,,, 52 ,.__ 4 H + . . 59211?" fl“). ........ 316 342 321 352 +36 1.114 1.02 3 (mg 7L) ............. 29.3 ...5 2315.0 350 233.5 +537 1.357951 1.04 SiO (mg/L) _ ........ 240 2 0 — . .___ 2 /L) "e ,,,,,,,, 1,500 , ._ 60 A, 20 —1,480 n.a. ____ A: (ED (11%) _ ........ 43(5) as 52 36 —429 n.a. 69.4 A ( ( ) ........ , _.__ e..- _,_ n.a. e-" A1(Ul)(litgg/L) ........ 595 45 39 . __ 39 —556 11.3. Fe Fgll) (fig/ILL) ........ 63.3 12.9 8.4 +2.1 n.a. 17.2 ) ) ........ _ H __ __ fl" __ n.a. "fl Men (Fl) (ti/L) _. ........ .61 W, 2.01 2.02 +1. 4 31.8. 7.69 Mn (Ul) (M /L) , ________ 4.3 2.7 2.6 2.5 2. 6 -1. 7 n.a. “.1 As (V) (1.1g ................ 30 . 27 __ 810 +780 n.a. _.__ As (V) + A3 (111) (pg/L) ........ 1,430 1,505 1,520 1540 1 615 +135 1.09 Sb(p.g/L) .................... 81 "1 ___ .V. 1.1. _.__ n.a. ____ P‘mg/L) ...................................................... 3 5 ._.- "H 3 ____ n.a. A-.. *Filtered samples, except as noted. Stations 2 and 4 were unfiltered. ‘Ff = Filtered and field extracted; Fl — filtered and lab extracted; U1: unfiltered and lab extracted. 2Concentration at station 5 minus that at station 1 “Concentration ratio of station 5 to station 1 l to 1’ ”I‘emperature variation in response to discharge periodicity was measured at selected stations only. 5Not a plicable. 6Solublle reactive phosphorus as P (Strickland and Parsons, 1968.) . . . . . 1" 9° 7 1 1 1 1 1 T 500 — Independent (hence addltive) 1n reckonlng the total .5, f A. octopus spams uncertainty in solute concentration at any point along 5 8°- .1 ‘400 -100 . Lu I.” the drainage channel. 0 5 g 3 70 - «300 .- 1y 1:: VI 0 . 5 li' '_'1 1— STATISTICAL DEFINITIONS 9.0-3 60— g -200 g 3 z E 3’ g ' _ 3.5-1-‘50-3 -1ooE <5og LetR(S),-,j be the who R of concentratlon for solute S g :1 3 11‘ . . . . . 3 E ‘ .— at statlons l and J, respectlvely, along a drainage chan- 5 8.0-1; 4o— ; - 0 fi 3.. . . . m . < nel. Let R,(S) be the who where the base statlonJ = 1. 7 5 :1 30 g m . . ' '2 ' ‘— For mlxed water systems, let P,(S) denote the fraction :2 of water at station 1' which is contributed by source 1 7-0— 20— —1oo 4 a calculated (equation 2) using the conservation of mass DRA'NAGE STAT'ON NUMBER principle for solute S (equation 3): 90_ 7 _ T . 1 ‘ 1 50° _ F3 \“~~ B. AZURE SPRING Pi(S) = (Ci—C2” (Cl—Cg) (2) E 30— cc 6— «400 -100 t“, E E" 1- ‘ Z P * Cl + (1—p)* Cg 2 Ci, (3) “Pg 70’ g5" DH —3003 fig 1: a. .1 mo . - 0 where C denotes the concentratlon of solute S, and p 1S 9.0— g 60— g4- p“ o -200; 3g . . . Q j; the m1x1ng fraction. 85 E 50 g 3 ‘Q r g < - ' ‘ 3 — / 1;. -1oo ~50 13 I LLI _1 / 1:. Z ”J z a m — 5/ 3 -_ e — 8.0—240— :2- 3/ 9‘0 5 -0 i 5: OCTOPUS AND AZURE SPRINGS g - ,1 $6 3 5 m o ,I 1— c1: . - °' 30— D 1 \ -. Z OXIDATION AND DIFFUSION PROCESSES 75 g ‘9: °° ‘°° 8E 70 20 0M As(V) Sulfide J0 E5 . . . . . — _ , .200 Several ox1dat1ve and d1ffus10n-controlled processes 1 1 2 2 4 s U’ . . . . DRAINAGE STATION NUMBER have been 1dent1fied in these two Lower Geyser Basm hot-s rin draina es fi s. 3 and 4 . Ox en concentra- . . _ . p d iih . g ( gth t t. ) YE b f FIGURE 3.—Phy51cal and chemlcal parameters of two hot springs tlon an . 1_ncrease W1 S a Ion nun} er ecause. 0 and their drainages. (Vertical bars on curves represent the OXYgen d1ffus10n from the atmOSphere Into the COOhng range in observed values. Sulfide was below detection in Oc- turbulent flow; pH increases 1 to 2 unlts, presumably topus Spring.) F8 1.110 1.100 1.090 1.080 1.070 1.060 1.050 CONCENTRATION FACTOR 1.040 1.030 1.020 1.010 1 .000 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS I I I I I A. OCTOPUS SPRING El «°° I /o\ ‘ \ I // \.‘ \\\ //¢’\ \5 A \WQ . I II/ \\\‘ K \ ‘ /%o D \I ______ WI 7/ “’ c /’ > / 0/, / o / 70 T l I 7/ I r I // / — [so ’// [I D / *I’ \;\/' II _ V3,) / / I I I I _ OALKALINITY _ E1 ARSENIC TOTAL I l I I I 2 3 4 5 6 7 STATION NUMBER ALONG DISCHARGE CHANNEL FIGURE 4,—Concentration factors of several parameters for two hot springs and their drainages. (The vertical bars on the T curves represent limits on the concentration factors which are attributed to the observed temperature range. The concentration factors for the satellite pool in figure a; are inverted to economize on graphical representation). because of C02 losses accompanying evaporative cool- fluctuations because gaseous diffusion is sensitively re- ing. Analytical dissolved 02 concentrations in the lated to water temperatures, turbulence, and residence upper channels are particularly sensitive to discharge time in the channel. TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK F9 I | I I B. AZURE SPRING 1.110 1.100 1.090 1.080 1.070 1.060 1.050 1 .040 1.030 CONCENTRATION FACTOR 1.020 1.010 1.000 0.990 0.980 0.970 U ARSENIC TOTAL (As) 0.960 — 0.950 I I I 1 1' 1 2 3 4 5 STATION NUMBER ALONG DISCHARGE CHANNEL FIGURE 4.—Continued. If dissolved O2 is detectable by Winkler titration, implies near-zero dissolved 02 levels, negative Eh val- S(—II) is below detection, Eh is positive, and the ues, and large As(III)/As(V) ratios (in some cases >50). As(III)/As(V) ratio is much less than 1 (figs. 3, 4). Con- Arsenite (As(III) ) is not oxidized in significant quan- versely, the presence of analytically detectable S(—II) tities until after the S(—II) has been either volatilized F10 or oxidized. The S(—II)-dissolved O2 relationship (fig. 3) suggests that at 90°C the rate of diffusion of molecular 02 into the hot pools is the rate-controlling step in the oxidation of S(—II). Traces of dissolved S(—II) some- times accompany nonzero oxygen levels in cooler down-drainage water (tables 2, 3); this thermodynamic disequilibrium probably reflects delayed oxidation at the lower temperature. The abiotic oxidation of S(—II) is known to be approximately first order with respect to both S(—II) and 0.2, to be relatively insensitive to pH, and to be relatively slow compared to the expected hy- draulic residence times (minutes) in the lower drain- age channels (O’Brien and Birkner, 1977). Brock, Brock, Bott, and Edwards (1971) noted that sulfur bacteria are active in oxidizing S(—II) (using molecular 02) even at boiling temperatures (approximately 92°C on the Yellowstone plateau) in the alkaline hot springs. Some HZS volatilizes from the channels, yield- ing the characteristic odor. The first dissociation con- stant of H28 is near 7.0. Thus, HZS volatilization is decreased at the pH values near 8 observed in the lower channel reaches. PRECIPITATION REACTION The hypothesized evaporation-solute concentration model (T curves in fig. 4) adequately describes changes in F, CI, As, and Na in the two drainage channels (fig. 4). Statistical inferences based on the R (S ) ratios lead to acceptance of the null hypothesis (conservative be- havior) for Na, K, Li, F, alkalinity, and total As in the downdrainage of Octopus Spring. Because of the very high concentrations of Na in Azure Spring (13.5 mil- limolar) relative to other constituents, the anticipated low reactivity of Na in this situation, and the superior analytical precision for Na as compared to CI in the high-804, high-F water, Na is the element of choice for evaluating the conservatism of other solutes in the Azure Spring system. There is no convincing statistical evidence of As precipitation in either the satellite pool or along the down drainage of Azure Spring. Applying the statistical properties of the solute ratio statistics, with probability ~95 percent, the losses of As along the drainageway are less than 1.8 percent of the spring flux. The Azure Spring K ratios increase prior to station 4, as expected of a conservative solute, and then decrease significantly in the latter part of the channel. Approx- imately 95-percent confidence limits on the K losses prior to station 5 are 35:09 percent, or 0.48 mg/L of K. Li is nonconservative in the Azure Spring system. A 7.5 percent loss of Li is indicated prior to station 2 of the downdrainage and an 8 percent loss in the satellite as compared to the main upwelling pool (fig. 4). After station 2, the Li ratios increase at a rate which is not GEOHYDROLOGY OF GEOTHERMAL SYSTEMS statistically distinguishable from the T, Cl, Na, As, and F curves, indicating conservative behavior after initial Li deposition in the immediate vicinity of the upwelling. The filtered and unfiltered HN03-acidified sample sets yielded similar results (8 percent loss of Li in the downdrainage, all of it prior to station 3; 91/2 percent low concentration anomaly in the satellite). The Li losses occurred in two sharply contrasting re- gimes of temperature and pH. The Li losses in the Azure Spring system are likely to be the result of coprecipitation with Al, because 430 ug/L, or 92 percent of the original soluble Al, was lost from solution prior to station 3 and >395 ug/L, or 85 percent, was lost in the satellite pool as compared to the main pool. Total (solute + suspended) Al losses were 525 and 555 [Lg/L in the satellite pool and prior to station 3, respectively. The field-extracted Al data for the main pool and channel show the same trend as the laboratory-extracted Al data (table 3); however, the field—extracted A1 estimate is only 19 percent of the laboratory-extracted estimate for the main pool. One explanation for this contrast is that the field-extracted Al (filtered through a 0.1 ,um membrane) includes a colloidal fraction which« is not extractable with 8-quinolinoI-MIBK but which becomes soluble during sample storage at pH 2 1. The difference between the field- and laboratory-extracted soluble A1 estimate is greatest for the main pool, where the soluble Al is pre- sumably in the process of being precipitated. In Octopus Spring the field-extracted AI decreases from 47 to 16 ug/L, concomitant with a pH increase of about 1, between stations 1 and 7; no significant change occurs in the laboratory-extracted Al along the drainage way. Although the main pool Al level is an order of magnitude higher for Azure as compared to Octopus Spring, at the ends of the two drainage chan- nels the Al levels have converged to similar values. The Li losses in the Azure Spring system may be due to precipitation of a Li—bearing aluminosilicate min- eral. The Li mica, lepidolite, was identified as a deposi- tion mineral at a depth depth of 25—40 m (below sur- face) in the Y—3 drill core (Barger and others, 1973), about 75 m northwest of Ojo Caliente Spring. Like Azure Spring, Ojo Caliente has an elevated A1 concen- tration (1.0 mg/L, Barger and others, 1973). Both the scientific drilling experience (White and others, 1975) and the typically high CI/Li ratios for the River Group hot springs (R. E. Stauffer and J. M. Thompson, unpub. data, 1975) suggest that Azure Spring and Ojo Caliente are derived from a common hot-water upflow regime where Li is being precipitated in the epithermal zone. If lepidolite is, in fact, forming in Azure Spring, the Li losses are probably not confined to the formation of this mineral. The Al/Li molar loss ratio for Azure TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK Spring is estimated to be 0.42, in contrast to the ratio 1.27 for lepidolite found by Barger, Beeson, Fournier, and Muffler (1973). Furthermore, the stoichiometry of lepidolite (K2Li3A13(AlSisOm)2(OH F)4) calls for K losses in the Azure Spring system twice as large as those inferred from figure 4. The concentrations of Mn and Fe in Azure and Oc- topus Spring waters are representative of the very low levels of these two elements in typical alkaline high-Cl (>250 mg/L) hot springs in the Upper, Lower and N or- ris Geyser Basins (tables 4 and 5). Significantly, water collected from the base of Porcelain Terrace at Norris has the lowest levels of both Fe and Mn of all the wa- ters tested; this spring also has the highest Cl level, hence the minimum dilution of hot-spring water, in Yellowstone Park. Ear Spring, representative of the high-Cl, low-HC03 waters of Geyser Hill, Upper Basin, also features low levels of both Fe and Mn. An un- named spring north of the Old Faithful interchange is a warm spring formed by near-surface dilution of hot high-Cl water with meteoric water. By contrast, this diluted spring has relatively high levels of both Fe and Mn. Steady Geyser is a dilute (low-Cl) hot spring in the Lower Basin characterized by low Fe levels and a large Mn/Fe ratio (table 5). The elevated Mn/Fe ratio indi- cates preferential oxidation of Fe(II) and losses from solution following dilution near surface waters. Little Whirligig water is a mixture of high-Cl, low-H003 Norris deep water with acid-sulfate water (low-Cl) of surficial origin. The acidic pH of the mixture preserves the relatively high levels of both Fe and Mn and the small Mn/Fe ratio to be expected from leaching of coun- try rock by the acid-sulfate water component. Brock, F11 TABLE 5.—Chloride, aluminum, iron, and manganese in selected geysers and hot springs Cl Al Fe Mn Sample Source (mg/L) (ALE/L) (pg/L) (Mg/L) Upper Basin: YT—73-71 ______ Ear Springé Geyser Hill *403 <5 1.1 1.0 YT— 73—62 ______ Sapphire, iscuit Basin" *297 7 7.9 .9 YT—73—76 ...... Excelsior, Midway Basin *266 23 1.5 18.3 74WA122U' ____Seep near Old Faithful _, 365 45 24 1.3 74WA123U -__ .Tortoise Shell __________ 389 26 2.4 .5 74WA124U ---.Unnamed, largest geyser near Sawmi 1 e“. 382 56 6.7 .5 74WA125U __V_Witches Cauldron "e 377 27 1.6 1.2 74WA126U ____Unnamed, east of Old Faithful, across Firehole River ____________________ 272 <6 5.2 1.5 74WA127U _-__Spouter ,,,,,,,,,,,,,,,,,,,,,,,,,,,, 317 29 3,5 .7 74WA128U ____Unnamed, new, below Old Faithful Interchange ______________ 242 16 154 14.3 74WA129U ____Unnamed, Black Sand Basin ________ 314 23 64 7.4 74WA130U ____Unnamed, north of Old Faithful Interchange __________ 353 36 20 1.4 Lower Basin: YT—73—68 ______ Gentian, Fountain Group ,,,,,,,,,,,, *300 11 2 7 .7 N YT—73—69 ______ Steady, at Firehole Lake __ *46 11 5 1 54.7 orris: YT—73~61 ______ Little Whirli ig __________ *587 10 160 20.8 YT-73— 74 ,,,,,, Base of Force ain Terrace ............ *704 <5 <7 .4 Mammoth: YT—73—65 ______ New Highland ______________________ *166 6 7.9 15.0 74WA146U .11_New Highland ...................... 171 18 7.1 15.8 74WA147U _.__Unnamed, new, blue colored near New High and ______________ 166 53 14.3 12.7 *Chloride anal sea by J. M. Thompson on HNO; acidified sam 195. ‘The letter U fo lowing the sample number indicates these samp es are unfiltered; the other samples were filtered. Cook, Peterson, and Mosser (1976) have found pure “acid sulfate type” (White, 1957; White and others 1971) thermal waters in Gibbon Meadows and Norris Basin with total soluble Fe levels (Fe(II) and F e(III)) exceeding 50 mg/L. UNNAMED SULFIDE-BEARING SPRINGS IN YELLOWSTONE CANYON Generically, both the 46°C source waters in the un- named sulfide-bearing spring system in Yellowstone Canyon would be classified as “warm” springs of TABLE 4.—Selected physical and chemical properties of additional hot waters [Temperature, Eh and 02 are field determinations; other analyses performed after return to Menlo Park Laboratory] As (V) Temperature Eh 02 (sq) S(vII) + As (V) As (III) Sample Source (°C) (mV) (mg/L) (ug/L) As (1111) (fig/L) (fig/L) As (III)/As (V) (ug ) Lower Ge ser Basin: 74W 119 White Creekl ,,,,,,,,,,,,,, ,51 274 H. , , n 345 315 ~30 0.10 Upper Geyser Basin: 74WA122U Seep near Old Faithful ,,,,,,, ,_.. ,,,,, ,895 —170 .0 630 1,280 $35 ~1,245 235 74WA123U Tortoise Shell 3. W" . . "A. ”93.5 A 156 .0 550 1,620 $30 ~1,590 253 74WA124U Unnamed, largest geyser near Sawmill . , "89.5 — 37 .22 <15 1,590 ~1,260 ~330 .25 74WA125 Witches Cauldron .W. H, ,91.0 ~108 .0 150 1,500 $40 1,460 236. 74WA126U Unnamed, east of Old Faithful across Firehold River H, N". .87 H n .0 390 850 S20 ~830 241. 74WA127U Spouter, Black Sand Meadow ,,,,,,,, , .85 *116 .0 190 1,495 $70 ~1,425 220 74WA128U Unnamed, new, below Old Faithful Interchange ,"735 +185 .60 <15 895 860 ~35 .04 74WA129U Unnamed, Black Sand Meadow . . , ,885 - 109 .0 90 1,455 $280 ‘> 1,175 *4.2 74WA130U Unnamed, north of Old Faithful Interchange . .e. W, .,,82 H .0 130 1,255 S200 \1,055 35.3 Mammoth: 74WA146U New Highland2 W, n" H, . W, , 690 $15 2675 ‘45 74WA147U Unnamed, new blue-colored, near New Highland‘ ,. W, H W. H, 680 S70 2610 >9 ‘Filtered sample; the other samples are unfiltered. 20n HNOu—acidified samples. F12 “mixed” origin (see Fournier and others, 1974; Four- nier and Truesdell, 1974; Fournier and others, 1976). The intermediate Cl levels and high HCO3, Mg, and Ca concentrations (table 6) identify the two source waters as high-Cl, high-enthalpy deep water which has been extensively diluted by meteoric water and has reacted with underlying carbonate rock strata. The high S(—II) and 804 concentrations of the springs, and the exten- sive solfataric activity in the surrounding region of the Yellowstone Canyon, may have their explanation in hot water leaching underlying sulfur-rich sedimentary rock. Similar conditions also exist in the Mammoth area, where some warm springs have reduced Cl levels and concentrations of major constituents which resem- ble the 46°C source water studied here (Rowe and others, 1973). The low F levels and high Cl/F ratios in the Yel- lowstone Canyon Springs (derived from data in table 6) probably result from fluorite equilibria acting on these high-Ca waters (Mahon, 1964; Nordstrom and Jenne, 1977). The sharp increases in pH, Eh, and dissolved 02 con- centration below the confluence and the decreases in S(—II) reflect gas exchange (C02, HZS ) with the atmos- phere and S(—II) oxidation, respectively. The high- carbonate alkalinity of these waters dominates the pH despite the release of protons accompanying S(—II) oxidation. Inferences based on the solute mixing fraction statis- tics P(S) (equation 3) must account for the solute- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS specific variances. The Li and K statistics are the most precise because of the large concentration differences between the two source waters for these two elements, and the high analytical precision at the concentration levels analyzed. The high SO, concentrations increase the analytical c.v. for C1 to ~2.5 percent. The large SO4/F equivalence ratio insures adequate analytical precision for the SO, statistics. However, the high con- centrations of S(—II) species in the source 1 water, and their potential for oxidation to S0,, add uncertainty to the 804 statistics. The P(F) statistics are imprecise be- cause of the low F concentrations and small concentra- tion difference between the two source waters. The pre- cision and accuracy of the P(As) statistics in this sytem suffer from the relatively high P concentrations, and the high and variable levels of S(—II) species among the samples (Stauffer, 1980b). The P3(S) (subscript identifies station) for S(—II), Cl, Na, K, and Li are statistically equivalent. The P4(S) statistics for these four solutes are all significantly smaller than the P3(S) values, indicating that the water sample obtained from immediately below the confluence was biased in favor of source 1 water. The sampling bias reflects incomplete mixing of the two source waters. The steep gradient and adequate mixing length between stations 3 and 4 insure that P4(S) val- ues are unbiased (best estimate P, — 0.375- 0. 007). The equivalence of the Cl, Na, K and Li mixing- fraction statistics indicates conservative behavior for the three alkalis in this system. Because of the high TABLE 6. —Chemical and physical parameters of Calcite and two other unnamed sulfide-bearing springs in Yellowstone Canyon with the drainage of the mixed spring system [Temperature, Eh, pH, and 0., are field determinations; other analyses performed after return to Menlo Park Laboratory, except as noted] Stations along common discharge drainage Concen- 0 two unnamed springs’“ Mixing fractionZ tratioan Parameter‘ Calcite 1 2 3 4 5 P3131 P.(S) 235% spring Spring Spring Sample 74WA ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 144 142 141 143 140 139 138 __________________ Time (hr) ,,,,,, . . ...... 1,700 ~1,400 ~1,300 ~1,500 ~1,215 ~1,130 ~1 030 ‘n.3 n.a n 8 Temperature (“0) _ ,,,,,,,,,,, 9 46 40 14 23— 26 19 14.5 n.a n.a n a pH ,,,,,,,,,,,,,,,,,,,,, 7 7 5.98 6 90 7.45 7 22 8.05 8.35 n.a n.a n a Eh(mV) .......... ~166 —94 —66 +14 -30 +185 +139 n.a n.a n3 0 (a ) (mg/L) ..... . . 4.60 80 5 5.8 n.a. n.a n a Alkdlinity .. 61 798 keoo 207 ...... 399 ______ n.a. n.a n a Na (m /L) 196 _, W" 116 36 6 75 3 63 0 64.8 .487 .333 103 K (mg ) _._ .958 ...... 56.4 14 1 34 8 29 9 29.5 .489 374 987 Li (mg/L) ............. 2.24 ...... 1.09 22 65 54 .56 .495 368 1 03 Mg (mg/L) ....... , 112.6 28.5 29.1 16 4 21 4 19 4 20.3 .394 236 1 05 Ca (mg/L) ,,,,,,,,,,,,,,,,,,, "28 2 227 232 62 0 141 120 123 .465 341 1 02 SO. (mfijL) ........................ 203 251 279 86 187 159 171 .523 378 1 08 Cl (mg l ...... .242 112 127 25 78 64 63 .487 382 987 F(mg/L).." ".43 2.2 25 16 20 2.0 20 .550 470 100 $102 (mg/L) . 169 72 78 68 68 69 66 S: (u /L) 800 3,210 4,800 55 2,250 <15 <15 n.a n a ,,,,,, Al (F ) (Mg/L) .113 239 243 11 128 94 12 n.a n a ,,,,,, Al (F0 (fig/L) . ......... 328 415 <4 69 ...... <4 n.a n a Fe (Fl) (Mg/L) 1,, 3 5 190 215 35 113 108 6 n.a r1 3 Mn (Fl) (g ,,,,,,,,,,,,, 7 4 283 98 89 88 88 37 n a n a §§(V)HL +ALS(lLII)) (fig/Ll ,1,450 742 710 150 395 280 270 n a n a (ug/L ................................................................... P6 (fig/L) ...................................... 2 61 57 57 53 56 60 n a n a ...... *Filtered sample, the other sam ‘Ff = filtered and field extrac les are unfiltered. ;Fl - filtered and lab extracted, Ul - unfiltered and lab extracted. 2PAS) is defined as the apparent fraction of water at station 3 which' 15 source 1 water; similarly, P. (S). “Concentration ratio of station 5 to station 4. ‘Not applicable. ”Best estimate; sample “Soluble reactive phospo rly mixed orus as P (Strickland and Parsons, 1968). TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK K/Al atomic ratio (93 for source 1, higher for source 2) and the very short hydraulic residence time in the sys- term, the conservation of K is to be expected. The mean mixing fractions (153,134) for the conservative elemental suite (Na, K, Li, C1) are used to quantitatively evaluate the behavior of Ca, Mg, Al, Fe, Mn, As and P in the mixed-water system. The concentrations of Mg and Ca recovered at sta- tions 4 and 5 indicate minor losses of both elements below the confluence (1.3 mg/L Mg, or 6 percent; 6 mg/L Ca, or 5 percent). Temperature-dependent min- eral equilibria calculations performed using WATEQF (Plummer and others, 1976) revealed supersaturation. with respect to carbonate and silicate minerals at sta-' tion 4. The saturation indices (1, = log“, (IAP/KT) * at station 4were: calcite (Is = 1.07); magnesite (I8 = 0.25); and well-ordered sepiolite (Is = 0.35). The pH increase and temperature decrease below station 4 (table 6) re- sult in increased supersaturation with respect to sepiolite at the lower station. Neighboring boiling "Calcite” spring is also strongly supersaturated with respect to both well-ordered and poorly ordered sepiol- ites (I, = 4.22;], = 2.63). The “Calcite” spring was near saturation with respect to calcite (I, = —0.14) and magnesite (Is = 0.37). The Yellowstone Canyon thermal springs are also supersaturated with respect to a broad class of alu- minosilicates, especially the montmorillonites. Clearly, the potential exists for Ca and Mg solute losses via a variety of mineral reactions. However, be- cause of the great stoichiometric excesses of Mg and Ca over Al in these springs, carbonates and pure silicates are likely to be the more important solubility controls. Wollast, MacKenzie, and Bricker (1968) noted that the *IAP/KT is the ratio of ion-activity product to thermodynamic solubility constant.) F13 precipitation of sepiolite is not kinetically hindered at low temperatures (25°C). Field-extractable soluble Al decreased rapidly below the confluence (table 7), whereas laboratory- extractable Al did not show a decrease until at least station 4, continuing the lag pattern noted earlier for the Azure Spring drainage channel. Iron precipitation in the drainage (table 7) likely re- sults from at least two mechanisms: amorphic FeS dep- osition from the high S(—II) waters accompanying the rapid pH rise, and Fe precipitation as amorphic hy- drous oxide from the oxygenated waters as they attain positive Eh values (Hem,1977). Large quantities of black precipitate (presumed to be FeS) were observed above station 3 in the mixed water, and especially above the confluence in the source 1 tributary. Reddish-brown hydrous oxide coatings were abundant in the drainage channel of the oxygenated 14°C source. The oxidation of both Fe(II) and S(—II) is probably mediated by the luxuriant growths of white filamen- tous bacteria observed in the drainage channel above station 4 (Castenholz, 1973). Calculations using WATEQF confirmed supersaturation with respect to amorphic FeS at station 1 (I, = 0.90) and with respect to amorphic Fe(OH)3 at station 4 (I, = 2.12). Soluble Mn and Fe concentrations were charac- terized by minor losses prior to station 4 and losses of 60 and 95 percent, respectively, by station 5 (table 7). The increases in the Mn/Fe ratio with distance below the confluence indicate a less rapid oxidation rate of Mn(II) than of Fe(II) in the aerated water. Singer and Stumm (1970) discussed the enormous kinetic en- hancement of Fe(II) oxidation associated with bacterial oxidation. The rate of oxidation of Mn(II) increases with increasing pH and increasing MnOz surface area, a catalytic effect (Morgan, 1967; Delfino and Lee, 1968). The observed delay in oxidation of Mn(II) and TABLE 7.—Estimated solute losses following mixing in Yellowstone Canyon mixed spring and Madison River systems Station 32 Station 4 Station 5 Projected Concentration Loss from Concentration Loss from Concentration Loss from concentration found confluence found confluence found confluence at *— Constituent confluence‘ (Mg/L) (pg/L) (percent) (Mg/L) (ng/L) (percent) (Mg/L) (fig/L) (percent) Yellowstone Can- yon mixed springs: S (—11) ,,,,,,,,,,,,,,,, 1,782 0 1,782 100 0 .1” Al (Ff)1 ,.., <4 151 100 94 l 1 12 83 87 108 A 7 — 7 6 95 95 88 4 4 37 55 60 280 74 21 270 84 24 ,.__ .n. ___, 56 1 2 60 -3 -5 67 39 37 56 50 74 62 44 42 1 1 1 — 9 — 9 97 5 5 90 1 2 12 16 2 11 13 5 28 14 22 7.6 .9 11 58 2.7 32 16 69 81 2 12 — 6 - 3 200 6 3 204 1 6 -.6 -11 8 ~26 —37 2 34 63 ‘Computed from source waters using mean mixing fractions for Na, K, Li and Cl from tables 6 and 8. 2Station in Yellowstone Canyon Mixed S ring System omitted due to nonrepresentativeness of collected sample. “Ff = filtered and field extracted; Fl = fi tered and lab extracted. ‘Soluble reactive phosphorus as P (Strickland and Parsons, 1968), F14 hence the increase in the Mn/F e ratio down drainage fits the Mn oxidation model. The relatively low concentrations of S(—II), Fe, and Al found at station 2 (14°C source) are probably residu- als of higher concentrations prevailing before interac- tion with the atmosphere. The high Mn/Fe ratio for source 2 is also evidence of the delayed oxidation of Mn(II) relative to Fe(II). Soluble reactive P shows no statistical evidence of precipitation losses below the confluence. Whereas Fe(II) and S(—II) are potential energy substrates for the filamentous bacteria in the channel, P is used only for steady—state maintenance of the bacterial popula- tion. The precipitation of 2102 pg of Fe/L between sta- tions 4 and 5 did not result in a significant reduction in P or As concentration. The high pH (>8.0) is unfavor- able for P and As(V) sorption on hydrous ferric oxides (see later discussion). An estimated 2 percent of the total solute As was lost between stations 4 and 5, the region where the principal losses of Fe and Mn through oxidation occurred. The As loss is not statistically significant. MADISON RIVER SYSTEM The date/time group is different for each of the sam- pling sites in the Madison River system. Time of day can be expected to affect temperature, pH, and dis- solved 02 at a site because of diurnal effects of the solar cycle on heating and photosynthesis. The estimated fraction of Madison River water at station 3, which is derived from the Firehole River and calculated using solute concentrations (equation 2), is biased unless the solute concentrations were independent of time during September 18—21. Because the period September 18—21 was precipitation free for the water, and was a low-flow period for the Madison River, the assumption of time independence in solute concentrations is equiv- alent to assuming that the cold-spring and hot-spring fluxes in the Gibbon and Firehole River drainages were invariant during the 3-day span. One indication of the near correctness of that assumption is provided by the close comparisons (mean difference <2 percent) in two sets of total As data for stations 1 and 2, collected 11 to 14 days apart during late September-early October 1974. The steadiness of the geothermal solute flux over short timespans has been noted by others (Fournier and others, 1976). The best estimate of the Firehole River mixing frac- tion P3 (fraction of Madison River discharge at station 3 derived from the Firehole Tributary) lies in the interval 0.70 to 0.75 (table 8). The geothermal indica— tor element, Li, defines the fraction most precisely be— cause of the high analytical precision and the large concentration difference between the two tributaries. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS The P3(As) value, 0.76, indicates that As being transported in the Gibbon and Firehole Rivers remains in solution below the confluence and above station 3. Using solute concentration ratios defined analogous- ly to those for the Lower Basin, the geochemical behav- ior of Li and As can be contrasted with that of Cl in the Madison Canyon region. Let Ri/;3(S ) where i = 4, 5, 6, be the ratio of solute S concentration at station i to the concentration at station 3, immediately below the con- fluence. The R4/3(S) and R 5,/;,(S) ratios apply to the Madison River between West Yellowstone and Madi- son Junction. The R6/3(S) ratios reflect the extensive dilution of Madison River water by stored water and other tributary water flowing into Hebgen Lake. Ad- ditions of geothermal water between Madison Junction and West Yellowstone are considered here to be negli- gible. This assumption is only approximately correct; however, inferences based on the solute ratio statistics are rather insensitive to minor unaccounted geother- mal fluxes within Madison Canyon. Elements which are well represented in nonthermal waters (particularly K, Ca, and Mg) can be expected to have R4, R5 and R6 ratios higher than for C1. This was, in fact, observed (table 8). TheRm and R 3/3 ratios for K, Ca, and Mg are only slightly higher than for Cl because of the very slight tributary development of the Madi- son River within the Madison Canyon. The R 5,3(Cl) statistic shows that the water at Madison Junction has been diluted by approximately 5 percent when it arrives at West Yellowstone gaging station. Elements of geothermal origin which behave conser- vatively in the river will have concentration ratios which are not statistically distinguishable from the Cl ratios. Nonconservative behavior is identifiable by concentration ratios significantly lower than for C1. Thus, the concentration ratios indicate that both Li and total As are conservative in the Madison Canyon (table 8). The mean percent difference between the As/Cl and Li/Cl ratios for stations 4 and 5 is only -0.3 percent. A -2.0 percent mean difference between the As/Cl ratios, or between the Li/Cl ratios, is necessary to reject the null hypothesis (equivalent) at the 5 percent level. The statistical test is thus capable of discerning an ~2.0-percent loss of either Li or As in the Madison River above West Yellowstone. The R6 ratios for Cl, As, and Li also show quite close agreement (table 8). Because the Cl concentration at station 6 was determined by a different analytical method and different analyst, the C1 ratio has an ad- ditional error component. Furthermore, the analytical coefficients of variation for As and Li increase at the low concentration levels characteristic of station 6. Hence, the small negative differences between R6/3(Li) and R 5/3(Cl) and between R6/;,(As) and R.;/3(Cl) are TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK F15 TABLE 8.—-Chemical and physical parameters of the Madison River drainage ['I‘emperature. Eh, pH, and 02 are field determinations; other analyses performed after return to Menlo Park Laboratory, except as noted] Stations along Madison River drainage* Concentration Firehole Gibbon Madison Mixing ratio3 fraction Parameter‘ 1 2 3 4 5 6 P; (S)2 5/3 6/1 Sample 74WA ,,,,,,,,,,,,,,,,,,,,,,,, 107 108 10 110 111 120 We- ,___ .___ Time _____________ .H ,1.030 1,230 1 600 1 730 1,400 1,600 n.a.‘ n.a n.a. Date (1974) ,,,,,,, 9/18 9/1 9/18 9/18 9/1 9/21 Temperature (°C) _ , 15.0 14.0 18.5 1 .5 16.5 14.0 n.a. n.a. n.a. H _______________ , 8.40 7.04 7.50 7.68 8.25 ,___ n.a. n.a. n.a. Eh (mV) ......... , 452 402 358 394 4 ,___ n.a. n.a. n.a. 01 (sq) (mg/L) ..... , 7.60 9.10 8.40 8.15 8.20 H" n.a n.a. n.a. Alkalinity (mg/L) . 106 115 115 108 109 ,___ .18 .948 n.a. Na (mg/L) _________ 73.6 55.4 68.1 67.5 66.4 __._ .70 .976 n.a. K (mg/L) ,,,,,,,,,,,,,, 6.75 8.40 7.25 7.15 7.15 4.30 .70 .986 .601 Li (mg/L) ”,7 A .528 .323 .477 .457 .450 .186 .75 .947 .412 Mg (mg/LL... .488 1.23 .704 .691 .762 1.83 .70 1.08 2.40 Ca (mg/L) .___ 5.24 8.29 6.37 6.04 6.12 11.9 .63 .961 1.94 50‘ (m /L) H 10 18 13 12 ___ .60 .954 1... Cl (mg ) "v. ......... 54 33 47 44 20 .66 .949 .449 F (mg/L) ___. 7 2 4.1 6 4 6 4 6.3 ".1 74 .984 n.a. $102 (mg/L) __ 100 75 92 87 1 n a n.a. n.a. Al (Fl) (#g/L) 52 235 111 90 15 68 .811 n.a. Al (Fl) (as/L) 57 235 67 56 62 .fi. 94 .910 n.a. Fe (Fl) (pg/L) H. ................ 14 30 16 14 24 87 .875 n.a. Mn (Fl) (p. /L) .__ 2.6 24 7.6 5.8 16 24 78 .205 n.a. As(V) (pg ) ............ 245 ~92 ~185 _ _. As (V) + As (Ill) (pg/L) ... 252 83 212 200 204 88 76 .962 431 Psmga.) ............................ 6 4 6 7 ’Filtered sample, the others samples are unfiltered. ‘Ff = filtered and field extracted; Fl = filtered and lab extracted; Ul = unfiltered and lab extracted. 1?; (S) is defined as the apparent fraction of water at station 3 which is source 1 water. 3Concentration ratio station 5 to station 3, station 6 to station 3. ‘Not a plicable. 5Solub e reactive phosphorus as P (Strickland and Parsons, 1968). inconclusive evidence of Li and As losses in Hebgen Lake. The ratio comparisons suggest that only a minor fraction of the soluble Li and As fluxes in the Madison River are lost in Hebgen Lake during fall base-flow conditions. The near conservatism is not surprising since the elevated river water temperature and low suspended-solids load of the Madison should lead to short circuiting through the surface waters of Hebgen Lake. Similarly, Axtmann (1974) noted the near con- servatism of total As in the Waikato River and Lake Taupo below Wairakei in New Zealand. The mean P concentration for the Firehole and Madison Rivers (5.5 ug/L) is typical of low levels found in unpolluted rivers derived from forested drainage basins; the present estimates are smaller by a factor of 20 than the As(V)-biased P data previously reported by Boylen and Brock (1973) and Zeikus and Brock (1972) for the Madison River system. The field-extracted Al estimates reveal a significant loss of Al immediately following the mixing of the Gib- bon and Firehole Rivers (table 7). However, following the pattern described for the other drainages studies, the laboratory-extracted Al data reveal smaller de- creases, occurring more slowly as a function of distance below the confluence. The methodological difference may the result of the formation of nonextractable A1 polymers (Smith, 1971), which are depolymerized dur- ing storage in 0.1 molar HNO3. The loss of Al is not accompanied by a detectable loss of Li. The Gibbon River contains higher Al, Fe, and Mn concentrations than the Firehole River. The concentra- tion of soluble Mn decreases gradually below the con- fluence to ~16 [Lg/L near West Yellowstone; however, no significant change was detected in soluble iron (ta- ble 7). The large increases in Mn, Ca, and Mg concen- tration (1,400, 94, 140 percent, respectively) between stations 5 and 6 are a result of nonthermal waters en- tering Hebgen Lake from other drainage areas. SORPTION CONTROLS ON ARSENIC IN GEOTHERMAL WATERS Amorphic hydrous Fe oxides are the most efficient As scavengers in the class of amorphic precipitates formed by Al, Fe, and Mn in natural waters (Gulledge and O’Connor, 1973; O’Connor, 1974; Singer, 1974). The almost quantitative removal (>99 percent) of As(V) from neutral to acidic solutions using ferric hy- droxide coprecipitation has been the chemical basis of many concentration schemes for analyzing As at low levels in the environment (Portmann and Riley, 1964). Furthermore, it is well known that the sorption affinity of ferric hydroxide for As(V) is greatly reduced under alkaline conditions (Portmann and Riley, 1964; O’Connor, 1974). Trivalent As is inefficiently sorbed at all pH values. For example, Logsdon, Sorg, and Symons (1974) found only about 50-percent removal of As(III) (initially at the 0.3 mg/L level) using a 30 mg/L ferric sulfate dose; the percentage removed was inde- pendent of acidity in the pH range 6 to 9. For As(V) (initially at 0.05 mg/L) O’Connor (1974) reported 88 percent removal at pH 8.0 in the presence of 10 mg/L Fe(III). Al is less effective than Fe at scavenging both As(III) F16 and As(V), the differences increasing above pH 7.0. Logsdon, Sorg, and Symons (1974) found that 30 mg/L alkali aluminium sulfate (alum) treatments removed about 90 percent of As(V) and 10 percent of As(III) in the pH range 5 to 7. The initial As levels tested were 0.30 mg/L, very similar to the total As concentration in the lower Firehole River. Above pH 7.0, the percent removal of As(V) by 30 mg/L alum declined almost linearly to 15 percent at pH 8.5. The authors are not aware of quantitative studies on As sorption by MnOz. However, in light of the chemical similarity between H3AsO4 and H3PO4 (the two acids have nearly identical pK values for all three dissocia- tions, Sillen and Martell, 1964), and the inefficiency of phosphate sorption on MnOz, Mn may play a relatively minor role in the sorption chemistry of As. As sorption on MnOz is likely to be insignificant for two additional reasons: (1) The acid leaching of typical igneous rock solubilizes great stoichiometric excesses of both Fe and Al as compared to Mn; (2) the oxidation and precipita— tion of Mn is often delayed in geothermal waters until the pH is far too high to effect efficient removal of As(V). The conservative behavior of total As in Octopus Spring is an expected consequence of the high pH and very high As/Fe atomic ratio of 570. Furthermore, very large As/Fe ratios are typical of alkaline hot springs throughout the major hot-spring basins. In Azure Spring, the initial soluble As/Al atomic ratio is only 1.20, in contrast to the As/Fe atomic ratio of 183. How- ever, Al precipitation occurs prior to station 3 in the drainage channel, and prior to the oxidation of a significant amount of As(III). This oxidation lag can be expected to greatly reduce the effectiveness of As re- moval by oxidic Al precipitates. Coprecipitation of A1 with SiOZ to form montmorillonite (R. O. Fournier, oral communication, 1975) may also reduce As losses onto oxidic Al in the high-Al springs of the River Group. Thermodynamic calculations indicate Azure Spring is supersaturated with respect to several montmoril- lonites. The levels of both P and As in the Yellowstone Can- yon warm spring indicate that, even in the presence of anomalously high Fe levels (2200 ug/L), the removal of As by sorption is relatively insignificant in this al— kaline spring water. The low Fe/As molar ratios, the very large As(III)/As(V) initial ratios, and the rapidly rising pH after the emergence of the spring water at the ground surface all combine to reduce sorption by hydrous Fe oxides as an important mechanism affect— ing As in the drainages of alkaline hot springs. As behaves nearly conservatively during the long base flow period (losses <2 percent of flux) in the Madison-Firehole River system. In addition to evi— GEOHYDROLOGY OF GEOTHERMAL SYSTEMS dence previously discussed, samples obtained from the Firehole River at Madison Junction and just below Lower Basin (early October 1974) had Cl/As atomic ratios of 466, and 456, respectively. Given the probable analytical errors in the Cl and As determinations and the lag time in sampling, the ratio difference is not statistically significant. The Cl/As atomic ratio in the Firehole River is close to the median ratio for hot- spring groups in the Firehole River drainage (R. E. Stauffer and J. M. Thompson, unpublished data, 1975). Because of inadequate spring flow data, a flow- weighted mean Cl/As ratio for the hot springs in the Firehole River drainage cannot be precisely calculated. Based on Fournier, White, and Truesdell (1976) and the difference between the Upper, Midway, and Lower Basin Cl/As ratios, the Firehole River Cl/As ratio is close to a weighted mean for the springs of the drain- age basin. Near conservatism of As during low flow occurs be- cause of the following factors: (1) The soluble Fe/As ratio is too low to significantly affect As by sorption or coprecipitation; (2) although the suspended solids may be assumed to have surface-active coatings of Fe oxides (Jenne, 1968, 1977), the concentration of suspended in- organic solids in the rivers is low during the low-flow period; (3) the alkaline pH of the river during daylight hours is suboptimal for As(V) sorption on Fe oxides. During high runoff periods (principally the May- June snowmelt period, Fournier and others, 1976) the pH of the Madison River drops and the riverborne flux of sorption-active clastics increases dramatically. Dur- ing such periods the increased contact opportunity be- tween As (V) and the amorphic Fe oxide coatings on the particulates might be expected to result in significant losses of soluble As in the Madison River. The most likely sink for sorbed As(V) during spring runoff is the sediments of Hebgen Lake. The Cl/As atomic ratio for the Gibbon River at Madi- son Junction is 760, a value significantly higher than the ratios for key indicator springs in both the Gibbon (Cl/As=400 : 25) and Norris (Cl/As=543 : 11) hot- spring basins (R. E. Stauffer and J. M. Thompson, un- published data, 1975). The Gibbon and Norris thermal basins constitute most of the geothermal activity in the Gibbon River drainage basin (Fournier and others, 1976). The Cl/Li atomic ratio for the Gibbon River at Madison Junction is 20.1, intermediate among the mean ratios for the Norris (23.8 i 0.5) and Gibbon (Gibbon Meadows area: Cl/Li = 19.2 t 1.4; Artist Paint Pots: Cl/Li = 16.4 i 0.4) geyser basins (R. E. Stauffer and J. M. Thompson, unpublished data, 1975). On the basis of the Cl/Li ratios, it is possible to infer that in excess of 50 percent of the Gibbon River Cl flux origi- nates in the Gibbon Geyser Basin. An immediate im- TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK plication of this inequality is that 60 percent or more of the geothermal As flux in the Gibbon River drainage basin during base-flow conditions is removed by pre- cipitation and sorption processes. The high Gibbon River Cl/As ratio probably results from As(V) sorption on hydrous Fe oxides and oxidic Al above Gibbon Can- yon as the river’s Cl/As ratio is also elevated in the canyon. Some As may also be precipitated as FeAsO4 and realgar (ASS) in the acidic spring environments (depending on redox potential). Acid-sulfate waters containing high levels of Fe and Al mix with poorly buffered high-Cl, high-As waters in both the Norris and Gibbon thermal basin (Allen and Day, 1935; White, 1957); the pH of the weakly acidic water rises, and large amounts of Fe and Al precipitate. The mix- ing of two geochemically distinct water types in the Norris-Gibbon region thus produces conditions which are optimal for the sorption of As(V) on both Fe and Al precipitates. The time delay required for mixing and F e(II) oxidation (Brock and others, 1976) helps insure that As initially in the form of As (III) has been oxidized to the more readily sorbing As(V). ARSENIC FLUX OF THE MADISON RIVER Using the September concentration estimate of 202 ,ug/L of As and available earlier discharge measure- ments during September low flow at the West Yel- lowstone gaging station (12.7 m3/s, US. Geological Survey, 1974; flow measurement was discontinued after 1973), the estimated base flow total soluble As flux for the Madison River is 220 kg/day, 90 percent of which is contributed by the F irehole River. An ad- ditional 25 to 30 kg/day of As is precipitated in the Gibbon River drainage basin. Because the September geothermal As flux can be assumed to be representa- tive of the daily flux on an annual basis (Fournier and others, 1976; R. E. Stauffer and J. M. Thompson, un- published data, 1975), the annual total (sol- uble+particulate) As flux of the Madison River at West Yellowstone is 290,000 kg. It is likely that particulate As deposited in the Gibbon drainage basin is flushed out during peak discharge in May and June and depos- ited in the sediments of Hebgen Lake. The annual As flux for the Madison River is less than the flux estimate by Axtmann (1974) for Wairakei (158 metric tons) and the atmospheric As flux estimated for the ASARCO Copper smelter in Tacoma, Wash. (150,000 kg/year, Crecelius, 1975). The Tacoma smel- ter acquired notoriety as a result of its high As flux (Lawson, 1975). Swain (1949) reported an atmospheric As flux of 22.5 metric tons/day for the Anaconda Smel— ter at Anaconda, Mont., a value which is nearly 100 times the Madison River flux. The reason for the large discrepancy between the present ASARCO and histori- F17 cal Anaconda fluxes is unclear; however, the high Anaconda figure applies to a period prior to any pollu- tion abatement, and As is currently recovered as a by- product at the Tacoma plant. Because the smelter As flux is almost exclusively the highly toxic particulate As203 (Crecelius, 1975), and the Madison River flux is dominantly soluble As(V) at the 200 ug/L level, an im- portant distinction exists between the environmental impacts of the two As sources. At levels above 1 mg/L in drinking water, As is associated with long-term tox- icity effects in humans, As(III) being notably more toxic than As(V) (Penrose, 1974). During low-flow conditions the total As concentra- tion at the Hebgen Dam spillway is about 85 to 90 ug/L, or about 75 percent, above the US. Public Health rejection limit for municipal water supplies (50 ug/L, Environmental Protection Agency, 1972). With in- creasing distance downriver, this level is presumed to be progressively reduced by dilution and sorption on clastics. POTENTIAL ECOLOGICAL SIGNIFICANCE OF ARSENIC The ecological role of As in the microbiological com- munities of the hot-spring drainages, and in the receiv- ing rivers, is mainly a subject of conjecture. As is known to be oxidized and reduced (between As(III) and As(V)) by microorganisms under a variety of natural conditions (Johnson, 1972; Myers and others, 1973; Pilson, 1974). As is also concentrated by a factor of 5,300 by the dominant macrophyte in the Waikato River (Axtmann, 1974, 1975). However, as Penrose (1974) noted, the concentration factors are likely to be highly dependent on aquatic species, as well as redox state of the As. Nothing is known at present about the As concentration factors for algae and macrophytes in- habiting geothermally derived waters in Yellowstone Park. Because of the chemical similarity between A80;3 and PO4‘3, and the importance of P as a critical nutrient for all plant forms, both the low levels of P and the very high As(V)/P ratios for many of the hot springs and receiving waters may be of enormous ecological significance in Yellowstone Park. Limnetic environments enriched in As have become dominated by the relatively tolerant blue—green algae (J. Shapiro, oral communication, 1978), suggesting that the As/P gradients in Yellowstone thermal waters may also act to select blue green. Oceanographers have noted that the As/P atomic ratio approaches 1.0 in the oligotro- phic waters of the ocean (Portmann and Riley, 1964; Johnson and Pilson, 1972b). Penrose (1974) has noted that marine fish have relatively high As contents, probably as a result of concentration within the food chain and the effects of the As/P ratio on the phyto- plankton. As concentration in phytoplankton is also F18 the probable cause of the high As concentration in or- ganically rich marine shales (Onishi and Sandell, 1955). In Octopus Spring the As(V)/P molar ratio reaches 350, two orders of magnitude higher than the upper bound ratios likely to occur in the ocean. Even in the Madison River the As(V)/P ratio is about 15. CONCLUSIONS 1. Soluble A1, Fe, and Mn approach low limiting val- ues of <5.0, 1.0, and 1.0 ug/L, respectively, in the least diluted high-Cl waters of the Norris and Upper Geyser Basins. The springs so charac- terized represent zones of highest enthalpy flux in the park. The levels of Al, Fe, and Mn do not appear to correlate with the presence or absence of dissolved S(—II). 2. The soluble-Mn levels are strikingly elevated (ap- proximately two orders of magnitude) in alkaline springs of “mixed water” origin. These springs in- clude the Mammoth hot springs, Steady Geyser, the unnamed spring north of the Old Faithful highway interchange, and the warm springs in the Yellowstone Canyon. The Mn/F e atomic ratio is >1 in almost all of the alkaline springs of mixed origin, the ratio tending to increase with temperature. 3. The Mn/F e ratio increases along the drainage chan- nels of S(-II)-containing springs of mixed origin. The increase in the ratio probably reflects the slower rate of Mn(II) vs Fe(II) oxidation and the less effective solubility controls on Mn(II) at low temperatures. 4. Al precipitates rapidly in the relatively rare boiling springs containing elevated levels of the element. Al precipitation apparently occurs in Azure Spring prior to emergence of the water at the sur- face; the reaction proceeds virtually to completion at 89°C and pH <75. 5. Li, and to a lesser extent K, behave nonconserva- tively if Al is being precipitated from hot-spring waters. The most likely mechanism of Li removal is through the formation of Li-bearing aluminosilicate minerals. 6. P is present in very low concentrations (:2 ,ug/L) in undiluted hot-spring waters of the Upper, Lower, and Norris Basins. Higher concentrations of P are associated with warm springs of mixed-water origin. 7. The As(III)/As(V) ratio in these thermal waters var- ies over three orders of magnitude and is strongly bimodally distributed according to whether or not S(—II) is present in detectable quantities. 8. Arsenite is rapidly oxidized to As(V) in the hot- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS spring drainages, but only after the volatilization and oxidation of S(—II). 9. Total As is essentially conservative (losses under 2 percent) in all of the surface-water systems studied, with the prominent exception of the Gib- bon River. The major losses of total dissolved As in the Gibbon River occur above Gibbon Canyon and are hypothesized to result from As(V) sorp- tion on oxidic Fe and Al formed as a result of the mixing of acid-sulfate and alkaline high—Cl wa- ters in the upper Gibbon drainage basin. 10. The high levels of As, the variable As(III)/As(V) ratios, and the extraordinarily high As/P ratios in Yellowstone geothermal waters are potential adverse ecological factors in the park. REFERENCES CITED Allen, E. T., and Day, A. L., 1935, Hot springs of the Yellowstone National Park: Carnegie Inst. Washington Pub. No. 466, 525 p. American Public Health Association, 1971, Standard methods for the examination of water and wastewater: Washington, DC, 874 p. Axtmann, R. C., 1974, An environmental study of the Wairakei power plant: Physics and Eng. Lab. Report. 445, Dept. Sci. and Indus. Research, Lower Hutt, New Zealand, 38 p. 1975, Environmental impact of a geothermal power plant: Sci., v. 187, p. 795—803. Ball, J. W., Jenne, E. A., and Burchard, J. M., 1976, Sampling and preservation techniques for waters in geysers and hot springs, with a section on gas collection by A. H. Truesdell: Workshop on Sampling Geothermal Effluents, lst, Proc., Environmental Pro- tection Agency-600/9—76—011, p. 218—234. Bargar, K. E., Beeson, M. H., Foumier, R. 0., and Muffler, L. J. P., 1973, Present-day deposition of lepidolite from thermal waters in Yellowstone National Park: American Mineralogist, v. 58, p. 901-904. Barnes, H. L., 1967, Geochemistry of hydrothermal ore deposits: New York, Holt, Rinehart and Winston, 670 p. Baumann, E. W., 1974, Determination of parts per billion sulfide in water with the sulfide-selective electrode: Anal. Chemistry, v. 46, p. 1345—1347. Bowen, F. S., 1926, The ratio of heat losses by conduction and by evaporation from any water surface: Phys. Rev., v. 27, p. 779— 787. Boylen, C. W., and Brock, T. D., 1973, Effects of thermal additions from the Yellowstone geyser basins on the benthic algae of the Firehole River: Ecology, v. 54, p. 1282—1291. Brock, T. D., 1967, Relationship between standing crop and primary productivity along a hot spring thermal gradient: Ecology, v. 48, p. 566—571. 1969, Vertical zonation in hot spring algal mats: Jour. Phycologia, v. 8, p. 201—205. Brock, T. D., and Brock, M. L., 1966, Temperature optima for algal development in Yellowstone and Iceland hot springs: Nature, v. 209, p. 733-734. 1967, Methodology for measurement of chlorophyll, primary productivity, photophosphorylation, and macro-molecules in benthic algal mats: Limnology and Oceanography, v. 12, p. 600—605. 1968a, Measurement of steady-state growth rates of a ther- mophilic alga directly in nature: Jour. Appl. Bacteriology, v. 95, p. 811—815. TRACE ELEMENTS IN HOT-SPRING DRAINAGES, YELLOWSTONE NATIONAL PARK 1968b, Relationship between environmental temperature and optimum temperature of bacteria along a hot spring thermal gradient: Jour. Appl. Bacteriology, v. 31, p. 54—58. 1969a, Effect of light intensity on photosynthesis by thermal algae adapted to natural and reduced sunlight: Limnology and Oceanography, v. 14, p. 334—341. 1969b, Recovery of a hot spring community from a catas- trophe: Jour. Phycologia, v. 5, p. 75—77. Brock, T. D., Brock, M. L., Bott, T. L., and Edwards, M. R., 1971, Microbial life at 90C: the sulphur bacteria of Boulder Spring: Jour. Bacteriology, v. 107, p. 303—314. Brock, T. D., Cook, Susan, Petersen, S. and Mosser, J. L., 1976, Biogeochemistry and bacteriology of ferrous iron oxidation in geothermal habitats: Geochim. et Cosmochim. Acta, v. 40, p. 493— 500. Browne, P. R. L., 1969, Sulfide mineralization in a Broadlands geothermal drillhole: Econ. Geol., v. 64, p. 156— 159. Castenholz, R. W., 1973, The possible photosynthetic use of sulfide by the filamentous phototrophic bacteria of hot springs: Limnology and Oceanography, v. 18, p. 863—877. Crecelius, E. A., 1975, The geochemical cycle of arsenic in Lake Washington and its relationship to other elements: Limnology and Oceanography, v. 20, p. 441—451. Delfino, J. J., and Lee, G. F., 1968, Chemistry of manganese in Lake Mendota, Wisconsin: Environmental Sci. and Technology, v. 2, p.1094—1100. Ellis, A. J ., 1963, The effect of temperature on the ionization of hy- drofluoric acid: Jour. Chem. Soc., London, p. 4300—4304. 1967, The chemistry of some explored geothermal systems, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Holt, Rinehart and Winston, p. 465—514. 1969, Present-day hydrothermal systems and mineral deposi- tion: Commonwealth Mining and Metallurgical CongJ, 9th, 1969, United Kingdom, Proc., p. 1—30. 1970, Quantitative interpretation of chemical characteristics of hydrothermal systems: U.N, Symposium on the Development and Utilization of Geothermal Resources, Pisa, Italy. 1973, Chemical processes in hydrothermal systems—A re- view: International Symposium on Hydrogeochemistry and Biogeochemistry, Japan, 1970, Proc., v. 1, Hydrogeochemistry, p. 1—26. Ellis, A. J., and Wilson, S. H., 1960, The geochemistry of alkali metal ions in the Wairakei hydrothermal system: New Zealand Jour. Geology and Geophysics, v. 3, p. 593—617. Environmental Protection Agency, 1972, Water Quality Criteria: Washington, U.S. Govt. Printing Office, 56 p. F ournier, R. 0., and Truesdell, A. H., 1974, Geochemical indicators of subsurface temperature—Part 2, Estimation of temperature and fraction of hot water mixed with cold water: U.S. Geol. Sur- vey Jour. Research, v. 2, p. 263—269. Fournier, R. 0., White, D. E., and Truesdell, A. H., 1974, Geochemi- cal indicators of subsurface temperature—Part 1, Basic assump- tions: U.S. Geol, Survey Jour. Research, v. 2, p. 259—262. 1976, Convective heat flow in Yellowstone National Park: U.N. Geothermal Symposium, 2d, San Francisco, 1975, Proc., p. 731— 739. Gulledge, J. J., and O’Connor, J. T., 1973, Removal of arsenic (V) from water by adsorption on aluminum and ferric hydroxides: Am. Water Works Assoc. Jour., v. 5, p. 548—552. Hem, J. D., 1977, Reactions of metal ions at surfaces of hydrous iron oxide: Geochem. et Cosmochim. Acta, v. 41, p. 527—538. Hogg, R. V., and Craig, A. T., 1970, Introduction to mathematical statistics [3d ed.,]: London, Macmillan, 415 p. Jenne, E. A., 1968, Controls on Mn, Fe, Co, Ni, Cu, and Zn concen- trations in soils and water; the significant role of hydrous Mn F19 and Fe oxides: Adv. Chemistry Ser. 73, p. 337—387. 1977, Trace element sorption by sediments and soils—sites and processes, in Chappel, W. and Petersen, K., eds., Symposium on molybdenum in the environment, v. 2: New York, M. Dekker, Inc., p. 425—553. Johnson, D. L., 1972, Bacterial reduction of ar'senate in sea water: Nature, v. 240, p. 44—45. Johnson, D. L., and Pilson, M. E. Q., 19723, Spectrophotometric de- termination of arsenite, arsenate, and phosphate in natural waters: Anal. Chim. Acta, v. 58, p. 289—299. 1972b, Arsenate in the western North Atlantic and adjacent regions: Jour. Marine Research, v. 30, p. 140— 149. Kennedy, V. C., Zellweger, G. W., and Jones, B. F., 1974, Filter pore-size effects on the analysis of Al, Fe, Mn, and Ti in water: Water Resources Researvh, v.10, p. 785— 790. Kennedy, V. C., Jenne, E. A., and Burchard, J. M., 1976, Backflush- ing filters for field processing of water samples prior to trace- element analyses: U.S. Geol. Survey Open-File Report 76— 126, p.1—12. Lawson, H. G., 1975, Smelter’s emissions threaten populace, but so does possible loss of 1,000 jobs: Wall Street Jour., 16 July 1975, p. 30. Logsdon, G. S., Sorg, T. J., and Symons, J. M., 1974, Removal of heavy metals by conventional treatment: Water Quality Confer- ence, 16th, Proc., v. 71, no. 108, p. 111—133. Mackereth, F. J. H., 1963, Some methods of water analysis for lim- nologists: Freshwater Biol. Assoc. Scientific Pub., v. 21, 71 p. Mahon, W. A. J ., 1964, Fluorine in the natural thermal water of New Zealand: New Zealand Jour. Sci., v. 7, p. 3—28. Marler, G. D., 1973, Inventory of thermal features of the Firehole River geyser basins and other selected areas of Yellowstone Na- tional Park: NTIS PB—221 289, Springfield, VA, 648 p. Morgan, J. J ., 1967, Chemical equilibria and kinetic properties of manganese in natural waters, in Faust, S. D., and Hunter, J. V., eds., Principles and applications of water chemistry: New York, John Wiley, p. 561—622. Myers. D. J., Heimbrook, M. E., Osteryoung, J., and Morrison, S. M., 1973, Arsenic oxidation state in the presence of microorganisms: examination by differential pulse polarography: Environmental Letters, v. 5, p. 53—61. Nordstrom, D. K., and Jenne, E. A., 1977, Fluorite solubility equilib- ria in selected geothermal waters: Geochim. et Cosmochim. Acta, v. 4, p. 175—188. O’Brien, D. J., and Birkner, F. B., 1977, Kinetics of oxygenation of reduced sulfur species in aqueous solution: Environmental Sci. and Technology, v. 11, p. 1114—1120. O’Connor, J. T., 1974, Removal of trace inorganic constituents by conventional treatment processes: Water Quality Conf., 16th, Proc., v. 71, n. 108 p. 99— 110. Onishi, Hiroshi, and Sandell, E. B., 1955, Geochemistry of arsenic: Geochim. et Cosmochim. Acta, v. 7, p. 1—33. Penrose, W. R., 1974, Arsenic in the marine and aquatic environ- ments: Analysis, occurrence, and significance: Chem. Rubber Co. Critical Rev., Environmental Control, v. 4, no. 4, p. 465-482. Pilson, M. E. 0., 1974, Arsenate uptake and reduction by Pocillopora verrucosa: Limnology and Oceanography, v. 19, p. 339—341. Plummer, L. N., Jones, B. F., and Truesdell, A. H., 1976, WATEQF, A Fortran IV version of WATEQ, a computer program for cal- culating chemical equilibria of natural waters: U.S. Geol. Sur- vey, Water Resources Inv. 76— 13, 61 p. Portmann, J. E., and Riley, J. P., 1964, Determination of arsenic in sea water, marine plants, and silicate and carbonate sediments: Anal. Chim. Acta, v. 31, p. 509—519. Reay, P. F., 1973, Arsenic in the Waikato River system: Pollution Research Conference, Wairakei, 20—21 June 1973, Proc., New F20 Zealand Dept. Sci. and Indus. Research Inf. Ser. No. 97, p. 365— 376. Ritchie, J. A., 1973, A determination of some base metals in Broad- lands geothermal waters: Physics and Eng. Lab. Rept. 2164, Dept. Sci. and Indus. Research, Lower Hutt, New Zealand, 24 p. Rowe, J. J., Fournier, R. 0., and Morey, G. W., 1973, Chemical analysis of thermal waters in Yellowstone National Park, Wyoming, 1960—65: US. Geol. Survey Bull. 1303, 31 p. Sillen, L.G., and Martell, A. E., 1964, Stability constants of metal-ion complexes: Chem. Soc., London, no. 17, 754 p. I Singer, P. C., 1874, Chemical processes for the removal of trace met— als from drinking waters: Water Quality Conf., 16th, Proc., v. 71, no. 108, p. 91—98. Singer, P. C., and Stumm, W., 1970, Acidic mine drainage: The rate-determining step: Sci., v. 167, p. 1121—1123. Smith, R. W., 1971, Relations among equilibrium and non equilib- rium aqueous species of aluminum hydroxy complexes, in Gould, R. F., ed., Nonequilibrium systems in natural water chemistry: Adv. Chemistry Ser. 106, p. 250—279. Stauffer, R. E., 1977, Measuring total antimony in geothermal wa- ters by flame atomic absorption spectrometry: U.S. Geol. Survey Jour. Research, v. 5, p. 807—809. 1980a, Bias evaluation of the cation exchange method for chloride and sulfate: Canadian Jour. Fisheries and Aquatic Sci. (in press). 1980b, Molybdenum-blue applied to arsenic and phosphorus determinations in fluoride and silica rich geothermal waters: Environmental Sci. and Technology (in press). Strickland, J. D. H., and Parsons, T. R., 1968, A practical handbook of seawater analysis: Canadian Fish Resources Board, Ottawa, 211 p. Swain, R. E., 1949, Smoke and fume investigations: A historical review: Indus. and Eng. Chemistry, v. 41, p. 2384—2388. GPO 689-14 3 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Thompson, J.M., Presser, R. S. Barnes, R. B., and Bird, D. E., 1975, Chemical analysis of the waters of Yellowstone National Park, Wyoming from 1965 to 1973: US. Geol. Survey Open-File Re- port 75—25., 53 p. US. Geological Survey, 1974, Water Resources Data for Montana; Part 1, Surface Water Records 1973: US. Geological Survey, 278 p. Walter, M..R., Bauld, J ., and Brock, T. D., 1972, Siliceous algal and bacterial stromatolites in hot spring and geyser effluents of Yel- lowstone National Park: Sci., v. 173, p. 402—405. Weissberg, B. G., 1969, Gold-silver ore-grade precipitates from New Zealand thermal waters: Econ. Geology, v. 64, p. 95— 108. White, D. E., 1957, Thermal waters of volcanic origin: Geol. Soc. America Bull., v. 68, p.’1637— 1658. 1967, Mercury and base-metal deposits with associated ther- mal and mineral waters, in Barnes, H. L., ed, Geochemistry of hydrothermal ore deposits: New York, Holt, Rinehart, and Winston, p. 575—631. White, D. E., Fournier, R. 0., Muffler, L. J. P., and Truesdell, A. H., 1975, Physical results of research drilling in thermal areas of Yellowstone National Park, Wyoming: US Geological Survey Prof. Paper 892, 70 p. White, D. E., Muffler, L. J. P., and Truesdell, A. H., 1971, Vapor- dominated hydrothermal systems compared with hot-water sys- tems: Econ. Geology, v. 66, p. 75—97. Wollast, R., MacKenzie, F. T., and Bricker, O. P., 1968, Exper- imental precipitation and genesis of sepiolite at earth-surface conditions: Am. Mineralogist, v. 53, p. 1645— 1662. Yanagisawa, M., Takeuchi, T., and Suzuku, M., 1973, Flameless atomic absorption of antimony: Anal. Chim. Acta, v. 64, p. 381— 386. Zeikus, J. G., and Brock, T. D., 1972, Effects of thermal additions from the Yellowstone Geyser Basins on the bacteriology of the Firehole River: Ecology, v. 53, p. 283—290. Klamath Falls Géothermal Area § Oregon GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—G , {ificumgm bEPARFMEEE’f} , j APR25 980 C I ' I‘BRAR/ . UN WERSiTY 0F (INFORM 1 , Hydrogeologic Appraisal of the Klamath Falls Geothermal Area, Oregon By EDWARD A. SAMMEL GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—G A description and evaluation ofa moderate—temperature hydrothermal convection system UNITED STATES GOVERNMENT PRINTING OFFICE. WASHINGTON : 1980 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Sammel, Edward A. Hydrogeologic appraisal of the Klamath Falls geothermal area, Oregon. (Geohydrology of geothermal systems) (Geological Survey professional paper ; 1044-G) Includes bibliographical references. Supt. of Docs. no.: I 19.1621044-G 1. Geothermal resources—Oregon—Klamath Falls region. 2. Water, Underground—0regon—K1amath Falls region. I. Title. II. Series. 111. Series: United States. Geological Survey. Professional paper ; 1044-G. GBll99.707525 553.7 80-607034 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock number 024-001-03300-3 CONTENTS Page Page Conversion of units of measurement ________________________ IV Chemistry of ground water—Continued Abstract __________________________________________________ G1 Reservoir temperatures estimated from chemical geother- Introduction ______________________________________________ 1 mometers __________________________________________ G25 Background and scope of study ________________________ 1 Isotopes in ground water ______________________________ 28 Methods of investigation ______________________________ 2 Oxygen-18 and deuterium isotopes __________________ 28 Acknowledgments ____________________________________ 2 Tritium concentrations ____________________________ 30 Numbering system for wells and springs ________________ 2 Isotope mixing models ____________________________ 30 Units of measurement ________________________________ 4 Reservoir temperature estimated from oxygen isotopes Regional setting __________________________________________ 4 in dissolved sulfate ______________________________ 32 Regional geology and geothermal activity ______________ 4 Temperature distribution and heat flow ____________________ 33 Climate and drainage __________________________________ 8 Areal distribution of temperature ______________________ 33 Precipitation and temperature ______________________ 8 Vertical distribution of temperature ____________________ 33 Evapotranspiration ________________________________ 9 Conductive heat flow __________________________________ 37 Surface water ____________________________________ 10 Convective discharge of heat __________________________ 38 Occurrence and movement of ground water __________________ 10 Interpretations and conclusions ____________________________ 38 The shallow reservoir __________________________________ 10 Review of evidence relating to the geothermal resource -_ 38 Deeper basaltic aquifers ______________________________ 12 Geologic evidence __________________________________ 38 Movement and discharge of ground water ______________ 12 Geophysical evidence ______________________________ 39 Hydraulic properties of the rocks ______________________ 14 Chemical evidence ________________________________ 40 Estimates from specific capacities of wells __________ 14 Conceptual model of the geothermal system ____________ 41 Estimates from pumping tests in wells ______________ 15 Recharge to the geothermal system ____________________ 42 Chemistry of ground water ________________________________ 16 Magnitude of the geothermal resource __________________ 43 General chemical characteristics ________________________ 16 References cited __________________________________________ 44 Relations among dissolved constituents ________________ 20 ILLUSTRATIONS [Plates are in pocket] PLATE 1 Map showing altitudes of water levels and lines of equal altitude on the water table. 2. Map showing specific electrical conductance and Stiff diagrams of the principal dissolved constituents in ground water. 3. Map showing temperature of ground water, depth of wells, and lines of equal temperature. 4 Graph showing temperature profiles measured in unused wells and heat-flow holes. Page FIGURE 1. Index map showing the Klamath Falls area, locations of "hot-well” area, National Weather Service stations, and stream-gaging stations ____________________________________________________________________________________ G3 2. Diagram showing subdivision of township and range sections for well-numbering system __________________________ 4 3. Map showing areal geology of the Klamath Falls area __________________________________________________________ 6 4. Diagram showing profiles of wells at the Oregon Institute of Technology and Presbyterian Intercommunity Hospital, showing discordant water levels and assumed relations to faults ______________________________________________ 13 5. Graph showing percent reacting values of major ions in ground water ____________________________________________ l7 6. Graphs showing log molal ratios of dissolved constituents in ground waters arranged in order of increasing chloride concentration ________--_______-_______-_-______-_-________--_______-__________-_______'_ __________________ 23 7 Graphs showing silica and chloride concentrations in ground water and their relations to observed water temperature 24 8. Graph showing temperature of ground water versus log silica concentration ______________________________________ 27 9. Graph showing temperature of ground water versus log (Na/K) + Blog (VIE/Na) __________________________________ 28 10. Graph of 80xygen—18 versus Sdeuterium in ground water and lake water __________________________________________ 29 11. Graph of chloride concentration versus 80xygen-18 and Bdeuterium in ground water and lake water ________________ 30 12. Graph showing mixing proportions of thermal and meteoric waters based on 80xygen-18 and 5deuterium relations and hypothetical geothermal reservoir waters ______ ____________________________________________________________ 30 13. Geologic sketch map of the Klamath Falls area (after Peterson and McIntyre, 1970), showing locations and categories of wells in which temperature profiles were obtained. __________________________________________________________ 34 Ill IV TABLE 3" FPS°P°>195$WPPJN HI—l CONTENTS TABLES Mean annual precipitation and temperature measured at National Weather Service stations in the vicinity of Klamath Falls ____________________________________________________________________________________________________ Evapotranspiration in the Upper Klamath River basin, Oregon and California ____________________________________ Summary of ground-water discharge and use in the Klamath Falls area __________________________________________ Specific capacities of wells and estimates of transmissivity in aquifers of the Klamath Falls area ____________________ Chemical analyses of water from wells and springs ______________________________________________________________ Ratios of dissolved constituents in ground water ________________________________________________________________ Estimates of reservoir temperature based on chemical relations in ground water __________________________________ Analyses of the isotopes oxygen-18, deuterium, and tritium in ground water and Klamath Lake water ______________ Calculated silica and chloride concentrations and temperatures in the geothermal reservoir ________________________ Data related to temperature profiles obtained in unused wells and heat-flow holes Mass and thermal-energy discharge in ground water from thermal springs and wells in the Klamath Falls area ____ CONVERSION OF UNITS OF MEASUREMENT Used in this report inch (in) inch (in) foot (ft) mile square foot (ftz) acre square mile (mi?) acre-foot (acre-ft) cubic mile (mia) gallons per minute (gal min“) million gallons per day (Mgal d") feet per mile (ft mi") feet uared per day «$11 ,) gallons per minute per foot (gal min“ ft") degrees Celsius (°C) degrees Celsius per kilometer (°C km") Multiply by Length 25.40 2.540 0.3048 1.609 Area 0.0929 0.4048 2.590 Volume 1.233 X 10‘3 4.168 Flow 0.06308 3.785 x 103 Topographic gradient 0.1894 Transmissivity 0.0929 Specific capacity (wells) 0.2070 Temperature 1.8 (and add 32) Thermal gradient .05486 To obtain millimeter (mm) centimeter (cm) meter (m) kilometer (km) square meter (m2) hectare (ha) square kilometer (kmz) cubic hectometer (hma) cubic kilometer (kma) liters per second L 5“) cubic meters per day (m3 d“) meters per kilometer (m km“) meters squared per day (m2 d“) liter per second per meter (L s“ m“) degrees Fahrenheit (°F) degrees Fahrenheit per undred feet (°F/ 100 ft) Used in this report millicalories per centimeter per second per degree Celsius (mcal cm‘1 5‘1 °C“) calories per cubic centimeter per degree Celsius (cal cm‘3 °C“) microcalories per square centimeter per second (Heat flow unit = HFU) (acal cm‘2 s“) CONTENTS Multiply by Thermal conductivity 0.4184 Specific heat (volumetric) 4.184 X 106 Heat flow 4.184 x 10‘2 To obtain watts per meter per degree Celsius (W m‘ °C") joules per cubic meter per degree Celsius (J 111‘3 °C“) watts per square meter (W m‘”) GEOHYDROLOGY OF GEOTHERMAL SYSTEMS HYDROGEOLOGIC APPRAISAL OF THE KLAMATH FALLS GEOTHERMAL AREA, OREGON BY EDWARD A. SAMMEL ABSTRACT Geothermal phenomena observed in the vicinity of Klamath Falls include hot springs with temperatures that approach 96°C (the ap- proximate boiling temperature for the altitude), steam and water wells with temperatures that exceed 130°C, and hundreds of warm- water wells with temperatures ranging from 20° to 40°C. Although warm water is produced from wells scattered throughout most of the 350-square-mile area studied, waters with temperatures exceeding 60°C are confined to three relatively restricted areas, the northeast part of the City of Klamath Falls, Olene Gap, and the southwest flank of the Klamath Hills. The hot waters are located near, and are presumably related to, major fault and fracture zones of the Basin and Range type. The displaced crustal blocks are composed of basaltic flow rocks and pyroclastics of Miocene to Pleistocene age, and of sedimentary depo- sits and basalt flows of the Yonna Formation of Pliocene age. Dip-slip movement along the high-angle faults may have been as much as 6,000 feet at places. Average annual precipitation in 7,300 square miles of the upper Klamath River basin surrounding the study area is about 18.2 inches, of which between 12 and 14 inches is estimated to be lost through evapotranspiration. Ground water of local meteoric origin moves through the shallow sedimentary deposits and volcanic rocks at relatively slow rates. Within the older basaltic rocks of the area, hydraulic conductivities are greater than in the shallow sediments, and ground water may move relatively freely parallel to the northwest-southeast structural trend. A small amount of ground water, perhaps 100,000 acre-feet per year, leaves the area in flow toward the southwest, but much of the ground water is discharged as evapotranspiration within the basin. The local meteoric water that is assumed to be the source of the thermal water in the area has low concentrations of dissolved-solids in which calcium and bicarbonate are the dominant ions. During its passage through the geothermal reservoir, concentrations of dis- solved solids increase to about 900 milligrams per liter, and sodium and sulfate become the dominant ions. Chloride concentrations re- main relatively low, and silica concentrations increase from an aver- age of about 35 milligrams per liter to about 100 milligrams per liter. The evidence from cation ratios, silica concentrations, and oxygen and deuterium isotopes in the hot waters indicates that tempera- tures in the near-surface geothermal reservoir are relatively low. The estimated minimum reservoir temperature, based on the quartz geothermometer and mixing models, is 150°C. In contrast to these indications, the sulfate—oxygen isotope geothermometer indicates that, at some time in their history, the thermal waters have been exposed to temperatures approaching 200°C, possibly in a deeper reservoir. Temperature distributions and heat flows in the shallow rocks of the area are strongly influenced by convective flow of water. Most observed temperature gradients are unreliable indicators of depths to the geothermal reservoir. Heat flow in the vicinity of the geother- mal areas has not been determined, but evidence from temperature profiles suggests that heat flow in the Lower Klamath Lake basin is about 1.4 microcalories per square centimeter per second (1.4 Heat flow units), a value that is near the minimum expected for the Basin and Range province. The net thermal flux discharged from springs and wells in the area is estimated to be on the order of 2 X 106 calories per second. Dis- charge by thermal waters into the shallow ground-water system be- neath land surface may be many times this amount. Reportedly, at present only about 1.4 X 106 calories per second (6 megawatts) of thermal energy is beneficially used in the area. A conceptual model of the geothermal system at Klamath Falls suggests that most of the observed phenomena result from transport of heat in a convective hot-water system closely related to the re- gional fault system. Temperatures at shallow depths are elevated both by convective transport and by the blanketing effect of rocks of low thermal conductivity. Circulation of meteoric water to a depth of 15,000 feet could account for the estimated temperatures in the thermal reservoir, assuming conductive heat flow and a temperature gradient of 30°C per kilometer in the shallow crustal rocks. Circula- tion to shallower depths may be sufficient to warm the water to the required temperatures under the more probable conditions of convec- tive transport of heat and the insulating effect of overlying sedi- ments. Heat content in the shallow hot-water system (<10,000 feet depth) is probably in the range 15 x 10‘“ to 190 X IO'Rjoules, but is likely to be in the lower part of this range. This thermal energy may be stored in two or more separate reservoirs having a total volume not much greater than 12 cubic miles. INTRODUCTION BACKGROUND AND SCOPE OF STUDY Geothermal phenomena in the vicinity of Klamath Falls consist of at least 7 thermal artesian springs and about 800 thermal wells. More than 400 of the wells are used for space heating, a use that began in the area about 50 years ago and is increasing rapidly at the present time. The amount of geothermal energy recov- ered from these wells and applied to nonelectrical use in the Klamath Falls area is probably one of the largest such applications in the United States. G1 G2 Rapid growth in the use of geothermal energy at Klamath Falls and plans for expanding this use have led in recent years to a need for increased knowledge of the physical properties of the geothermal system. As one of the steps toward meeting this need, the investi- gation described in this report was begun by the US. Geological Survey in 1973. General objectives of the study were to bring together the large amount of exist- ing data and to add new information in such areas of deficiency as chemical characteristics, geothermal gra- dients, and heat flow. An open-file report containing data and preliminary conclusions from the study has been released (Sammel, 1976). The area of intensive study for this report covers approximately 350 square miles of the Klamath River drainage basin (fig. 1), and includes all known thermal anomalies in the immediate vicinity of Klamath Falls. The major area of thermal use, the so-called “hot-well area” of northeastern Klamath Falls (fig. 1), was not studied in detail, however. Because of time constraints, an adequate study of this area of intensive use would have severely limited the overall view that might otherwise be gained in a wider coverage of the geoth- ermal areas. Staff members of the Geo-Heat Utiliza— tion Center, Oregon Institute of Technology (OIT), are now investigating the urban geothermal area, sup- ported in part by a grant from the Geothermal Re- search Program of the US. Geological Survey. METHODS OF INVESTIGATION This report is based to a large extent on information obtained in an inventory of more than 300 wells and 20 springs in the Klamath Falls area. In most of these wells and springs measurements were made, as appro- priate, of depth, water level, discharge, temperature, and specific electrical conductance. Driller’s logs are available for about two-thirds of the wells. Chemical analyses were obtained for 35 samples of water from wells and springs. Concentrations of car- bonate, bicarbonate, and pH were determined in the field and the remainder of the analyses were performed in Geological Survey laboratories. Relatively complete analyses of water from 22 additional sources were ob- tained from owners or from published data. Water was also collected from 24 sources for analysis of stable isotopes and tritium. A small amount of detailed geologic mapping was done, but in general a map by Peterson and McIntyre (1970) provided adequate geologic information for this reconnaissance. Four shallow hydrologic test holes were drilled in the Lower Klamath Lake valley to depths ranging from 223 to 463 feet. Two 600-foot holes were drilled for heat-flow information. Temperature profiles were ob- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS tained by means of a thermistor probe in the 6 drill holes as well as in 41 unused water wells. Additional geophysical data obtained during this in- vestigation include gravity measurements at 103 sta- tions in the Upper and Lower Klamath Lake basins and thermal infrared imagery obtained in aircraft flights over about 300 square miles of the area. The gravity measurements are described in the open-file report by Sammel (197 6). The infrared imagery has not been completely processed and is not discussed in this report. Preliminary visual examination of computer- generated contour plots based on the imagery suggests that properly calibrated high-resolution infrared im- agery may be useful as an exploration tool in areas of surficial convective thermal phenomena. ACKNOWLEDGMENTS The author wishes to acknowledge the assistance of many colleagues in the Geological Survey who pro- vided services, advice, and inspiration. Aid received in the fields of water chemistry, isotope analysis, heat flow, and gravity investigations was particularly help- ful. Gratitude is expressed to numerous other individ- uals and organizations who provided assistance and facilities during the study; special mention is made of the US. Bureau of Reclamation, Klamath Falls Proj- ect, for making facilities available and for help in ways too numerous to describe; the US. Forest Service, Winema Forest Headquarters, for facilities and aid; Jack Hitt and Professors J. W. Lund, G. G. Culver, P. J. Lienau, and L. S. Svanevik at the Oregon Institute of Technology for facilities and aid during a study of wells at OIT; J. B. Koenig and M. C. Gardner, Geothermex Co., for making data available and for providing initial insights into the geothermal system at Klamath Falls; and the hundreds of well owners who allowed us to visit, measure, and sample wells or, in some instances, to drill test holes on their property. Finally, special thanks are due to N. E. Voegtly who collected and com- piled many of the data for this report. NUMBERING SYSTEM FOR WELLS AND SPRINGS In this report, wells and springs are numbered ac- cording to a system based on their location within townships, ranges, and sections referred to the Willamette Baseline and Meridian. Thus, the number 39/9—21 bcdl specifies a well in Township 39 South, Range 9 East, and Section 21. The letters which follow indicate respectively the 1/4, 1/16, and 1/64 quadrants within the section according to the scheme illustrated in figure 2. A number following the letters indicates that more than one well was inventoried in the same 1/64 section 42°30‘ 122°00' 121°45' HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON I40 To Chiloquln l 6 ml. Modoc Point \hiteline Reservoir 8w 4M L4’T5 V 444 5y % “Hot-well" area A Stream-gaging station 0 Climatological station 121°30’ 42°15' —- 9 I Hogback mm. 6198, H KLAMATH FALLS Y Round Altamont V40-4 Lake LLEy HO \' 39 Diane Gap 3.1 Q 0 Olene \ ‘4 66 ”$2; 'i i g 1 '( Q~ < KMQSISY A“ Nuss - 6275- Fleld 4? Lake [0" Canal p05 V4 Midland ” L K Keno ___ a ,6, LE), \ ‘(V . ° V ‘ LOWER » Stukel Mtn ‘9' Spring I . ' ‘i' Lake 6525 ' ' KLAMA TH 42 IHamaker Mtn. ~~~~~ ‘7 6598’ ‘32, 471; 39 9’ \e, LAKEBED ‘7 ‘ ‘f ’57 O Worden I ((5. L4; \gf‘; Merrill I ‘9 IE. 42°00' — I: —* I I J SCALE 0 1 2 3 4 SMIkES O 1 2 3 4 SKILOMETERS FIGURE 1.-The Klamath Falls area, locations of “hot-well” area, National Weather Service stations, and stream- gaging stations. G3 G4 \ b a ‘\ \ C ‘ WELL39/9—21bcd1 c d / ¢ \ \ \ \ \ \\ \\ \\ \\ \ \ ‘\ \\ \\ b\\ a \ \ \\ \ )o a \ \ \\ C\ d \ \ 21 c d FIGURE 2.——Subdivision of township and range sections for well— numbering system. quadrant; such wells are numbered serially in chronological order of inventory. Each well number, therefore, specifies a location within a quadrant ap- proximately 600 feet on a side. Most wells shown on the accompanying maps are located with a probable error of less than 300 feet. UNITS 0F MEASUREMENT Most of the data obtained during this study were measured in the inch-pound system of units. For reasons of accuracy, as well as the convenience of most readers, these units have been retained in this report. A table of multipliers is provided for use in converting the inch-pound units to SI (metric) units. Field measurements of water temperature were ob- tained in units of degrees Celsius (°C), and for the sake of consistency, the few reported temperatures have been converted to this system. Air temperatures re- ported by the National Weather Service are in °F. In the fields of geothermal gradients and heat flow, most current scientific reports use SI units or other estab- lished metric equivalents. In order to retain tempera- ture measurements in degrees Celsius and provide a convenient basis for comparison of calculated results, sections of this report dealing with these subjects con- tain commonly used metric units rather than inch- pound units. Thus, geothermal gradients are reported in degrees Celsius per kilometer, thermal conduc- tivities are in milicalories per centimeter per second per degree Celsius, and heat flow is in microcalories GEOHYDROLOGY OF GEOTHERMAL SYSTEMS per square centimeter per second (heat flow units). Conversion factors for these units are provided in the table of units. The metric unit, kilometer, is used in the section, “Magnitude of the Geothermal Resource,” in order to facilitate comparison of estimates made there with previously published estimates. REGIONAL SETTING REGIONAL GEOLOGY AND GEOTl-IERMAL ACTIVITY Geologic structures in the Klamath Falls area are typical of the Basin and Range province (Gilbert, 1928). Large grabens form the Upper and Lower Klamath Lake basins; these basins are flanked by uplifted blocks bounded by steeply-dipping normal faults. The regional structure has a northwest- southeast alinement, but numerous faults have north- south strikes, and a few strike northeast-southwest. Figure 3 shows geologic units and major faults in the Klamath Falls area as mapped by Peterson and McIn- tyre (1970). The oldest rocks exposed in the area are of latest Tertiary age; they include lacustrine and fluvial tuffaceous siltstone, sandstone, ashy diatomite, basal- tic tuff and breccia, and a few thin basalt flows. These rocks were grouped by Newcomb (1958) in the Yonna Formation of Pliocene age. Later investigators have been reluctant to use this formation name because of uncertainties regarding the stratigraphic boundaries of the formation and the specific rock sequences that should be included. The name is used in this report for convenience in referring to groups of rocks of the type and age described by Newcomb. Rocks of the Yonna Formation underlie the lower slopes of many upraised fault blocks in the area and occur at depth beneath most of the lacustrine and allu- vial sediments in the structural basins. Ashy diato- mites of the formation are conspicuous in roadcuts and quarries and are penetrated by many wells. The most prevalent rocks of the Yonna Formation are thick, massively bedded, coarse-grained palagonit- ic sediments and pyroclastic rocks which are not clearly related to specific eruptive centers. Examples of these rocks are exposed in two large fault blocks south and east of Klamath Falls, Stukel and Hogback Mountains. Interbedded with the volcanic sediments are thin basaltic lavas which are generally dense, black, glassy, and vesicular. Some, such as those ex- posed at the Klamath Rock Products quarry (T39, R8, Sec. 11), are brecciated and altered owing to extrusion into water or wet diatomaceous ooze (Peterson and McIntyre, 1970). These flows probably issued from nearby fissures which were subsequently covered. The occurrence of local volcanic activity is also indicated by a few scattered maar and tuff-ring deposits. HYDROGEOLOGIC APPRAISAL ()F KLAMATH FALLS GEOTHERMAL AREA, OREGON Massive diatomites of the Yonna Formation have counterparts in Butte Valley, California, about 20 miles southwest of Klamath Falls, where they underlie Pliocene volcanic rocks of the High Cascades (Moore, 1937; Wood, 1960). Near Klamath Falls, diatomites are interbedded with Pliocene volcanic rocks (Meyers and Newcomb, 1952). The entire area over which diato- mites are found is not known to have been covered by a continuous water body, but the lake, or lakes, must have been extensive. Thickness of the Yonna Formation beneath the val- ley floors is largely unknown, but in adjacent ridges it ranges from a few feet to at least 850 feet. The maximum thickness is inferred from the driller’s log of a well at the Presbyterian Intercommunity Hospital, located about 1 mile north of Klamath Falls. Rocks older than the Yonna Formation do not crop out in the vicinity of Klamath Falls. The characteris- tics of these older rocks are known, however, from drillers’ logs, and the probable ages of the rocks may be inferred from geologic mapping in adjacent areas. A few miles east of Klamath Falls in Swan Lake, Yonna, and Poe Valleys, the Yonna Formation lies uncon- formably on olivine basaltic rocks of probable Pliocene age (Peterson and McIntyre, 1970; Leonard and Harris, 1974). The basalt is blue-gray to gray-brown, vesicular, and columnarly jointed. Individual flows are as much as 30 feet thick and are interbedded with tuff and scoriaceous flow breccia (Leonard and Harris, 1974). Maximum observed thickness of the unit is about 300 feet in the Yonna Valley. Pliocene basaltic flows also underlie extensive areas of the Warner Range, about 80 miles east of Klamath Falls. They dip westward for the most part and prob- ably underlie' tuffs and lacustrine deposits of the Sprague River valley about 35 miles northeast of Klamath Falls (Peterson and McIntyre, 1970). These large volumes of basaltic lavas and pyroclastics were extruded onto relatively level surfaces in the region surrounding Klamath Falls during late Pliocene time. Similar flows probably also covered the Klamath Falls area. They are recorded in drillers’ logs as massive gray to brown vesicular basalt and tuffaceous sediment that underlie the Yonna Formation at places such as the Oregon Institute of Technology, about 2 miles north of Klamath Falls, at the Oregon Water Corpora- tion wells in Klamath Falls, and possibly at a 1,300 foot well on Hamaker Mountain south of Keno. Similar rocks crop out in the Klamath River canyon southwest of Klamath Falls where their thickness is at least 800 feet (Newcomb and Hart, 1958). The Pliocene olivine basalts described above prob- ably correlate with extensive basaltic lavas that occur in adjacent areas of northern California (Wood, 1960). I G5 These rocks are older basalts of the High Cascades mapped in Butte Valley, California (Wood, 1960), and in the Macdoel Quadrangle, Calif. (Williams, 1949). According to Williams, the large shield volcanoes of the High Cascades that dominate the eastern margin of the Cascade Range were probably extruded from north—south trending fissures during a period extend- ing from late Pliocene to Holocene time. The flows are composed largely of olivine-augite basalt, although andesite flows and related pyroclastic rocks are com- monly interbedded. The major flows dip eastward and are buried by younger basalts and pyroclastic rocks in the hills immediately west of Klamath Falls. West of the High Cascade volcanoes, a belt of vol- canic rocks from 30 to 40 miles wide forms the Western Cascades. These rocks, consisting largely of pyroxene andesites, olivine basalts, and rhyolite tuffs and lava domes, were uplifted, folded, and deeply dissected prior to the High Cascade eruptions. With ages ranging from Eocene to Pliocene (Williams, 1949), the rocks are equivalent to the Clarno Formation, John Day Forma- tion, Columbia River Basalt Group, and the Cedarville Series of Russell (1928) in the Warner Range of north- eastern California. At places, such as Oregon-Calfornia border west of Klamath Falls, the thickness of the Western Cascade rocks is at least 12,000 feet (Williams, 1949, p. 20). Although these rocks dip to the east beneath the High Cascades, they do not crop out in the vicinity of Klamath Falls and have not been penetrated by wells there. Depth of occurrence and characteristics of pre- Tertiary rocks in the Klamath Falls area are unknown. The oldest rocks exposed in adjacent regions are metacherts and quartzites of probable Paleozoic age intruded by quartz monzonites of Jurassic age near the foot of Mount Shasta (Williams, 1949, p. 14). The Chico Formation (Cretaceous), consisting of marine arkosic sandstone, conglomerate, and shale, underlies parts of the Macdoel Quadrangle in northern California. Hot springs that occur in the Macdoel Quadrangle about 30 miles southwest of Klamath Falls have high chloride concentrations that may indicate the presence of the Chico Formation at this location (Williams, 1949, p. 55). It is possible, therefore, that marine sedimentary rocks also underlie the Klamath Falls area, although no evidence of the presence of such rocks is known. The structural development of the Klamath Falls area may have begun with normal faulting as early as Pliocene time (Peterson and McIntyre, 1970, p. 28). Subsequent to the deposition of the Yonna sediments and contemporaneous basalt flows, the entire region was apparently folded into broad anticlines and synclines with generally north-south axial trends. The GEOHYDROLOGY ()F GEOTHERMAL SYSTEMS G6 ’4 n... p bx \\ V7/////////////4 4 L .. , / V. , z? \ ?fl////V\fl\l%&r ,,.,/// //M.w./ \\ \ \ H 9% \ a /.s % 1/ ‘/ % // .1 mu I “.x/fix/x/fl // // 42°00 ' SCALE 0 1 2 3 4 SMILES 0 1 2 3 4 5 KILOMETEFIS Areal geology of the Klamath Falis area. (Modified from Peterson and McIntyre, 1970.) FIGURE 3. HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON G7 CORRELATION OF MAP UNITS Holocene Pleistocene Pleistocene OUATERNARY Pliocene and }TERTIARY AND } l } Pliocene } TERTIARY DESCRIPTION OF MAP UNITS Qal Alluvial deposits including diatomite and peat Fluvial terrace and lacustrine deposits Basalt flows Basalt flows, pyroclastic breccia, and agglomerate, of uncertain stratigraphic relationships. May in part be contemporary with Tst and Ob Basaltic eruptive centers 2 Basaltic cinder cones § Lacustrine and fluviatile tuffaceous siltstone, sandstone, ashy diato— § \ mite, basaltic tuff and breccia, few thin basalt flows (=Yonna Formation) Maar and tuff—ring deposits —-——- Contact — Dashed where approximate —s———‘- Fault — Dashed where approximate. Ball on downthrown side Note: Description of map units and map symbols are quoted from Peterson and McIntyre, 1970 FIGURE 3.—Continued. Sycan Marsh-Bly-Beatty area lay in a broad syncline, The volcanic activity and faulting that produced the and the present Summer Lake-Goose Lake graben was present topography in the Klamath Falls area occurred an anticline. The Klamath Falls area may also have mostly during. the Pleistocene. To the west, new vol- been an anticline whose crest subsequently dropped to canic vents continued to form in the High Cascades, form the present graben (Peterson and McIntyre, 1970, and stratovolcanoes such as Mts. Shasta, McLoughlin, p. 29). and Mazama were built. Near Klamath Falls, wide- G8 spread thin basalt flows were extruded over the Pliocene lacustrine sediments. Exposures of these ve- sicular, medium to dark gray, diktytaxitic Pleistocene basalts occur in the Klamath Hills, on Stukel Mountain south of Olene Gap, and on Hogback Mountain east of Klamath Falls. The cataclysmic eruption of Mount Mazama, about 6,700 years ago, produced thick deposits of ash and pumice that mantle the northern part of the area. Large volumes of Quaternary and Tertiary basalt flows and breccias of uncertain stratigraphic relation- ships underlie some of the highest ridges near Klamath Falls. These rocks may in part be contem- poraneous with the Pliocene Yonna Formation, but in part are Pleistocene. Normal faults cut all the above units, as well as the rocks of basalt eruptive centers that form some of the higher isolated hills. The present-day basins of the Klamath Falls area were probably established early in Pleistocene time as the result of tensional fracturing of the entire region. Fluvial-terrace and lacustrine sediments deposited in these basins are now found at altitudes as high as 200 feet above the present valley floor, presumably as the result of high lake levels in late Pleistocene time. Some of these deposits are deformed by faults, indicating that deformation continued into late Pleistocene time (Peterson and McIntyre, 1970, p. 19). It is not certain, therefore, that the present levels of these deposits rep- resent original altitudes of lake shorelines. Basin and Range faulting in the Klamath Falls area reached a climax during Pleistocene time. Movement was almost entirely dipslip along steeply-dipping (>60°) normal faults. The maximum displacement is at least 980 feet and may be as much as 6,000 feet if current interpretations of gravity data are accepted (D. L. Peterson, 1976). Some seismic activity has occurred in the region during historical times, but no historical surface breaks have been recorded (Couch and Lowell, 1971, p. 67). Thermal wells and springs are widespread in the Klamath Falls-area. In this report, a thermal water is defined as one having a temperature 220°C, or about 8°C above the average temperature of shallow ground water. Known hot waters, with temperatures greater than 60°C, are confined to the following three locations: (1) an area of about 2 mi2 along the northeast edge of the city; (2) a stream-cut gap in the hills about 8 miles southeast of the city at Olene Gap; and (3) the south- west flank of the Klamath Hills, about 12 miles south of the city. These three areas are referred to in sub- sequent sections of the report as the “principal geo- thermal areas.” The highest-temperature geothermal manifestations include several springs with temperatures that are GEOHYDROLOGY OF GEOTHERMAI. SYSTEMS near the boiling point (less than approximately 96°C), a few hot-water wells with subsurface temperatures that exceed the boiling point, and a few steam wells. One exceptionally hot well had a bottom-hole tempera- ture when drilled of about 140°C. Most of the wells having temperatures greater than 60°C are located in the city’s hot-well area. In hundreds of other wells scat- tered through the lower Klamath Falls basin, water temperatures range from just above normal for the re- gion (15°C) to about 40°C. The principal geothermal phenomena in the area are spatially related to fault zones. All waters known to have temperatures greater than 65°C, for example, occur within 1 mile of major faults. The association of the Klamath Falls geothermal phenomena with Basin and Range type faults suggests an anology with other geothermal areas of the Basin and Range province such as Warner Valley, Ore., Surprise Valley, Calif, and the many large fault-controlled hot-water systems in Nevada. Each of the principal geothermal areas in the vicinity of Klamath Falls contains extensive areas of silicified rocks as evidence that hydrothermal activity was formerly more widespread than at present. Some of these rocks, such as the silicified palagonite tuff that crops out above the Presbyterian Intercommunity Hospital, occur at altitudes as much as several hundred feet higher than any present surface flows of hot water. This evidence suggests that hydraulic heads in the major hydrothermal system were formerly much higher than at the present time or, alternatively, that hydrothermal activity preceded most of the displace- ment along the fault zones. The latter alternative ap- pears to be the more likely, but the question has not been resolved by this study. The ultimate source of heat for the geothermal wa- ters is not known. Peterson and McIntyre (1970, p. 37) refer to the presence of dikes, intercalated sill-like masses, bleached and silicified rocks, deposits of calcite and gypsum, and a halo of mercury mineralization as suggesting the existence of a hot igneous mass at depth. An alternative explanation, proposed by the au— thor, that involves only deep circulation of meteoric water is discussed below under the heading "Concep— tual Model of the Geothermal System.” CLIMATE AND DRAINAGE PRECIPI'I‘A'I'ION AND 'l‘EMPERA'l‘L'RE Mean annual precipitation at six stations in the vicinity of Klamath Falls is shown in table 1 for the period of record available. The range of annual precipi- tation at a given station is relatively great. The range between stations is also great with a low of 11.8 inches HYDROGEOLOGIC APPRAlSAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON G9 TABLE 1.—Mean annual precipitation and temperature measured at National Weather Service stations in the vicinity of Klamath Falls Altitude Mean annual Mean annual (ft above Period of precipitation temperature Station NGVD of 1929) record used (in) (°F) Chiloquin ......... 4,198 1948—738 18.13 42.9 Keno _______________ 4,116 1948—73b 19.49 ______ Kingsel Field _____ 4,085 1949-73 1 1.93 46.6 Klamat Falls _____ 4,098 1948—73 14.37 48.0 Merrill _____________ 4,080 1953—66h 1 1.80 ______ Rocky Point _______ 4,150 1967—73" 23.15 44.6 IPartial records for years 1964, 1968, and 1971 completed by estimates based on correla- tions with other stations. . . I’Record for 1964 estimated by correlation with other stations. cPartial record for 1968 completed by internal correlation. at Merrill and a high of 23.15 inches at Rocky Point. Outside the study area, extremes in this part of the Klamath River basin range from less than 10 inches at Tule Lake, Calif, to more than 60 inches on the high- est peaks north and west of Klamath Falls. Local orographic effects largely determine the rela- tive amounts of precipitation. Thus, of the moisture that reaches the area after passing over the Cascades on the westerly winds, most falls on the highlands and a relatively small amount reaches the valley floors. In most years, between 60 and 70 percent of the total precipitation falls during the months October through March. Most of this moisture is in the form of snow, although midwinter rains are frequent in the valley. Based on rainfall maps published by the Oregon State Water Resources Board (1971), average annual precipitation in the drainage basin above Klamath Falls, including subbasins of the Williamson River, Sprague River, and Upper Klamath Lake, is calculated to be 22 inches over 3,820 mi2. For the basin below Klamath Falls, precipitation is calculated to be about 15 inches over 3,497 mig. The area of this estimate includes the Lost River subbasin of Oregon and California but excludes the Meiss Lake closed basin southwest of Dorris, Calif. On the basis of the above calculations, the average annual precipitation in the four subbasins near Klamath Falls is 18.6 inches over 7,317 mi2. Mean annual air temperatures at stations near Klamath Falls are given in table 1. Stations with re- ported air temperatures are located at relatively low altitudes and mean annual temperatures at greater al- titudes are undoubtedly considerably lower than those reported at the weather stations. No estimate of mean annual air temperature has been made for this part of the basin. ' Measured temperatures of nonthermal ground water in shallow aquifers near Klamath Falls range from 7° to 19°C. Temperatures at individual sites may vary 5°C or more annually. The mean annual ground-water temperature at most places is probably several degrees higher than mean annual air temperatures, and the average temperature of shallow groundwater in the area is estimated to be about 12°C. EVAPO'I‘RANSPIRATION Estimates of evapotranspiration in the Upper Klamath River basin are given in table 2. Rates of evapotranspiration in the region are low owing to the relatively cool average temperatures and the low rain- fall. Minimum rates may occur in the high dry valleys of the eastern part of the basin, where evapotranspira- tion from sage and prairie vegetation is probably less than 1 foot per year (virtually the total precipitation). Maximum evapotranspiration may be as much as 3 feet per year from widespread marshes and ponds in the upper basin. For the upper Klamath Lake basin, including the three subbasins above Klamath Falls (3,802 mi2), evapotranspiration is estimated to be 1.1 feet per year. This estimate includes the effects of an estimated ground-water outflow from the upper Klamath Lake basin of 25,000 acre-feet per year (Illian, 1970). If, as a lower limit, ground-water outflow is assumed to be one-half Illian’s estimate, the calculated evapotranspi- ration is 1.2 feet per year. For the Lost River subbasin in Oregon and Califor- nia (3,497 miz), evapotranspiration is calculated to be 15 inches per year. The calculation includes Illian’s estimate that ground-water flow from the upper Klamath Lake basin into the lower basin is 200,000 TABLE 2.—Evapotranspiration in the Upper Klamath River basin, Oregon and California Source of Type of area Area Evapotranspiration estimate (mi!) (ft yr") High, dry valleys—sage and prairie grass. ____________________ <1 A Marshland ______________________ 56 3.6 O 100 3 L Low valleys—irrigated crops "“422 '1.3 0 Upper Klamath Lake (altitude 4,139 ft) ______________ ‘125 31/2—4 0 Clear Lake (Calif) (altitude 4,470 ft) ______________ 69 217/ —31/2 0 Upper Klamath River basin: Sprague, Williamson, and Upper Klamath Lake subbasins ________________ 3,820 2 1.1 A Lost River subbasin (Ore. and Calif.) ______________ 3,497 3 1.3 A A. Author (this report). 0. Oregon State Water Resources Board (1971) L. Leonard and Harris (1974) ‘Range 100 to 140 miz, depending on seasonal change in lake stage. 2Calculated from precipitation data (table 1), published streamflow records (US. Geologi- cal Survey), and unpublished canal-flow records (U.S. Bureau of Reclamation). Estimates of ground-water outflow from the Upper Klamath Lake and Sprague River subbasins (Illian, 1970) have been included in the calculation. 3Calculated from precipitation data (table 1) and published streamflow records for the Klamath River at Keno and Boyle Power Plant (US. Geological Survey). Estimates of ground-water flow into the Lost River subbasin (Illian, 1970) have been included. along with the author’s estimate of ground-water outflow below Keno. G10 acre-ft per year and the author’s estimate that about 100,000 acre—ft of water leaves the lower basin as ground-water flow. It is assumed that ground-water flow from the south end of the basin in California is negligible and that ground-water discharge occurs al- most entirely through fractured rocks in the Klamath River gorge below Keno. SURFACE WATER Three major streams, the Sprague, Williamson, and Wood Rivers, drain the upper Klamath River basin above Klamath Falls. The Williamson and Wood Riv- ers flow into Upper Klamath Lake and the Sprague River is a tributary to the Williamson. Upper Klamath Lake is also fed by springs and seeps in or near the lake and by small tributaries on its periphery. The average annual flow from all the above sources for the period 1930—68 is estimated to about 1,400,000 acre-ft (Ore- gon State Water Resources Board, 1971, table 21). Upper Klamath Lake, the largest natural lake in Oregon, has a surface area that varies between 100 and 140 miz, depending on seasonal changes in lake stage, and a length of about 22 miles. Agency Lake, 61/2 miles in length, is appended to the north end of Upper Klamath Lake. The average depth of Upper Klamath Lake is only 8 feet, although depths reach 40 to 50 feet in a narrow trench between Eagle Ridge and Bare Is- land (Hubbard, 1970). Because of its shallow depth, regulated outflow, and generally warm water, Upper Klamath Lake is undergoing relatively rapid eutrophi- cation. Water from Upper Klamath Lake flows through a narrow strait on the western edge of the city of Klamath Falls, meanders through marshlands toward the southwest, and then, after passing over a dam near the town of Keno, leaves the area through a narrow gorge cut into bedrock. Below Klamath Falls, the stream becomes the Klamath River. Average annual flow at Keno for the period 1921—75 is 1,234,000 acre-ft (US. Geological Survey, 1976). More than 300,000 acre-ft of water is diverted annu- ally from Upper Klamath Lake and the Klamath River above Keno. This water is conveyed through three ca— nals, the “A,” the North, and the Ady, for use in irrigat- ing lands in the Lower Klamath Lake basin. The ca- nals are part of a major drainage and irrigation scheme built and, for the most part, operated by the US. Bureau of Reclamation (USBR). The project was begun in 1904 and is one of the oldest USBR projects in exis- tence. Southwest of Klamath Falls, the Lost River enters the Lower Klamath Lake valley through Olene Gap and flows into California. This stream has been incor- porated in the USBR irrigation project, and its fl0w is GEOHYDROLOGY OF GEOTHERMAL SYSTEMS now totally regulated. From a diversion dam north of Stukel Mountain, water is either diverted to the Klamath River or is returned to the Lost River channel for use downstream. About 116,000 acre-ft of water from the Lost River is annually distributed in the lower valley. The network of canals and laterals that now serve the USBR Klamath Project totals about 700 miles in length and irrigates about 200,000 acres of land in Oregon and California. Tailwater flows southward to a closed basin in the vicinity of Tulelake, Calif. The ex- tent of the water surface in this swamp area is closely regulated; excess water is pumped back to the north through the Klamath Straits Drain into the Klamath River. About 108,000 acre-ft of water is annually pumped out of the Tulelake area in addition to an av- erage net flow of 141,000 acre-ft diverted to the Klamath River from the Lost River through the Lost River Diversion Channel. In 1909 the Southern Pacific Railroad dike was com- pleted along the west side of the lower lake valley, separating the marshland in the valley from the Klamath River. Since this time the Lower Klamath Lake in Oregon and California has been essentially a closed basin. It is likely, in fact, that the Tulelake area acted as a sump for some thousands of years pre- viously. At present, surface water leaves the lower sump area only by evapotranspiration or by pumping to raise the water approximately 15 feet to its dis- charge point in the Klamath River. The above description of the USBR Klamath Project is an oversimplification of an extremely complex sys- tem. The average figures for inflow and outflow mean little in attempts to evaluate the hydrology of the Lower Klamath Lake valley. Only the major canals and drains are gaged, and the patterns of irrigation use and drainage are not readily available for quantitative evaluation. In the absence of detailed data, only the crudest generalizations are possible regarding con- sumptive use, evapotranspiration, or recharge to the ground-water reservoir. OCCURRENCE AND MOVEMENT OF GROUND WATER THE SHALLOW RESERVOIR The shallow ground-water reservoir in the Upper and Lower Klamath Lake valleys is, for the most part, a continuous, largely unconfined body to depths of more than 1,500 feet. Although few data are available in the highland areas, shallow cold ground water be- neath the upraised rock masses also appears to be gen- erally unconfined. Water levels in the valleys range from land surface to 10 or 15 feet below land surface, HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON and annual fluctuations of the water table are gen- erally less than 10 feet. Water levels in terrace deposits and alluvial fans are commonly 15 to 20 feet below land surface, whereas at places under the higher bed- rock ridges the water table is as much as several hundred feet below the surface. Fluctuations of the water table in the highlands are probably even smaller than those observed in the valleys. The generalized surface of the shallow nonthermal ground-water reservoir is shown on plate 1. The water table appears to be a continuous, fairly smooth surface which forms a subdued replica of the topography. At numerous places, however, water levels in individual wells show departures from the smooth surface. As ex- plained in the map text, some of these discrepancies result from errors of measurement or reporting. Many, however, reflect differences in geologic structure, stratigraphy, or well depth. Effects of ground-water pumping are believed to be negligible in most of the area. Compartmentation of the shallow aquifer by faults undoubtedly occurs both in the valley sediments and the bedrock ridges. These discontinuities may account for water-level discrepancies such as those shown by wells on the southwest flank of the Klamath Hills and the edge of the hills near Keno. The effects are minor, however, throughout most of the volume of Pleistocene and Holocene sediments. Compartmentation that oc- curs at places in the Tertiary rocks is discussed below in the section on “Deeper Basaltic Aquifers.” Wells in the valley-fill deposits commonly show a decrease in static water level with depth; that is, the deeper wells have lower water levels. The lower water levels in deep wells are the result of partial confine- ment of the deeper zones of the aquifer and a loss of hydraulic head between the recharge areas and the aquifers tapped by the wells. Hydraulic heads in wells are above land surface at several places in the area, indicating that the aquifer is confined at these places. Examples are wells north and east of Lake Ewauna (T38, R9, Sections 32 and 33), most wells north of Stukel Mountain (T39, R10, Sec- tions 28—34), and some wells in Lost River valley south and west of Stukel Mountain. Many of these wells, par- ticularly those north of Stukel Mountain, have large flows, but shut-in pressures indicate that artesian heads are generally less than 10 feet above land sur- face. In the southeastern part of Klamath Falls, hy- draulic heads in some hot wells formerly were more than 20 feet above land surface (J. W. Lund, oral com- mun., 1976). In recent years, however, the heads have declined to near land surface, possibly as the result of increased use of water from hot wells in the Klamath Falls geothermal area. G11 Few artesian wells are more than about a mile from upraised fault blocks, and it is assumed that the re- charge areas for all artesian wells are in the nearby ridges. Examination of drillers’ logs generally suggests that the water is confined in permeable zones between basalt flows, or in granular sediments overlain by clay and silt of the lake deposits. A few shallow artesian wells in the flood plain of the Lost River southeast of Klamath Falls (T39, R10, Sec- tions 17—20) penetrate fine black sand that is overlain by silt and clay. Drillers’ logs and data from a US. Geological Survey test hole indicate that the sand may be hydraulically connected to recharge areas in Hog- back Mountain through permeable deltaic or fan de- posits which form a ridge extending southward from the mountain. The relatively high water levels in this ridge are apparent on plate 1. Logs of wells in the area suggest that the permeable deposits just described are also connected to the permeable terrace deposits on the north face of Stukel Mountain, as well as to subsurface stream gravels'fan- ning out from Olene Gap. If so, the hot-water areas of Hogback Mountain, Stukel Mountain, and Olene Gap may be hydraulically connected through the valley sediments in a zone that extends to depths of 1,000 feet or more. Springs are not abundant in the Klamath Falls area, and most springs appear to have no connection with hotter water at depth. The 15 springs shown on plate 1 constitute nearly all the identifiable perennial springs that issue at land surface. Additional springs issue from the bottom of Upper Klamath Lake, Spring Lake, Nuss Lake, and probably other small ponds, but these have not been separately identified on the map. Most springs in the area produce cold water that is supplied by gravity flow from precipitation on adjacent recharge areas. Examples are Hummingbird, Shell Rock, Barkley, and Neubert Springs north of Klamath Falls, and an unnamed spring in the southwestern part of the area (T41, R8, Section 5cbb), all of which issue from basalt rocks near the base of fault scarps. Eagle Point Spring and two springs in Olene Gap are warm artesian springs whose waters are partly derived from the geothermal reservoir. Temperatures of thermal. springs in Upper Klamath Lake are not known but are assumed to be fairly low. Springs with temperatures near the boiling point formerly issued along the base of the slope below the hot-well area of Klamath Falls, but flows from these springs gradually diminished during the past several decades and the springs no longer exist. The general absence of springs on the flanks of upraised fault blocks confirms other evidence that water tables are relatively deep beneath the surfaces of G12 these blocks. Seeps occur following precipitation on the mountains, but these are mostly short lived and the ground water is only temporarily perched above them. DEEPER BASALTIC AQUIFERS A few deep wells located on ridges near Klamath Falls penetrate basaltic rocks of probable Tertiary age. Four such wells clustered at the north end of the city of Klamath Falls are among the deeper hot-water wells in the area. A diagram showing profiles of these wells, along with three nearby cooler wells (fig. 4), suggests possible relations to faults and illustrates the com- partmentation of the basaltic aquifers at this location. As the diagram illustrates, static water levels in the hot wells are accordant with each other but are at depths ranging from 170 to 200 feet below static levels in the three cooler wells and about 100 feet below the average level of Upper Klamath Lake. The water level in well OIT 3 is anomalously high, suggesting the presence of a local source of recharge and a high degree of compartmentation at this site. An aquifer test has shown that the four hot wells (left side of diagram, fig. 4) are hydraulically connected through permeable basalt rocks. (See section on “Hy- draulic Properties of the Rocks”) Water levels in these wells, therefore, probably represent the hydraulic head in the local geothermal system, which is assumed, from chemical evidence, to be largely sealed off from the shallow cold water. Similar relations apparently occur in the hot-well area of Klamath Falls, where water-level altitudes in most hot wells are also lower than water levels in nearby cool wells (Lund and others, 1974; J. W. Lund, oral commun., 1976). It is concluded, therefore, that hydraulic heads in the shallow bedrock aquifer are generally higher than those in the geothermal aquifer over an extensive area of the Klamath Falls thermal area, although local departures from this generaliza- tion may occur. One additional well that penetrates deep Tertiary basaltic rocks was inventoried in the Klamath Falls area. This is a 1,305-foot well on the north flank of Hamaker Mountain (T40, R7, Section 11ccc). The well produces water at a temperature of 16°C from fractured basalt flows and fragmental interflow zones at depths between 1,260 and 1,288 feet below land surface, cor- responding to altitudes of 3,600 to 3,572 feet above sea level. Static water level in the well was reported by the driller in 1959 to be 1,193 feet below land surface at an altitude of 3,670 feet above sea level. The reported water level is 400 feet lower than water levels in shal- low wells that penetrate bedrock at locations 2 to 3 miles to the northeast. GEOHYDROI.OGY ()F GEOTHERMAL SYSTEMS The low static head in the well at Hamaker Mountain is probably due to compartmentation of the aquifers by parallel faults that follow the regional trend through the area. A major fault has been mapped as striking through Bear Valley, just northeast of the well (Murray Gardner, written commun., 1975). The fault block is downthrown to the southwest, and numerous faults west of the main fault also have downthrown blocks to the southwest. Thus, water levels may be stepped down toward the southwest in partly isolated compartments created by the faults. Little information is available on aquifers in Ter- tiary basaltic rocks in the highlands immediately north and east of Klamath Falls. Tertiary rocks that occur outside the study area in the Sprague River, Swan Lake, Langell, Poe, and Yonna Valleys were de- scribed by Leonard and Harris (1974). Hydraulic characteristics of these rocks have been evaluated by Illian (1970), and they appear to be similar to those of the deeper basaltic rocks underlying the Klamath Falls area, as confirmed by the few data obtained in the present study. MOVEMENT AND DISCHARGE OF GROUND WATER The water-level contours on plate 1 show that ground water in shallow aquifers near Klamath Falls moves generally southward toward Tulelake in California. Some of the flow, perhaps on the order of 80,000 acre-ft per year, is intercepted by Lake Ewauna and the Klamath River, and leaves the basin as streamflow. Most of the remainder enters the closed basins of Lower Klamath Lake and the Tulelake sump. On the flanks of the northwest-southeast trending fault blocks, water-table gradients are steep, ranging from about 0.027 (140 ft mi“) on the southwest side of Stukel Mountain to about 0.038 (200 ft mi“) west of Lake Ewauna. These gradients reflect the steep surface topography as well as apparently low vertical per- meabilities within the fault blocks. South of Altamont, the moderate gradient of about 29 ft mi'1 suggests that rather permeable rocks extend into the valley from the southern flank of Hogback Mountain. Lithologic descriptions in drillers’ logs con- firm this suggestion. Similar or even smaller gradients were observed along the northwest edge of the valley between Klamath Falls and Keno. These gradients in- dicate that ground water moves freely parallel to the regional structure. In areas near the principal hot-water bearing faults, thermal waters may move upward, discharging both water and heat to the near-surface environment. In larger areas surrounding the faults, the flow of hot water appears to be predominantly lateral. One exam- ple is a region of several square miles bordering the HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON DISTANCE FROM WELL 1, G13 IN FEET ALONG SECTION PERPENOICULAR TO STRIKE OF FAULTS 1600 1400 1200 1000 800 600 400 200 0 1400 I I I 1 I I I l T I I I 460 I I I l I I I 7a #— 8 / Presb terian g? ’ "I “’Y E V #4 // 3 Hospital#6 ”#5 “L, ..— .l—v- f/ -440 E '— ,_._’—.4U‘_/——-“1-PHI‘—’—r-‘ E < I o1300— // MI 414245 // ‘ < = 4205:. _ Q < WY /‘5 1® ‘I 2 YM MW 4080 B !@ (<3 I— 414030 414030 40301.- .14035 Y/ 8 — o: / / g —4oo I- “-' 1200 — /E ‘1 m > / va : 5 I; o /s [E w p l: b < ~380 0 Lu = S p: :1 1.5 /; ® Level at Upper Lu 0 /§ 5 \- KIamIth Lake(4140) , D Lu 1 100 - 5 / "-360 O ‘5 5 ® Highest groom-water L” --I Y”'/ I / levels in Lam ‘3 ‘2’: I Klamath Lakebed _ 2‘ o J, / (4085) 340 2 fl _ <20°cp 9 <2: 1000 / I / 320 ’3: Lu / / (28°C) / (310) z > / Lu 0 (60°C) > on / —3oo o < 900 - / — m U: I < c: / Y~v Inferred base of the Yonna |_ E / / Formation —280 m I-I-I (91°C,) 1.. Static water level. u— E / Altitude in feet 2 z 800 - (90°C) , ~260 — —‘ (88°C) / E hrforatrons u; Lo“ \LI/ Bottom of cesing g D (91°C) Temperature of water in well ~240 I: I: I— ... 700 _ ,— / Land surface (approximate) _ _I _I < Strike of butts "climate” H.30°W. _220 < I I I I I I I I I I J 550 500 450 400 350 300 250 20 150 10 50 0 DISTANCE FROM WELL I, IN METERS ALONG SECTION PERPENOICULAR TO STRIKE OF FAULTS FIGURE 4.—Profiles of wells at the Oregon Institute of Technology and Presbyterian Intercommunity Hospital, showing discordant water levels and assumed relations to faults. south edge of the city in which warm water from the hot-well area is believed to flow southward in perme- able zones of layered basalt. Similar lateral flow may occur in the area west of the Klamath River and north of the Weyerhauser sawmill. The movement of thermal water is discussed further in the section on "Tempera- ture Distribution and Heat Flow.” Large amounts of ground water probably move from one subbasin to another within the Upper Klamath River basin. Illian (1970, p. 55) estimates that 200,000 acre-ft per year enters the Lost River subbasin from the Upper Klamath Lake subbasin, and that 150,000 acre-ft per year enters from the Sprague River sub- basin. Leonard and Harris (1974) conclude, however, that, although a significant amount of flow does occur, the available data do not permit an accurate quantita- tive evaluation of the interbasin movement. A crude estimate of the amount of ground—water flow from the Upper Klamath lake basin into Lower Klamath Lake has been made on the basis of data from G14 Illian and estimates derived from the present study. From data on wells penetrating Tertiary basaltic aquifers, Illian (1970, p. 13) calculated an average specific capacity of 145 gallons per minute per foot of drawdawn. A rough estimate of transmissivity may be obtained from values of specific capacity in efficient wells by multiplying gallons per minute per foot by 2000. On this basis, and converting to square feet per day, the transmissivity of the Tertiary basalt aquifers is estimated to be 40,000 ft2 d". Assuming that flow occurs through a 20-mile section extending along the 4,100 foot water-level contour on plate 1, and that the average gradient is 26 ft mi'l, the volume of ground- water flow into the Lower Klamath Lake basin is cal- culated to be about 175,000 acre-ft per year. The estimated transmissivity, based on data from highly permeable basalt aquifers in subbasins north and east of Klamath Falls, may be an overestimate for the Klamath Falls area. A small amount of pumping- test data from Klamath Falls suggests that the basalt aquifers underlying the city are less permeable than those in nearby basins. The assumed gradient is prob— ably within 25 percent of the actual value for the upper 1,000 to 1,500 feet of the aquifer, but may be significantly higher than the actual gradient in a deep, regional flow system. The 20-mile section is a maximum probable length. The estimate of ground- water flow, therefore, may be near the upper limit for the actual value. A similar approach has been used to estimate ground-water flow out of the Lower Klamath Lake basin, but the uncertainties involved are much greater. Southwest of Klamath Falls, in the direction of the regional topograpic gradient, ground-water conditions in the closely-spaced parallel fault blocks are virtually unknown, and it is not possible to make realistic esti- mates of either transmissivities or gradients in this direction. The figure of 100,000 acre-ft per year used in the section on “Evapotranspiration” is, therefore, little more than a guess at a probable maximum ground- water flow from the basin. Within the area studied for this report, ground water is discharged at land surface by pumping from wells, by artesian flow from wells and springs, and by seepage to the land surface or into lakes and streams. No at- tempt has been made to estimate the total discharge into lakes and streams, but rough estimates of the re- mainder of the ground-water discharge have been made. The results are shown in table 3. The total discharge to the land surface of ground water from springs and wells in the Klamath Falls area is estimated to be about 21 Mgal d", or about 22,000 acre-ft per year. This amount is an extremely small fraction of the water that enters the area either GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 3.——Summary of ground-water discharge and use in the Klamath Falls area Discharge and Percent of use total (Mgal d") Public Su 1 ': KlamathpFlalls, Altamont, Merrill, and Falcon Heights __________ 7.3 35 Irrigation .............................. 5.8 29 Industrial and commercial ,,,,,,,,,,,,,, 3.0 14 Spring discharge ______________________ 2.3 11 Space heatingz ________________________ 1.9 9 Domestic (individual wells) ,,,,,,,,,,,, .2 1 Stock watering ________________________ .1 <1 Total _____________________________ 2 0.6 'As of 1975. 2Interpreted in part from data in Lund, (1978). in streamflow or precipitation. (See section on "Climate and Drainage”) HYDRAULIC PROPERTIES OF THE ROCKS ESTIMATES FROM SPECIFIC CAPACITIES OF WELLS The specific capacities of efficient wells afford crude estimates of the transmissivity of the aquifers pene- trated by the wells. Specific capacity, expressed as yield in gallons per minute divided by the drawdown in feet, has been calculated from data provided by drillers and owners for about 170 wells in the Klamath Falls area. These data and estimated transmissivities de- rived from them are shown in table 4. Because of widely differing well efficiencies and depths of aquifer penetration, estimates based on specific capacity are undoubtedly imprecise. They are used in this report solely for purposes of comparison. In table 4, wells are assigned to one of several categories of rock type on the basis of drilling informa- tion. The rock types listed do not conform to the geologic units of Peterson and McIntyre (1970), but have been placed in generalized categories correspond- ing to commonly used terms in drillers’ logs. The as- signment of many wells to specific categories of water- yielding rocks is difficult, owing to uncertainties in drillers’ logs or in geologic mapping, and the calculated numbers in table 4 thus contain some uncertainty. Where the number of wells is small, reassignment of one high-yielding well would make a large difference in the average specific capacity for the rock unit. The estimated transmissivities seem reasonable, however, in the light of reported yields of wells and observations of rock characteristics in outcrops. The data of table 4 show that several types of rocks in the area provide similar low yields to wells. Wells in alluvial terraces, Holocene and Pleistocene lake depos- its, “older” lake deposits, and rocks of volcanic eruptive centers have specific capacities in the narrow range 1.6 HYDROGEOLOCIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON TABLE 4.—Specific capacities of wells and estimates of transmissivity in aquifers of the Klamath Falls area Average Estimated Water-yielding Number Average specific transmissivity rock unit of wells de th capacity of aquifer ( t) (gal min'lft”) (ftzdr') Alluvial terrace 4 279 1.8 500 deposits. Holocene and Pleis- 24 194 1.6 400 tocene lake deposits. Rocks of Quaternary 6 406 19 5,000 basaltic flows. Fragmental basaltic rocks of 27 187 2.1 600 Quaternary and Tertiary eruptive centers. Older lake deposits 21 334 1.9 500 (Yonna Formation?) Yonna Formation 70 450 29 8,000 Rocks of Quaternary 18 869 37 10,000* and Tertiary basaltic flows. *Aqu‘ifer’tests at three sites yielded values ranging from 11,000 to 20,000 ft2 d‘I at prevailing temperatures. (See text for temperature-corrected estimates.) to 2.1 gal min“ ft“. Although these rock types differ greatly in physical characteristics, it is not possible to distinguish one from the other on the basis of well per- formance. The distinction between the categories “Rocks of Quaternary basaltic flows” and “Fragmental basaltic rocks of Quaternary and Tertiary eruptive centers” in table 4 is especially arbitrary because of disagreements among published maps and uncertainties in drillers’ descriptions of the rocks. However, fragmental rocks of either Quaternary or Tertiary age appear to have simi- larly low permeabilities, whereas the basalt flows, re- gardless of age, are apparently much more permeable. As noted above, the permeable zones in the Yonna Formation are also associated with basalt flows. Rocks labeled “Older lake deposits” in table 4 appear to be partly consolidated clay, silt, and diatomite de- posits. These rocks have been identified only in drillers’ logs and they are distinguished from rocks of the Yonna Formation primarily on the basis of their low transmissivities. Most wells identified as penetrating the Yonna Formation have fairly high transmissivities which are apparently due to the presence of basalt flows and breccias. Thus, the presence or absence of volcanic rocks may be the only significant feature that distinguishes the category “Yonna Formation” from “Older lake deposits” in table 4. Drillers’s logs indicate that the moderately high transmissivities in basaltic rocks of the Klamath Falls area are probably due largely to fractures rather than to permeable interflow zones such as occur elsewhere in many basaltic sequences. Few of the drillers’ logs describe the exact nature of the water—bearing zones, however, and it is possible that “caving” or “soft lava” G15 rocks referred to by the drillers represent scoriacious or vesicular tops of flows which may also be relatively permeable. Two of the specific capacities listed in table 4 may be compared with those given by Illian (1970) for wells in the northern Klamath River basin. For wells in sedimentary aquifers, Illian reported an average specific capacity of 0.45 gal min“ ft“, a value only about 1/4 as great as the corresponding value in table 4. For the “Lower Basalt Aquifer,” Illian reports an average specific capacity of 145 gal min" ft‘1 a value almost 4 times greater than the value for Quaternary and Tertiary basaltic rocks in table 4. Although these discrepancies may represent real differences between aquifers, it seems more likely that they reflect statisti- cal discrepancies resulting from small sample popula- tions, uncertain geologic classifications, and error re- sulting from inefficient and partly-penetrating wells. ESTIMATES FROM PUMPING TESTS IN WELLS Estimates of transmissivity in Tertiary basaltic aquifers were derived from three pumping tests in the Klamath Falls area. The first test was conducted in 1956 by Robinson and Roberts, Ground-water Geologists, Tacoma, Wash., for the Weyerhaeuser Timber Company. The test was run in Weyerhaeuser well 4 (T39, R9, Section 8cbb) which is open in the lower basalt aquifer from 276 to 545 feet below land surface. Analysis by the writer of data from this test suggests that transmissivity of the aquifer at the Weyerhaeuser location is about 11,000 ft 2 d—1 at pre- vailing temperatures. A second aquifer test was performed in 1974 by the US. Geological Survey, assisted by staff and mainte- nance personnel of the Oregon Institute of Technology. Three thermal wells penetrating Tertiary basaltic rocks were used for the test: well OIT 5 was pumped while observations of water levels were made in well OIT 6 and a well at Presbyterian Hospital (fig. 4). The three wells range in depth from 1,584 to 1,805 feet and had temperatures at the time of the test ranging from 88° to 91°C. The responses of the observation wells demonstrate that all three wells are hydraulically con- nected in a relatively permeable formation. The draw- down data provide a poor fit to type curves based on standard analytical models, but the general shapes of the curves and the magnitudes of the drawdowns sug- gest that the apparent transmissivity is about 22,000 ft2 d“. A factor of approximately 0.3 should be applied to this value in order to reduce temperature, viscosity, and density to standard values for comparison with nonthermal aquifers. Thus, the actual transmissivity of the aquifer may be less than 7,000 ft2 d“ for an aquifer at 16°C. G16 A third pumping test was performed in 1976 by the staff of the Geo-Heat Utilization Center, OIT, in an artesian well at the County Museum, Klamath Falls. During this 28-hour test, water levels were measured in 12 nearby wells. The Museum well, which is reportedly 1,234 feet deep, penetrates the Yonna Formation and underlying Tertiary basaltic rocks at a location near the south- western limit of the main hot-well area of Klamath Falls. Observation wells for which data could be analyzed penetrate the Yonna Formation (including some basalt flows) at depths shallower than 500 feet. Water temperatures in the observation wells ranged from 56 to 85°C. Distances from the pumped well to the 12 observation wells ranged from 123 to 1,425 feet. Analysis of the drawdown data by the author shows that all the wells measured are hydraulically con- nected through rocks of the Yonna Formation and the underlying basalts. Apparent transmissivity is about 20,000 ft2 (1‘1 at prevailing temperatures, although some of the less reliable data suggest that the value might be significantly higher. Converting to a stan- dard temperature of 16°C, transmissivity is estimated to be about 10,000 ft2 d“. The storage coefficient of the aquifer, a dimension- less number that expresses the volume of water re- leased from, or taken into, the aquifer per unit surface area per unit change of head, was also estimated from the drawdown data. The storage coefficient is probably in the range 0.001 to 0.01. Estimates of transmissivity derived from the three aquifer tests are close to the average value calculated from specific capacities in 18 wells (table 4). The data indicate, therefore, that the average transmissivity of shallow Tertiary basaltic aquifers in the vicinity of Klamath Falls is moderately high (about 10,000 ft2 (1“), but is significantly lower than that calculated for Tertiary basaltic aquifers in adjacent basins (Illian, 1970, p. 13). Estimates from both thermal and non- thermal wells at Klamath Falls are similar, indicating that for depths less than about 1,500 feet, no distinc- tion can be made between thermal and nonthermal aquifers on the basis of well performance. CHEMISTRY OF GROUND WATER GENERAL CHEMICAL CHARACTERISTICS Ground water of the Klamath Falls area has gen- erally low concentrations of dissolved—solids and is of good quality for most uses. The thermal waters are of poorer quality, although concentrations of dissolved solids in the hotter waters, estimated from specific electrical conductance, are generally no greater than GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 1,000 milligrams per liter (mg L“). Both thermal and nonthermal waters have pH in the range 7.5 to 8.5 and are, therefore, mildly alkaline. Chemical analyses of water from 59 wells and springs in the vicinity of Klamath Falls are listed in table 5. Differences in the quality of ground water in the area are shown by contours on the values of specific electrical conductance on plate 2. The chemical charac- ter of the water is indicated on plate 2 by Stiff dia- grams, and the dominant chemical characteristics are shown in a diagram of percent reacting values of major cations and anions (fig. 5). In figure 5, the open circles represent cold, dilute waters of the area. These samples are typical of cold water derived from basaltic rocks and shallow volcanic sediments. Temperatures of these samples range from 10° to 14°C. The crosses in figure 5 represent hot wa- ters, nearly all of which issue from basaltic rocks and associated volcanic sediments. Temperatures of hot water sampled in this study ranged from 61° to 93°C. Solid circles in figure 5 represent samples collected from 25 of the hundreds of warm wells and three warm springs scattered throughout the area. These waters, produced mostly from the Yonna Formation and allu- vial and lacustrine deposits, appear to be mixtures of recent metoric water and warm water of deeper circu- lation. Temperatures of these samples ranged from 15° to 37°C. Waters in this category are referred to in this report as transitional waters. A fourth category, warm waters from the Klamath Hills, comprises transitional waters whose chemical characteristics differ greatly from other warm waters of the area. These waters are discussed in following sections of this report. The principal ionic constituents of the cold waters are calcium, magnesium, and bicarbonate (figure 5). Concentrations of dissolved solids are generally low, and specific electrical conductance in the waters sam- pled ranged from 110 to 430 itth cm". The transitional waters are also predominantly calcium-magnesium-bicarbonate waters which, as shown in figure 5, contain greater amounts of sodium plus potassium than the six cold waters. In actuality, this increase is due almost entirely to higher concen- trations of sodium. The trend of increasing sodium cor- responds generally to increasing temperature in the transitional waters. It is apparent in figure 5 that there is no corresponding increase in sulfate and chloride in the transitional waters as would be expected in a sim- ple mixing trend between cold and hot waters. The hot waters shown in figure 5 are clearly distin- guishable from both cold and transitional waters. Their positions on the graph reflect greatly increased concentrations of sodium, sulfate, and chloride relative to the cold waters. The samples of hot water obtained G17 HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON .3th 33 $3 $33 9.5 an 939% EEMGES :95“ 9: 59¢ @3883 v.3 ~35 ”macaw ~33 m3a £858 :38 acwmwawh Ewe vamsmézoEflv 95 5. manm .333 939% E mag .8qu we m03m> mafia?— Ewfiok'h 359m mmD._<> 02....Q335 #2:. may 83:? Efiom 2?. do €8on on. 3332: v 3580.0 03 w? 836: v 5550.0 :35 AUomnIUo—Nv mi: :«uEa-u— a... as 9.8-3 5:3 mm“ mm 51: — dam fiver—ow aEaNdE mm. 1. £82.03?" Emmy 3:3. .4 uh annmmAtwm m stow E0..— Mh uuuwméEm :omwtn— .m A. a summémm bio .3532 B 218.2% 16 “a £583 Db mm mm Bow—6:2” .mnm mac 2.30 ha mm; 288%? w .2 z»; .50 «6 mm nuamdem _oc:um 2:2 09 avohNéBv ESmO .I .0 mm» aoavméév hog—m5 v.05. APSAV 28.3 3: 0° 5583.. «Eu: 5 3:30 dEoP G18 $6.25. was $33 EEK .333 kc $9356 BSENEOIm mamfir Tusuuuwafifi mfimdouuwv “a EVE E Em: mum moi—Ema .bomgmu :23 55:5 £630va $23350 $25. .82— .:K 2.33:1: E mcosuhaaoccg G19 8.: :3 .H. .33 ”H :35 315 Sam SHEER—ad 12..qu .mwism HuoHMBOoO .w .3 “H .Hmeamv €an 2ch $956 .3823 .m .3 Nu .3550 n .5131 Ho 28m 33m :omeO u .w 23$ .32 {am can nEooBoZ ”w «SH :EHEEoo :SHHHB cognac hut-HE ”v ._oEEww 1:» bamwo> JESS—53H: Em: no .vraH rag—5:8 H532? £26.50 zeta: .anwthzame Eu: 5 ”83H 35355 as much—om 8szon mm @8858 .EEmmSoa m2 EHHHHSm «.3253? $3. How N 95%. gm .Efiuom and 695% .HHEmEsoE :o «imam. HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON <3 HH 3.8.8 w? on? 1 1 1 1 1 my. 1 1 1 1 1 1 1 1 1 oH. H. .5 8.8.8 N5. 1 1 1 1 omv 1 NQN HH. H NV 9. 8.. NH 3 N5 H.NH NH. .35.me mam 5862 m .1. 3.8.3 1 m2. 1 8v: 1 1 .- 1 1 1 : 1 1 .. : 1 .. HH. H. .2 3.8.8 m.w. .. 1 1 1 E. 1 8 H. H NV 9. «w; «H Qw md Nd HF 12585 mam mac—Em H. H .0 8.8.8 5. 8H}. 1 1 1 8H.H 1 1 mH 1 1 1 1 1 1 1 1 E. _ ..... 5.88 3.. SQH. 1 1 1 1 1 9. 1 H: Hm H. ON». 3. ONN mN NN NH. HENHQHH. 9.3 H232 Nm F. 28.8 1 8H. 1 1 1 1 .- NH... 1 1 1 1 1 1 1 1 1 oH. H. .3 «5.8.8 NP .. 1 1 1 omv .. NHN H. H NV 9. ma. NH os N.HH 93 NH: £555ka wnm EEuEEEHi m H .9 8838 H6. SH. 1 -. 1 1 1 8.. 1 1 1 1 1 1 1 1 1 h. H. a 3.3.3 1 1 .. 1 1 onv 1 En H. Hv NV 9. on. mH ah Qm H,w NH. gunmfikfi mam x82 :26 N 9 8.8.8 1 mHH 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 NH 89 H. .n 8.8.8 a? 1 1 1 1 8v 1 VN S. H NV 9. 8.. N H E 3. N.HH 2.. £8.32. 25¢ 55:5. 2. H H... 3.8.8 ed. 82.. SN 1 cm . om c Q. m H m H. «N H. EN. NH. HN HUN 8 «H. “388:8 52 5m 5 H .0 3.5.8 3w. o3”. oNH 1 o8 m a m Hm o. .3 mN m. 8N. 3 2 NH VN mH. uafiéov Ease: Em> NH. H 9 3.8.3 v.5 own. on 1 omH 2: H mm m me n H. own. ma mm mH wH mH. Hvuuoméam 32m Ewm 9 a 3.8.8 1 omN 1 1 1 1 1 1 1 1 1 1 : . : 1 1 1 mH a HH.. .8 Qw 1 1 1 1 .. 1 S 8. 3. NH 1 1 1 NN.‘ ma «va vH nuaomdsm .980 .353 €qu m HH .9 3.3.8 N6. 39 1 ONV; 1 1 1 1 - 1 1 1 1 1 1 1 1 pH. _ a N>.H.H.oH 1 mNN oHv 1 S 1 1 NH. H. EN Ham o. HS. 3.. wH HH vH cH 8.2.2:. 88m HH< .m .3 HH. _ a $8.8 ms. o8. ow 1 08 E H 8 m. on o m 9 can. 9m NH. wH NH 2. pfiNNéHam .on HHmeme 8 H .o E.8.8 Haw. 8N. o 1 ON on m 2. o. NN oH H. SH. 3. 2 w w 2 B. HquNdHam Ba; .3 Ban «m H a E.Hm.8 vs. 8N. ow 1 own on m 8 m. H.m HN 9. 8H. 9m HUN 8 HH wH. £32.28 .3th ii :5 Nm H .9 3.8.8 2. 8H. 2 1 can on H. 2. N. NH 2 o. oHH. 3. H.H 05 HH 2. 2.38:8 HEEGH 3m Hm 6.3V. ESE: 3.5 H .u E.8.8 5.. A3N. on .. 3. ON o 8 o. NH. 8 9 3H. I. H: mH H8 8. HmoanéHsv :5 En 8 .o Emmofio 5.. SN. 1 1 1 on 1 1 N. 1 1 , 1 1 1 1 1 11 NN. .0 3.8.8 5. 8H. 1 1 1 1 1 8 1 Ham 3 9 8H 3. Nu H.H hm HN. BumNéHam bammows mm a H5. .8 0w 1 oHv 1 8 1 1 vN Ho. HF. NH 1 1 1 8.. H3 N.HH HN NaBNmémm .3»? E820 HH . 3.8.8 1 8N .. 1 1 1 1 1 .. 1 1 1 1 1 1 1 1 E m 8.8.8 1 1 1 .. 8v 1 1 8 1 m o o :H . . mm. H. H: 1 $3888 65 3.33 cameo oH H a «58.8 3. 2:. o 1 ON 3 o 8 o. 3. 0.8 N. 8H. N...” on H..H H; NN. NuauméHHHH. .oo :82 page; mm. H .o EéN.8 3. 8H. 8} 1 8H 8 N 2. N. 3 as 9 SH. Wm E 8 NH NN. 3.28:8 .meHm Nam EN 9 8.2.2 2 SN 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 . v 62 :35 H. a NWNN.8 1 1 1 . 1 1 8 1 m: HH. H. H. 3... mHH; mm Nm mH o.oH NN flowgam 3.52533 wH a 8.8.8 1 SN 1 1 1 1 1 1 1 1 1 1 1 1 1 1 11 NN .u H5. .8 EH 1 EV 1 on 1 1 8 N. Ham w. 1 1 1 8.. mm N.HH 8 838.98 ES .355 585 NH H .9 3.2.8 5.5. mom. 1 1 1 on 1 1 N. 1 .. .. 1 1 .. .. 1 «N. H .9 :.8.8 5. 8H. 1 1 .. 1 1 on 1 QN 2 H. S: H.m 8 On H.o mN. BowméHam HHozb .x .H. 3 H .. 3.28 3. SN. 1 : 1 ON 1 1 H. 1 1 1 1 1 1 .. 1 «N. H .u «55.8 3. SN. 1 1 1 1 1 on 1 9N E H. oHH. Hum NN m6 8H 8. 883:8 E83 fiBSH on H u E.8.8 3. SN. o 1 ON 3. 1 NF 0. 3. «N 8. 8H. 3 ow H.m H.w vN. 83.6st .8523 830 S G20 for this study are typical of water in the more than 400 hot wells used for space heating in Klamath Falls (Lund, 1978) as well as hot water from Olene Gap and the Klamath Hills. In two areas of the lower Klamath Lake valley, ground water occuring at depths of 1,000 feet or more has generally high values of specific electrical conduc- tance and is, for the most part, unfit for drinking. These areas are delimited on plate 2 by two 1,000 umho contours located north and south of Miller Hill. No chemical analyses are available from these areas, but owners’ reports and field observations indicate that much of the water contains large amounts of iron, sul- fate, and methane gas, as well as traces of hydrogen sulfide gas. Temperatures of water in the northernmost area are generally higher than normal. Drillers’ logs indicate that the water is contained in fine-grained, organic-rich sediments depleted in oxygen. These parts of the basin quite possibly were marshy sumps and areas of evaporative ground-water discharge during long periods of late Pleistocene and Holocene time. RELATIONS AMONG DISSOLVED CONSTITUENTS Concentrations of dissolved constituents in ground waters of the Klamath Falls area differ widely, depend- ing on the lithology of the aquifer, the temperature of equilibration, and the amount of mixing that has oc- curred between thermal and nonthermal waters. Some of these relations are discussed in this section of the report. The relations of common cations and silica to temperature are discussed more fully in a separate sec— tion below. Cold ground water derived from shallow alluvium and basaltic rocks in the Klamath Falls area is a calcium-magnesium bicarbonate water which origi- nates in local precipitation and acquires small concen- trations of dissolved solids mainly by solution of feldspars and other silicate minerals in the volcanic rocks. Silica concentrations are relatively high, com- monly as great as 40 to 50 milligrams per liter, indicat- ing that even in basaltic rocks abundant silica is avail- able for solution. The chemical character of water in the basaltic rocks does not change significantly even after penetration to great depths. For example, water obtained at a depth of about 1,300 feet in a well on the northeast flank of Hamaker Mountain (sample number 49, table 5 and pl. 2) has acquired slightly increased concentrations of dissolved solids relative to meteoric water but is not greatly different in quality from waters that presuma- bly have penetrated less than a few hundred feet before emerging at the surface in gravity springs (samples 2, 3, 4, and 5). GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Hot waters from the three principal geothermal areas are represented by 10 samples having tempera- tures greater than 65°C. An additional sample, number 17, with a temperature of 61°C, plots below the main group in the diagram of figure 5 but is sufficiently simi- lar to warrant inclusion with the hot waters in most of the discussions below. The hot waters are obtained from wells and springs that tap both the Yonna Formation and underlying basaltic rocks. These waters probably originate in local precipitation. If this supposition is true, the meteoric waters, in their passage through the geothermal reser- voir, acquire increased amounts of calcium, sodium, potassium, chloride, sulfate, and silica, while concen- trations of bicarbonate decrease somewhat and mag- nesium is reduced almost to zero. Changes of this kind are commonly found in hot-water geothermal systems (White, 1970), and they are compatible with the as- sumption that the Klamath Falls thermal waters are locally derived. In hot waters from the principal geothermal areas, concentrations of major ions generally differ more widely among samples from the same area than among the three areas. Ratios of the major ions in hot waters show no consistent differences from one geothermal area to another. Some of the ratios that are of particu- lar interest in geothermal waters are given in table 6. Molal ratios of Ca/Mg, Na/Ca, and Na/K are significantly higher in the hot waters than in the cold waters and ratios of Na/Cl and Ca/Cl are lower. High ratios of calcium to magnesium are generally charac- teristic of high-temperature geothermal systems, but may occur at temperatures as low as 100°C (White, 1970, p. 70). High ratios of sodium to potassium also imply elevated reservoir temperatures, but suggest that the waters equilibrated at temperatures not much higher than the measured temperatures (Mariner and Willey, 1976; White, 1970). Ratios of Na/K'are strongly influenced by the amount of available calcium in the water and cannot be relied upon as geothermometers in waters, such as those at Klamath Falls, for which molal ratios of VC—a/Na are greater than one and in- dicated reservoir temperatures are low (Fournier and Truesdell, 1973, p. 1272). Molal ratios of Ca/H003 are low in the cold waters (range 0.04 to 0.2) and increase, as would be expected, in typical hot waters (range 0.5 to 1.2). The values of ' this ratio in the thermal waters are not high enough, however, to suggest that the waters equilibrated at ex- tremely high reservoir temperatures (White, 1970, p. 69). Ratios of total carbon species (mainly HC03 and CO3) to chloride are low in the hot waters and these ratios clearly distinguish the hot waters from waters of lower HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON G21 TABLE 6,—Ratios of dissolved constituents in ground water [Molal ratios except where weight ratio is indicated] . Temp. SiOZ/Cl No. Owner or name °C VCa/Na Na/K Ca/Mg Call-1C0; Ca/Cl Na/Cl B/Cl F/C] SO4/Cl (HC03+C03)/Cl (wgt. ratio) Hot waters 46 Liskey 93 2.2 85 > 91 0.5 0.2 5.2 0.04 0.05 2.3 0.5 1.5 44 Osborn 90 2.3 78 > 91 .6 .2 5.2 .05 .05 1.8 .5 1.6 13 Mills School 89 1.1 209 > 82 .9 .4 11 .06 .04 3.1 .5 1.4 6 OIT #6 88 1.5 85 >147 .8 .4 5.2 .06 05 2.5 .4 1.6 28 Olene Gap Sp 87 2.7 44 121 1.2 .6 4.9 .06 04 2.5 .5 1.7 58 Asaembl ongod 87 2.9 79 170 .7 .5 6.1 .05 AAAAAA 2.7 .7 2.0 7 Medo-Be 1 Dairy 81 1.8 87 >140 1.1 .4 6.0 06 .04 2.8 .4 1.5 8 Friesen 73 1.8 85 >152 1.2 .4 6.0 05 .05 2.8 .4 1.6 14 Serruys 71 1.8 92 >133 .8 .4 6.3 05 .05 2.8 .5 1.7 29 Jones (Sp ) 65 2.2 73 212 1.1 .5 5.1 1 .04 2.5 .4 1.5 17 Mazama hool 61 .8 70 33 .07 .1 1.0 ______ .05 2.6 2.0 3.4 Warm waters of the Klamath Hills 50 O’Connor 38 2.4 71 7.6 1 1 4.0 .01 .01 10 1.3 1.3 56 O’Connor Livestock 30 4.2 51 5.5 4 3 2.0 .01 .0 5 .9 .5 45 Boehm 25 3.2 46 2.3 2 9 4.3 03 .002 6 5.0 .6 51 Liskey 22 1.2 82 1.3 03 2 6.3 .04 .005 03 6.4 .7 57 Carter 22 6.2 20 7.9 5 3 1.5 .01 .0 4 .6 .4 53 Suburban Water Co. 21 7.8 22 1.8 4 9 2.2 .01 .003 5 2.3 1.0 54 Town of Merrill 21 6.2 20 1.6 2 4 2.5 .001 .007 3 2.5 1.4 Miscellaneous warm waters 22 Falcon Hts. School 37 5 6 14 1.0 .07 .7 9.9 04 .04 1.0 11 9.0 1 Eagle Point Spg 35 1.0 19 >36 .01 .03 6.0 03 .09 <.04 4.9 4.5 20 Air Force 31 3.4 8.8 1.9 .05 .6 12 .09 .5 12 8.8 23 Jacobsen 30 6.0 7.0 1.5 .06 .6 9.6 .02 .6 12 8.2 24 Ore on Water Corp. 26 9.2 ...... .9 ,,,,,, .3 1.9 ,,,,,, .2 .08 ______ 2.1 39 M. ehlin er 26 21 5.6 .9 .2 5.0 9.4 .02 .0 .7 26 15 15 Holliday pg 25 15 16 2.4 .2 2.0 8.2 .03 .04 .5 14 5.3 25 McCollum 25 28 815 1.1 .3 10 10 .02 .1 6.2 23 17 36 Brookshire 25 20 11 3.9 .2 5.0 12 .01 .07 1.3 20 13 38 Stacy Co. 25 21 8.5 3.9 .2 5.0 15 .02 .0 2.1 27 17 26 Shuck 24 14 6.8 1.5 .1 3.3 14 .02 .1 1.9 23 20 43 O'Connor 24 13 6.8 1.1 .1 1.1 5.4 ,,,,,,,,,,,, .6 8.3 4.6 47 C. Dehlinger 24 8.5 11 1.6 .08 1.4 13 .03 0 .2 18 15 30 Langley 23 16 6.8 1.0 .1 3.3 13 .03 ...... 1.9 25 19 40 O'Neil 23 11 8.7 .7 .08 2.0 14 .04 .1 1.7 23 18 12 Oregon Water Corp. 22 15 ______ 1.1 ,,,,,, 2.0 7 6 ,,,,,, .04 .04 ______ 4.7 18 We erhaeuser #4 22 7.8 16 3.2 .1 5.0 12 .08 .05 .4 17 8.0 27 Bix er 22 25 8.2 .8 .2 5.0 13 .03 2 1.5 34 26 55 Pope's Meat Co. 22 11 16 3.7 .1 1.7 10 .01 0 .5 13 5.7 10 Oregon Water Corp. 21 13 ...... 1.5 .1 1.1 __________________ 0 8.5 5.8 11 Oregon Water Corp. 21 15 ,,,,,, 1.0 ______ 2.0 ,,,,,,,,,,,, .004 1.0 ______ 4.5 35 Dobry 21 8.5 12 3.2 .08 1.4 14 .03 ______ 1.5 17 19 48 Hill 20 18 5.1 .4 .1 3.3 69 .02 0 .6 19 6.2 Cold waters 31 Kinney 19 23 6.0 1.0 .2 5.0 15 .2 2.2 38 29 32 Kirkpatrick 18 16 7.8 .7 ,1 3.3 11 .2 2.5 24 22 34 Sharp (Spg) 18 22 8.2 1.2 _2 5.0 13 .0 1.7 29 20 33 Gabriel 16 9.2 7.5 .4 _04 3.3 22 .2 .4 76 32 41 Air Force 16 17 8.0 .8 ,2 5.0 10 .07 .5 30 16 9 Oregon Water Corp. 15 20 ______ .9 ,,,,,, 2.5 ______ .004 .1 ,,,,,, 6.0 19 Shaw 15 7.1 22 .7 .08 2.5 15 .8 .2 31 9.3 42 Howard 15 33 19 .8 .2 10 11 .0 .4 51 13 37 Adair 14 26 8.5 .6 .1 5.0 7.2 .5 .2 35 10 49 Tulana Farms (Spg) 14 43 8.2 1.2 .2 10 8.8 .1 <7 35 48 2 Shell Rock Spg 12 39 6.6 1.3 .2 >10 >90 .3 Indet. 29 >60 (small) 5 Hummingbird Spg 12 37 9.9 .8 .2 10 11 .3 <.7 56 42 (small) 52 Lon 12 2.5 94 .5 .05 1.3 22 .2 .2 1.2 26 28 4 Barfiley Spg 11 31 11 .8 .2 10 12 .2 .3 <.7 50 36 3 Neuben Spg 10 37 9.9 1.0 .2 10 12 Indtfi. .2 <.7 55 48 (srna ) temperature. Cold waters have uniformly high ratios, and transitional warm waters have ratios that are generally similar to or slightly lower than those of the cold waters. Ratios of total carbon to chloride show no significant differences in 10 of the 11 hot waters de- spite temperature differences of almost 30°C among these waters, whereas in nonthermal and mixed wa- ters in which temperature differences are also about 30°C, ratios of total carbon to chloride differ by an order of magnitude. The nearly constant ratios of carbonate species to chloride in hot waters of greatly differing temperature suggest that the temperature differences in these wa- ters are due to conductive heat loss rather than mixing with cooler waters. The apparent lack of mixing in the hot waters implies, therefore, that these waters rise through conduits that are largely isolated from the shallow ground-water reservoirs of the area. It seems clear that most of these conduits occur in the major fault zones of the principal thermal areas. The wide- spread occurrence of warm water in the lower Klamath Lake basin suggests, however, that significant amounts of thermal water also arise from undiscovered faults underlying the valley floor. G22 Ratios of SO4/Cl are slightly higher in the hot waters than in the cold waters, suggesting that the deeply circulating thermal waters are enriched in sulfate, either through contact with gypsum deposits or by oxi- dation of hydrogen sulfide gas which is known to be present in the hotter waters of the area (J. W. Lund, oral commun., 1975; Peterson and Groh, 1967). The increased SO4/Cl ratios in the hot waters have been interpreted as suggesting that a steam reservoir underlies the Klamath Falls graben (Murray Gardner, written commun., 1975; Peterson and Groh, 1967). The interpretation is based on the possibility that hydrogen sulfide gas rising from a deep steam reservoir may be oxidized to the sulfate as it enters shallower aquifers thereby increasing the ratio of SO, to the nonvolatile chloride. If this hypothesis is correct, volatile con- stituents such as boron (B) and fluoride (F) might also be expected to be transported in the vapor from the underlying steam reservoir, thereby increasing the ratios of these elements to chloride in the thermal wa- ters (Truesdell, 1976a). In the samples obtained for this study, however, the ratios of fluoride to chloride are smaller in the thermal waters than in the cold waters, and ratios of boron to chloride in the thermal waters are similar to those in the five cold water samples. Thus, although these ratios are somewhat unreliable owing to the small concentrations of the three con- stituents in the cold waters, they provide no support for the theory of an underlying steam reservoir. Most of the warm transitional waters in the lower Klamath Lake valley have temperatures in the range 20° to 40°C. These waters are produced by wells, gen- erally no more than 500 feet deep, that penetrate allu- vial and lacustrine deposits and interbedded basaltic rocks of the Yonna Formation. The unconsolidated de- posits include diatomaceous earths, mixed-layer clay-mineral assemblages, kaolinite, feldspathic and organic-rich silt, fine siliceous sand with large per- centages of volcanic glass and mafic materials, and coarser-grained sand and gravel also derived from vol- canic rocks. A few wells and three warm springs prob- ably derive water almost entirely from volcanic rocks and breccias of basaltic composition. These diverse sources of warm water are represented by 25 samples which show, as expected, wide variations in chemical composition. Log molal ratios of several major ions in cold, tran- sitional, and hot waters are shown graphically in figure 6. The ratios selected are those that best illus- trate the diversity of the transitional waters and the trends, or absence of trends, between the hot and cold waters. These ratios do not coincide with those selected for table 6. In each graph, samples are arranged from left to right in order of increasing chloride concentra- GEOHYDROLOGY ()F GEOTHERMAL SYSTEMS tion. This also turns out to be the general order of in— creasing sodium concentration and increasing temper- ature. For reasons discussed below, warm waters from the Klamath Hills area are placed in a separate group. If the logical assumption is made that higher tem- peratures in the transitional waters result from larger amounts of thermal waters mixing with local meteoric waters, it would be expected that ratios of sodium to chloride, which do not differ greatly between the cold and hot waters, would remain fairly constant as tem- peratures increase. However, other ratios, such as sodium to calcium, calcium to magnesium, and total carbonate to chloride, which differ greatly between cold and hot waters, should show trends toward their values in the hot waters. The data of figure 6 show these expected trends in the sequence of transitional waters and thus tend to support the assumption that, in general, the warm waters in the Klamath Falls area represent varying mixtures of thermal and meteoric waters. The trends in sodium/calcium, calcium/ magnesium, and total carbonate/chloride ratios are fairly distinct, despite effects of local lithology which are the probable cause of significant departures from the trends in individual waters. Concentrations of silica are extremely variable in the warm waters (table 5) and ratios of this constituent to other major ions (not shown in figure 6) have little correlation with temperature trends. A graph of log molal silica concentrations in ground water (figure 7A) demonstrates the variability of silica concentrations in cold and transitional waters in relation to the trend of increasing chloride concentration shown in figure 6H. It can be seen in figure 7A that the variability of silica in five of the coldest and most dilute waters is greater than the variability in 10 hot waters from the principal geothermal areas. The relation of silica concentration is shown explicitly in figure 7B, where two trends are indicated for Klamath Hills warm waters and the hot waters. The suggested trend, A, between cold and hot waters is not supported by any field evidence, however, and the silica concentrations in both groups appear to be largely independent of the chloride concentrations. The relation of chloride concentration to tempera- ture might be expected to define mixing trends in the warm waters if such trends exist. As shown in figure 7C, however, the trends are not well defined and the relation does not show a high degree of correlation. It is not surprising, therefore, that silica concentrations show an even poorer correlation with temperature (figure 7D) and fail to suggest any clear trends. The wide scatter in the concentrations of the various ions and the consequent scatter in plots of ionic ratios indicate the difficulties that may be encountered in at- itempts to apply chemical geothermometry to low- LOG MOLAL RATIO HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON G23 3.0 2.0 . A. Ca/Mg ratios °o° - 2'5_ _ 1'5?) 0.... o. a o . . ' _ T .C 7 + I . Io. '. 2.0— I - 1.0— u' '3. — G) I © 1-5- - 0.5— —I © H © *17 © 10— ( ) © 9 " OI- ® _ >056 O ‘1' Q + 0.5— ‘ ’ . - ' \[+] — —0.5- i ‘ . .' . .- “7 Q 9 G’ E. (HC03+CO3I/Cl ratios —° 0' ' ' . I G) _ _10lIIIIIllllllllllIIIIIIIIIIIIILlJlJlII IIIIIIllIIl 0°00...,".".°,‘ I||||IIIIIII|IIIIIIIIII D —o'5IIIIIIJI‘IIIIIIIIIIIIIIIIIIIIIJIIIIIIIJIIIIIII“ 8 2'0 + g F. Na/Ca ratios [#17] 1.5— © © — _I 15 <1: ° - .1 ° . A 10 . ‘ 7 I 1.0-)(000 ......o... .... a? . 7 ..o ' I— _ g . ..o . o . 0 © © © \( o . . . . o .0 + (9 (D 5_ ' . I 0 '°' . ®_1 >035) 'L o O O '." o ' . ' ® 0.5- G) — -J o . B.Na/CI|‘8tIOS. G) @ 0TOIOAIII‘TIIIIIIIIIIIILIIIIIIIIIIIIIIlllllllllll ‘9 I I I I I I I I I I I I I I I I I I I I I I I oillllIlll|lllllIllllIllllllllllllllllIllllllllll IFIIIIIIIIIIIIIIIIIITII 2.5 . G. Na/SO4 ratios ° © 2.0— ‘ 2.5 ' o . C C. Na/K ratios I 1 5 ,_ . o 0 . — 2.0— .+ — . . . . 0 © © 000 . O .7 e I (93 1.0—0 o . o . " © _ 15'— "‘ 0". . o . © . o o o . . ® ©© . ©©© © —o°o . 7 . 7 ' 7? ?. —- 0.5— T 1'0 o O 0'. .0. ' .o ‘ ' . . t . .. 0IIIIIIllllllIlllllIIIIIlJlllllillllllI|l|il|llll IllliillIillllilllllI|Ii|lllllI_LIll||I|J_llllll 0.5 ,HHHHHHHHI HM, llllllllllllllllll II II ‘2.0 H. Cl concentrations ©©© 1.5 ”2.5— 5) " D. Ca/Cl ratios >- T l— I <9" 1,0 —xoooo o ‘ 3 ‘3.0 — Q _ 510).... o . no 0 j o. 0.5- ' ' " ' — O-35— - o .. Oi. . E 0.... . . . . . o ..'.000 0' '7 o o“ C “4-0" (<-4.55) "Nu-"fl _ . C _1 .0 o T J .00. -0.5" . I Q Q G) " ’45 oooo _ @ 3317+] _l.0 Illlllllllllllllllllllll‘lllll|lll ll IQIIJLII -5.o _ 2|4I49|27|34130125|4o|33|35I13|55|37[15I12|19[1o|43|52 5AI53I5||45 2 4 49 273430 26 40 333518 55I371512191043|52 54I53 51I45 3 5 31 38 42 41 36 2‘ 32 39 48 9 47 11 20 23 22 l 50 57 56 3 5 313842 41 36 24 32 39 48 9 4711 20 23 22 I 50 57 56 SAMPLES, IN ORDER OF INCREASING CHLORIDE CONCENTRATION EXPLANATION I Median value and range of values for 11 hot waters (>60°CI. @ Warm water from the Klamath Hills A When sample 17 is shown separately, median and range are for - Warm transitional (mixed) water 0 Cold water (<15°CI derived from basaltic rocks x Value greater or less than number in parentheses remaining 10 samples [+]Hot water sample 17; shown when ratio differs significantly “17 from other hot waters FIGURE 6.—Log molal ratios of dissolved constituents in ground waters arranged in order of increasing chloride concentration. G24 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS —2'ollllllllllllllllllllllllllllllllllll| IIIIIII A. 2 O — —2.5 e - :5 >- If t w _l o + 9 O E j 0 9 . 1' © 0 I O @ U o ‘3 0 9' o . o 0 ® — g E O O 0 . . o . 0 . o C I 0 © 3 Q o o . 0 o . < O o O o _l O 2 —3.5 ~ . —- _l O o :7: T . . I Median value and range of values In 10 samples _4.ollIiIllllllJlllIIIIIIIllllIIIIIIlIlIl lLlllIl 315I31I38I42I41[36|24|32139|4a|9I47I11|20|23,Izzl1T 54|53I51|45 4 49 27 34 30 26 4O 33 35 18 55 37 15 12 19 IO 43 52 50 57 56 SAMPLES, IN ORDER OF INCREASING CHLORIDE CONCENTRATION 180 I I I I I I I I I I I I 140 I T I I / @45 160l— B. ®56 / — 5130— D — I: 140— / ©51 — .1120— +17 - Q7/ a: 120 — / - E 110 ~ +58 4 7 m 100 — @57 / — E 100— 51@©45 +28 — I / a: .33 63:3 I.I..I _ _ (D l— _ I: 80 / b j 90 +3 +\46 -' 60 _ +29V 7 46 / :1 80 __ 56@ .22 +14 +7 _ 33 /©53 (9 13+\+ 8%; 28 +58 2 1' +13 0. / j 44 z .47 w 40 — / — — we ,3 — 2 5w ’7 +17 2‘ 019.35.24.23 @50 g 20 _ /.,4310 4, .— 9 60— 02 — :5 12 15 37 QI‘JZO/o/Lza 47’1 I— 026 20 _ " 9'26 0. °’ .22 < 53 o I .1 0 Ill/49H P " 593v? 24 ‘ I I L '33I I I I u: 50 — 313535.21 ate - :I o 10 20 30 4o 50 so 70 80 90 100 110 120 130 140 I— 23,. .Io3°"° +29 2 Z 5- 0' 43 z SILICA, IN MILLIGRAMS PER LITER 3 4°“ 57 @5'4 35 — " Z 38 . o 30_ 042 _ z u 9. . 015 2 < 49-12}! 2180 g 20— _ “I: :' I" v: 10F _ E 160 I- U o I L I I z o 20 40 so 80 100 o 140 — 3 TEMPERATURE, IN DEGREES CELSIUS E 120 — CC C " 100 — a EXPLANATION 80 _ . . I o + Hot water from prInCIpaI geothermal areas (>60 Cl 60 _ © Warm water from the Klamath Hills - Mixed and fresh waters of low to moderate temperature 4° ‘ 0 Cold water (<15°) derived from basaltic rocks 20 _ * Probable error in silica analysis 54 Sequence number in table 7 0 l o 10 20 30 4o 50 60 7o 80 90100 TEMPERATURE, IN DEGREES CELSIUS FIGURE 7.—Silica and chloride concentrations in ground water and their relations to observed water temperature. A. Log molal S‘iO2 concentrations in samples arranged in order of increasing chloride concentration. B. Chloride versus silica concentration, showmg separate trends for warm waters of the Klamath Hills (line B) and hot waters from the principal geothermal areas (line A). C. Chloride concentration versus temperature, showing two possible dilution trends (lines A and B). D. Silica concentration versus temperature. HYDROGEOLOGIC APPRAISAL ()F KLAMA'I‘H FALLS GEO'I‘HERMAL AREA, OREGON temperature waters from volcanic terranes. (See the section on “Reservoir Temperatures Estimated from Chemical Geothermometers.”) It is clear that both silica and cation geothermometers may be unreliable in these settings and should be applied to low- temperature waters only when most of the chemical relations indicate that the waters have cooled conduc- tively or are simple mixtures of thermal and regional background waters. A group of seven samples obtained from warm wells in the vicinity of the Klamath Hills thermal area are chemically anomalous with respect to other warm wa- ters of the region. (See tables 5 and 6.) Despite the relatively low temperatures of these waters (maximum 38°C), concentrations of sodium, calcium, sulfate, chloride, and silica are among the highest encountered in the area, and electrical conductances are generally high, ranging from 640 to 2,700 umho cm‘1 in six of the seven samples. The distinctive chemical character of the Klamath Hills warm waters is shown by the graphs of ionic ratios in figure 6 and by graphs of chloride concentra- tion versus temperature (fig. 7C). The chloride concen- trations in figure 7C form a trend for the Klamath Hills warm waters (line B) that differs greatly from the trend inferred for the hot waters (line A). Graphs of other ions, notably sodium and sulfate, versus temper- ature (not included in this report) also show distinctive trends similar to line B. These data are consistent with the isotope compositions (discussed in a subsequent section of this report) in suggesting a separate origin or history for warm waters of the Klamath Hills area. The cause of the anomalous chemical characteristics in the Klamath Hills warm waters is uncertain, but the isotopic compositions strongly suggest that these characteristics are the result of low-temperature evap- oration of the hot waters of the area. If so, the charac- teristics of these waters may be genetically related to their occurrence in the closed basin of Lower Klamath Lake. RESERVOIR TEMPERATURES ESTIMATED FROM CHEMICAL GEOTHERMOMETERS Because the solubility of silica in water of less than pH 9 is affected only by the temperature of the water, silica concentrations in thermal waters have been used as a basis for estimating reservoir temperatures (Fournier and Row, 1966). The estimates of reservoir temperature are based on the following assumptions: (1) Silica is freely available inthe reservoir rocks; (2) the water has remained in contact with the rocks for a sufficient length of time to reach chemical equilibrium with respect to quartz or other silica minerals; (3) the silica concentration attained at depth remains con- stant as the water moves from the reservoir to the sampling point. G25 For waters of the Klamath Falls area, assumption (1) is probably true. Although virtually no quartz occurs in the upper Tertiary rocks of the region, silicate min- erals are abundant in the basaltic rocks and deeply circulating thermal waters may encounter rhyolites of Eocene or Oligocene age, or quartz-rich sediments of Eocene age, both of which underlie younger Tertiary rocks in adjacent ares (Williams, 1949; Peterson and McIntyre, 1970). The high silica concentrations in the cold spring waters indicate that silica in some form is abundantly available even in the shallow aquifers. In spite of a general availability of silica, it is possi- ble that the geothermal system at the present time is largely restricted to flow paths and conduits that have been depleted in silica. There is no evidence, for exam- ple, for recent silicification comparable to the older silicification that pervades the rocks surrounding the geothermal areas, and this may indicate that silica concentrations, as well as flow volumes, have de- creased during the life of the system. Although this question has not been resolved by the present study, the uniform silica concentrations of hot waters occur- ring over a wide area suggest that silica saturation has been generally attained by the thermal waters. In light of this conclusion, the second assumption relating to the time required for equilibration with quartz or other silica minerals is probably also a reasonable one for the Klamath Falls geothermal system. The assumption is further supported by results of tritium analyses, dis- cussed below, which indicate that the thermal waters have had residence times on the order of 10 years. The third assumption, specifying that silica concen- trations remain constant as the thermal waters ascend to the surface, is probably not completely met. The rel- atively constant silica concentrations of the hotter waters of the region suggest, however, that the amount of wallrock interaction and dilution by mixing is min- imal. Supporting this conclusion are the constant (HCO3 + C03)/Cl ratios, the high Na/K and Ca/HC03 ratios, and the fairly small increases in oxygen isotope concentrations (see next section of this report), all of which are indicators of relatively low reservoir tem- peratures. As the geothermal waters ascend toward the surface, therefore, silica concentrations would not be expected to change appreciably through either precipi- tation or further solution of silica. Estimates of the probable extent of mixing in the thermal waters are given in the next section. Several silica geothermometers are given by Four- nier and Rowe (1966) and Fournier and Truesdell (1974). Briefly, the four models used for this study are based in the following assumptions: (1) Quartz satura- tion with adiabatic cooling (Qtz. Adiab.), (2) quartz saturation with conductive cooling (Qtz. Cond.), (3) chalcedony saturation with conductive cooling, and (4) quartz saturation and mixing with cold, meteoric wa- G26 ter. Computations were made by means of Geotherm, a computer program written by Truesdell (1976b). Re- sults are listed in table 7 and shown in the graph of figure 8. Input data for the mixing model include a silica con- centration of 45 mg L‘l and a temperature of 20°C for the cold water. The temperature chosen for the cold- water component is typical of cooler waters in the vicinity of the geothermal areas. Slightly lower esti- mated reservoir temperatures would be obtained if the coldest ground—water temperatures in the region were used for the analysis. The silica concentration chosen for the cold-water component is tyical of concentrations in basaltic aquifers of the area. Somewhat higher esti- mated reservoir temperatures would be obtained if lower silica concentrations were used. Estimated reservoir temperatures based on the quartz geothermometer have a narrow range for most of the hot waters but a much broader range for the warm waters. Reservoir temperatures calculated for nine of the hotter waters by means of the quartz con- ductive model range from 124° to 143°C; those calcu- lated by means of the chalcedony model range from 94° GEOHYDROLOGY OF GEOTHERMAL SYSTEMS to 108°C. When the silica mixing model is applied, temperatures range from 126° to 184°C for chalcedony. Athough nearly all the estimated temperatures fall within reasonable limits, they represent an uncom- fortably large range of values for each of the ap- proaches. In evaluating the various approaches, an im- portant criterion is the fact that for geothermal waters having temperatures less than about 150°C, silica equilibration may occur with respect to chalcedony rather than quartz (R. O. Fournier, oral commun., 1976). However, the reservoir temperatures estimated from chalcedony concentrations are lower than ob- served temperatures in thermal wells, indicating that these direct estimates are inadequate and suggesting that mixing models are probably required. Final con- clusions regarding these approaches and the best esti- mate of reservoir temperature are deferred to the sec- tion “Isotopes in Ground Water.” Estimated reservoir temperatures for warm tran- sitional waters in the area (table 7) are extremely erra— tic and are probably unreliable. Although it is virtually certain that these are mixed waters, mixing models fail to provide reasonable estimates of reservoir TABLE 7.—Estimates of reservoir temperature based on chemical relations in ground water Calculated reservoir temperature (°C) Seq. Well Measured' Qtz.2 Qtz.2 Chalced.2 Na/K +BVU5/Na“ Mixing models‘ fine in No. Owner or name def1t3th tegéip. Date (adiab) (cond) (cond.) 13:1/3 3:4/3 Quartz Chalced. sulfates Hot wells and springs 6 OIT Well #6 1805 88 0&05—72 127 131 103 108 71 150 112 185 7 Medo—Bel Dairy 765 81 017 24-55 122 126 96 109 75 150 109 185 8 J. E. Friesen 564 73 02719—55 124 130 101 109 75 163 127 ,,,,,, 13 Mills School 800 89 0&07‘75 121 124 94 83 67 135 99 179 14 Lois Serruys 528 71 12_22A54 122 127 98 106 73 126 122 ,,,,,, 28 Olene Gap Spg 777777 74 03709—74 130 136 108 130 80 142 127 196 44 O. H. Osborn 418 90 0531—74 127 131 103 114 83 154 114 135 46 Jack Liskey 282 93 0&09—74 128 131 103 111 82 152 111 138 58 Assembly of God 363 87 07—177 74 137 143 117 111 74 184 144 ,,,,,, Warm wells and springs 1 Eagle oint Spring 777777 35 04—06—75 1 16 124 94 190 170 >200 200 17 Mazama School 972 61 ...... 138 147 120 129 126 197 163 20 US. Air Force 500 31 10~17—72 99 105 73 209 129 163 123 22 Falcon Hts. School 1050 37 05—07474 117 125 96 181 104 >200 >200 23 Alfred Jacobsen N155 30 0&20— 74 108 115 84 221 131 >200 >200 50 Jack O’Connor 700 38 0509— 74 108 115 84 116 83 196 163 56 O‘Connor Livestock 770 30 06— 0&74 117 125 96 123 72 >200 >200 ,,,,,, Slightly warm waters 15 Howard Holliday Spg. 777777 25 05—22— 74 72 76 42 157 56 ____________ 18 We erhaeuser #4 545 22 10— 1&73 79 83 49 159 65 ...... 25 Mellvin McCollum 223 25 05—28—74 98 104 72 178 50 ,,,,,, 26 Claude Shuck 1060 24 05— 30—74 101 108 77 203 84 ,,,,,, 27 Ra Bixler 460 22 05—29—74 95 101 69 182 55 30 R0 ert Langley 350 23 05—04—74 96 102 70 200 78 35 Len Dobry 612 21 05404—74 107 115 84 180 88 36 Lester Brookshire 1210 25 0504—74 83 87 53 173 57 111111 38 Geo. Stacy Co. 916 25 06—05-74 81 86 52 183 61 ,,,,,, 39 Monte Dehlinger 148 26 06—06- 74 99 106 74 207 73 ,,,,,, 40 J. K. O’Neil 33 23 05—04—74 96 102 70 194 88 ______ 43 Dan O’Connor 85 24 05—07—74 90 95 62 207 91 ______ 45 Abe Boehm 155 25 05-30— 74 127 137 109 132 87 ...... 47 Cl de Dehlinger 100 24 06—05—74 112 120 90 189 95 ...... 48 Bi 1 Hill 820 20 06—06—74 71 74 40 217 83 ...... 51 Jack Liskey 95 22 0&09—74 126 137 109 121 115 ............ 54 Town of Merrill 1088 21 02‘ 18—55 86 90 57 155 71 ............ 55 Pope’s Meat Co. 1128 22 0&06—74 69 73 39 161 67 ............ 57 George Carter 67 22 06.06—74 89 94 61 156 83 ............ ‘For sources of measurement and chemical analyses, see table 5. lFournier, R. 0., 1976. “Fournier and Truesdell, 1973, 1974. 'Mixing models based on quartz or chalcedony equilibria and a cold-water component havin a silica concentration of45 mg L " and a temperature of 20°C (Truesdell and Fournier, 1977). "Based on 8‘“0(H20)/6"‘0(SO,) relations and assumed conductive heat loss (McKenzie an ’I‘ruesdell, 1977). HYDROGEOLOGIC APPRAISAL OF KLAMAT H FALLS GEOTHERMAL AREA, OREGON G27 TEMPERATURE, IN DEGREES CELSIUS 100 80 Sp 410 zp 2.40 . ' . . ' . 2.00 _5 +7 “+14 +29 fi @564 (950 23024.7: I35 30 2°"?%57 277 40 1.60 - 3 4 _ N 36\9/15: 018 3 O 38’f 55. 12 "’ 1.20 - <5 3 + Hot water from principal 0.80 _ geothermal areas (>60°C) - Transitional water 040 6) Warm water from the Klamath Hills 27 Sequence number in table 7 o I I l I I I I I I I 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3 10 /To,( FIGURE 8.—Temperature of ground water versus log silica concentration. Straight lines show empirical relations of temperature to equilibrium concentrations of quartz, chalcedony, and silica gel (Fournier, 1977). temperatures. The most likely causes for the failure of the mixing models are the general availability of amorphous silica and the general presence of nonequilibrium chemical conditions in the deposits permeated by the warm waters. A second quantitative geothermometer that has been applied successfully to geothermal systems is the Na-K-Ca geothermometer (Fournier and Truesdell, 1973). Based on worldwide data from springs and wells, the following empirical expression has been for- mulated: log (Na/K) + [3 log (VCa/Na) 1647 273+T — 2.24 = K* where concentrations of Na, K, and Ca are in moles, T is temperature in degrees Celsius, B = 4/3 for waters equilibrated at less than 100°C and 1/3 for waters equilibrated at more than 100°C. Data from table 5 were used to obtain a value of K* for each sample, and this value was then plotted versus the reciprocal of the measured temperature in degrees Kelvin. The results are shown in figure 9 along with a line representing the Fournier-Truesdell expression. Projection of the K* values horizontally onto the line gives the apparent temperature of equilibration in the reservoir. Calculated equilibration temperatures based on the relations shown in figure 9 for two values of ,8 are listed in table 7. The graph shows that the majority of water samples fall into two groups with roughly similar estimated reservoir temperatures. The left-hand group comprises the hot waters of the area; projection of these points onto the line produces estimated reservoir tempera- tures that, in most cases, are lower than actual water temperatures. This result is caused by Na/ K ratios that are higher than expected in relation to mNa ratios and observed temperatures. ‘ The group of data points on the right side of the graph in figure 9 represents warm, and presumably mixed, waters with wide ranges of concentrations of Na, Ca, and K. Estimated temperatures of last rock- water equilibrium for many of these samples are simi- lar to estimates derived for the hot waters (table 7). Five samples give anomalously high estimates. The two major groups of samples in figure 9 are dis- tributed so that a line connecting their centers is nearly horizontal. Thus, values of K* along this line would vary only slightly with large changes in temper— ature. It might be expected, therefore, that mixing models using typical hot and warm waters would change estimated reservoir temperatures only slightly from those estimated for the hot waters alone. In order to test the effect of mixing, trials were made using mixtures of a typical hot water (No. 58: Na=213 mg L“; Ca=28 mg L"; K=4.6 mg L") and averages for warm water between 15° and 25°C (Na=25 mg L": Ca=12 mg L"; K=4 mg L“). With mixtures ranging G28 TEMPERATURE, IN DEGREES CELSIUS 120 100 so so 40 20 o 4-5 I i II I l I i l i I II I 4 + Hot water from principal geothermal areas (>60°C) ‘0 _ 6 Warm water from the Klamath Hills 4 ° Transitional water M 9%“ _ 3'5 _ .36 Sequence number in table 7 3 ‘6 65“” 3.0 — K*'2'5 - +17 022 .1 [21' 1.0 - ‘23 — o l: B= 4/3 (Equilibrium temperatures <100°Cl .5 - _ l:l {3= 1/3 (Equilibrium temperatures >100°Cl 0 | 2.5 2.6 2.7 2.8 2.9 3.4 3.5 3.6 3.0 3.1 103/10,. 3.2 3.3 FIGURE 9.—Temperature of ground water versus log (Na/K) + Blog (VCa/Na). Straight line shows empirical relation to equilibrium reservoir temperatures (Fournier and Truesdell, 1974). from 15 to 45 percent warm water, reservoir tempera- tures estimated by the Na-K-Ca geothermometer ranged only from 74° to 75°C. Additional trials using an average composition for five cold waters (< 5°C) in basaltic aquifers (Na=25 mg L"; Ca=12 mg L"; K=4 mg L“) produced only a slight improvement in estimated reservoir temperatures. For mixtures of 15 to 45 percent cold water, estimated res- ervoir temperatures ranged from 78° to 86°C. Analysis of the terms incorporated in the geother- mometer shows that molal ratios of Na/K decrease by an order of magnitude between the hot waters and the cold or warm waters, but that ratios of m/Na have an offsetting increase of nearly an order of magnitude. When the cation concentrations of hot and warm or cold waters are combined in mixing models and logs of the resulting ratios are added in the geothermometer expression, the net effect of mixing is very small. For this reason, simple mixing models used with the cation geothermometer do not appear to be capable of provid- ing reliable estimates of reservoir temperature for these waters. ISOTOPES IN GROUND WATER Samples of water from a number of wells and springs in the vicinity of Klamath Falls were analyzed for con- centrations of isotopes of oxygen (“‘0) and hydrogen (deuterium and tritium) (table 8). Analyses of these isotopes have been used elsewhere to ascertain the ori- gin and the history of thermal waters. (See for exam- GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 8.—Ar_ralyses of the isotopes oxygen-18, deuterium, and tritium in ground water and Klamath Lake water snot (0/00) 8'“O in sulfate (0/00) Tritium (TU) SDeuterium' (0/00) Sefience Name 0. Hot waters (>60°C) 6 OIT #6 7 Medo Bel Dairy 13 Mills School —14.4G —118.7LW —5.45N 1.6i0.5TW —14.6G —119.4LW —5.45N —14.83”—121.8Lw —4.94N 17 Mazama School —14.2C —117.0C ____________ 28 Olene Gap Spring —13.3G —111.6C —4 82N 7.1i0 6T“ 29 L. Jones spring —13.5C —110.2(’ ____________ —14.6G —117.0LW -1.85N -14.6G —116.6LW —2.13N 1.3105TW —14.75C—120.Oc 44 O. Osborn 46 J . Liskey 58 Assemb y of God Warm waters from the Klamath Hills (21°C—25°C) 45 A. Boehm —6.78N —73.20 +1542N ______ 51 J. Liskey —8.67"' —87.20 +8.69N ______ 53 Suburban Water Co. —10.79"' —96.70 +9.56N ______ Warm waters (20°C—37°C) 1 Eagle Point Spring — 14.3G — 109.70 ____________ 18 Weyerhaeuser #4 —13.9G —109.10 ____________ 22 Falcon Heights School— 14.9G —120.50 ____________ 25 M. McCollum — 14.1G — 113.2O ____________ Cold waters (<20°C) ‘ 2 Shell Rock Spring - 15.35C- 113.1c ____________ 4 Barkley Spring — 15.1 lN— 115.00 ____________ 9 Oregon Water Corp. —14.4C —109.4C ______ 2 4:05” 33 Gabriel Brothers — 16.050— 123.6‘J ____________ 41 US. Air Force — 14.52“— 110.00 ____________ 42 V. Howard — 14.360— 10800 ____________ 49 Tulana Farms Spring — 13.95C— 105.8C ____________ __ Klamath Lake — 12.35" -97.5" ______ 25.7: 1.4TW ‘Referred to Standard Mean Ocean Water (Craig, 1961b) Anal sts C . B. Coplen, Reston, Va. G Geochron Laboratories, Denver. Colo. N N. L. Nehring, Menlo Park, Calif, 0 James O’Neil, Menlo Park, Calif. LW L. D. White, Menlo Park, Calif. TW T. A. Wyerman, Heston, Va. ple, International Atomic Energy Agency, 1974; Taylor, 1974.) OXYGEN-18 AND DEL'TERILTM ISO'I‘OPES The correlation between concentrations of oxygen-18 and deuterium in precipitation is generally given by the expression 8D=85"‘O+d where 6D and 8‘“O are ratios of deuterium and 1“O to H and 160, respectively, expressed in parts per thousand (°/oo) as departures from the ratios in Standard Mean Ocean water (SMOW) (Craig, 1961a, b), and d is typi- cally 10. The "meteoric line” resulting from the above expression is shown in figure 10 along with data from 25 water samples from the Klamath Falls area. The cold, fresh waters of the area, represented by samples 4, 9, 41, 42, and 49 in figure 10 appear to establish a trend parallel to but offset from the theoret- HYDROGEOLOGIC APPRAISAL ()F KLAMATH FALLS GEOTHERMAL AREA, OREGON ical meteoric line. Although one cold-water sample, number 2 (Shell Rock Spring), is anomalous, the trend established by the five remaining samples may be taken as a probable baseline for most locally derived recharge waters. A seventh cold-water sample, number 33, derived from a 1,004-foot well located near Olene Gap, also appears to be anomalous but may represent high-altitude recharge from Stukel or Hogback mountain. Transitional waters are represented by only four samples, two of which seem to have anomalous 8180/5D ratios. The trend of these samples from left to right parallel to the meteoric line is in keeping with a gen- eral trend of decreasing temperature in the four sam- ples and also with their relative positions in the trends of ionic ratios in figure 6. Uncertainties resulting from possible errors of analyses and the small number of samples would seem to preclude attempts to draw gen- eral conclusions from the isotope concentrations in these transitional waters. Three warm waters from the Klamath Hills, samples 45, 51, and 53, show large displacements from the meteoric trend in figure 10 and demonstrate further that these waters differ significantly from both hot waters and other warm waters in the area. (See section on “Relations Among Dissolved Constituents,” this re- port.) The positions of these samples are consistent with the supposition that the Klamath Hills warm waters are mixtures whose components have been sub- jected to low-temperature evaporation. Nine samples of hot water from the three principal geothermal areas are included in the graph of figure 10. These waters have apparently been enriched in 1"0 during their passage through the geothermal reser- voir, with increases in 6180 ranging from 0.8 to 1.4 “/00. The increase is small compared to that observed in high-temperature geothermal systems such as Steam- boat Springs, Nev. or The Geysers, California, which have increases in 6180 as great as 3 to 5 0/00 (White, 1968). The isotope data suggest, therefore, that the Klamath Falls hot waters have not equilibrated at ex- tremely high reservoir temperatures, or, alternatively, that they have been diluted by mixing with cooler wa- ters. These alternatives have been discussed in the preceding section of this report and will be discussed further in the next section. The hotter thermal waters from both the Klamath Falls hot-well area (samples 6, 7, 13, and 58) and from the Klamath Hills (samples 44 and 46) have a fairly small range of 8‘80 values. They differ more widely, however, in their deuterium concentrations and in their displacement from the presumed meteoric line. Assuming these differences to be significant, they in- dicate that two groups of hot waters originate in G29 slightly differing meteoric waters and are subjected to slightly differing reservoir conditions. Thus the isotope data are consistent with the chemical analyses and temperature data for these waters, which together suggest that although the reservoirs and sources of re- charge are generally similar for the two groups, small variations occur as the result of differing local condi— tions. Two samples of hot water from the geothermal area at Olene Gap (samples 28 and 29) differ significantly from other hot waters, principally in having higher concentrations of calcium, I"O, and deuterium. One ex- planation for the high concentrations of "‘0 and deuterium in these samples is that the altitude at which recharge occurs is lower than that of other re- charge waters. An alternative explanation for the composition of samples 28 and 29 is that either the recharge water or the thermal water has undergone low-temperature evaporation. Although neither of these explanations is clearly supported or excluded by the available data, both would imply that the geother- mal reservoir at Olene Gap is separated to some degree from other geothermal reservoirs in the area. Some of the relations suggested by deuterium and 1"0 ratios in figure 10 may be substantiated and clarified by plotting both deuterium and 8‘80 versus a conservative ion such as chloride (figure 11). The ap- parent mixing trend of Klamath Hills warm waters is clearly shown in figure 11, as are groupings of thermal waters on differing trend lines. Samples 28 and 29 (Olene Gap springs) are separated from other thermal waters, and there is the suggestion that sample 17, a _ 40 I I l I l I I l I | + Hot water from principal geothermal areas (>60°C) ‘ 5° “ (9 Warm water from the Klamath Hills I ‘ - Mixed water of low to moderate temperature egstfi/ ‘60 ' 0 Cold water (<20°C) ‘ _ ’3 A Klamath Lake water 0 _ _ . . o\ 70 I8 Sequence numberin tables 7 and 10 6 945/ V “as // 2 {we /’ 2 " 80 ' 9°\'5//’ - n: ‘2, 92‘?” l“: _90_ «EL/em _ 3 LLI 9 0° —1oo —- ._ ~110 — +28 Theoretical ”meteoric” trend - 5 — — Apparent local ”meteoric” trend ‘120 7 Values of 8180 and 5 deuterium relative ' to SMOW (Craig, 1961b) —130 I l l I l l l l I l -—12 —11 —1d 5180 (°/oo) -17 '9 -8 —7 FIGURE 10.—8oxygen-18 versus fideuterium in ground water and lake water. —6 G30 130 I I I I I I I I l I I _ _ ms 160 _ +Hot water from prInCIpaI _ O geothermal areas (>60 ) 14o — ®Warm water from the Klamath Hills 951 - -Mixed water of low to moderate 12° ' temperature ' O 100 °Co|d water (<20 C) 1 r AKIamath Lake water 80 _ sasequence number in tables 7 and 10 _ é +29 3 60 - #2 +23 @53 _ a: 13 , )44 31' 4o _ 53 a u, +17 2 g 20 — .15 _ <5 2 022 3 0 0133 xxl‘loogh‘a ‘ l I l l I I I E -17 —16 —15 —14 -13 -12 —11 -10 —9 -8 —7 —6 —5 z 5130 (°/oo) z. 2 180 I I I I I 5—. 45© é ,_ 160— — 2 Lu ‘2’ 140L— @51 — C) (.3 g 120 — - —-L a: 3 100 — - I U 80 — — 46 +29 60 ~ — 13+5f§6%:28 @53 8 4o — - +17 20 _ ol - 022 41 18 O 033 I 0402 042043025) A r I —130 —120 —1 10 —100 —90 -80 —7o 5 DEUTERIUM (°/oo) FIGURE ll—Chloride concentration versus 60xygen-18 and 8deuterium in ground water and lake water. water having somewhat anomalous chemical charac- teristics, falls on a possible mixing trend with the Olene Gap springs. TRI'FI UM CONCENTRATIONS Four samples of ground water and one sample of Klamath Lake water were analyzed for tritium concen- trations (table 8). These analyses represent only one point in time and hence do not provide a basis for reli- able estimates of residence time in the aquifer. How- ever, some qualitative conclusions are possible. Samples 6 and 46, from OIT 6 and the Liskey hot well, respectively, are clearly older waters, possibly de- rived from pre-1954 meteoric water. Sample 9, from a cold, deep well used for municipal supply, is also an older water which may predate the 1954 atomic-bomb input. Sample 28 (Olene Gap hot spring), has a GEOHYDROLOGY OF GEOTHERMAL SYSTEMS significant amount of tritium, indicating the presence of post-1954 meteoric water. On the basis of tritium concentrations in thermal and nonthermal waters, a mixing model can be con- structed from which the apparent reservoir tempera- ture for the Olene Gap springs can be estimated. Utiliz- ing the tritium concentration in sample 28 (7.1 TU), and assuming that the tritium concentration in the geothermal reservoir is zero, and that of the cold-water component is the same as Klamath Lake (26 TU), the mixing proportions are determined and applied to ob- served temperatures. The calculated reservoir temper- ature is 112°C. When observed silica concentrations and geothermometers are used in this tritium mixing model, estimated reservoir temperatures are 121° and 146°C, respectively, for the chalcedony and quartz con- ductive geothermometers. Although these results are reasonable, no great weight should be attached to them in view of the large uncertainties involved. In particular, the choice of a tritium concentration for the cold-water component is arbitrary and appears to be lower than would be ex- pected for meteoric water in this region (T. A. Wyer- man, oral commun., 1978). It is clear, however, from the form of the mixing expression, that any increase in the assumed tritium concentration of the cold water would result in even smaller calculated fractions of cold water and consequently in lower apparent reser- voir temperatures. Conversely, trials with assumed tritium concentrations in the cold water smaller than about 15 TU result in calculated reservoir tempera- tures that are much higher than those estimated by other means and are probably unreasonable in the light of the chemical nature of the water. If other as- sumptions of the mixing model are reasonable, there- fore, it is probable that the tritium concentration in the cold-water component of the Olene Gap springs is in the range 15 to 30 TU, and that the calculation cor- rectly indicates the presence of a significant fraction of fairly young water in the mixture. A cold-water frac- tion as old as that in sample 9, a cold deep water de- rived from Tertiary basaltic rocks, is apparently ruled out by the mixing model relations. lSO'I‘OPE MIXING MODELS Isotope concentrations in thermal waters can be used to gain insight into several questions that are of paramount importance in assessing the energy content of a geothermal system. These questions are: (1) Do the waters rise from a single extensive reservoir? (2) to what extent are the thermal waters mixtures of shal- low cold water and hotter water from the geothermal reservoirs? and (3) what is the temperature of the re- HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON servoir? Unequivocal answers to these questions prob- ably cannot be obtained from the data at hand, but the graphical analysis described below may point toward probable answers for the Klamath Falls geothermal system. Figure 12 is a large-scale version of a part of the graph of figure 10 showing 80xygen-18 versus 6deuterium. In figure 12, simple mixing models for the geothermal waters have been tested by trial and error. The procedure is based on the following assumptions: (1) Recharge waters for the geothermal reservoir have positions along the observed cold-water trend line; and (2) thermal waters are simple mixtures of these cold waters with a single hypothetical reservoir water hav- ing isotope compositions specified by arbitrarily chosen points, P1 and P2, for example. Hot-water and cold-water fractions of each thermal water can be calculated from the geometric proportions of straight lines connecting hypothetical cold waters, thermal water, and the hypothetical reservoir waters, as is shown for sample 29 and reservoir water P2. Based on an assumed cold-water temperature of 20°C, appar- ent temperatures of the reservoir water are calculated by means of the expression, Tm + frTr + fi‘Tc (1) where Tm is observed temperature in the mixed water, T, and Tr are temperatures in the reservoir and cold water, respectively, and f, and f, are fractions of reser- voir water and cold water. 6 DEUTERIUM (°/oo) -92 I I I I —96 _ + Hot water from principal geothermal areas (>60°C) _ 0 Mixed water of low to moderate temperature // A 0 Cold water (<20°C) 7 -100 - A Klamath Lake water Q’QNx/f) ' ‘ D Hypothetical geothermal reservoir water C8// _104 P I Hypothotical recharge water 0/ ///$ :-Hot water _ 58 Sequence number m tables 7 and 10 :5 // fraction —108 — — -1 12 - oz - /04 -1 16 )— // _Cold-water / fraction ’ +7 ‘120 " / 22. +53 _ // l _124 _ / 330 _ // —128 —/ _ / _132 I I I I -17 —16 -15 -14 -13 -12 5180 (°/ooI FIGURE 12.—Mixing proportions of thermal and meteoric waters based on 8oxygen-18 and 8deuterium relations and hypothetical geothermal reservoir waters. G31 Solving equation (1) for TT, the apparent tempera- ture of reservoir water is determined for each observed thermal water (table 9). For a reservoir water in posi- tion P2, representing an enrichment in 5‘80 of about 31/2 0/00, apparent reservoir temperatures range from 130 to 324°C. This wide range indicates that the simple mixing model fails to account for the observed distribu- tion of thermal waters on the graph in relation to P2. As a further test of the model, chloride concentrations in the hypothetical reservoir water may be calculated in the same manner as the temperatures, on the as- sumption that chloride concentration is a conservative property that is likely to be relatively unaffected by factors other than mixing. The resulting differences in calculated reservoir chloride concentrations (table 9) confirm the conclusion from the temperature calcula- tions that simple mixing with a reservoir water at or near the position of P2 is not an acceptable model. No further calculations were made using this model. More acceptable results were obtained in a second trial that applied the above procedures to a hypotheti- cal reservoir water at position P1 (figure 12). As was true for reservoir water P2, however, the range of calcu- lated chloride concentrations shows that a single re- servoir fluid at or near Pl cannot account for chloride concentrations in all thermal waters. Furthermore, on the basis of additional trials (not indicated in figure 12), it is concluded that no single point of origin for a reservoir water could provide consistent results under the assumed conditions of the model. TABLE 9,—Calculated silica and chloride concentrations and tem- peratures in the geothermal reservoir [Based on oxygen-18 and deuterium mixing models derived from figure 12. See text for explanation] Calculated Calculated Calculated reservoir temperature silica chloride Isotope Chalcedony Quartz concen- concen- mixin geother- geother- tration' tration2 model" mometer‘ mometer‘ (mg L") (mg L") (°C) (°C) (°C) Hypothetical Geothermal Water, P, Sample No. 29 136 114 100 131 155 1 7 1 14 50 80 1 1 9 145 (44, 46) 147 128 186 136 160 6 1 1 1 8 1 1 20 1 1 7 143 7 1 02 84 1 16 1 1 2 1 39 58 155 89 130 140 163 13 95 80 126 108 135 Hypothetical Geothermal Water, P2 Sample No. 29 169 157 130 ____ ____ 17 179 94 137 ____ ____ (44, 46) 232 239 324 1", ”fl 6 187 150 209 u“ -1“ 7 154 161 204 ____ ____ 58 262 173 237 ____ ____ 13 139 149 220 ____ ____ ‘From isotope mixing mode]; assumed silica concentration in cold water is 45 mg L". 2From isoto mixing model; assumed chloride concentration in cold water is 3 mg L‘ '. “From grap, re 12; assumed temperature of cold water is 20°C. ‘Based on calc uluted silica concentrations in first column G32 In this regard, it may also be noted in figure 12 that as the trial points chosen for hypothetical reservoir waters approach the meteoric trend line so as better to satisfy the criteria of consistent and reasonable reser- voir temperatures, required position of recharge wat- ers become more dispersed along the meteoric trend line. For a hypothetical reservoir water near point P1, for example, recharge points are required to range from the vicinity of the Klamath Lake sample at one extreme to points beyond sample 33 at the lower left representing exceptionally high altitude or dilution. On the basis of the above graphical procedure, there- fore, it would seem most probable, although by no means certain, that two or more reservoir waters of slightly differing isotopic compositions exist. One of these waters may be similar in composition to P1. A second reservoir water with significantly different 180 and deuterium concentrations and of differing meteoric origin could then account for the observed compositions of samples 28 and 29, the Olene Gap springs. It is further speculated that Klamath Hills hot waters (samples 44 and 46) and a fairly low-temperature thermal water such as sample 17 (Mazama School) would have to be accounted for by more complex pro- cesses involving mixtures of previously warmed waters or conductive cooling. The data in table 9 show that an isotope mixing model based on hypothetical reservoir water Pl pro- duces consistent estimated chloride concentrations and a fairly small range of reservoir temperatures (116° to 130°C) for Klamath Falls hot-water samples 6, 7, 13, and 58. Although these temperatures are similar to those in table 7 calculated directly from a mixing model based on the chalcedony equilibrium curve (range 99° to 127°C), they are generally lower than the highest temperatures observed in the hot waters. As a further test of the isotope mixing model, silica concentrations calculated from this model for hypothetical reservoir water Pl may be used with the silica geothermometer (Fournier and Rowe, 1966) to estimate reservoir temperatures. Results of these cal- culations, given in table 9 for both chalcedony and quartz geothermometers, compare well with the silica mixing-model results in table 7. These results appear to confirm the isotope mixing model, and they suggest that a reservoir water with an isotopic composition in the vicinity of point P. in figure 12 could be the source of the Klamath Falls hot waters. It is apparent that only the mixing models based on quartz equilibria produce estimated minimum reser- voir temperatures that are generally higher than ob- served water temperatures. For the four Klamath Falls samples that appear to be most reliable (6, 7, 13, and 58), estimated reservoir temperatures range from 135° / GEOHYDROLOGY ()F GEOTHERMAL SYSTEMS to 184°C for the quartz mixing model (table 7) and from 135° to 163°C for quartz equilibrium in the isotope mix- ing model (table 9). These data are not sufficiently numerous to justify a statistically derived best esti- mate, but it appears reasonable to suppose that reser- voir temperatures in all three geothermal areas have a minimum value in the range 145° to 155°C. In the light of all chemical and isotope data, a best guess of 150°C is used in the final section of this report for calculations of heat content in the reservoir. RESERVOIR 'I‘EMPERA'I‘URES ESTIMATED FROM OXYGEN ISO'I‘OPES IN DISSOLVEI) SULFATE The concentrations of oxygen-18 in dissolved sulfate can be used as a geothermometer in thermal waters that have equilibrated for long periods of time in a geothermal reservoir (McKenzie and Truesdell, 1977). It is assumed that exchange between the oxygen isotope in the sulfate ion, Ix0(804), and the oxygen isotope in water, l"0(H20), will have attained equilib- rium after residence times that range from 2 years at a temperature of 300°C to 18 years at 200°C (McKenzie and Truesdell, 1977, p. 53). Concentrations of 18O(SO4) were determined for six hot waters from the Klamath Falls geothermal areas (table 8). As residence times for these waters probably exceed 18 years (see previous section, “Tritium Con- centrations”), equilibration between "‘O(SO4) and '“O(H20) may be assumed to be virtually complete if reservoir temperatures are not far below 200°C. Based on this assumption, calculated reservoir temperatures for the six hot waters range from 135° to 196°C, assum- ing only conductive heat loss in the rising thermal waters, and from 130° to 179°C, assuming single-stage steam loss at the surface temperatures of the hot wat- ers. The two lowest values in each case are for the Liskey and Osborn hot wells in the Klamath Hills. Values for the Olene Gap and Klamath Falls samples range from 179° to 196°C for the conductive model. These estimates of reservoir temperature are clearly at variance with the estimates derived from concen- trations of silica and water-oxygen isotopes (table 7) and explanations for the discrepancy must be sought. Two types of explanations can probably be discarded immediately. The bulk of the chemical and isotope data favors the concepts of low reservoir temperatures and largely undiluted thermal waters. Any explanation for the high sulfate-oxygen temperatures involving dilu- tion of the thermal waters, therefore, would not only be opposed by this evidence but would also tend to in- crease the apparent reservoir temperatures and widen the gap between the estimates. Furthermore, the con- cept of mixing and'dilution by nonthermal waters is HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEO'I‘HERMAL AREA, OREGON also opposed by the fact that ll"0(SO4) concentrations are independent of chloride concentrations in the thermal waters. Explanations involving the solution of gypsum from shallow evaporite deposits may also be ruled out, inasmuch as the sulfate from such deposits would have high oxygen-isotope concentrations that would result in lower apparent reservoir temperatures. One possible explanation for the high reservoir tem- peratures estimated from the sulfate-oxygen isotope could involve the formation of isotopically light 804 in volcanic rocks during their emplacement at high tem- peratures. This 804, taken into solution by geothermal waters circulating in faults of much later origin, would react with the lower-temperature thermal water in a process that possibly has not yet reached equilibrium at the low indicated temperatures. Although this ex- planation could account for the shift toward the ob- served isotope concentrations, it is somewhat unsatis- factory in that it requires the presence of high- temperature sulfate-bearing minerals which are not prevalent in volcanic rocks. A second speculative explanation is that the sulfate-oxygen isotope has equilibrated in a deep res- ervoir that underlies the shallower zone tapped by thermal wells and springs. Hot water, presumably at or above the temperatures indicated by the sulfate- oxygen geothermometer, rises into the shallow, cooler reservoir where most constituents, including silica, equilibrate at the lower temperatures. The'sulfate- oxygen isotope fails to attain equilibrium with the water-oxygen isotope in this reservoir as the waters rise more quickly along faults and fractures in the shallow geothermal system. Calculations based on ex- pressions given by McKenzie and Truesdell (1977) show that complete equilibration at 150°C would re- quire more than 77 years. A completely satisfactory explanation of the discrep- ancy between the estimated reservoir temperatures is not available, and, therefore, no final conclusion has been reached in this study. The phenomenon described here is similar, however, to others observed in low- temperature thermal waters in the High Cascades and in the Gila River area of Arizona where high sulfate- oxygen isotOpe temperatures are also indicated (R. H. Mariner, oral commun., 1979). These discrepancies will apparently require further study of the chemical relations in thermal waters derived from thick volcanic sequences such as those at Klamath Falls and the areas mentioned above. Three additional waters analyzed for l"0(804) (table 8) are warm waters from the Klamath Hills. Results of these analyses, which show high concentrations of l80(804), further confirm the belief that the waters are mixtures containing evaporation residues. Estimates G33 of reservoir temperature have not been made for these waters. ‘ TEMPERATURE DISTRIBUTION AND HEAT FLOW AREAL DISTRIBUTION OF TEMPERATURE The distribution of temperature measured in ground water of the Klamath Falls area is shown on plate 3. The isotherms of the map are generalized from data that have considerable scatter in many areas. Much of the scatter is due to differences in temperature be- tween deep wells and shallow wells in the same area, the deeper wells having generally higher tempera- tures. The isotherms, therefore, are at best a rough guide to areas where wells of moderate depth (gen- erally 150 to 500 feet) may be expected to encounter water with temperatures similar to those indicated by the contours. The map shows that although warm water occurs at a large number of isolated places, it is found princi- pally within the three closed 60°C contour lines that contain the Klamath Falls urban area, Olene Gap, and the southwest flank of the Klamath Hills. VERTICAL DISTRIBUTION OF TEMPERATURE Temperature profiles measured in 41 unused, non— flowing wells and two test holes are shown on plate 4. The locations of these holes are shown in figure 13, and data pertaining to the profiles are given in table 10 where the profiles have been divided into three groups on the basis of their dominant characteristics: Mainly Conductive Profiles, Cold Convective Profiles, and Ir- regular Profiles. Most of the wells in the first group, Mainly Conduc- tive Profiles, penetrate fine-grained lacustrine sedi- ments that range from silt and clay of Holocene age to shale and diatomite of the Pliocene Yonna Formation. Only three of the holes are known to penetrate basaltic rocks. The upper parts of the individual profiles are difficult to trace on plate 4, but nearly all show evi- dence of downward convective transport in the upper tens of feet. In addition, many have temperature ir— regularities at depths that, in some instances, corres- pond to known permeable zones in the rocks. Thermal gradients in the conductive profiles are generally high, ranging from 50° to 180°C per kilome- ter (°C km"), even though temperatures are relatively low. The highest temperature observed is 38°C at 500 feet in CW 3, located at 39/9—3ddd, just south of Al- tamont. The second category in table 10, Cold Convective Profiles, includes 10 wells characterized by low tem- G34 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 121°45 122°00 \ \l‘\ a ' a” V W3 Klamath Falls 3% ‘1 Geo hermal Zone cows 42°15 _ o PH20 Klamath Hills Geothermal Zone OC|(3) (1.44) + 42°00 CATAGORIES 0F PROFILES v Downward convective © Upward convective 0 Mixed convective + Quasi-conductive Lacustrine and alluvial sediments of Quaternary age [:1 Sedimentary and volcanic rocks of Quaternary and Tertiary age —— Major faults 121°3O CM18 (1.4l+ .JG17 LC25+ (2.2) OCI O +SW1 (1.3) FlK15. O——O 10M|LES 4| l 10 KILOMETERS FIGURE 13.—Geologic sketch map of the Klamath Falls area (after Peterson and McIntyre, 1970), showing locations and categories of wells in which‘ temperature profiles were obtained. For wells with quasiconductive (linear) profiles, a lower limit of heat flow, based on an assumed thermal conductivity of 1.8 meal cm“s"°C“, is shown in parentheses. Heat-flow determinations at JLl and 0C 1(3) are also shown. (Adapted from Sass and Sammel, 1976.) peratures and small thermal gradients. Most of these wells are located in the lowlands of the area and are relatively shallow. The profiles show that downward flow of cool water occurs to depths of 300 feet at places, indicating that the lake-bed sediments are relatively permeable at these locations. Most of the areas dominated by convective flow and cold water are ground-water recharge areas. Significant amounts of recharge may also occur elsewhere as indicated by profiles that, although they are dominantly conductive, show evidence of down- ward circulation in their upper 50 to 100 feet. An example is the US. Geological Survey test hole, 0C1 (3). One of the deeper wells, OIT 3, shows downward convection to a depth of more than 900 feet. This well, with a maximum temperature of 28°C, is located less than 2000 feet from one of the three hot wells at the Oregon Institute of Technology (fig. 4). The tempera- ture difference between these wells (about 60°C) and the difference in static water levels (210 feet) demon- strate the profound effect exerted by local structure and stratigraphy on subsurface temperatures and ground-water regimes. The third category in table 10, termed “Irregular,” includes all hot wells in which profiles were obtained as well as several cold wells. Most of these wells pene- trate basalt flows of the Yonna Formation; two of them, HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON G35 TABLE 10.—Data related to temperature profiles obtained in unused wells and heat-flow holes [*Gradient used in heat-flow calculations. See text for explanation] Altitude De th to Thermal Maximum Depth land surface basa t rock“ gradient.4 Intervals temper- Name Symbol' Location ft. CompletionZ ft. above MSL °C Km“ ft. ature °C Remarks and heat- flow estimates Mainly conductive profiles *Robert Puckett ____________ TRP 39/8—31acd 168 OH(30) 4130 145 *55 50—168 14 On terrace below rid e. Diatomite and shale overfilyjing basa t. Minimum heat flow 1. 0 H . *Cass Windsor .............. CW3 39/9—3ddd 500 0H (<300?) 4115 N *170 80—490 38 "Gray shale.” Minimum heat flow 3.1 HFU. *George Her esheimer GH12 39/9— 12bcd 635 OH(19) 4140 N *138 260—490 36 “Gray shale." Minimum heat flow 2. 5 HFU. *E. D. Danie .............. ED25 39/9—25aaa 231 OH(20) 4090 N *90 130—230 16 Safig‘émd clay. Minimum heat flow 1. 6 *Joe Bair __________________ JB35 39/9—35cbb 171 OH 4095 N? *86 33— 165 13 "Green shale" reported in hole V4 mile north. Minimum heat flow 15 HFU. H. Waggoner ______________ HW18 39/10-18bcb 375? OH(140) 4175 N 50 100—245 15 Downward convective above 100 ft. “‘11. Campbell ______________ ”r HC 39/10—21bba 453 OH 4120 N *47 33— 158 12 USGS test hole. Profile measured 6 days after drilling com leted. Hole caved to 158 ft. Mostly cla low 60 ft. Minimum heat flow 0. 8H HFU *Ben Adair ________________ +BA 39/10—30ada 238 OH 4095 N? *80 65—238 16 Downward convective above 60 it. Probably penetrates lake deposits. Minimum heat flow 1.4 HFU *Edwin Jenkins ____________ TEJ30 39/10—30add 178 OH 4090 N? *~80 65—175 14 Upward convective curve over entire pro- file. Probably penetrates lake deposits. Minimum heat flow 1. 5 HFU *V & W Ranch ____________ VW3 40/8—3cd32 236 ,,,,,, 4090 ______ *~150 65—236 19 On irrigated valley floor. Irregular profile, downward convective above 40 ft. Prob- ably penetrates lake deposits. Minimum heat flow 2.6 HFU. *Charles Matney ____________ CM18 40/10— 18cbb 470 OH(60) 4080 ...... *~75 39—460 21 Irregular profile; upper 60 ft cooled by con— vection. Probably penetrates lake depos- its. Minimum heat fiow 1.4 HFU *Larry Clark ______________ LC25 40/10—25ccb 130 ______ 4120 ...... * ~l20 65— 125 16 Downward convective, above 40 ft; adient decreases below 120 ft. Probab y pene— trgtglsFlfike deposits. Minimum heat flow *Bill Hill __________________ BH26 40/10—26bca2 350 ______ 4155 ,,,,,, ‘~60 195290 16 Irregular profile; downward convection upper 140 ft. Probably penetrates Yonna Formation. Minimum heat flow 1.0 HFU. *U.S. Bureau of OCI(3) 41/9—3dbb 608 Plug 4080 N *78.6 180—580 25 USGS heat-flow hole. Uniform gradient Reclamation (JD. below 140 ft; cooled b convection, ilrfi‘per O’Connor, lessee). 25 ft. Silt and clay. eat flow 144 *Suburban Water Co. ...... SW1 41/10—1cdd2 335 OH(35) 4065 285 *74 80—285 20 Downward convective above 70 ft; profile offset at 285 it at top of basalt. Minimum heat flow 1. 3 HFU. Ray Kramer ______________ RK15 41/10—15abc 241 OH(42) 4090 80 ~80 10—240 15 Gradient given for overall (profile although profile shows downwar convection be- tween 10 and 200ft. Mostly clay and sand; two thin basalt layers above 100 0ft. *U.S. Fish and Wildlife ,,,,,, F & W 40N/2E— 13d 1100 OH 4080 N *90 195—655 25 Downward convection above 110 ft; uniform (California) gradient, 150—630 ft. Silt, clay, some sand in lake deposits. Minimum heat flow Cold convective profiles 1'6 HFU William Smith ______________ TWS 38/9—17bbb 135 OH(19) 4240 100 ____________ l4 Convective 80—120 ft; gradient increases 120-130 ft (gray basalt). Oregon Institute of OIT 3 38/9—17cdd 1130 OH(700) 4353 205 ____________ 28 High gradient begins at about 940 ft; de- Technology #3. creases toward bottom. (“Brown-shale" between basalt flows.) Bill McConron ____________ fBM 39/8—28dcb 105 P(84—104) 4160 3 ,,,,,,,,,,,, 14 Nearly isothermal profile. Sam Shaw ________________ tSS 39/9—30ccd2 110 OH(49) 4090 N ____________ 11 Temperature decreases, 40—90 ft; gradient increases, 90—110 ft. (top of “green clay"). H. Waggoner ,,,,,,,,,,,,,, 1‘HW 39/10—18bcd 100 P(30—62) 4170 N ____________ 12 Nearly isothermal profile. Clay, sand, OH(62) " een shale," “black sandstone." Klamath Irrigation KID19 39/10—19aca 450 OH 4120 N ~30 300—390 16 US test hole. Nearly 1sothermal 20—300 District. it; gradient increases at 300 ft (silts); col- der zone in sands at 160 ft. Ed Gardner ________________ EG17 40/8—17bad 175 OH(99) 4110 30 ____________ 10 Small gradient above 100 ft; becomes isothermal below 140 ft. Stratified vol— canic rocks. J. E. Frost ________________ ’rJF 40/9— 5dda 310‘? ______ 4115 N ____________ 13 Profile becomes colder, 30—90 ft. Ray Dillon ________________ RD6 40/9—6aba 930 OH(48) 4090 490? ~20 330—490 19 Convective cooling above 300 ft. Probably Yonna Formation. ’I‘ulana Farms ______________ TF19 40/9—19dcc 223 OH 4080 N ~40 100- 160 13 USGS test hole. Conductive gradient begins Uin silty cla s below sandier zones. *U,S. Bureau of JLI 41/8—13acd 605 Plug 4080 N "16.8 245—605 15 S heat- ow hole. Upper 190 ft cooled Reclamation (Jack Uby convection. Clay, silt some sand. Heat Liskey, lessee). flow 030 HFU Irregular profiles Presbyterian Inter- PH20 38/9—20acb3 1584 OH(1003) 4370 841 ,,,,,,,,,,,, 92 Gradient very high, 550—650 ft; decreases community Hospital. to isothermal below 1,000 ft. Yonna For< mation overlying Tertiary basalts. W. A. Hughes ,,,,,,,,,,,,,, BH28 38/9— 28dcd 266 OH(256) 4160 95 ,,,,,,,,,,,, 94 Gradient very high above 150 ft; isothermal below 160 ft. Yonna Formation; clay, gravel, interbedded basalt. Richard Po e and PF29 38/9—29aad 194 OH(194) 4200 M. _______________ 90 Isothermal 96— 150 ft; temperature de- Fred Fan on. creases below 150 ft. Hottest tempera- tures at probable top of basalt. Assembly ofGod Church "HAG33 38/9—33aba 363 P(170‘ 185) 4150 162 ____________ 88 Very high gradient, 42—180 ft; isothermal, (333—363) 190—350 ft. Probably Yonna Formation. *Wiard Park District ________ WP2 39/9—2cdd 1018 OH(19) 4140 N ~160 580—948 51 Convective cooling above 90 ft; relatively uniform 1gradient below 580 ft except for bulge be 800 ft. Bulges correspond to occurrences of water-bearing sand layers. Sands, silts, clays. Minimum heat flow 3.0 HFU J, D. O’Connor ............ KLADZ 39/9—9bcb 275 ______ 4085 ,,,,,, ~180 90. 275 28 Convective cooling above 90 ft; upward (ave) convective bulge, 100—260 ft. Probable lake deposits J, D. O’Connor ____________ KLADI 39/9—9cbd 225 ...... 4115 __________________ 20 Convective cooling above 180 ft. Probable lake deposits G36 GEOHYDROLOGY ()F GEOTHERMAL SYSTEMS TABLE 10.—Data related to temperature profiles obtained in unused wells and heat-flow holes—Continued Altitude Depth land surface basa t rock“ gradient4 ft. above MSL Name Symbol‘ Location ft. Completion2 Maximum lntervaP temper- ft. oC Km’I ft. ature 0C De th to Thermal Remarks and heat-flow estimates Oregon Water Corp ,,,,,,,,, OWCG 39/10—60db 732 OH(147) 4310 Marion Barnes ,,,,,,,,,,,, MB14 39/10—14ccb 380 OH(41) 4120 Klamath Falls Oil Co. #1 _.KF01 40/9—2daaZ 1965 4125 Leon Andrieu ______________ 1' LA 40/9— 28adb 128 OH(23) 4120 Vern Peugh 40/10—6cca 120 0H(31) 4080 Joe Glodoski ,,,,,,,,,,,, .. JJG 40/10~17ccb 189 OH(60) 4075 O’Connor Livestock Co. “HOCI 41/9—1bcd 105 ,,,,,, 4155 O’Connor Livestock Co. __.,OC7 41/10—7dba 770 OH(79) 4120 Very low adients above 520 ft; warm bulge beow 520 ft; nearly isothermal, 7W730 ft. Top of basalt layer at 514 ft. Very high gradient above 190 ft. WaterA bearin sands in "green claystone” ac: count or bul es below 190 ft. Basalt reporte at 600 ft (questionable— may be Yonna Formation). Nearly isothermal, 740—1320 ft. Convective cooling to 60 ft (water-bearing gravel); top of conductive gradient in "shale." Convective cooling, upper 40 ft; gradient increases, top of "blue-green clay" at 110 ft. Sand and clay. Convective cooling. upper 50 ft; gradient decreases in blue clay (60»160 ft); isothermal in coarse to fine sand (160— 189 ft). Convective cooling above 60 ft. Probably Yonna Formation Moderately high gradient to 100 ft ("hard green shale, water-bearing"); isothermal, 100—490 it ("hard green shale” and "green clay"). 280‘ 380 55 120—620 31 'Location on plate 3 and profile on plate 4 identified by this symbol, 1’ Shallow, cold well, not identified on plate 4. ’OH:Open hole below depth given (ft); P=Perforated in interval given (ft ). “From driller’s log; N:Not encountered. ‘In general, the maximum gradient measurable over a significant depth interval; in wells designated by *, the conductive gradient for the given interval. "Interval over which thermal gradient was determined PH 20 and OWC 6, also penetrate older Tertiary basalt. Three irregular profiles in the upper right corner of plate 4 probably represent conditions at many places in and near the principal geothermal areas. These pro- files, PF 29, BH 28, and AG 33, show typically high thermal gradients through relatively impermeable shale and clay of the Yonna Formation. Beneath these zones of high gradient, the profiles became isothermal in layers of permeable basalt and volcanic sediments. Profiles such as these suggest that hot water moves laterally in the permeable basalt layers and that this lateral movement may account for the widespread dis- tribution of warm ground water in places such as the lowlands south of Klamath Falls and Altamont. Most hot wells in the vicinity of Klamath Falls and Olene Gap have irregular but generally convex- upward temperature profiles. To the extent that these profiles represent formation temperatures they indi— cate that both upward and lateral convective transport of heat occur in the aquifers. It is assumed that wells penetrating the Yonna Formation in the Klamath Hills geothermal area would show similar profiles. On the basis of the temperature profiles and the known discharge of thermal waters, these three areas are be- lieved to be the principal thermal discharge areas. In the foregoing discussion it is tacitly assumed that the temperature profiles measured in most wells repre- sent temperatures in formations surrounding the wells. Some of the data make it clear, however, that many profiles do not accurately represent formation temperatures. Discrepancies between temperatures measured in wells and actual formation temperatures are usually due to one or both of two common occur- rences: vertical convective flow in the well itself, and in cased wells, vertical flow in the annular space between the casing and the formation. Such vertical flow occurs in response to both temperature gradients and hydrau- lic gradients in the formation penetrated by the wells. An isothermal profile caused by convective flow was measured in well OC 7 (pl. 4). The temperature profile in this uncased well is isothermal at 30°C between depths of 120 and 500 feet, the greatest depth to which measurements were made. Subsequently, when the well was pumped, the temperature of the water was found to be 37°C. It is likely, therefore, that under non- pumping conditions, water with a temperature of about 30°C enters the well at shallow depths and flows downward with sufficient velocity to maintain an isothermal profile. Thus, temperature gradients meas- ured in this well have little relation to geothermal gra- dients in the formation. A second example can be seen by comparing temper- ature profiles obtained in two of the deepest wells in the area, PH 20 (1,584 feet) and OIT 6 (1,805 feet). These wells are located about 500 feet apart and drillers’ logs indicate that they penetrate the same formations. The temperature gradient measured in well PH 20, however, ranges from 330°C km'l at about 575 feet depth to nearly isothermal at 1,475 feet, whereas the gradient calculated from the bottom-hole measurements in OIT 6 over the same interval is more nearly uniform at about 180°C km“. The bottom-hole measurements are believed to provide a good estimate HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON of the actual thermal gradient at this location in spite of small errors in absolute temperature that may have been introduced by the cable-tool drilling. Therefore, both the high gradient and the nearly isothermal gra- dient measured in PH 20 are almost certainly the products of convection in the well and are not reliable measures of gradients in the formation. It is of interest to note also that the position of the sharp break in bottom-hole temperatures in the profile of OIT 6 near the 1,740 foot depth (pl. 4) corresponds to that of a gravel layer between basalt flows noted in the driller’s log. The thermal discontinuity is probably due to the rapid movement of lower-temperature water through the gravel layer. Small lithologic and thermal discontinuities of this kind are difficult to detect in temperature profiles obtained in water wells, but they may strongly affect the measured geothermal gra- dients at some places. Several examples of stratigraphic discontinuities , were obtained in the temperature profiles. Wells OIT 3, OWC 6, and MB 14 had lower temperatures when pumped than the maximum temperatures measured under static conditions. These wells appear to derive much of their discharge from cool strata at fairly shal- low depths. The decreases in temperature observed during pumping may be taken as evidence of low verti- cal permeabilities in the formations and the stratifica- tion of the thermal and nonthermal waters at these locations. The examples cited are believed to represent condi- tions that occur widely in the shallow hydrothermal system at Klamath Falls. They illustrate the difficul- ties that may be encountered in attempts to obtain re- liable geothermal gradients in holes not especially con- structed for this purpose, and they serve as a caution that the thermal gradients reported in this study should not be used uncritically in evaluating geother- mal prospects in the area. CONDUCTIVE HEAT FLOW Two 600-foot holes were drilled by the Geological Survey in the Lower Klamath Lake area in order to obtain heat-flow information. These holes and the data obtained from them are described in detail in Sass and Sammel (1976). The temperature profiles from these holes are shown on plate 4. The first hole, JL 1, is on reclaimed marsh land in Lower Klamath Lake near the California border at T41, R8, Sect 13 acd. The hole penetrates lacustrine silt and clay to a depth of 605 feet. The mean thermal conductivity of cores from the hole is 1.82 meal cm" sl °C". For the interval between 167 and 587 feet the average gradient is about 16.75°C km". The heat flow at the site, therefore, is calculated to be 0.30 heat flow units (HFU). G37 The second hole OC 1 (3), was drilled at a location northeast of JL 1, about 1 mile southwest of the Klamath Hills. For this hole, the mean thermal con- ductivity is 1.83 mcal cm“s‘l °C", the temperature gra- dient in the interval 174 to 577 feet is 786°C km", and the heat flow, therefore, is 1.44 HFU. The heat flow calculated for hole J L 1 (0.30 HFU) is too low to represent actual heat loss from the earth’s interior over a broad area. Convective cooling by re- charge of ground water to depths greater than 600 feet must be invoked in order to explain the observed heat flux, which, in the light of data discussed below, may be anomalous, even for the supposed recharge areas of Lower Klamath Lake. Heat flow in the second hole, OC 1 (3), (1.44 HFU), is near the expected lower limit for conductive heat flow in the Basin and Range province (Sass and others, 1971). Data obtained from other conductive tempera- ture profiles in the area suggest that hole OC 1 (3) probably represents near-normal conductive heat flow over much of the Lower Klamath Lake basin. The basis for this conclusion is discussed in the following para- graphs. A possible range of heat-flow values in the Lower Klamath Lake basin was determined by using temper- ature gradients from 15 conductive temperature pro- files obtained in the basin (table 10). These gradients were combined with the mean thermal conductivity de- termined for heat-flow holes J L 1 and OC 1(3), which is believed to be a representative value for much of the lower basin. Each of the wells in which these profiles were obtained penetrates unconsolidated to semicon- solidated lacustrine deposits in the Lower Klamath Lake valley. The resulting estimates of heat flow (table 10) are regarded as minimum values for the sites. Calculated heat flows for the 15 wells range from 0.8 to 3.1 HFU, with the highest values coming from wells nearest the Klamath Falls and Olene Gap thermal areas. Seven of the wells in the southern part of the valley appear to represent average conditions over a wide area. Heat flows calculated for these wells range from 1.3 to 1.6 HFU. The calculated heat flow for DC 1(3), 1.44 HFU, therefore appears to represent near- normal heat flow over much of the Lower Klamath Lake graben. ‘ The wide range of the estimated minimum heat flows for the remainder of the wells (0.8 to 3.1 HFU) tends to confirm the conclusion, based on variations in the tem- perature profiles, that localized geologic and hydrologic conditions determine the temperature regime at most places. Many of the higher temperatures attained in the profiles appear to be caused by the insulating effect of poorly conductive unconsolidated deposits overlying basalt aquifers. Others are apparently influenced by G38 lateral convective flow of heat and water in permeable zones within the basalt sequences. In general, the ob- served temperature profiles do not provide reliable es- timates of regional temperature gradients and regional heat flux for areas outside the Lower Klamath Lake basin. CONVECTIVE DISCHARGE OF HEAT Estimates of heat discharged from springs and hot wells in the three principal geothermal areas have been calculated from measured and estimated ground-water discharge at land surface. An average base temperature of 12°C is assumed for ground water near the land surface. Results are given in table 11. In the absence of detailed studies, the estimate of spring discharge into Upper Klamath Lake is little more than a guess. The remainder of the figures in table 1 1 are probably reasonable approximations in the light of presently available data. Heat discharged at the water table or by the mixing of thermal and nonthermal ground water in the vicinity of the three geothermal areas would probably add large heat fluxes to those listed in table 11. In assessing this thermal discharge, one may compare the temperature distribution shown on plate 3 with the ’ contour map of the shallow ground-water surface (pl. 1). In the vicinity of Klamath Falls, for example, shal- low ground water apparently flows from the highlands surrounding the city, traverses the zone of highest temperatures, and flows southward across the valley floor into an area where ground-water gradients are very low. The temperature contours suggest that the thermal waters mix and cool rapidly in the immediate vicinity of the principal thermal area and that slow convective upwelling and additional cooling occur in an extended area to the south. This simple conceptual model is undoubtedly compli- cated in actuality by the stratified flow of hot water in basalt layers in the Yonna Formation. The general patterns of fluid flow and thermal mixing postulated above, however, are in accord with the evidence of the chemical data, and these donvective flow patterns are believed to occur in the vicinity of each of the‘three principal thermal areas. Because the patterns of lateral and vertical convec— tion are complex in detail and are not well known, it is not possible to make a reliable estimate of the total convective heat discharge. On the basis of several reasonable assumptions, however, a crude estimate has been made of probable upper and lower limits for the total heat discharged by convective flow. For this estimate, it is assumed that ground-water flow within 500 feet of land surface in the vicinity of the geother- mal areas occurs mostly in the Yonna Formation. The GEOHYDROLOGY ()F GEOTHERMAL SYSTEMS TABLE 11.—Mass and thermal-energy discharge in ground water from thermal springs and wells in the Klamath Falls area Mass flow Thermal energy‘ (gal d“) (105 cal s") Springs 0 ene Gap ______________________________ 30,000 1.2 Eagle Point and Upper Klamath Lake.2 __43,000 .4 Wells Residences, schools, and commercial establishments? ______________________ 895,000 18.1 Klamath Hills (irrigation and stock wells) _______________________________ 1,900,000 1.8 Totals ______________________________ 2,868,000 21.5 xHeat discharge above datum of 12°C assumed for shallow ound water. 2Spring discharge into Upper Klamath Lake estimated by 0 of lake surface. Estimate not reliable. 3Based in art on numbers of wells, flow rates, and temperatures described by Lund, Culver, and vanevik, 1974. rvation of winter thawing transmissivity of the aquifer is estimated to be in the range 1,000 to 5,000 ft2 (1“. Total length of the three thermal discharge areas is probably greater than 3 and less than 5 miles, and the ground-water gradients away from these discharge areas may range from 20 to 50 ft mi". Assuming, finally, that ground-water tem- perature decreases from 95 to 12°C, convective heat dissipated in the shallow ground-water body or dis- charged at the water table is calculated to be in the range 2 X 106 to 30X 106 cal s“. Combining these figures with the estimate of thermal flow from springs and wells (table 11), the total convective flux in the vicinity of Klamath Falls is estimated to be in the range 4 x 106 to 32 x 106 cal s". The range of values given above may be compared with an estimate of geothermal heat being utilized at present for space heating and other purposes. Lund and others (1978, p. 45) estimate that about 6 megawatts per year (1.4 X 106 cal s") are currently utilized in some manner. The geothermal heat actually used is, therefore, a small fraction of the heat now being dis- charged by natural or artificial means in the Klamath Falls area. INTERPRETATIONS AND CONCLUSIONS REVIEkV OF EVIDENCE RELATING TO THE GEOTHERMAL RESOURCE GEOLOGIC EVIDENCE Known geothermal reservoirs in the Basin and Range province appear to attain their high tempera- tures in two ways: (1) by conduction of heat from a magma body emplaced at fairly shallow depths, and (2) by circulation of ground water to depths at which the normal thermal gradient in crustal rocks provides the required temperatures. These two mechanisms may exist in combination. HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON At most places where the first of these two mecha- nisms provides heat for known geothermal systems, the magma bodies are composed‘of silicic melts, often of crustal origin, that are both large enough and young enough to retain significant amounts of heat at the present time. In southeastern Oregon, silicic volcanic activity has occurred during the last 10 million years in a westward-trending progression which culminated in the extrusion of rhyolitic domes and obsidian flows at Newberry Volcano, near Bend (MacLeod and others, 1975). The youngest silicic volcanic rocks at this loca- tion are less than 10,000 years old. In the Klamath Falls area, however, no silicic volcanic rocks occur at the surface and none has been penetrated by wells. When the trends of silicic volcanic activity in south- eastern Oregon are extrapolated into the Klamath Falls area these trends suggest that if silicic intrusives existed, they would probably be about 4 million years old. On the basis of calculations made for the Long Valley, California, geothermal area (Lachenbruch and others, 1976), it is concluded that 4-million-year-old buried domal masses of average size would have cooled to near ambient temperatures by the present time. Thus, the available geologic evidence makes it appear improbable that the geothermal system at Klamath Falls is heated by silicic intrusive rocks, although this possibility is not ruled out. Young volcanic rocks in the Klamath Falls area, some of them dating from late Pleistocene time, are of basaltic composition. At depth they probably occur as dikes and plugs in conduits that formerly conducted basaltic magma to eruptive centers and flow vents. These conduits probably extend deep into the crust, but there is little likelihood that they contain high- temperature rocks at depths shallow enough to be reached by circulating ground water. The second of the two processes by which geothermal fluids are heated, circulation of ground water to great depths in a temperature field created by the earth’s geothermal gradient, requires the presence of perme- able channels through which flow can occur. The large Basin and Range faults that bound the Klamath Falls graben could provide the required channels as similar faults are assumed to do elsewhere in the Basin and Range province (White, 1968; Olmsted and others, 1975). The depth that would have to be reached by the Klamath Falls thermal waters in order to attain a temperature of 150°C is 15,000 feet if the geothermal gradient is as low as 30°C per kilometer. If the gradient is higher, as would be expected if rocks of low thermal conductivity overlying the reservoir created a blanket- ing effect, required depths would be smaller. Evidence from earthquakes in southeastern Oregon (Couch and G39 Lowell, 1971) suggests that faults in this area may extend to depths far greater than 15,000 feet, and it is probable, therefore, that these faults and associated fracture zones form the channels in which the Klamath Falls thermal waters circulate. GEOPHYSICAL EVIDENCE Temperature measurements in springs, wells, and test holes provide the only direct geophysical evidence relating to the geothermal resource at Klamath Falls. The temperature measurements show that, although warm ground water occurs throughout the lower Klamath Lake valley, hot water (>60°C) is discharged principally in three elongate zones associated with major faults. Hot water may also be discharged from concealed faults beneath the lacustrine deposits of both upper and lower Klamath Lake valleys, and the pres- ence of water with temperatures greater than 30°C at a few places in the Lower Klamath Lake valley may be evidence of this occurrence (see pl. 3). Temperature gradients observed in water wells are generally higher near the principal geothermal areas than elsewhere, but few of these gradients are reliable indicators of crustal temperature regimes. Most of the temperature profiles obtained for this study are strongly affected by the convective transfer of heat in moving ground water and hence cannot be used for estimates of conductive heat flow. A few conductive thermal gradients obtained in the Lower Klamath Lake Basin suggest that heat flow through the lake sediments may be on the order of 11/2 HFU. This value is at the low end of expected values for the Basin and Range province. Although this evi- dence cannot be considered highly reliable, it greatly decreases the probability that a source of magmatic heat exists at shallow depth beneath the Lower Klamath Lake graben. Temperature measurements from the upper Klamath Lake basin and from the upraised blocks surrounding the basins are too scanty to permit conclusions regarding possible magmatic sources of heat in these areas. Indirect geophysical evidence obtained from gravity and aeromagnetic surveys has been described and in- terpreted by D. L. Peterson in a previous report (Peter- son, 1976). In general, the gravity and aeromagnetic trends coincide with the dominant northwest- southeast fault trend through the area. The data indi- cate that the lowlands are sunken blocks bounded by steeply dipping normal faults and covered by great thicknesses of low-density sediments. Gravity model- ing suggests that the graben structure may be 3,000 feet deep in the area southeast of Klamath Falls and may be more than twice as deep southwest of Klamath Hills. G40 The gravity data also indicate the presence of struc- tures that may be only suspected from surficial evi- dence. Examples are a possible northeast-southwest trending fault passing through Olene Gap and an east-west trending fault near the north end of Upper Klamath Lake. A major gravity anomaly in the area southwest of Keno may be related to rocks of the adja- cent eruption center. Of these anomalies, only the pos- tulated fault through Olene Cap is related to known geothermal occurrences. Peterson’s interpretation of the aeromagnetic map indicates that most magnetic anomalies are related to the same structures that cause the gravity anomalies. For example, gravity and magnetic lows both occur in the deep structural depression in the Lower Klamath Lake basin. In addition, magnetic anomalies suggest the presence of buried basaltic rocks in the valley-fill deposits at Squaw Point in Upper Klamath Lake and at a location near Keno in Lower Klamath Lake. The gravity and magnetic data suggest that subsur- face structure exists at many places in marked contrast to the nearly featureless valley floor. Transecting faults apprently offset the basement topography in a pattern of tilted blocks which have a wide range of depths beneath the valley surface. Basalt flows interca- lated in the valley-fill deposits add to the complexity of the geophysical patterns. The hydrologic consequences of these discontinuities are many. Locally, the boundaries of the lithologic units may be relatively impermeable, thereby contri- buting to the segregation of waters of differing chemi— cal quality. Elsewhere, basalt flows and fault zones may form distribution channels for thermal waters. These phenomena are not well known in detail, but they will have important consequences in the future if significant quantities of thermal water are withdrawn from and reinjected into the thermal reservoir. CHEMICAL EVIDENCE Hot waters from the three geothermal areas of Klamath Falls have similar, although not identical, chemical characteristics. The similarities suggest that the waters from the three areas have equilibrated in nearly identical rocks at similar depths and tempera- tures. On the basis of the chemical data, the thermal wa- ters appear to circulate in a hot-water system of moderate temperature (< 150°C). Ratios of boron and fluorine to chloride indicate that the water is heated mainly by conduction rather than by condensation of steam from an underlying vapor-dominated system. Hot water rising in the principal conduits probably cools largely by conduction, with little loss of volatile elements and little exchange of ions with the surround— GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ing rocks. Thus, the chemical compositions of the hot- ter waters observed at the surface may be closely simi- lar to the compositions of water in the reservoir rocks. Sources of the meteoric recharge waters and flow paths through the geothermal system have not been identified in this study, but the chemical data are con- sistent with the hypothesis that local ground water furnishes recharge for the system. Primary evidence for this hypothesis consists of similar concentrations of deuterium and similar ratios of SO4/Cl and B/Cl in the hot and cold waters as well as expected changes in concentrations of sodium, chloride, sulfate, silica, and bicarbonate. Most shallow ground waters in the region have similar chemical compositions, and the available chemical analyses do not make it possible to distin- guish separate recharge waters or reservoir waters. Ratios of ionic constituents and concentrations of stable isotopes in the thermal waters are qualitative indicators of the probable range of reservoir tempera- tures at Klamath Falls. Application of the quartz geothermometer to mixing models based on silica and stable-isotope concentrations indicates that reservoir temperatures are greater than 145°C, but the moderately high Ca/HC03 ratios, high Na/K ratios, and the small shift of 8‘80 from the levels measured in local meteoric waters suggest that temperatures are probably not much higher than 150°C. The estimates of reservoir temperature derived from both silica and '“O(H20) concentrations are reassuring in their general consistency and are supported by the patterns of ionic ratios in the thermal waters. These results must be weighed, however, against the higher estimates derived from 18O(SO4) concentrations. (See section “Reservoir Temperature Estimated from Oxy- gen Isotopes in Dissolved Sulfate”) If the concentra- tion of "‘0 in dissolved sulfate is a reliable indicator, the thermal waters have, at some point in their his- tory, equilibrated at temperatures that could have been as high as 196°C and were at least 180°C. Two possible models are offered in explanation of the high reservoir temperatures indicated by the sulfate oxygen isotope. In model (1), a two-stage reservoir is postulated, with the two stages separated in either space or time or both. In this model, 180 (SO.,) in deep thermal water is assumed to equilibrate with l"0(H2O) at temperatures greater than 190°C for a period of time which may be as long as several tens of years (McKen- zie and Truesdell, 1977, p. 53). Other constituents, in- cluding silica, would also equilibrate at these tempera- tures. At some later time and at a shallower depth, the water re-equilibrates at lower temperatures, acquiring the chemical characteristics observed at the surface and leaving only the sulfate-oxygen isotope as a ves- tige of its former chemical state. In model (2), isotopi- HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON cally light SO4 is formed in volcanic rocks during their emplacement at high temperatures. Subsequent ex- change with circulating thermal waters tends to shift the l80(SO4) concentrations toward their equilibrium values at much lower temperatures in a process that is still incomplete. These explanations are merely speculative and others undoubtedly are possible. Until such alternative explanations can be either confirmed or ruled out by additional evidence, the apparent high reservoir tem- peratures indicated by the sulfate-Oxygen isotope should probably be viewed with scepticism in the light of the remainder of the chemical and isotope data. Analysis of the stable-isotope concentrations sug- gests rather strongly that thermal waters in the three principal geothermal areas near Klamath Falls rise from three similar but distinct reservoirs. The oxygen and deuterium relations indicate separate sources of recharge for waters of the Klamath Falls hot-well area and those of Olene Gap, and the sulfate-oxygen geo- thermometer suggests yet a different history or origin for the Klamath Hills hot waters. Although these in- terpretations lack confirmation by other lines of evi- dence, they cast doubt on the existence of a single large reservoir beneath the Klamath Falls graben. CONCEPTUAL MODEL OF THE GEOTHERMAL SYSTEM The deep geothermal system at Klamath Falls has not been explored, and only near-surface indications and geophysical data provide indirect evidence of the nature of the system. Based on the evidence obtained during this study, a highly generalized conceptual model has been formulated. Verification and refine- ment of this model probably will require the acquisi- tion of new data, and in particular, data from deep drill holes. Geothermal phenomena at the earth’s surface have the following fundamental causes: (1) the presence of an intrusive mass of molten or still hot rock; (2) high heat production from radiogenic sources; (3) convective transport of heat and fluids from depth in a cyclic sys- tem; and (4) the insulating effect of layers of rock hav- ing low thermal conductivities. Although each of these types of systems may produce high-temperature gra— dients near the earth’s surface, only the first two are necessarily accompanied by higher than normal con- ductive heat flows. Most geothermal systems are pro- duced by a combination of at least two of the possible causes. On the basis of the evidence gathered during this study, the geothermal system at Klamath Falls is best explained as a convection system with deep circulation of meteoric water. Thus, the conceptual model proposed for this system is a fairly simple one. Meteoric water in G41 the Tertiary basaltic rocks of the region is assumed to percolate downward along the vast network of faults and fracture zones that characterize the regional struc- ture. After attaining depths on the order of 15,000 feet the temperatures of at least 150°C, the water rises to shallow depths through conduits closely associated with major faults. If the two-stage reservoir postulated on the basis of sulfate-oxygen isotope data exists, the shallower of the two reservoir zones may occur at depths less than 10,000 feet. This zone may be supplied by slow upward leakage through a few conduits from depths as great as 15,000 to 20,000 feet. The postulated depths are based on the assumption that geothermal gradients are uniform and are no greater than 30°C km", a probable minimum value for the region. It is likely, however, that actual depths of circulation are less than those calculated above on the assumption of uniform thermal gradients in the crustal rocks. Normal heat flow in the Klamath Falls area is probably impeded by rocks of low thermal conductivity that overlie the sunken blocks in the Klamath Lake grabens. Within these grabens, isotherms are probably displaced upward, and assuming that offsetting ther- mal effects in adjacent upraised blocks are smaller than the insulating effect of the sunken blocks, tem- perature at a given depth will be higher than that pre- dicted by a uniform normal gradient. The high thermal gradients (nearly 80°C km“) and low heat flows (<1.5 HFU) observed in the Geological Survey heat-flow hole, OC 1 (3), and in well FW in the Lower Klamath Lake basineare evidence for this concept. Although the structural and lithologic conditions postulated above probably also exist in several of the basins adjacent to Klamath Falls, conditions for the upward flow of heated water presumably are especially favorable at Klamath Falls. These favorable conditions may include both the elevation of isotherms in a large volume of rock and the presence of relatively perme- able flow channels in parts of the area. The fault- related conduits that permit reasonably free convective flow of hot water at three separate locations in the vicinity of Klamath Falls may be vestiges of the exten- sive and active hot-water system that must have oper- ated to silicify some thousands of cubic meters of rock during Pleistocene or earlier time. At present, the rising hot waters are probably largely confined to conduits in the three principal geothermal areas, with only minor amounts of upward flow occurring in the interior of the valley. If conduits exist in concealed faults beneath the valley bottoms they may be effectively sealed off at most places by the fine-grained deposits that occupy the valleys. Hot waters from the principal geothermal areas originate and equilibrate under generally similar res- G42 ervoir conditions. Differing oxygen and deuterium isotope concentrations suggest, however, that the res- ervoir underlying the three areas is separated into at least three distinct zones. The boundaries of these pos— tulated reservoir zones are unknown and their overall lateral extent can only be surmised from the extent of the surface phenomena. Based on surface manifesta- tions, a single reservoir could extend over an area of at ‘ least 100 square miles. If, as seems more likely, the reservoir consists of more or less separate permeable zones, the total areal extent of these zones may be as small as 20 square miles. Depth to permeable reservoir rocks beneath the val- ley floors may be as great as 6,000 feet based on in- ferred depths to basaltic rocks. In the upraised basaltic rocks adjacent to the valley, depth to areally extensive thermal reservoirs may be significantly smaller, perhaps no more than 2,000 or 3,000 feet at some places. Most of the heated water rising from the reservoir does not reach land surface, but spreads through permeable zones in basaltic rocks beneath or within the Yonna Formation. In these strata, which occur at depths of a few tens of feet to at least 1,500 feet, the flow is predominantly lateral in response to local gra- dients in the shallow ground water reservoir. The hot water mixes with local meteoric water in these strata, and within a half mile of the major fault conduits, tem- peratures may decrease about 60°C. The spread of hot water near the geothermal areas produces notable effects on temperature gradients in these areas. Most thermal gradients measured in the Klamath Falls area represent only local perturbations in the normal geothermal gradient that occur as the result of convective transport of heat in moving ground water. Gradients may be extremely high (> 150°C km") in impermeable rocks of the Yonna Formation where these rocks overlie basalt or permeable gravel bearing hot water. At other places where the Yonna Formation is more permeable, temperature profiles may be nearly isothermal, and gradients may even be reversed beneath hot-water zones. These phenomena are observed at depths as great as 1,800 feet (plate 4) and they probably persist to even greater depths in the principal geothermal areas. The conceptual model described above suggests that heat flow may be generally low in the thick valley-fill deposits of the area as the result of the concentration of the regional heat flow in a few small areas. The rela- tively low estimated heat flow observed in the lower Klamath Lake basin may be evidence of this effect. The model also implies that heat flow may be disturbed on the periphery of the basin as the result of displacement of temperature contours within the basins. This effect GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS is probably masked by topographic effects and by con- vective heat flow in the major geothermal areas; con- sequently, it may be difficult or impossible to detect. As a rough check on the areal heat-flow balance, as- sume that the regional conductive heat flow is near the upper limit of normal values for the Basin and Range province, or about 2 HFU. Over the approximately 100 square miles of the Klamath Falls thermal area, total conductive heat flow is therefore calculated to be 5.2 X 106 cal s“. This figure may be compared with an esti- mate given earlier which indicates that the total con- vective thermal discharge from the Klamath Falls thermal areas is probably in the range 4 X 106 to 32 x 106 cal 3". According to these estimates, therefore, most of the normal heat flow through the crust in this area may be concentrated by convective discharge in the three geothermal areas. Uncertainty in the esti- mate of convective discharge is large, but even if the actual discharge were at the low end of the estimated range (4 X 106 cal s"), the regional conductive heat- flow patterns must be significantly altered by the con- vective discharge of heat at Klamath Falls. It is likely, therefore, that observations of heat-flow patterns within and adjacent to the Klamath Falls graben could provide a test of the conceptual model presented here. Digital models may be used to obtain insight into permissible boundary conditions and heat-flow rates in the geothermal system. A model study for this purpose is currently (1979) underway in the USGS geothermal research program in Menlo Park. Results of this study will be presented in a subsequent report. RECHARGE TO THE GEOTHERMAL SYSTEM Postulated modes of recharge depend greatly on the assumed depth, nature, and extent of the geothermal reservoir. In the absence of definitive data on the reser- voir, only general conclusions have been reached re- garding the source and magnitude of recharge. Re- charge may occur through the same fault systems that, at places, conduct the hot water to the surface, al- though it seems unlikely that downward flow would be confined to isolated channels as the thermal discharge appears to be. Rather, the recharge may occur as slow, diffuse seepage through extensive segments of the fault zones, perhaps at relatively great distances from the discharge sites. Quantities of water available for recharge are large if the area of potential recharge is assumed to extend to a radius of, say, 25 miles from the center of the three geothermal areas. Over this area precipitation is about 2.4 X 10“ acre-ft yr“, of which about 2.1 X 106 acre-ft yr" is lost through evapotranspiration. The remainder, about 0.3 X 106 acre—ft yr" is available for recharge or HYDROGEOLOGIC APPRAISAL OF KLAMATH FALLS GEOTHERMAL AREA, OREGON runoff. It is likely, of course, that only a small part of the potential recharge can be captured in the areas that directly feed the geothermal system. Ground-water flow into the Lower Klamath Lake basin from upper basins may be as much as 170,000 acre-ft yr". If so, this flow would represent a major source of recharge to the geothermal reservoir. Be- cause the Klamath Falls area is a large structural de- pression, shallow ground water may enter the area from all directions, although scanty observations sug- gest that there is an overall gradient through the area toward the southwest. A deep regional flow system could bring additional ground water into the area from the High Cascades, but there is no evidence to indicate whether or not such flow occurs. MAGNITUDE OF THE GEOTHERMAL RESOURCE On the basis of data available in early 1978, Brook and others (1979) estimated the total heat content of geothermal systems at Klamath Falls within 3 km of the surface to be 33.1 X 1018 joules (J). The estimate was based on an assumed total reservoir volume of about 125 km3 (30 mi3) in two separate areas, Klamath Hills and Klamath Falls. The Klamath Falls area in- cluded the springs at Olene Gap. Employing the figures given by these authors for the probable maximum and minimum errors of estimate, the probable maximum heat content is calculated to be 49 X 10‘“ J and the probable minimum heat content is 17 X 10‘“ J. On the basis of additional temperature meas- urements, isotope mixing models, and differing esti- mates of reservoir volumes obtained in the present study, the writer has made new estimates of probable maximum and minimum heat contents. Unlike the previous estimates, which are based on differing mean reservoir temperatures for the Klamath Hills and the Klamath Falls systems (124°C and 111°C, re- spectively), the present estimate assumes a single res- ervoir temperature for all systems (150°C). This tem- perature represents a significant increase over those used previously, but it is clear from the calculations given below that the most significant parameters used in the estimation of heat content are those related to the volume of the reservoir, and, in particular, the es- timates of area. The probable maximum heat content is calculated by assuming that the reservoir is a continuous volume of basaltic rock in which hot water moves through frac- tures, permeable interflow zones, and fault zones. The reservoir is assumed to occupy an irregular but roughly rectangular area aligned northwest—southeast and extending from Upper Klamath Lake to the south end of the Klamath Hills, a distance of about 17 miles. The reservoir area is assumed to be bounded by the G43 west face of Stukel Mountain and the west face of the Klamath Hills, to extend east into Olene Gap, and to narrow to about 21/2 miles in the vicinity of the city of Klamath Falls. The total surface area is approximately 108 mi2 (275 km2). Additional assumptions are as follows: (1) The reser- voir temperature is 150°C, or 138°C above a base tem- perature of 12°C assumed for shallow ground water in the region; (2) the reservoir is 6,600 feet (2 km) thick above a depth of 13,000 feet (4 km); and (3) the specific heat of the reservoir rocks is 0.6 cal cm'3 °C", based on averages for crustal basaltic rocks (John Sass, oral commun., 1974). Combining these assumptions with the estimate of reservoir area given above, the heat content is calculated as follows: Heat content = 138(°C) X 2(km) X 105 m (km) x 0.6 __(C_al>__x 275 (km2) x 10'0 (cm3)(°C) (cm2) (ka) = 46 x 10l8 (cal) X 4.184 joules per calorie = 192 X 1018 J. For an estimate of a probable minimum reservoir volume, it is assumed that the reservoir is confined to three separate elongate areas aligned with the main faults in the principal geothermal areas at Klamath Falls, Olene Gap, and the Klamath Hills. Widths of the areas range from 1 mile at Olene Gap and the Klamath Hills to 1% miles at Klamath Falls. Lengths range from about 41/2 miles at Olene Gap to 6%. miles at Klamath Falls and the Klamath Hills. The total sur- face area is about 19 mi2 (50 kmz). It is further assumed that the reservoir is no more than 3,000 feet (1 km) thick. Thus, in this conceptual model, the reservoir is restricted to rocks immediately adjacent to the princi— pal faults that conduct the hot waters to the surface. Total volume is assumed to be 12 mi3 (50 km3). Based on the assumptions given above, and using the temperature and specific heat values employed for the previous estimate, the probable minimum amount of stored heat is calculated to be 3.5 X 10‘” cal (14.6 X 10‘“ J). The difference of an order of magnitude between the probable limits of reservoir volume given here reflects a serious deficiency in actual knowledge of subsurface conditions at Klamath Falls. The estimates can be re- garded only as attempts to assign reasonable limits to the overall size and type of reservoir that might pro- duce the observed geothermal phenomena. G44 If the actual heat content were determined to be in the upper part of the estimated range, the geothermal reservoir at Klamath Falls would be among the largest nongeopressured reservoirs in the United States. If, as seems more likely, the reservoir at Klamath Falls comprises two or more permeable hot-water zones that are separated by relatively unproductive rocks, the total reservoir volume may be near the probable minimum estimate given above. A most likely range of heat content, therefore, may be two to three times the probable minimum, or 7 X 10‘8 cal to 10 X 10'8 cal (30 x 10‘“ J to 40 x 10[8 J). Compared to heat contents estimated for other hydrothermal convection systems in the United States having temperatures near 150°C, this is still an impressively large number. The above estimates of heat content, based on the concept of deep convective flow, would not be changed by assuming the presence of a hot igneous mass at greater depth. Although the ultimate recoverability of geothermal heat and the methods of exploiting it would differ depending on whether a heat source were present or not, the total heat stored to depths of 10,000 feet or more may fall within the estimated limits in either case. As is true for most low-temperature hydrothermal systems, utilization of the Klamath Falls geothermal resource for production of electrical power is not eco- nomically feasible under present conditions. Future utilization for any purpose will probably be highly sen- sitive to the actual nature and extent of the reservoir. The facts can be ascertained only by additional explo- ration, by the acquisition of more complete chemical and isotope data, and by subjecting the data to more sophisticated analysis. Exploratory drilling to depths of at least 6,000 feet will probably be required in order to provide definitive answers to the many questions that remain unresolved by this study. REFERENCES CITED Brook, C. A., Mariner, R. H., Mabey, D. R., Swanson, J. R., Guffanti, Marianne, and Muffler, L. J. P., 1979, Hydrothermal convection systems with reservoir temperatures 2 90°C, in Muffler, L. J. P., ed., Assessment of geothermal resources of the United States— 1978: US. Geological Survey Circular 790, p. 18—85. Couch, R. W., and Lowell, R. P., 1971 Earthquakes and seismic en- ergy release in Oregon: The Ore Bin, v. 33, no. 4, p. 61~84. Craig, H., 1961a, Isotopic variation in meteoric waters: Science, v. 133, p. 1702. 1961b, Standard for reporting concentrations of deuterium and oxygen-18 in natural waters: Science, v. 133, p. 1833. Fournier, R. 0., 1977, Chemical geothermometers and mixing models for geothermal systems: Geothermics, v. 5, p. 41-50. F ournier, R. 0., and Rowe, J. J ., 1966, Estimation of underground temperatures from the silica content of water from hot springs and wet steam wells: American Journal of Science, v. 264, no. 9, p. 685—697. GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Fournier, R. 0., and Truesdell, A. H., 1973, An empirical Na-K-Ca geothermometer for natural waters: Geochimica et Cos- mochimica Acta, v. 37, p. 1255— 1275. 1974, Geochemical indicators of subsurface temperature— Part II, Estimation of temperature and fraction of hot water mixed with cold water: US. Geological Survey Journal of Re- search, v. 2, no. 3, p. 263—270. Gilbert, G. K., 1928, Studies of Basin Range structure: U.S. Geologi- cal Survey Professional Paper 153, 92 p. Hubbard, L. L., 1970, Water budget of Upper Klamath Lake, south- western Oregon: U.S. Geological Survey Hydrologic Investiga- tions Atlas HA—351. Illian, J. R., 1970, Interim report on the ground water in the Klamath basin, Oregon: Salem, Oregon State Engineer, 110 p. International Atomic Energy Agency, 1974, Symposium on isotope techniques in ground-water hydrology: Vienna, March 1974, Proc., 2 vols. Lachenbruch, A. H., Jr., Sass, J. H., Munroe, R. J., and Moses, T. H., Jr., 1976, Geothermal setting and simple heat-conduction models for the Long Valley Caldera: Journal of Geophysical Re- search, v. 81, no. 5, p. 769—784. Leonard, A. R., and Harris, A. B., 1974, Ground water in selected areas in the Klamath Basin, Oregon: Salem, Oregon State En- gineer Ground Water Report No. 21, 104 p. Lund, J. W., Principal Investigator, 1978, Geothermal hydrology and geochemistry of Klamath Falls, Oregon, urban area: Klamath Falls, Oregon Institute of Technology, Geo-Heat Utilization Center, Final report to the US. Geological Survey (Grant 14— 08—0001-G—291), 123 p. Lund, J. W., Culver, G. G., and Svanevik, L. S., 1974, Utilization of geothermal energy in Klamath Falls: International Conference on Geothermal Energy for Industrial, Agricultural, and Com- mercial Uses, Klamath Falls, Oregon Institute of Technology, Proc., p. 146—178. McKenzie, W. F. and Truesdell, A. H., 1977, Geothermal reservoir temperatures estimated from the oxygen isotope compositions of dissolved sulfate and water from hot springs and shallow drillholes: Geothermics, v. 5, p. 51—61. MacLeod, N. S., Walker, G. W., and McKee, E. H., 1975, Geothermal significance of eastward increase in age of Upper Cenozoic rhyolitic domes in southeastern Oregon: US. Geological Survey Open-File Report 75—348, 22 p. Mariner, R. H., and Willey, L. M., 1976, Geochemistry of thermal waters in Long Valley, Mono County, California: Journal of Geophysical Research, v. 81, no. 5, p. 792—800. Meyers, J. D., and Newcomb, R. C., 1952, Geology and ground-water resources of the Swan Lake-Yonna Valley area, Klamath County, Oregon: US. Geological Survey Open-File Report, 151 p. . Moore, B. N., 1937, Nonmetallic mineral resources of eastern Ore- gon: US. Geological Survey Bulletin 875, 180 p. Newcomb, R. C., 1958, Yonna formation of the Klamath River basin, Oregon: Northwest Science, V. 32, no. 2, p. 41—48. Newcomb, R. C., and Hart, D. H., 1958, Preliminary report on the , groundwater resources of the Klamath River basin, Oregon: US. Geological Survey Open-File Report, 248 p. Olmsted, F. H., Glancy, P. A., Harrill, J. R., Rush, F. E., and Van Denburg, A. S., 1975, Preliminary hydrogeologic appraisal of selected hydrothermal systems in northern and central Nevada: US. Geological Survey Open-File Report 75—56, 267 p. Oregon State Water Resources Board, 1971, Klamath Basin: Salem, Oregon, State Water Resources Board, 288 p. Peterson, D. L., 1976, Preliminary interpretation of geophysical data, in Sammel. E. A., 1976, Hydrologic reconnaissance of the geothermal area near Klamath Falls, Oregon: US. Geological Survey Water Resources Investigations 76— 127, p. A1—A10 HYDROGEOLOGIC APPRAISAL ()F KLAMA’I‘H FALLS GEOTHERMAL AREA, OREGON Peterson, N. V., and Groh, E. A., 1967, Geothermal potential of the Klamath Falls area, Oregon; a preliminary study: The Ore Bin, v. 29, no. 11, p. 209—231. Peterson, N. V., and McIntyre, J. R., 1970, The reconnaissance geol- ogy and mineral resources of eastern Klamath County and west- ern Lake County, Oregon: Portland, Oregon Department of Geology and Mineral Industries Bulletin 66, 70 p. Renner, J. L., White, D. E., and Williams, D. L., 1975, Hydrothermal convection systems, in White, D. E., and Williams, D. L., eds, Assessment of Geothermal Resources of the United States— 1975: US. Geological Survey Circular 726, p. 5-57. Russell, R. J ., 1928, Basin Range structure and stratigraphy of the Warner Range, northeastern California: California University Publications in Geological Sciences, v. 17, no. 11, p. 387—496. Sammel, E. A., 1976, Hydrologic reconnaissance of the geothermal area near Klamath Falls, Oregon: US. Geological Survey Water Resources Investigations 76— 127, 129 p. Sass, J. H., Lachenbruch, A. H., Munroe, R. J., Greene, G. W., and Moses, T. H., Jr., 1971, Heat flow in the western United States: Journal of Geophysical Research, v. 76, p. 6376—6413. Sass, J. H., and Sammel, E. A., 1976, Heat-flow data and their rela- tion to observed geothermal phenomena near Klamath Falls, Oregon: Journal of Geophysical Research, V. 81, no. 26, p. 4863—4868. Taylor, H. P., Jr., 1974, The application of oxygen and hydrogen isotope studies to problems of hydrothermal alteration and ore deposition: Economic Geology, v. 69, p. 843—883. GPO 689-14 3 G45 Truesdell, A. H., 1976a, Chemical evidence of subsurface structure and fluid flow in a geothermal system: International Symposium on Water-Rock Interaction, Prague, Czechoslovakia, 1974, Czechoslovak Geological Survey, Free. 1976b, Geoterm, a geothermometric computer program for hot-spring systems: in Second United Nations Symposium on the Development and Use of Geothermal Resources, San Francisco, 1975, Proc., p. 831—836. Truesdell, A, H., and Fournier, R. 0., 1977, Procedure for estimating the temperature of a hot-water component in a mixed water by using a plot of dissolved silica versus enthalpy: US. Geological Survey Journal of Research, v. 5, no. 1, p. 49—52. US. Geological Survey, 1976, Water resources data for Oregon, Water year 1975: US. Geological Survey Water-Data Report OR—75—1, 586 p. White, D. E., 1968, Hydrology, activity, and heat flow of the Steam- boat Springs thermal system, Washoe County, Nevada: US. Geological Survey Professional Paper 458—C, 109 p. 1970, Geochemistry applied to the discovery, evaluation, and exploitation of geothermal energy resourcs: U. N. Symposium on the Development and Utilization of Geothermal Resources, Pisa, 1970, in Geothermics, Special Issue 2, p. 58—80. Williams, Howel, 1949, Geology of the Macdoel Quadrangle: San Francisco, California Department of Natural Resources: Divi- sion of Mines Bulletin 151, p. 7—60. Wood, P. R., 1960, Geology and ground-water features of the Butte Valley region, Siskiyou County, California: US. Geological Sur- vey Water-Supply Paper 1491, 150 p. UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—G PLATE 1 Eagle Point H 19.71% g; TRUE NORTH EXPLANATION O Nonflowing well 0 Nonflowing well, sampled for chemical or isotope analysis (0) Well which normally flows at or above land surface © Flowing well, sampled for chemical or isotope analysis 0 Well in which temperature profile was obtained q\ Perennial spring. Tail indicates direction of flow .51 Perennial spring, sampled for chemical or isotope analysis “<‘0 . . . 70\ LIne of equal altitude on the water-table surface (contour Interval— 10 feet); dashed where approximately located. Altitudes in feet above National Geodetic Vertical Datum. Water levels measured during 1973—74 4165 Altitude of water surface in feet above mean sea level, measured in spring or well The contours shown in this figure represent the generalized surface of water in the shallow aquifers. In drawing the contours, apparent discrepancies among water-level altitudes in neighboring wells have been ignored. These differences may be due to errors in the measured or reported depths to water, to errors in the estimated altitudes of land surface, or to actual differences in water levels resulting from local variations in recharge, discharge, or seasonal variations. Topographic contours have been relied upon in interpolating water levels between data points. (ELM- 3000 0 3000 5000 9000 12000 15000 10000 21000 rm mun ._. ._4 .—. H OREGON 1 9 (1 1 z 3 4M|LES 0 1 2 3 4 5 KILOMETERS 1 .5 HHHHH CONTOUR INTERVAL 40 FEET S I APPROXIMATE MEAN DOTTED LlNES REPRESENT ZO'FOOT CONTOURS DECLlNATlON, 1957 NATIONAL GEODETIC VERTICAL DATUM OF 1929 StUdY area «a; Base from U.S. Geological Survey Merrill and Klamath Falls, Oregon~ California; Modoc Point, Oregon 1:62,500, 1957 MAP SHOWING ALTITUDES OF WATER LEVELS AND LINES OF EQUAL ALTITUDE ON THE WATER TABLE I DEPTH BELOW LAND SURFACE IN FEET UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSION PAPER 1044—G GEOLOGICAL SURVEY PLATE 4 Assumed land surface temperature TEMPERATURE, IN DEGREES CELSIUS for reference gradients (12°C) 00 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 1000 | I I I I I I I I I I I I | 100 - — 50 200 — R — LL 0. 6) [996’ 300 — _ —100 AG33 400 — — I I I I __ 500 I I I I I I l 150 Im II) \)> 52 Im Z I—l 600 —— :1 ll _ -= lo: Co In I — 200 700 — _ 800 — E L— 250 900 — — :r l I I j —300 1000 ‘ I I l I I SYMBOL OWNER LOCATION I 1100 _ AG33 Assembly of God Church 38/9-33 aba I w_ _ BH26 Bill Hill 40/10-26 bca a: I BH28 W.A. Hughes 38/9-28 dcd 6 I CM18 Charles Matney 40/10-18 cbb I I—350 CW3 Cass Windsor 39/9-3 ddd I 1200 _ ED25 E.D, Daniel 39/9-25 aaa I __ _I EG17 Ed Gardner 40/8-17 bad I F &w US. Fish and Wildlife 4UN/2E-13 o I (California) l l GH12 George Hergesheimer 39/9-12 bcd l l 1300 _ HW18 H.R. Waggoner 39/10~18 bcb I .__- _ JB35 Joe Balr 39/9-35 cbb lb _400 JL1 US. Bureau Reclamation ILiskey) 41/8-13 acd 5 KFOl Klamath Falls Oil Company 40/9-2 daa2 : KlDl9 Klamath Irrigation District 39/10-19 aca 1400 _ KLADl J.D.O'Connor 39/9-9 cbd ——— — KLADZ JID.O'Connor 39/9—9 bcb LC25 Larry Clark 40/10—25 ccb M314 Marion Barnes 39/1044 ccd OCl O'Connor Livestock Company 41/9—1 bcd ’ _450 1500 —— OCl(3) u.s. Bureau of Reclamation I I I I (O'Connor) 41/9-3 dbb OC7 O’Connor Livestock Company 41/10=7 dba O|T3 Oregon Institute of Technology 38/947 cdd OlTB Oregon Institute of Technology 38/9-20 acb2 EXPLANATION 1600 _ OWCB Oregon Water Corporation 39/10~6 cdb __ _ PF29 Richard Pope, Fred Foulon 38/9-29 aad M CI t 1: H — easure e PHZO Presbyterian Intercommunity 38/9-20 acb3 mpera ure pro I e Hospital . / -— 500 - — — f d t / ROB Raymond Dillon 40/9-6 aba Re erence thermal gra len s // 1700 _ RK15 Ray Kramer 41/1045 ab“ __ &---—£ Point measurement in well __ 65\ L SW1 Suburban Water Company 41/10-1 cdd2 (Data from Van Orstrand, 1938) \ ‘Q TF19 Tulana Farms 40/9-19 dcc \ VW3 V and W Ranch 40/8-3 cda2 G} ————— © BOILOI’TIHIIOIG temperatures from ____________ (EU->6 ______________ C.) . . . drlller 5 log (91%} ____ WP2 Wlard Park Water Dlstrlct 39/9-2 cId ‘‘‘‘‘‘ ~—— ____ Q 1800 — —— —— I '——- 550 l l | l l l l l l l l l l l I ~69 l 0 5 10 15 20 25 3O 35 40 45 50 55 60 65 70 75 80 85 90 95 TEMPERATURE, IN DEGREES CELSIUS TEMPERATURE PROFILES MEASURED IN UNUSED WELLS AND HEAT-FLOW HOLES DEPTH BELOW LAND SURFACE, IN METERS UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 1044i—G GEOLOGICAL SURVEY PLATE 2 EXPLANATION Nonflowing well Nonflowing well, sampled for chemical or isotope analysis Well which normally flows at or above land surface Flowing well, sampled for chemical or isotope analysis 0 Well in which temperature profile was obtained Qfl Perennial spring. Tail indicates direction of flow \\ M Perennial spring, sampled for chemical or isotope analysis 7000 Line of equal specific electrical conductance, in micromhos per centi— meter (pmho/cm) at 25°C. Dashed where approximately located 28/ Sequence number of well or spring sampled for chemical analysis (table 7) /1150 Specific electrical conductance of water sample, in [.lth/Cm at 25°C /X indicates no value available ®©OO Modoc Point SCALE FOR STIFF DIAGRAMS Na+K‘ — fol ”(Ah—l l l l l ‘ | | l | l lfiHco3 +003 g77730 25 2o 15 10 5 5 1o 15 2o 25 30780‘ Milliequivalents per liter The contour lines shown on the map separate areas in which wells of moder— ate depth (150-500 ft) may be expected to produce water whose specific electrical conductance is between the indicated values. The electrical conductance of water is closely related to the amount and nature of dissolved solids in the water; high conductance indicates a high concentration of dissolved minerals. The map is intended to indicate only the generalized areal pattern of spe— cific conductance. Contours are based on scanty data at some places. For ex~ ample, the position of the lOOO—pmho line west of the Klamath Hills is uncertain owing to a complete lack of data in the immediate area. The Stiff diagrams have three characteristic shapes corresponding to sodium sulfate waterslhot),sodium bicarbonate waters (warm), and calcium bicarbonate waters (cold).The example shown on the scale above represents a typical hot water containing about 9 miliequivalents per liter (meq L'I) Na, 1 meq L“ Ca, 0.0 Mg, 2 meq L'1Cl, l meq [1 HCO3, and 9 meq E1804. .29.? fiflfl l I L75 l l o 1 2 a A MILEs >— g m HH H .—I , g 5 3000 0 3000 6000 9000 12000 15000 18000 21000 FEET OREGON m i: 1 .5 0 1 z 3 4 5 KILOMETERS «II :1 Lu m—IHHH E 5 § CONTOUR INTERVAL 40 FEET APPROXIMATE MEAN DOTTED LINEs REPRESENT 207mm CONTOURS ~ “WW“ DECLINATION 1957 I NATIONAL GEODETIC VERTICAL DATUM OF 1929 StUdV area may R'Illflllfh {ml «c .35in Base from U.S. Geological Survey Merrill and Klamath Falls, Oregon- California; Modoc Point, Oregon 1:62,500, 1957 MAP SHOWING SPECIFIC ELECTRICAL CONDUCTANCE AND STIFF DIAGRAMS OF THE PRINCIPAL DISSOLVED CONSTITUENTS IN GROUND-WATER PROFESSIONAL PAPER 1044—G UNITED STATES DEPARTMENT OF THE INTERIOR PLATE 3 GEOLOGICAL SURVEY EXPLANATION Nonflowing well Nonflowing well, sampled for chemical or isotope analysis Well which normally flows at or above land surface Flowing well, sampled for chemical or isotope analysis Well in which temperature profile was obtained Modoc Poin Perennial spring. Tail indicates direction of flow Perennial spring, sampled for chemical or isotope analysis ‘ lsotherms in degrees Celsius, based on maximum temperatures measured in springs and wells; dashed where approximately located Measured temperature, in degrees Celsius Water reportedly cold Water reportedly warm Depth of well, in feet below land surface The isotherms on this map define areas in which wells of moderate depth (150-500 ft) may be expected to encounter waters in the indicated temperature ranges. Temperatures reported for wells in which profiles were obtained are the maximum temperatures measured at depth. The data have been generalized over most of the map, and no confidence should be placed in the exact locations of contour lines. The map can be used, therefore,only as a guide to the general distribution of thermal waters. lsotherms placed around isolated warm wells, for example those in T39, R8, Sect. 11 and T39, R9, Sect. 11, serve mostly to call attention to the occurrence of warmer water; their location and configuration have little significance. Acknowledgement: The 60°C isotherm in the urban area of Klamath Falls has been located in general conformity with data provided by the Geo—Heat Utilization Center,0regon Institute of Technology,Klamath Falls. The assistance provided by Professors John Lund, Gene Culver, and Paul Lienau is gratefully acknowledged. IL" .211" 1 2 3 4 MILES . 3000 6000 9000 12000 15000 15000 21000 FEET H 1—1 H 1—4 1—4 H a »—1 ,_. u—a OREGON TRUE NORTH s O Q l s l k . l 2 3 4 5 KILUMETERS s ~ 3 L7 \ V 3 CONTOUR INTERVAL 40 FEET \ APPROXIMATE MEAN DOTTED LINES REPRESENT ZO'FOOT CONTOURS L DECLINATION, 1957 NATIONAL GEODETIC VERTICAL DATUM or 1929 StUdY area 'C/132 17112013157 r901 .g. g16/2387 /,/.L ’ ‘ .15/17g, / 17/1036 (317/1045 5:055 7/1074 II”$5‘ 13/5? ._ . i1“ 15/800? 9 39/214 'w-~x__/ 024/85, . \ 25/610‘ D Base from U.S. Geoloical Survey Merrill and Klamath Fills, Oregon- California; Modoc POIIt, Oregon I, I MAP SHOWING TEMPERATURE OF GROUND WATER, DEPTH OF WELLS, AND LINES OF EQUAL TEMPERATURE Reconnaissance of the Hydrothermal Resources of Utah By F. EUGENE RUSH GEOHYDROLOGY OF GEOTHERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044-H A brief description of the hydrothermal resources of Utah UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1983 UNITED STATES DEPARTMENT OF THE INTERIOR JAMES G. WATT, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Rush, F. Eugene Reconnaissance of the hydrothermal resources of Utah . (Geological Survey Professional Paper 1044-H) Bibliography Supt. of Does. no.: I 19.16: 1044—H 1. Geothermal resources—Utah. I. Title. II. Series: United States. Geological Sur- vey. Professional Paper 1044—H. GBll99.7. U8R87 553.7 824500009 AACR2 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 CONTENTS Page Page Abstract __________________________________________________ H1 Discussion of prospects ____________________________________ 8 Introduction ______________________________________________ 1 Roosevelt Hot Springs and the Cove Fort-Sulphurdale Purpose and scope ____________________________________ 1 areas ______________________________________________ 11 Previous work ________________________________________ 1 Thermo Hot Springs __________________________________ 13 Well- and spring-numbering system ____________________ 1 Southwestern Escalante Desert ________________________ 18 Regional geologic setting __________________________________ 3 Monroe and Joseph KGRA ____________________________ 21 Middle Rocky Mountains ______________________________ 3 Crater Hot Springs ____________________________________ 28 Colorado Plateaus ____________________________________ 3 Navajo Lake KGRA __________________________________ 33 Basin and Range ______________________________________ 3 Meadow and Hatton Hot Springs ______________________ 33 Regional hydrologic setting ________________________________ 4 Vicinity of Salt Lake City ______________________________ 37 Geothermal relations ______________________________________ 5 Great Salt Lake Desert ________________________________ 39 Regional heat flow ____________________________________ 5 Other areas __________________________________________ 40 Relation of thermal waters to hydrogeologic framework __ 6 Summary and conclusions __________________________________ 40 Distribution of thermal waters ________________________ 7 References cited __________________________________________ 43 ILLUSTRATIONS Page FIGURE 1. Map showing land prospectively valuable for geothermal resources in Utah and index map of Utah ____________________ H2 2. Diagram showing location-numbering system ______________________________________________________________________ 3 3. Map showing generalized east-west patterns of Cenozoic igneous rocks and positive aeromagnetic anomalies and a generalized distribution of young igneous rocks ____________________________________________________________________________ 4 4. Diagram showing a conceptual model of a hydrothermal convection cell ______________________________________________ 7 5. Map showing geothermal areas of Utah ____________________________________________________________________________ 9 6. Map showing location of leased National Resource Land in Utah, September 1976 ____________________________________ 10 7. Reconnaissance geologic map of the Thermo Hot Springs area and audio-magnetotelluric configuration for the Thermo Hot Springs area ________________________________________________________________________________________________ 14 8.—11. Map showing 8. Distribution of orifices at Thermo Hot Springs ______________________________________________________________ 15 9. Temperature at a depth of 30 m in the Theme Hot Springs area and distribution of vegetation in the Thermo Hot Springs area, 1976 ____________________________________________________________________________________ 16 10. Shallow conductive heat flow in the Thermo Hot Springs area ______________________________________________ 18 11. Estimate heat flow in southwestern Utah __________________________________________________________________ 20 12. Reconnaissance geologic map of the Newcastle area; and ground—water levels and direction of flow in the alluvium of the Newcastle area, spring 1976 __________________________________________________________________________________ 22 13. Graph showing temperature profile of the Christensen Brothers thermal irrigation well near Newcastle, Utah __________ 23 14.— 17. Map showing 14. Temperatures at a depth of 100 m in the Newcastle area ____________________________________________________ 24 15. Distribution of heat-flow from the principal hot-water aquifer in the Newcastle area __________________________ 25 16. Helium concentrations in the Newcastle area ____________________________________________________________ 26 17. Estimated heat flow and measured spring temperatures in the MonroeJoseph area ____________________________ 26 18. Surficial geology of Monroe and Red Hill Hot Springs ______________________________________________________________ 28 19. Map showing Thomas, Keg, and Desert calderas, near Crater Hot Springs ____________________________________________ 30 20. Reconnaissance geologic map of the Crater Hot Springs area and a simple Bouguer gravity map of the Crater Hot Springs area 32 21. Generalized cross section of Crater Bench __________________________________________________________________________ 34 22. Map showing spring orifices and pools on the mound of Crater Hot Springs __________________________________________ 35 23. Map showing phreatophyte distribution in the Crater Hot Springs area ______________________________________________ 36 24. Map showing heat flow in the Crater Hot Springs area ______________________________________________________________ 37 25. Reconnaissance geologic map of the Meadow and Hatton Hot Springs area ____________________________________________ 38 26. Map showing estimated heat flow and water temperatures in the Meadow and Hatton Hot Springs area ________________ 40 27. Map showing areas of rapid snowmelt near Hatton Hot Springs, March 1976 ________________________________________ 41 28. Map showing areas of warm ground water in the Jordan Valley-__________-___‘ _______________________________________ 42 III IV TABLE 10. 11. 12. 13. 14. 15. 16. 17. 18. @F'PPJF’!‘ CONTENTS TABLES Page Relation of hydrothermal areas to mineral belts and to age of Cenozoic igneous rocks ________________________________ H5 Thermal—conductivity values used in this report ____________________________________________________________________ 7 Formulas for geothermometers used in this report __________________________________________________________________ 8 Known geothermal resource areas in Utah ________________________________________________________________________ 8 General characteristics of thermal spring groups __________________________________________________________________ 11 Estimated reservoir temperatures derived by geothermometer formulas for springs (and one well) with temperatures greater than 50°C and silica concentration greater than 50 mg/L ________________________________________________________ 12 Evapotranspiration of mixed water from Thermo Hot Springs hydrothermal system __________________________________ 19 Estimated conductive heat discharge from Thermo Hot Springs hydrothermal system—alluvial area only ______________ 19 Chemical analyses of water from Christensen Brothers thermal well near Newcastle, Utah __________________________ 21 Estimated conductive heat discharge from the Newcastle hydrothermal system—alluvial area only ____________________ 21 Measured flow and temperature of Monroe and Red Hill Hot Springs ________________________________________________ 27 Measured flow and temperature of Joseph Hot Springs ______________________________________________________________ 27 Measured flow and temperature of Crater Hot Springs ______________________________________________________________ 34 Evapotranspiration of ground water from Crater Hot Springs hydrothermal system __________________________________ 35 Summary of data for selected hydrothermal systems in Utah ________________________________________________________ 41 Inventory of Thermo Hot Springs ________________________________________________________________________________ 46 Inventory of Crater Hot Springs, February 1976 ____________________________________________________________________ 47 Selected subsurface temperature and heat-flow data not summarized on maps ________________________________________ 48 CONVERSION OF UNITS For use of those readers who may prefer to use inch-pound units rather than metric units, the conversion factors for the terms used in this report are listed below: Multiply metric unit By To obtain inch-pound unit calories (cal) 3.974 X 10‘3 British thermal units (B.t.u.) degrees Celsius (°C) 1.8°C + 32 degrees Fahrenheit (°F) milligrams (mg) 1.543 x 10‘2 grains kilograms (kg) 2.205 pounds (lb) liters (L) .2642 gallons (gal) meters (m) 3.281 feet (ft) millimeters (mm) 3.937 X 10‘2 inches (in) centimeters (cm) .3937 inches (in) hectometers (hm) 3.281 X 10‘2 feet (fiz) kilometers (km) .6214 miles (mi) square centimeters (cmz) .1550 square inches (in2) square kilometers (kmz) .3861 square miles (miz) cubic meters (m3) 35‘31 cubic feet (fta) liters per second (US) 15.85 gallons per minute (gal/m) GEOHYDROLOGY OF GEOTHERMAL SYSTEMS RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH By F. EUGENE RUSH ABSTRACT Geologic factors in the Basin and Range province in Utah are more favorable for the occurrence of geothermal resources than in other areas on the Colorado Plateaus or in the Middle Rocky Mountains. These geologic factors are principally crustal extension and crustal thinning during the last 17 million years. Basalts as young as 10,000 years have been mapped in the area. High-silica volcanic and intru- sive rocks of Quaternary age can be used to locate hydrothermal convection systems. Drilling for hot, high-silica, buried rock bodies is most promising in the areas of recent volcanic activity. Southwestern Utah has more geothermal potential than other parts of the Basin and Range province in Utah. The Roosevelt Hot Springs area, the Cove Fort-Sulphurdale area, and the area to the north as far as 60 kilome- ters from them probably have the best potential for geothermal devel- opment for generation of electricity. Other areas with estimated res- ervoir temperatures greater than 150°C are Thermo, Monroe, Red Hill (in the Monroe-Joseph Known Geothermal Resource Area), Joseph Hot Springs, and the Newcastle area. The rates of heat and water discharge are high at Crater, Meadow, and Hatton Hot Springs, but estimated reservoir temperatures there are less than 150°C. Ad- ditional exploration is needed to define the potential in three ad- ditional areas in the Escalante Desert. INTRODUCTION PURPOSE AND SCOPE The State of Utah has an abundance of thermal springs and probably is a promising area for geothermal exploration. This study, a 2-year reconnaissance of the geothermal resources on the public lands of Utah, was begun by the US. Geological Survey in the summer of 1975. The purpose of the reconnaissance was to describe the general geohydrologic framework for geothermal systems, and to provide more detailed descriptions and evaluations than were previously available for some of the more promising hydrothermal systems. Most of the data were gathered and evaluated during the summer of 1975 and in 1976. This report presents the results of the study. A data report (Rush, 1977) has already been released which contains subsurface-temperature data for 30 wells. PREVIOUS WORK The earliest known reference to geothermal systems of Utah is by Gilbert (1890, p. 332—335); he briefly described Fumarole Butte, gaseous discharges from the butte, and nearby Crater Hot Springs (fig. 1). Many years later Stearns, Stearns, and Waring (1937, p. 96, 108—109, 179—183) described about 60 thermal springs in Utah and summarized the literature about them. A similar summary was made by Waring (1965). At the East Tintic mining district, about 30 km northwest of Nephi (fig. 1), Lovering and Geode (1963) worked with geothermal gradient holes in their search for hydro- thermal ore bodies. In another mining area, the Iron Springs district about 16 km west of Cedar City, (fig. 1), Sass and others (1971, p. 6399—6400) described temper- ature measurements in eight drill holes. They con- cluded that the heat flow in that area is about 1.9 X 106 cal/cmz/s. Heylmun (1966) and Batty and others (1975, p. 233— 241) presented brief, general discussions of geothermal resources in Utah. However, both papers presented few data. A comprehensive data report on the thermal springs of Utah (Mundorff, 1970) contains an abun- dance of information for about 60 springs. Additional data have been published by Milligan, Marselli, and Bagley (1966). Additional estimates of reservoir tem- peratures were made for 47 hydrothermal systems in Utah by Swanberg (1974), using the Na-K-Ca geother- mometer developed by Fournier and Truesdell (1973). Olmsted and others (1975, p. 27—76) provided a dis- cussion of hydrothermal concepts and of exploration and evaluation techniques in a report describing hydro- thermal systems in the western part of the Basin and Range province. This discussion was useful as a guide in the study and other workers probably will find it of similar value. The University of Utah, Department of Geology and Geophysics, is currently (197 7) investigat- ing Roosevelt Hot Springs KGRA (Known Geothermal Resource Area) and other areas, primarily evaluating various geophysical techniques for geothermal explora- tion. In one of the resulting reports, Parry, Berson, and Miller (1976) describe the geology and water chemistry of Roosevelt and Monroe Hot Springs. WELL- AND SPRING-NUMBERING SYSTEM The system of numbering wells and springs in Utah, used herein, is based on the cadastral land-survey sys- tem of the US. Government. The number describes the position on the land net of the well, spring, or site where geothermal observations were made. In the land-survey H1 H2 GEOHYDROLOGY 0F GEO’I‘HERMAL SYSTEMS system, the State is divided into four quadrants by the Salt Lake base line and meridian, and these quadrants are designated by the uppercase letters A, B, C, and D, indicating the northeast, northwest, southwest, and southeast quadrants, respectively. Numbers designat- ing the township and range (in that order) follow the quadrant letter, and all three are enclosed in parenthe- section, and the section commonly is followed by three letters indicating the quarter section, the quarter- quarter section, and the quarter-quarter—quarter sec- tion (generally 4 hm2)1; the letters a, b, c, and d indicate, ‘The basic land unit, the section, is ideally 2.6 km“; however, many sections are irregular. Such sections are subdivided into 4 km2 tracts, generally beginning at the southeast corner, and the surplus or shortage is taken up in the tracts along the north and west sides of the ses. The number after the parentheses indicates the , section. 114° IDAHO 112° I —-- -- - --- TV F '3 Bear ake é ' COLUMBIA PLATEAUS l2 /"\ E l / 0 I l ' | l } WYOMING 110°BASIN 109,. 41°. . _- .. \ I “la;- HPMNATION Great Uinta Mountains '3 Salt MIDDLE ROCKY MOUNTAINS E Land prospectively ' m (1971) as deter- L , mined by uses \ 0 Known Geothermal Resource Area (KGRA) KGRA Crater Hot Springs Roosevelt Hot Springs Cove Fort-Sulphurdale Monroe-Joseph Thermo Hot Sprlngs Lund Newcastle OClVHO'lOO __\__________L Richfield Escal ante l Desert V 8 Plate'au COLORADO PLATEAUS 9599‘?P°N!‘ Navajo Lake Province Boundary l__/-‘———___ Blanding a we" Adapted from Godwin and others 1971. figure 2 0 50 100 150 l l 6 so 200K|LOMETERS l 160 MILES FIGURE 1.—Land prospectively valuable for geothermal resources in Utah, and an index map of Utah. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH H3 respectively, the northeast, northwest, southwest, and southeast quarters of each subdivision. The number after the letters is the serial number of the well or Spring within the 4 hm2 tract; the letter "S” preceding the serial number denotes a spring. If a well or spring can- not be located within a 4 hm” tract, less than three location letters are used and the serial number is omit- ted. Thus (C-29-8)9ba designates a well in the NWMLNW‘A sec. 9, T. 29 S., R. 8 W. The numbering system is illustrated in figure 2. REGIONAL GEOLOGIC SETTING Utah essentially includes parts of three physiograph- ic provinces as defined by Fenneman (1931): the Middle Rocky Mountains, the Colorado Plateaus, and the Basin and Range province (fig. 1). Each area is described briefly below, but more emphasis is given to the Basin and Range province because of its greater potential for geothermal development. MIDDLE ROCKY MOUNTAINS In Utah the Middle Rocky Mountains province in- cludes the Wasatch Range and the Uinta Mountains. The Wasatch Range rises to an altitude of 2,400 to 3,400 m above sea level, or between 1,200 and 2,000 m above Sections within a township Tracts within a section R 11 w Well Sec. 1 l 6 5 4 a 2 1“ b i a _____ a— --— 7 8 9101112 b /:b:-,a T 18 17 16 15 \14 13 / [Kai/d 29 \ ' ' s ‘19\2o 21 22 23‘ 24 / Well/ 30 2}\28 27 26 25 c d \ 31 32 33 34.35 36‘ : 6 miles 1’ // 1 mile (C—29—11) 1 add —2 SALT LAKE BASE LINE SALT LAKE MERIDIAN FIGURE 2.—Well- and spring-numbering system. the valley floors of the Basin-and Range province. The range is an uplifted block of folded and faulted strata, bounded on the west by a major fault zone, the Wasatch Fault. The Uinta Mountains are generally higher than the Wasatch Range, reaching altitudes greater than 4,000 m above sea level. They are described by Fenne- man (193 1, p. 177) as a flat-topped anticline. Most of the consolidated rocks that crop out in both mountain ranges are pre-Cenozoic sedimentary or silicic plutonic rocks. COLORADO PLATEAUS The province, as implied by its name, is an area of broad uplift with strata nearly horizontal in most places. The outcrops are mostly Mesozoic and older sedimentary rocks. Notable exceptions are Tertiary and Quaternary volcanic rocks in the southwestern part of the province (south-central Utah) and a few scattered Tertiary intrusive bodies in the southeastern part of the state. Land-surface altitudes are commonly between 1,500 and 3,000 m above sea level. A continuation of the Wasatch Fault zone marks the western boundary of the province. BASIN AND RANGE The Basin and Range province is characterized by elongated, mostly north-trending mountain ranges and narrow flat-bottomed valleys. The province contains rocks widely ranging in composition and age. The older rocks consist of a wide variety of Mesozoic and Paleozoic sedimentary rocks and their metamorphosed equiva- lents. Overlying the sedimentary and metamorphic rocks are Cenozoic volcanic rocks and valley fill. Valley fill, mostly alluvium, may be as thick as 3,000 min some basins. Lacustrine deposits are common. According to Stewart (1971), most or perhaps all of the major valleys in the Great Basin of the Basin and Range province can be considered to be grabens, and most or all of the mountains can be considered to be horsts or tilted horsts. The geometry of block faulting related to these structures requires sizable east-west extension of the thin crust under the province; the ex- tension was estimated by Stewart to be about 2.4 km for each major valley. Most of this extension took place in the last 17 million years, or perhaps even in the last 7—11 million years. In Utah, grabens which are not bounded by faults of equal displacement generally have the master fault on the east side. In western Utah, igneous rocks and hydrothermal mineral zones are in well-defined east-west belts (fig. 3), each successively younger to the south (table 1), accord- ing to Stewart, Moore, and Zietz (1977). Figure 3 shows 42° 40° 38° 36° H4 the distribution of igneous rocks less than 6 million years old; some are less than 10,000 years old. They are mostly basalt and crop out generally in southwestern Utah. These very young igneous rocks do not seem to be along an east-west belt but rather on an alinement parallel to Basin and Range structure. The implication is that Basin and Range structure controls the distribu- tion of these rocks; whereas the older belts predate Basin and Range structure. Stewart, Moore, and Zietz (1977) see genetic and age similarity between these belts and the belt of upper Cenozoic volcanic rock ex- tending along the Snake River Plain in southern Idaho eastward into the Yellowstone region of northwestern Wyoming. Very young volcanic rocks in Utah are reported by Rowley, Anderson, and Williams (1975, p. B18), and Smith and Shaw (1975, p. 82). Volcanic rocks probably less than 10,000 years old are found near Fillmore and 65 km southwest, 50 km south, and 30 km southeast of Cedar City (fig. 3). These young rocks are largely basal- tic, but scattered rhyolitic cones are known (Liese, 1957). Silicic intrusive rocks were emplaced at the same time as the silicic volcanic rocks (Whelan, 1970). The 120° __ __ 118° 118° _ ,- “‘7—'——'1————"—‘T'-——- , NEVADA | l” I 1 00 a x S 1 ' ARIZONA ‘0 100K|LOMETEHS \\ II 0 50 MILES i l i \2 l l EXPLANATION Generalized areas of outcrop of volcanic rocks within east-west belts m 17—6 my. 34—17m.y. [mm 4344 my. ————-— East-west trends of aeromagnetic anomalies related to Mesozoic and Tertiary rocks GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS largest exposure of such an intrusive body in Utah is the Mineral Mountains, 80 km north of Cedar City. REGIONAL HYDROLOGIC SETTING Precipitation on the semiarid valley floors of the Basin and Range province in Utah, as well as much of the Colorado Plateaus, averages less than 200 mm per year (US. Weather Bureau, no date). The higher mountains of the Wasatch Range and the Uinta Mountains generally receive precipitation of 1,000 mm or more per year, most accumulating as snow in the winter. The mountains of the Basin and Range province average about 500 mm or less yearly. The relatively large amounts of precipitation that fall in the mountains flow toward the ground-water reser- voirs and major streams in two ways: (1) Flow from the mountains in small streams. Part of this water infil- trates the stream beds and percolates to the water table. (2) Flow percolates directly into the fractures and pore space of consolidated rocks of the mountains; this water then flows in the subsurface across the consolidated rock —valley-fill contact. The latter is considered to be the smaller volume of water in most areas; an exception is in areas of carbonate rocks that have developed inter- connected solution channels. Most of the ground water . .. City— !I _____ 9 \ o 100 KILOMETERS ‘ $17 . f‘} 3 - '. 6° 1—————|—i $5. \\ ,-° . § ARIZONA - o 100 MILES ‘ ._ , . n . \ i ° . r .u a \ . . | I I ‘1 l .. .. Adapted from Stewart and others (1977, igur 1) EXPLANATION Basaltic flows W/A Rhyolitic flows - Intrusive roch Andesitic flows m Rhyolifictuffs and breocias FIGURE 3 . —Generalized distribution of young and old igneous rocks. A, East-west patterns of Cenozoic igneous rocks and positive aeromagnetic anomalies show, southward migration of igneous activity in California, Nevada, and Utah. B, Igneous rocks less than 6 million years old in western Utah, Nevada, and parts of adjoining states seem to parallel Basin and Range structure. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH H5 TABLE 1.—Relation of hydrothermal areas to mineral belts and to age of Cenozoic igneous rocks [The areas of western Utah are as shown in figure 4. The areas and mineral belts are listed in order from north to south. The ages of the rocks are from Stewart and others (1977, p. 71)] Approximate average age General Hydrothermal Mineral of rocks hydrothermal areas belt (m.y.) characteristics North of Cenozoic igneous Small, low-temperature rock area. _______________ >43 thermal springs. Northern Cenozoic igneous Oquirrh-Uinta __________________________________ 35 Small or low-temperature rock area. thermal sprin common. Deep Creek-Tintic ______________________________ 33 Includes Crater got Springs and hot water at east Tintic mining district. Between Cenozoic igneous rock Mid—Utah gap ____________________________________ Very few thermal springs areas. or wells. Southern Cenozoic igneous Wah Wah-Tushar ________________________________ 26 Includes Cove Fort-Sulphurdale rock area. area, Roosevelt, Thermo, Joseph, and Monroe Hot Springs. Iron Springs ____________________________________ 20 Includes Newcastle area. South of Cenozoic igneous rock area. Southern Nevada—Utah gap ________________________ circulates to depths of only several tens to several hundred meters, and generally has a temperature near or slightly higher than the ambient land-surface tem- perature for the lowlands (10°—16°C). Some ground water migrates through fault-created fractures to great depth where it absorbs heat from wall rock. This heated water returns to the land surface along with the shallow-circulating ground water, where it is ulti- mately discharged as springs, to streams, by evapo- transpiration, or from wells. In agricultural areas, a secondary source of ground-water recharge to the valley-fill reservoir is infiltration from fields, canals, and reservoirs. Summers are usually hot with low humidity. As a result, on lowlands, potential lake evap- oration greatly exceeds precipitation. Some of the valleys in the Basin and Range province are hydrologically isolated; that is, water that falls as precipitation remains within the basin until it is dis- charged back to the atmosphere. However, in areas where interconnected solution channels have devel- oped, ground water may follow complex flow paths be- neath interbasin divides and flow for tens or hundreds of kilometers and for thousands of years before discharg- ing to the land surface. Commonly, this interbasin flow involves moderately deep circulation beneath mountain ranges and, as a result, its discharge is significantly above ambient temperature. A probable example of dis- charge from an interbasin regional flow system is the Fish Springs group (Mundorff, 1970, p. 37), about 90 km northwest of Delta on the south edge of the Great Salt Lake Desert. The estimated discharge is 1.4 m3/s. Water temperatures reportedly range from 18° to 76°C. The recharge areas for these springs are probably to the south and west and probably include parts of Nevada. The principal sources of geothermal fluids are water stored in the hydrothermal reservoir and water enter- ing the geothermal-circulation system as recharge from precipitation. Because of the semiarid climate of much of the area, most geothermal development for genera- tion of electricity will remove fluids from storage at a higher rate than natural replenishment. GEOTHERMAL RELATIONS REGIONAL HEAT FLOW The average conductive heat flow to the earth’s sur- face is approximately 1.6 HFU (1 heat-flow unit [HFU] = 1 X 10‘6 cal/cmz/s or 1 ncal/cmz/s, according to Schubert and Anderson, 1974). Considerable variation in flow exists in Utah. Based on data from Sass and others (1971) and Sass and Munroe (1974), the area of highest heat flow in Utah is the Basin and Range prov- ince, which has heat-flow values commonly in the range of 1.5 to 2.5 HFU. By comparison the "Battle Mountain Hig ” (in Nevada) is an area of abnormally high heat flow where conductive heat-flow values are commonly in the range of 2.5 to 3.5 HFU. According to Lachen- bruch and Sass (1977), the "Battle Mountain High” may H6 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS be a part of a larger region of exceptionally high heat loss extending from western Nevada to Yellowstone Park, Wyo., and including the northwestern corner of Utah. No data were collected as part of this study to determine whether the northwestern corner of Utah is an area of very high heat flow. Data from this study, and the work of others, indicate that the Basin and Range province in Utah probably has an average heat flow of about 2 HFU. The Colorado Plateaus and the Middle Rocky Mountains provinces in Utah have heat-flow values near the average for the earth’s surface. Values pub- lished by Sass and others (1971) and by Sass and Mun- roe (1974) for these areas generally range from 1.3 to 2.0 HFU and average about 1.6 HFU. (The causes of variation in heat flow to the earth’s surface are complex and poorly understood. Some of the factors that contribute to the diversity are (1) variations in crustal thickness, (2) convection of magma beneath and possibly within the lower parts of the crust, (3) movement of ground water in hydrothermal convec- tion cells, (4) variations in the distribution of radioac- tive elements such as uranium, thorium, and potassium-40 in crustal rock, (5) intrusion into the upper crust of young magmas, and (6) general circula- tion of shallow ground water. RELATION OF THERMAL WATERS TO HYDROGEOLOGIC FRAMEWORK The present level of geothermal knowledge is in part presented by White and Williams (1975) and is briefly summarized as follows: (1) Some geothermal systems are supplied only by a "normal” geothermal gradient; some by magmatic heat. (2) Youngest igneous rocks have the best potential as heat sources. (3) Purely basic volcanic systems rarely form thermal anomalies of eco- nomic interest for generation of electricity, whereas silicic volcanic systems may do so if they are large enough. (4) Young basic volcanoes are produced by magma sources in the mantle and, under some condi- tions, are potential indicators of buried high-level silicic bodies with no obvious surface manifestations. (5) Silicic magmas are always erupted from high-level storage chambers, probably in the upper 10 km of the crust. (6) High-temperature convection systems can be sustained for many thousands of years with heat from high-silica magma bodies. Perhaps because of the very high viscosities of such magmas, these systems are as- sociated with magma chambers at shallow levels in the crust. (7) Cooling by hydrothermal convection tends to offset continued heating, but the rate of supply of magma from deep crustal or mantle sources is the domi- nant heat supply for both high-level magmatic and hy- drothermal systems. (8) Basic magmas rise through the crust to the surface through narrow pipes and fissures created by faulting; the individual magma pulses are volumetrically small, and such systems contribute little stored heat to the upper crust until magma chambers begin to form at high levels. (9) Fluid temperature is of critical importance in determining how a hydrothermal system may be utilized and is the most important single factor in evaluating a system. Hot-water convection systems can be divided into those of three temperature ranges: (a) Above 150°C; these systems may be consid- ered for generation of electricity; (b) from 90°C to 150°C; these systems are attractive for space and process heat- ing; and (c) below 90°C; these systems are likely to be utilized for heat only in locally favorable circumstances. (10) Natural geysers and active deposition of siliceous sinter (amorphous hydrous silica) are reliable indica- tors of subsurface temperatures at least as high as 180°C. On the other hand, travertine deposits (calcium carbonate) and opaline residues produced by sulfuric acid leaching have no reliable relation to reservoir tem- perature. (11) In the Basin and Range province, heat flows are sufficiently high that the existence of a thick blanket having low thermal conductivity (high-porosity clay beds, for example) could locally raise the tempera- tures to levels of economic interest. The above ab- breviated summary can be used as a partial guide to the general relation of thermal waters to the hydrogeologic framework. Most thermal springs and wells are in valleys near the margins of the mountains. Spring positions prob- ably are controlled by Basin and Range faults. Some springs are in valley bottoms; others are on upland slopes. Only a few thermal springs are in a mountainous setting; the most prominent example is Midway Hot Springs, 45 km southeast of Salt Lake City (Mundorff, 1970, p. 46). Recharge to the hydrothermal systems is by either meteoric water in the nearby shallow, ground-water reservoir or by percolation in the nearby mountains. The dominant driving force for deep circulation prob- ably is the difference in density between cold recharge water and hot upfiowing water, but head differences between recharge area and springs may contribute as shown in figure 4. The hot water rising from the hydrothermal reservoir may be greatly diluted by shallower-circulating cold water, lowering the temperature but increasing the flow of thermal springs above the hydrothermal reservoir. As shown in figure 4, only part of the upflow of thermal water may directly reach the land surface, because part may enter near-surface aquifers and cool by conduction as it flows laterally from the spring area. The model described above may be modified in several ways to approximate the variety of hydrothermal sys- tems: (1) No convecting magma may be present in the upper crust, but rather the heat source may be deeper in RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH the crust or in the mantle; (2) the downward flow of cold water to the hydrothermal reservoir may be through any deep permeable route; and (3) calcium carbonate and amorphous silica may deposit on the walls of the upflow zone, creating an isolated or semi-isolated con- duit or self-sealing cap. As a result, movement of water between the isolated part of the hydrothermal convec- tion cell and the surrounding rock, alluvium, or land surface would be reduced or eliminated. Vertical flow of shallow nonthermal ground water may modify conductive heat flow to the land surface. In areas of recharge (downward percolation of water), heat flow to the land surface is decreased; whereas in areas of ground-water discharge to the atmosphere (upward flow) the normal heat flow to the land surface is in- creased. The magnitude of the distortion is related to the velocity and quantity of vertical flow of water. Conductive heat flow is computed as follows: HFU = 10‘2KI where HFU is heat-flow unit, in microcalories per square centimeter per second, K is thermal conductivity of the rock material, in millicalories per centimeter per second per degree Celsius, and I is the geothermal gradient, in degrees Celsius per kilometer. The estimated K values used in this report are sum- marized in table 2. From the formula it has been seen Mountains Valley permeable rock downflow of cold water (recharge) Conductive heat flow in crystalline rock H? Heat source (Unspecified depth) FIGURE 4.—Conceptual model of a hydrothermal convection cell. H7 TABLE 2.—Thermal-conductivity values used in this report [Values based on work done in Nevada and Utah by Olmsted and others (1975, p. 64; Olmsted, oral commun., 1976) and Sass (written commun., 1975). See also Sass and others (1976)] Estimated conductivity values Lithology (mcal/cm/s"C) Saturated Unsaturated Clay ____________________________ 1.5—2.5 1.5—2.0 Silt ____________________________ 2.5—3.0 2.0—2.5 Sand ____________________________ 3.0—4.0 2.0-3.5 Gravel __________________________ 4.0-6.0 3.0—5.0 Basalt _________-__-______, _______ 4.5—6.0 3.0—5.5 Vesicular basalt ________________ . 3.5—5.0 2.5—4.5 Sedimen rocks ______________ 5.0—9.0 4.0—7.0 Igneous roc s of felsic to intermediate composition ______ 61- ____ that, with no variation in regional heat flow, the geo- thermal gradient generally will vary as lithology var- ies. As a result, heat flow to the land surface is a more . consistent index of geothermal-resource potential than shallow temperature-gradient data. The values of thermal conductivity for the various lithologies, as listed in table 2, are based on laboratory determinations using cores and drill cuttings. The val- ues listed in the table are the ranges into which most samples fall. The factors that control thermal conduc- tivity are: (1) texture of rock or alluvium, (2) mineral content, (3) layering within the rock, (4) porosity and pore size, (5) degree of water saturation, and (6) the dissolved mineral content of the saturating water. There may be additional factors. Chemical composition of thermal spring waters can be used to estimate hydrothermal-reservoir tempera- tures. The geothermometers used in this report are listed in table 3 and are from Fournier and Rowe (1966) and Fournier and Truesdell (1973). The basic assump- tions in using these geothermometers (Foumier and others, 1974) are: (1) Temperature-dependent reactions occur at depth; (2) there is an adequate supply of chemi- cal constituents; (3) water-rock chemical equilibrium occurs at reservoir temperature; (4) re-equilibration at lower temperatures as the water flows to the surface is negligible; and (5) hot water is not diluted by shallow cold water. If mixing occurs, the least adversely affected geothermometers listed in table 3 are the Na-K-Ca geothermometer and the graphic method. Therefore, where temperature calculations differ and mixing is suspected, the Na-K-Ca and graphic-method calcula- tions should be favored. DISTRIBUTION OF THERMAL WATERS In this report a thermal well or spring is defined as having a water temperature above the average land- surface ambient temperature, which, as stated pre- viously, commonly ranges from 10°C to about 16°C de- pending on altitude and geographic location. In the H8 TABLE 3.——Formulas for geothermometers used in this report [From R. 0. Foumier, written commun., 1975 and 1977. Concentrations: Na, K, and Ca in molality; $103 in mg/kg, which is approximately the same as mg/L] Formula identification used in table 6 Formula Remarks NKC Na-K—Ca Geothermometer _ 1647 tr ‘ log(Na/K + Blog(Ca/Na) + 2.24 ‘ 273 If magnesium concen- trations are in excess of 10 mg/kg or if travertine (tufa) is eing deposited, computed temperature may be too high. B=‘/3 in all computations for table 7. Computed temperature usable if spring dis- charge is lower than boiling. Computed tem- peratures between 120°C and 180°C are of questionable meanin , because either chal- cedony or quartz may be controlling the silica concentration. Used mostly in basalt areas. Computed tem- perature usable if <120°C, but may be us- able up to 180°C. S Amorphous Silica Computed temperature ( silica gel) Geothermometer should be considered if 731 opal deposits are pres- oc = m — 273 ent. Q Quartz (conductive) Geothermometer _ 1309 7 tr ‘ 5.19—logSiOg ‘ 2 3 C Chalcedony Geothermomeier 1032 W = 4.69—logSi02 ‘ 273 t GM Graphic methods for es- A plot of dissolved silica timating temperature of a hot-water component in a mix-3d water (Truesdell and Fournier, 1977, and Mariner, R. H., written and enthalpy are used. Valid for spring with temperature lower than about 80°C and flow rate greater than about 2 Us. commun., 1978). Use sol- ubility curve of quartz above 120°C and chal- cedony solubility curve below this temperature where reservoir is volcanic rock. low-altitude areas of southwestern Utah (near St. George), the highest ambient temperatures prevail. Elsewhere on valley floors, the range is commonly from 10°C to 14°C. A map compiled by the US. Geological Survey for Utah (fig. 1) shows the eight KGRA’s and lands pro- spectively valuable for geothermal resources. Not sur- prisingly, the general geographic distribution is the same as distribution of thermal springs described by Mundorff ( 1970, fig. 2). Table 4 summarizes information on the KGRA’s. The geographic distribution of thermal waters in Utah is shown in figure 5. This distribution is based on the thermal-spring report by Mundorff (1970) and on data collected during this study. The springs are GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 4.—Known geothermal resource areas in Utah {October 1976. Total area includes all land irrespective of ownership. Leased areas are 41 federal leases] Location Area (km‘) Total Leased KGRA County Township Range (8.) (W) Cove Fort-Sulphurdale __Beaver, Millard 24—26 6—7 100 79 Crater Hot Springs ______________ J uab 13 8 70 70 14 8—9 _ _ _ - Lund __________________ Iron 32 14 16 14 Monroe-Joseph __________ Sevier 25—26 3—4 66 2.9 Navajo Lake ____________ Kane 38 8 10 0 Newcastle ______________ Iron 36 1 5 4.3 0 Roosevelt Hot Springs ______________ Beaver 26 9 1 2 1 100 27 9 _ - - - Thermo Hot Springs ______________ Beaver, 29 13 105 54 Iron 30 1 1— 13 _ _ _ _ 3 1 1 2 - _ - _ Total (rounded) __________________________ 490 320 grouped into northwest, southwest, and Wasatch Range Front areas in figure 5. In table 5, the spring groups have been categorized on the basis of several hydro- thermal characteristics. The springs of the southwest area have the most favorable characteristics. This is also the area of geothermal leasing of Federal lands (fig. 6). Most of the following discussions will be concerned with hydrothermal prospects in the southwest area. Estimated reservoir temperatures for selected sites are given in table 6. The temperature estimates are based on chemical analyses of geothermometers (table 3). Six prospects may have reservoir temperatures above 150°C and, therefore, they may have potential for generation of electricity: Roosevelt, Thermo, and Joseph Hot Springs, Newcastle area, the Cove Fort- Sulphurdale area, and the Monroe-Red Hill Hot Springs complex. Six other hot springs listed in the table may have reservoir temperatures in the 90—150°C range, and therefore, they have value for space and process heating. The locations of the first group are shown in figure 1. The table contains location numbers for all the sites. In the table, graphic-method calculations are based on the quartz-solubility curve unless the chalcedony-solubility curve is indicated. DISCUSSION OF PROSPECTS In this section, geothermal systems that are consid- ered to have the best potential for development are discussed. These systems include seven of the eight KGRA’s in Utah and several other areas of interest. The Lund KGRA was not included because of lack of data. (See table 18 for temperature data for areas not dis- cussed in this section.) RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH 114° IT — \N EXPLANATION I \\\\\\\\ é ‘ l 0 Area where geothermal \\\\\\ 3 gradient exceeds 100°C/km, l % heat flow exceeds 3 HFU, or water in springs and : \ shallow-well discharge is warmer than 20°C ' NORTHWEST (See table 5) AREAS 110° 109° 41° , I . __ __+_______ l\ \ “l— Emendover v I \ \ Split Mountain Warm Spring ———r—————— OGVHO'IOO _l I <1: \ | E \\\\\\\ - %\\\\\\ \ i I 390 4 1 {wore 4 Seuier \\ . | Lake \Piighfieldx $00} I I ‘ -o . § I D 1 I Blanding D : I [I] 510 190 130 290 KILOMETEHS [3 5'0 160 MILES FIGURE 5.—Geotherma1 areas of Utah. H10 39° 38° GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS 114° 113° 112° EXPLANATION D Township in which National Resource Land is leased for geothermal exploration and develop- l I : ment Area I Leases Number (km 2) 1 Competitive 41 318' Noncompetitive 1_7_2 1227 Total 213 1545 I l 0 20 40 KILOMETERS l Il———_‘l—l—_|—' I 0 10 20 MILES FIGURE 6.——Location of leased National Resource Land in Utah, September 1976. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH H11 TABLE 5.—General characteristics of thermal-spring groups [Group data based mostly on data in Mundorfl' (1970) and Milligan and others (1966). For location of groups, see fig. 5] m Location discharge Bil-grills“ ”3:3 :Tand Sigh-t 3n 52138323112" groups rate concentra- surface water charges Lions ('6) (mg/L) Wasatch Alon the major Wide Wide 20—77; hottest 15—85; Travertine common. Range. fa t zone ranges. ranges. springs are at highest separatin the t e south end at the Basin an Range of the area. south end rovince of the area. m the C010- rado Plateaus and the Middle Rocky Mountains provinces. North- In the Basin and __do ______ __do ______ Generally low. 5—33; Travertine common; west Range province, 18—42, exce t Crater Hot some hydrogen area. an area of for Crater 01'. S rin sulfide. above normal Springs (87°C) ( 7-5 ) conductive near south end near south heat flow. of area. end of area. .. South- --do ______ __do ...... __do ______ Generally high, 28—400; Siliceous sinter, west 32—85, com- hi hest travertine, and area. monly above at se- hydrogen sulfide 70. gelt Hot aommon; sulfur prings. e its at Su phurdale. ROOSEVELT HOT SPRINGS AND THE COVE FORT-SULPHURDALE AREAS The locations of these two adjacent KGRA’s are shown in figure 1. Roosevelt Hot Springs (C-26-9)34ch, is on the west flank of the Mineral Mountains in Beaver County, about 20 km northeast of the town of Milford. Cove Fort, (C-25-7)30, and Sulphurdale, (C-26-7)7, are about 25 km northeast of Roosevelt Hot Springs near the northeastern corner of Beaver County and on the west flank of the Tushar Mountains and the Pavant Range. The surface geology of the two areas is quite different. At Roosevelt Hot Springs, an alluvial valley lies to the west and a large Tertiary pluton of granite to the east. At Cove Fort and Sulphurdale, the dominant lithology is young basaltic lava flows, commonly of Quaternary age. Lee (1908, p. 21) describes a silica deposit on the southwestern flank of the Mineral Mountains and about 25 km south of Roosevelt Hot Springs. He reports that cold water issues from a mound of silica 400 m in diame- ter. This geologic feature was not visited as part of this study, but if this is a hot-spring deposit, it may be an area of geothermal-resource potential. From the loca- tion description, the mound probably is at (C-29-10)24c. Phillips Petroleum 00., Thermal Power Corp., and the University of Utah are actively exploring the hy- drothermal system in the Roosevelt Hot Springs area. Phillips and Thermal Power, who have made many geophysical surveys, have drilled a total of nine deep exploratory wells (as of April 1977) to depths commonly less than 1,500 m. Results of well tests reportedly were very favorable, and additional exploration and devel- opment wells are planned. The University has made extensive geophysical surveys, funded by the National Science Foundation, the US. Department of Energy (formerly the Energy Research and Development Ad- ministration), and the US. Geological Survey, and the results were published in a series of reports. Because of these very extensive exploration activities, which have been continuing since the early 1970’s, no additional field data were collected in this area as part of this study. The geology of the Roosevelt Hot Springs area has been mapped by Petersen (1975), Parry, Berson, and Miller (1976, p. 22), and Liese (1957). Petersen mapped 10 lithologic units, including 4 units of hot-spring deposits. According to Patrick Muffler (written com- mun., 1976), the heat source is related to Pleistocene rhyolites that crop out on the west flank of the Mineral Mountains (fig. 1). Rhyolite as young as 490,000 years has been identified. Hydrothermally altered ground is common in the area. Brown (1977, p. 5) estimates the age of hydrothermally deposited opal near Roosevelt Hot Springs at roughly 350,000 years. Basin and Range faulting controls the location of Roosevelt Hot Springs. One of the Basin and Range faults, the Dome fault, is marked by an abundance of siliceous sinter(opa1) in mounds for a distance of nearly 5 km. The model shown in figure 6, with some H12 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS TABLE 6.—Estimated reservoir temperatures derived by geothermometer formulas for springs (and one well) with temperatures greater than 50°C and stlica concentrations greater than 50 mg/L [For formulas, see table 3. Chemical data mostly from Mundorfi‘, 1970, table 1. Silica: First number is concentration for thermal-water sample; remaining numbers are concentrations for nearby cold water sources. Discharge temperature: First number is for thermal-woman sample; remaining numbers are for cold-water sources and correspond to silica concentrations in a 'oining column] Concentrations in Concentrations Temperature Formula Estimated molality (x 10‘“) in mg/L (°C) used tf) reservoir Hydrothermal Potas— Magne- Dis— Computed reser-voir aturg, in Remarks source Location Calcium Sodium sium sium (SiOQ) charge reservoir temper- ° (Ca) (Na) (K) (Mg) ature (rounded) Christensen well, Newcastle, UT__(C—36—15)20bb 1.45 11.74 0.54 0.4 99 95 166 NKC 140—170 No springs present. Difference 34 13 138 Q in computed temperatures 62 12 110 C ma indicate mixin of ther- 150—160 GM ma and nontherma waters. Cove Fort- Sulphurdale ---_T. 25 and 26 S. ____ ____ ____ ____ ____ --__ ____ ____ 200: Estimate from Renner, White, R. 6 and 7 W. and Williams (1975, p. 21). Crater Hot Springs ____(C— 14—8)1OS 8.61 35.50 1.23 168 59 87 110 Q 110—140 Reservoir rock may not be 19 12 80 C basalt. Differences in com- 22 14 140 GM puted temperatures may Indicate mixing of thermal and nonthermal waters. C stal ot Springs __-_(C—4—1)11 and 2.54 ____ ____ 124 73 58 120 Q 90—120 Mixed water. Dissolved-solids 12$ 92 C are 1,665 mg/L. Hatton Hot Springs ____(C-22—6)35ddS 11.6 _--- ____ 89 44 36 66 C 70—110 Subsurface temperature of 37 14 100—110 GM 67°C measured in nearby shallow well. Mixed water. Joseph Hot Springs -___(C—25—4)23S 7.04 62.64 1.73 136 85 65 101 C 100—170 ____ 50 12 170 GM Meadow ' Hot Springs ____(C-22—6)26ccS 10.8 44.37 3.58 1114 47 41 69 C 70—120 Mixed water. 37 14 120 GM Monroe ‘ Hot Springs ____(C—25—3)10ddS 7.01 24.06 1.25 149 51 65.5 73 C 70—120 Reservoir temperature prob- 36 13 115 GM ably the same as for Red Hill 0 d 33 14 Hot Springs or 100°—160°C. en ot Springs __-_(B—6—1)23ccS 8.41 119.19 10.40 8 53 58 323 NKC 75—90 Reservoir tern erature com- 30 10 75 C puted with ormula NKC is 90 GM2 probably too high. Dis— solved solids are 8,820 mg/L. Red Hill Hot Spring _-__(C—25—3)11caS 5.99 26.88 1.35 ‘34 83 75 99 C 100—160 Difference in computed 36 13 160 GM temperatures may indicate 33 14 mixing of thermal and non- thermal waters. Roosevelt Hot Springs ____(C—26-9)34ch .475 90.48 12.07 3.3 405 85 293 NKC 260—290 Subsurface temperature as 234 Q high as 262°C reported by 109 S Philli s Petroleum Co. Opal eposits. Stinking Hot Springs --__(B—10-3)30bbS 22.40 487.2 16.82 1335 53 51 75 C 75—95 Dissolved solids are about 30 10 95 GM2 36,000 mg/L. Thermo Hot Springs ____(C—30—12)2IS 2.07 15.57 1.25 9 7 108 82.5 199 NKC 140—200 Travertine deposits indicate 23 14 115 C that temperature computed 49 14 141 Q with formula NKC may be 170—200 GM too high. Difference in com- puted temperatures may indicate mixing of thermal and non-thermal waters. IFormula NKC not used where magnesium concentrations exceeded 10 lug/L. “Based on chalcedony solubility curve. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH modifications, is perhaps representative of this hy- drothermal system. Recharge to the system probably occurs only in a permeable zone in the Mineral Mountains, the same general area where much of the recharge for Milford Valley originates. Nearly all the hot, saline water rising as part of the convection cell enters relatively shallow, fresh-water aquifers and then mixes with the fresh water as it flows westward toward the axis of Milford Valley (Mower and Cordova, 1974, pl. 4). Roosevelt Hot Springs, the only thermal spring in the area, has had a very small flow during historic time, on the order of 1 Us or less (Mundorff, 1970, p. 42). Since about the mid-1960’s, the only flow to the surface has been a small seep supporting a very small area of tules. The measured temperature of the spring was 85°C in 1950. The spring was sampled at that time, and accord- ing to Mundorff (1970, p. 16), silica had a concentration of 405 mg/L and the dissolved solids were 7,040 mg/L. The dominant ions were sodium and chloride. The estimated reservoir temperature (table 6), on the basis of geothermometer calculations and reported well temperatures, is 260°—290°C. If the reservoir tempera- ture is dependent entirely on regional heat flow, the maximum depth of circulation would be about 6—7 km. This calculation is based on estimated reservoir tem- perature, a conductive heat flow of 2 HFU, mean ther- mal conductivity of 5 X 10‘3 cal/cm/s°C, and an ambient land-surface temperature of about 10°C. If a shallow, magmatic-heat source is present, the local heat flow could be much higher and the depth to the hydrothermal reservoir would be less. The amount of heat and water discharged by the hydrothermal system under native conditions was not estimated. No data have been collected as part of this study at the Cove Fort-Sulphurdale area for two reasons: (1) Phillips Petroleum Co., Union Oil Co. of California, and other companies are actively exploring the area, and (2) the only surface manifestations of hydrothermal activity are sulfur deposits, hydrothermally altered ground, and gaseous emissions at Sulphurdale. Mundorff (1970, p. 50) lists mine drainage, sampled by Lee (1908, p. 19—20) as having 10,810 ppm (parts per million) of dissolved solids, sulfate concentrations of 7,600 ppm, and iron concentration of 1,360 ppm. Most of the Federal land in the KGRA was leased to Union Oil Co. of California, but no deep holes have been completed to date (1977). In 1976, Union Oil Co. failed to penetrate more than about 300 m in a hole scheduled to be drilled much deeper. Additional deep-well drilling attempts are planned. The reservoir temperature has been estimated by Renner, White, and Williams (1975, p. 21) to be approx- imately 200°C. On the basis of the estimated reservoir H13 temperature, a conductive heat flow of 2 HFU, esti- mated mean thermal conductivity of 5 x 10”3 cal/cm/s°C for the rock, and an ambient land-surface temperature of 10°C, the maximum depth of circulation to the hydro- thermal reservoir is computed to be about 5 km. If a shallow, magmatic-heat source were present, the local geothermal gradient may be much higher and the depth of circulation could be much less. Recharge for the area and for the hydrothermal sys- tem is believed to come from the Tushar and Pavant Ranges to the east. Ground water in the area probably flows generally westward or northwestward, and it probably is the only significant means of heat and water discharge from the hydrothermal system. At Neels, a railroad siding about 60 km north of Roosevelt Hot Springs at (C-20-8)28b, Lee (1908, p. 32) reports that a well was drilled to a depth of 609 m, and hot water was encountered at a depth of 549 In. An entry in Lee’s log of the well (1908, p. 33) states that gas under pressure was sufficient to raise 2,800 kg of drilling tools 122 m up the well bore. According to Kenneth Bull (oral commun., 1977), gas-discharging vents have been found in the young lava-flow area northwest of Cove Fort and north of the Mineral Mountains. The triangular area including Roosevelt Hot Springs and the Cove Fort- Sulphurdale area and extending northward to Neels probably has the best potential for geothermal devel- opment in Utah. THERMO HOT SPRINGS Thermo Hot Springs is about 50 km southwest of Roosevelt Hot Springs along the axial drainage of the northern part of the Escalante Desert (Milford Valley) at (C-30—12)2IS and 288. Schmoker (1972) interprets aeromagnetic and gravity data to indicate the general area to be underlain by a large Tertiary, intrusive plu- ton of tabular form having a thickness of about 8 km. The mountains to the northwest, the Shauntie Hills, are a horst dipping to the southeast under the Tertiary volcanic rocks and extending at least as far south as the Thermo Hot Springs area. Schmoker interprets Milford Valley as a graben with an alluvial thickness ranging from 760 to 1,070 m. The Black Mountains, southeast of the springs, are mostly volcanic rocks associated with a possible caldera (Crosby, 1973) that range in age from 19 to 26 million years (Rowley, 1978). Rowley mapped a rhyolite about 3 km east of Thermo Hot Springs having an age of 10.3 million years (fig. 7). Rhyolites and other quartz- bearing volcanic rocks of Pliocene age have been mapped by Erickson (1973). Their occurrence is wide- spread in both the Shauntie Hills and the Black Mountains. H14 T308 38°1D’ T315 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS R 12W 113°10’ T1 Shauntie Hills | 6 Tvs /| _L J. / MKMH - Tm W l +3! />/ Th Mountains .4 3' Black .\?\ Base from aerial photographs Consolidated-rock geology adapted from Peter D. Rowley (1978) 2 KILOMETERS 0 1 MILE FIGURE 7 .——Reconnaissance geologic map of the Thermo Hot Springs area, and audio-magnetotelluric configuration for the Thermo Hot Springs area. RECONNAISSANCE OF THE HYDRO’I‘HERMAL RESOURCES OF UTAH H15 comm-"ON 0,.- MAP Uan Fault-controlledIThermo Hot‘ Springs flow from two — - north-trending spring-controlled mounds each of wh1ch 0"" Qma 09‘” de 0“) 0f _' mm is about 1 km long and ranges from about 50 to about 0c Old L QUATER- 200 m w1de (fig. 8). The mounds, wh1ch rise about 4—8 m — Pleistocene NARY above the surround1ng valley floor, are probably com- 0‘ _ posed mostly of windblown sand and lesser amounts of 0Tb l. mind" me _ siliceous sinter and travertine debris. On the spring - mounds, a total of 69 spring orifices were inventoried T" f “mm (table 16 in table section), 54 of which were on the west Th _ TERTIARY mound. The total observed spring flow was estimated to be about 2 Us and the maximum observed water tem- TVS F Mm“ perature was 825°C. Tm J Mundorff, (1970, p. 18) lists four chemical analyses of ‘ water samples from Thermo Hot Springs. The DESCRIPTION OF MAP UNITS - - - . HYDROTHERMALLy_RELMED Deposrrs max1mum 31llca concentratlon was 108 mg/l. andthe Om Spring-mound deposits—Windbbwn quartz sand collecting in d1ssolved sohds were 1,500 mg/L. The donnnant 1ons areas of spring discharge; mostly fine-to medium-grained, - llg'lt tan. Mound surface commonly white where salt has were SOdIPm and sulfate: accumulated from evaporated water. Minor amount of The est1mated reserv01r temperature (table 6), based travertine and siliceous sinter - Oma 5pmg_moum apron depoleMosfly windblown qua": sand on chermcal analyses of water samples and geother- m in mm: so? darkaray basalt fragments mometer (table 3) calculations, IS 140°—200°C. The vmy_FL68R DEangESmE ’ ta" maximum depth of circulation of water to the hy- Opw Wet-playadeposlls—Light-tan.sandyslltandclay-Landsurfaoe drothermal reservoir probably is between 3 and 4 km, soft due to saturation of water to the land surface. Surface . . . whim due to deposition {mm evaluated was, on the bas1s of a reglonal conductive heat flow of 2 HFU, 09" D’V'P'W‘ dmi‘S—Mwly '3" “"dysmand day- Landsu'lace mean thermal conductivity of the rock and alluvium of “midwmhem b t45 x 10-3 l/m/°C d b‘ tl d pr Flood-plain deposits-Mostly light-tan to light-brown silt and a 0“ ' ca c 05 ' an an am 1en an ' day, locally includes some sand. Deposits underlie hard surface temperature of 12 C. The calculatlon assumes mmmdm'wmked by “bunch” and “mm“? ‘m‘mnm' the absence of any shallow magmatic-heat source. Old Dasected lake deposits—Mostly light-tan to light-brown, fine- to medium-grained quartz sand and silt; minor amounts of dark-gray volcanic rock fragments up to 2 cm in diameter locally present includes sand dunes that form low hills 113°12'30" H 12 W southwest of the spring mounds , ALLUViAL—APRON DEPOSITS 1716 1615 Of Alluvial fan of Shauntie Hills drainage—Light-tan to light-brown 20 21 21 22 slty tine- to medium—grained sand. [and surface is commonly EXPLANATION 10 soft 0c Colluvium—Tan to brown fine- to medium-grained sand and 17 ~ silt, commonly with a poorly developed gravel pavement on a" Splgntgslzqg umber the land surface. Includes some gravel bars and sand dunes Malor drainage channels have dark-gray volcanic rock bouldersuprOcmin‘diameter EAST Qt Talus of Mount Dution Formation—Accumulations of angular 0 200 400 METERS PRIN blocis of rock *__‘|—'__J S G CONSOLIDATED ROCKS 0 1000 FEET MOUND 0Tb Basalt lava Horus _ 38° 9 Th Horse Valley Formation—Rhyodacitic lava flows and volcanic 12’ _ mudflow brwcia Trt Rhyoiite of Thermo Hot Springs Tvs Volcanic rods of Shauntie Hills—Volcanic mudflow breccia, lava flows, and ash-flow tuft Tm Mount Button Fonnation—Voicanic mudflow breccia and T minor lava flows 30 20 21 2122 27 28 28 27 —— Contact 3 WEST _T_ _?_ Fault, bar angereaggnbdownthmn side. dashed where inferred, SPRING queried w u tful ——-? Possible fault MOUND — ' — ' — Lineament --------------- Shore line of lake Bonneville 0— Hot spring —w— LINE OF EQUAL APPARENT RESISTNlTY—Line values in SOUTH MEADOW ohmmeters AREA ll-4) 0 Data point I FIGURE 7.—Continued. FIGURE 8.—Distribution of orifices at Thermo Hot Springs. H16 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 113°12'30” R 12W l ' ‘ 30 38° 10’ Basefrom aerial photographs: US. Geological Survey, 1953 Based on tempergtures measured at 0 .5 1 KILOMETER a depth of 30 m In wells, 1976 l ' l ' I 0 1/2 1 MILE FIGURE 9. ~Temperature at a depth of 30 m in the Theme Hot Springs area; and distribution of vegetation in Thermo Hot Springs area, 1976. RECONNAISSANCE OF THE HYDROTHERMAL‘ RESOURCES OF UTAH Recharge for the hydrothermal system probably oc- curs in the nearby mountains or seeps downward through a permeable zone from the saturated alluvium of the valley. To estimate the amount of mixing of the upflowing thermal water with nonthermal water, a graphic method developed by Truesdell and Fournier (1977) was used. The amount of mixing was estimated using temperature and silica-concentration data for thermal and nonthermal ground water in the area. For nonthermal ground water, the range of silica concen- trations used in the calculation was 23—49 mg/L and a water temperature of 14°C (table 6). The results suggest that as the thermal water rises, it mixes with nonthermal water at an approximate ratio of 40 percent thermal water to 60 percent nonthermal water and enters a shallow aquifer within the alluvial valley fill. Some of this mixed water is discharged by Thermo Hot Springs, but most is discharged from a shallow water table by evapotranspiration of phreato- phytes. The distribution of the mixed thermal-nonthermal water in the alluvium is shown on three maps. Figure 7 includes an audio-magnetotelluric map which shows DESCRIPTION OF MAP UNITS . HYDROTHERMALLY(?)-—REIATED PHREATOPHYTES 2 Mostly saltgrass and pickleweed; locally some greasewood, rabbitbrush, and saltbush. Land surface commonly covered with salt, and soil is clamp to land surface. Depth to water table generally less than 1 m. Includes tules around spring orifices and areas of ponded water Bare playa soil except for locally scattered mounds of greasewood and saltbush. Land surface white with abundant salt east of spring mounds. Smaller salt accumulation west of mounds. Soils damp and very soft. Depth to water table generally less than 0.5 m Mostly bare playa; locally some saltgrass and pickleweed. Land surface less salty and soils less damp than bare playa (preceding unit). Depth to , _ water table generally less than 2 m Greasewood and rabbitbrush mixed with lesser amounts of pickleweed and saltgrass. Depth to water table probably ranges between 1 and 3 m Greasewood growing on a playa; land surface locally salty. Depth to water table probably ranges between 2 and 5 m OTHER PHREATOPHYTES Greasewood, rabbitbrush, and saltbush. Mixed with big sage and shadscale on higher parts of the area. Proportion of big sage and shadscale increases with increase in altitude. Depth to water table ranges between 3 and 15 m XEROPHYTES Big Sage, shadscale, and associated xerophytes growing on upland areas where depth to water table exceeds about 15m; planls obtain moisture only from soils above water table ‘7" Hot spring — — — LINE OF EQUAL TEMPERATUREIN oC—Dashed where approximately located FIGURE 9.—Continued. H17 the low-resistivity, elongated body of thermal water underlying the general spring-mound area; this body has its long axis parallel to the northeastward direction of ground-water flow as defined by Mower and Cordova (1974, pl. 4). Figure 9 includes temperatures measured at a depth of 30 m below land surface. In addition to the thermal anomaly near the spring mounds, two ad- ditional thermal areas are shown to the southeast and east. These two areas are probably associated with permeable fault zones separate from the permeable zones underlying the spring mounds. Contours of con- ductive heat flow on figure 10 show similar patterns. Figure 9 shows the distribution of vegetation in the Thermo Hot Springs area. Three types of vegetation are shown: (1) The xerophytes in the northwestern and southeastern parts of the map that do not root to or use ground water, (2) greasewood, rabbitbrush, and saltbush that obtain much of their water from the shal- low ground water in the valley fill reservoir, and (3) phreatophytes near the spring mounds that use rising thermal water at rates greater than the surrounding phreatophytes. Net mixed-water discharge by phreatophytes from the hydrothermal system is com- puted in table 7 as the difference between the gross discharge of the area and the estimated discharge if no hydrothermal system were present. The estimated dis- charge of mixed water by phreatophytes is 1.4 x 10“6 m3/yr or equals a continuous flow of 44 US. The portion of this flow that is deep-circulating thermal water is about 40 percent or 18 US. Olmsted and others (1975, p. 66, p. 220—224) have developed two methods of estimating conductive heat discharge from hydrothermal systems. Their method (A) is based on a subsurface-temperature map and the thermal gradient from the map’s datum plane to the land surface. Method (B) is based on a heat-flow map for a hydrothermal system. Because both methods use the same data base, only method (B) is presented in table 8. Method (B) yields a conductive heat discharge from the system of 16 X 1013 cal/yr; method (A), 15 x 1013 cal/yr. A third method was also used, based on an esti- mated reservoir temperature of 200°C (table 6), convec- tive water flow through the deep hydrothermal‘reser— voir of 0.6 X 103m3/yr, and an ambient land-surface temperature of 12°C. This last method yielded the low- est of the three estimates, 11 X 1013 cal/yr, and was applied only to the area of evapotranspiration near the spring mounds. In that computation of convective flow, because no increment for warm ground-water flow to the northeast from the thermal areas was included, the estimate of total heat discharge is probably too small. The value 15 x 1013 cal/yr was selected to represent the probable heat discharge from the entire hydrothermal system. As depth increases, the three heat anomalies H18 shown in figures 9 and 10 probably expand and merge into a single hydrothermal system. SOUTHWESTERN ESCALANTE DESERT The southwestern part of the Escalante Desert is northwest of Cedar City (fig. 1). Three communities are GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS referred to in this section: Newcastle ([C-36-15]17), Beryl ([C-33-16132), and Lund ([C-32-14121) (fig. 11). Newcastle is in the southeastern part of the desert, about 40 km west of Cedar City. Beryl is northwest of the valley axis and 30 km northwest of Newcastle. Lund is 25 km northeast of Beryl along the Union Pacific Railroad tracks. 113°00' R13w 113°15' n12w 113°10' 113°05' RIUW 38°15’ 1 I 129 s // / I T 30 3 38°10’ — T31 s 5 KILOMETERS 3 4 MILES 38°05’ — l . EXPIANATION Contact Alluvium—Mostly clay, silt, and sand CONSOLIDATED ROCK—Mostly Tertiary and Quaternary volcanic rocks Fault, dashed where approximately located; bar and ball on downthrown side Lineament observed on aerial photographs LINE OF EQUAL HEAT FLOW—Dashed where approximately located. Unis are is ca] ems? secil, FIGURE 10.—Shallow conductive heat flow in the Thermo Hot Springs area. RECONNAISSAN CE OF THE HYDROTHERMAL RESOURCES OF UTAH H19 TABLE 7. —-Evapotranspiration of mixed water from Thermo Hot Springs hydrothermal system [1976 conditions. The combined and nonhydrothermal system discharge rates are based on research by Lee (1912) White (1932), Young and Blaney (1942), Houston (1950), Robinson (1965), and Harr and Price (1972) in other areas] Average annual evapotranspiration rates (approximate) Estimated average Depth Combined Nonhydrothermal Net Area annual net Phreatephyte area to water discharge system hydrothermal (X lO’m‘) discharge table rate discharge rate discharge (rounded (in) (m) (in) (m) x 103m“) / Mostly saltgrass pickleweed ________________ <1 0.5 0.06 0.44 1,700 750 Bare playa soil ________________ < .5 .6 .06 .54 430 230 Bare laya, saltgrass, npickl eweed ____________ <2 .4 .06 .34 520 180 Greasewood, rabbit- brush, ickleweed, and sa tgrass ______________ 1-3 .2 .06 .14 1,400 200 Greasewood on laya __________ 2—5 .1 .06 .04 420 20 Greasewood, ra bit- brush, and saltbush __________ 3—15 .06 .06 .0 __ 0 Total (rounded) __________ __ __ __ 1.3 4,500 21,400 (1.4 x 105m3/yr) 1Average for area. 1Mixed thermal and nonthermal water. Newcastle is on the floor of the Escalante Desert near the northwestern flank of the Pine Valley Mountains, a range composed mostly of Tertiary volcanic rocks (Hintze, 1963). The floor of the Escalante Desert is de- scribed by Crosby (1973, p. 28) as the central region of a probable large caldera about 50 km in diameter. The rim of the caldera includes the surrounding mountains and Table Butte as shown in figure 11. Crosby gives no descriptions of the age or structure of the caldera in his report, but it is probably Tertiary. The detailed distribution of young igneous rocks in this area is poorly known. Some basic igneous rocks less than 10,000 years old lie about 40 miles south of New- castle, near the town of Veyo (Smith and Shaw, 1975, p. 82, listed under Utah as Santa Clara). The reconnais- sance geology of the area is shown in figure 12. During December 1975, a newly drilled irrigation well (C-36-15)20bbd was test pumped at rates as high as 108 US. This well, owned by the Christensen Brothers of Newcastle, is 152 m deep, has a 40-cm-diameter casing, and a static water level of about 43 m. The water dis- TABLE 8.—Estimated conductive heat discharge from Thermo Hot Springs hydrothermal system—alluvial area only [Method B of Olmsted and others (1975, ph 67). The approximate area was determined from gure 10 Range in heat flow units Approximate mean Approximate ( heatfl flow area Heat discharge (X 10“ cal/cmz/s (HFU) (km‘) (X105 cal/s) > 15 ________ 20 3.3 6.6 10-15 ________ 12.5 3.3 4.1 5— 10 ________ 7.5 24 18 3— 5 ________ 4 50: 20: Total (rounded) _ _ _ _ 6 80 50 (16 x 10'3 cal/yr) charged was boiling, about 95°C at land surface (a1- titude = 1,605 i- m). A water sample was collected after 6 hours of pumping; the chemical analyses are given in table 9. The dissolved-solids concentration was only 1,120 mg/L; the silica concentration was 99 mg/L; and the dominant ions were sodium and sulfate. On January 20, 1976, after a period of several weeks during which the pump was idle, subsurface temperatures were measured in the well, resulting in the temperature pro- file in figure 13. The profile shows an alluvial aquifer containing hot water at a depth below land surface between 70 and 110 m. Subsurface temperatures were lower above and below this aquifer. The maximum tem- perature recorded in the well was 107 .8°C; the bottom- hole temperature was 4.1°C less. The well was pumped during the 1977 irrigation season; the water was cooled in two ponds and applied to cropland by sprinklers. The estimated reservoir temperature for the hy- drothermal system, based on chemical analysis and geothermometers, is 140°—170°C in table 6. The differ- ence in the calculations in table 6 probably results from mixing of thermal and nonthermal waters. The maximum depth of circulation in the hydrothermal res- ervoir required to produce these temperatures is esti- mated to be 3—4 km, on the basis of regional heat flow of 2 HFU, thermal conductivity of the volcanic rocks that underlie the area of 5 x 103 cal/cm/s°C, and an ambient land-surface temperature of about 14°C. The calcula- tion assumes the absence of a shallow magmatic-heat source. Using the graphic method of Truesdell and Fournier (1977) to estimate the mixing of thermal with nonthermal water, the hot-water component of the mixed water is estimated to be about 60 percent on the basis of data in table 6. H20 Shape and size of the hot-water body in the valley-fill (alluvial) aquifer are indicated in figures 12, 14, and 15, as well as the temperature profile shown in figure 13. The recharge area for the hydrothermal system is probably in the Pine Valley Mountains southeast of the town of Newcastle. The upward flow of thermal water is through a permeable range-front fault zone 1 km south- GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS east of the thermal well (fig. 12). As hot water flows northward from the fault zone, its temperature declines owing to conductive-heat loss, mixing with cold, shallow-aquifer water, or both. Westward flow is rela- tively small compared to northward flow; this probably results from higher transmissivity toward the north than toward the west. Faulting and lithologic changes 114°00'R19W 818W 917W R15W R13W 113°15' RlZW I + + + ‘ -i' T 32 Rim of possible caldera %/ 3 (Crosby, 1973) \L % / " — \ 38°00’ _ «*6? / / 0t?— BPNGE - 1 E ° o$ / “E + 6‘ T P0 \\ 2‘1 _ o ’9 I 3: \ // w a? ' .<2 2 \ ////°/ / + T_96°C t /-.+ .1 // + T / -<1 AC‘F‘C depth 1.97am<2 Ovab'e Bull)" /(6 CT: 34 I ON 2 / 9' ar N‘ '13: EW / CW 3 l U l ‘5 //4/// <2 0 Modena|/" _+_ . + + l + + T /‘ + 0 <2 35 /_ .<2 -<2 a [.<1 8 / \ .9 .<2 “ 1 '<2 (go 37°45, ' \ I? 56‘ .<2 P~ For data in the / ‘23? \ p. 2 1 Newcastle area (0 s c .' Beryl see fig. 15 / 5 A _rL_ E 11. - Junction ‘rL +/ ’3); + 1.9 + To T \\ o3 / / Ye 0.2 .2 C313" ‘ ' 2 , 36 \ 1. /D Newcastle / 5 S , 6 \ \ %/ / . 2 59 \ / EXPIANATION \ \ \ ‘ <2. , ’F + + 4' + + +— 0 <2 ‘1 0: <2 2.4 T E Area of probable high ' E te. . o 3 $2 37 heat flow I n rpnse ’2‘ <1. S 1'70 Well. Number is estimated m C? Kannarraville 0 heat flow in y cal/cm2 sec ‘5’ l— + 5+ + + + + 3 4 g, T 0 5 10 KILOMETERS g?" 38 s 3’ U 5 MlLES Q L 3T9 JV 4" + 1 7 + + S a Central 0 ' | Base from US Geological Survey 1250.000 quadrangles FIGURE 11.—Estimated heat flow in southwestern Utah. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH TABLE 9.—-—Chemical analyses of water from Christensen brothers thermal well near Newcastle, Utah _ [Location ((3—36— 15)20bbd. mg/L, miligrams per liter; p /L. micro ams per liter; L/s, lite per second umho, micmmhos; ”C, egrees Geld-ins) H21 TABLE 10.——Estimated conductive heat discharge from. the Newcastle hydrothermal system—alluvial area only [Method B of Olmsted and others (1975, p539). 1331]: approximate area was determined from re Alkalinity, H, field __________ 7.6 total hosphate, (as CaC03) _-mg/L 53 Ortho, Aluminum, dissolved as dissolved - _ _ -ug/ L 40 phosphorus _ -mg/ L .04 Arsenic, Phosphate, dissolved - - _ _;Lg/ L 100 dissolved, Barium, Ortho ...... mg/ L .12 dissolved _ _ - -ug/ L 300 Potassium, Bicarbonate _ _mg/ L 64 dissolved _ _ _ -mg/L 2 1 Boron, Residue, dissolved _ _ _ -ng/ L 710 dissolved Calcium, calculated dissolved ____mg/L 58 sum ________ mg/L 1,120 Carbonate - _ - -mg/ L 0 Sodium Chloride, absorption dissolved - _ - -mg/ L 52 ratio ____________ 9. 7 Cobalt, Selenium, dissolved _ _ _ _p.g/ L 0 dissolved _ _ - _/.Lg/L 0 Fluoride, Silica, dissolved ____mg/L 7.3 dissolved --__mg/L 99 Hardness, Sodium, noncar- dissolved _ _ _ -mg/ L 270 bonate ______ mg/ L 95 Sodium ........ Percent 77 Hardness, Specific total ________ mg/ L 150 conductance, Iron, field ________ )1.th 1,550 dissolved _ _ _ _p.g/ L 10 Specific Lead, conductance, dissolved - _ - _ptg/ L 1 laboratory - -umho 1,600 Lithium, Strontium, dissolved -___p.g/L 460 dissolved,___-p.g/L 1,100 Ma esium, Sulfate, dissolved ____mg/L .4 dissolved __--mg/L 580 Manganese, Water dissolved - _ _ _pg/ L 70 tempera- Mercury, ture ________ °C 97 .0 dissolved “Hug/L .1 Yield-well ____L/s 95 Molybdenum, Zinc, dissolved _ _ _ _p.g/ L 13 dissolved _ - _ _p.g/ L 20 N 02 +NO3 as Nitrogen, dissolved _ _ _ -mg/ L .22 to the west may be factors reducing westward flow. A helium-concentration survey (fig. 16) made by Denton (1976) produced a pattern generally similar to those in figures 14 and 15 near the fault zone that produces the hot water. He concludes that the helium is released from solution in the water as a result of either temperature or pressure decline while the water flows laterally in the shallow aquifer (E. H. Denton, oral commun., 1977). Estimates of conductive-heat discharge to the land surface were made using method (B) of Olmsted and others (1975) (table 10) yielding identical results to method (A) of 7 x 10‘3 cal/yr. Calculations of method (A) are not presented. Essentially all the discharge from the hydrothermal system is by lateral ground-water flow from the fault zone. The flow rate was computed on the basis of an estimated conductive-heat discharge and a reservoir temperature of 170°C. The estimated flow from the hydrothermal reservoir is about 0.4 X 106m3/ Range in heat Geometric mean Approximate Heat discharge flow units heat flow area (x 10-5 cal/5) (X 10" cal/cm'ls) (HFU) (km’) >30 ________ 50 1.2 6.0 20—30 ________ 24 1.0 2.4 10—20 ........ 14 3.3 4.6 5— 10 ________ 7 1 7.8 5.5 3—5 ________ 3 9 9.1 3.5 Total (rounded) ____ 110 22 22 (7.0 X 1013 cal/yr) ‘Average for area. yr. The mixed water flows through the alluvium at an estimated rate of 0.7 x 106m3/yr. Figure 11 shows estimated heat flow based on meas- _ ured temperature gradients for many sites in and near the southeastern part of the Escalante Desert. In addi- tion to the Newcastle area, high heat flow is shown west of Beryl and northwest of Lund. The Lund KGRA (fig. 1) is a few kilometers southeast of Lund, but no data were available that indicated high heat flow in the KGRA. Another area east of Table Butte may have high heat flow, but data are sparse. In the irrigated area between Enterprise and Beryl, both upward and downward flow of shallow ground water may be distorting conductive heat flow to the land surface. MONROE AND JOSEPH KGRA The Monroe and Joseph KGRA lies along the Sevier River in Sevier County, south-central Utah (fig. 1). The Pavant Range is to the northwest, the Sevier Plateau to the southeast. The towns of Joseph ([C-25-4]14) and Monroe ([C-25-3]8 and 17) are on the flood plain of the river. Joseph Hot Springs is about 2 km southeast of Joseph (fig.17). Monroe Hot Springs is on the east edge of Monroe; Red Hill Hot Spring is about 1 km farther northeast. Tertiary volcanic rocks are dominant in the area. Monroe and Joseph are near the north edge of the Marysville volcanic area, where volcanism was exten- sive and prolonged from middle to late Tertiary. The volcanic rocks range in composition from basalt to rhyolite. Smith and Shaw (1975, p. 72) cite the age of the latest eruption, a rhyolite, as 20 million years. In the following discussion of the KGRA, Joseph Hot Springs will be described separately from Monroe and Red Hill Hot Springs because superficially it is a separate hy- drothermal system. Monroe and Red Hill Hot Springs are parts of a single system. Monroe and Red Hill Hot Springs flow from travertine mounds forming a bench at the western foot of the H 37°37’30” 22 113°37'30" 37°45' GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS R 16 W H15 W EXPIANATION l13°30’ O .__.> \ \ 7‘0 / /554 / l \ / / // l \ v / / Of \ \ //// \ an ‘ / \ \ 6/76 / \ 66 —I 5 60-— POTENTIOMETRIC CONTOUR—Shows altitude of water level in wells. Dashed where approximately located. Contour interval 2 meters. Datum is mean sea level Well Of \ Direction of ground-water flow \ / /I . I Qvf xnl—/ / / / Ofp O \‘3 \ Base from US. Geological Survey pr Qc ‘ 0c Newcastle 1:24,000, 1972 3 KlLOMETERS I O——D I 1 2 MILES Adapted from G. W. Sandberg. USGS,Cedar City, Utah, written communication 1976 FIGURE 12.—Reconnaissance geologic map of the Newcastle area; and ground—water levels and direction of flow in the alluvium of the Newcastle area, spring 1976. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH Sevier Plateau. The springs are near the Sevier Fault, as shown on figure 18. Parry, Berson, and Miller (1976, p. 67) have mapped the consolidated-rock units near the spring. The principal units are Pliocene andesite and Miocene volcanic flows and tuffs. A flat valley floor west of the spring bench is underlain by alluvium of un- known thickness (fig. 18). As part of this study, periodic measurements were made of all spring flow and water temperatures at Monroe and Red Hill Hot Springs (table 11). From March 1976 to March 1977 total flow varied from 11.0 to 21.1 US and temperature ranged from 72°——75°C. The cause of the variation is not understood because data CORREIATION OF MAP UNITS Oya Ofp Holocene QUATERNARY Of Qc Ql Pleistocene Tv TERTIARY DESCRIPTION OF MAP UNITS YOUNGER ALLUVIUM—Light—tan sandy silt and clay deposited in nearly horizontal, playalike areas. Land surface is hard and commonly has desiccation cracls. Deposit is thin and unsaturated. Ofp FLOOD DEPOSITS OF PINTO CREEK—Light-tan to light—brown sandy silt deposited by Pinto Creek on its alluvial fan during frequent flood flows. Sand is mostly fine-grained quartz and volcanic-rock fragments. Material is mostly reworked Qf, described below. Fomis moderately hard land surface with some desiccation cracks. Deposit is thin and generally unsaturated by ground water. lndudes sand and gravel deposits along Pinto Creek in the mountains. Of ALLUVlAL-FAN DEPOSITS OF PINTO CREEK—Mostly light-brown fine- to medium- grained silty sand and sandy silt deposited by Pinto Creek Sand is mosiiy quartz and volcanic—rock frag- ments. Material is derived mostly from volcanic rocks of the Pinto Creek drainage basin in the Pine Valley Mountains. Forms moderately soft land surface. Depth to ground-water saturation ranges from 15 m to 30 m beneath the land surface. Oc COLLUVIUM—Mosfly medium- to dark-gray poorly sorted silt, sand, gravel, and boulders derived from the adjacent volcanic rocks of the mountains. Under- lie a generally steep-sloping apron. Depth to ground-water saturation is generally greater than 30 an rn. va LAKE BONNEVILLE SEDIMENTS—Mostly light-tan sandy silt. Largely undissected. Land surface generally hard. Depth to ground-water saturation is generally less than 30 m. Tv VOLCANlC ROCKS—Generally dark-gray basalt; commonly vesicular. —— -- Contact; dashed where approximately located —'I— — Fault; dashed where inferred. Barand ball on down- thrown side —- - —- Lineament observed on aerial photographs FIGURE 12.——Continued. H23 were collected for only 1 year, 'probably a period too short to establish trends and causes of the trends. The annual spring flow was about 0.5 x 106m3. A small amount of additional water, about 0.05 X 106m3/yr, is estimated to be discharged by evapotranspiration by phreatophytes on the travertine bench. The flow of Red Hill Hot Spring, the largest spring in the complex, was at its maximum during October through March of the period of record. Generally, this was also the period of lower measured temperatures of the spring. This fact suggests that flow from the spring probably is a mixed water and that during the period of high flow the propor- tion of nonthermal water in the mixture is larger than during periods of low flow when the spring has higher temperatures. On the basis of the graphic method of estimating hot-water component in a mixed water (Truesdell and Fournier, 197 7) and data in table 6, it is concluded that thermal water probably is mixing with an equal amount of nonthermal water between the point where it leaves the hydrothermal reservoir and the springs. The thermal-water component in the mixed water that is discharged by the springs was about 0.2 X 106m3/yr. Samples of water from Red Hill and Monroe Hot Springs (Mundorff, 1970, p. 16) had sodium, sulfate, and chloride as the dominant ions as well as similar dissolved-solids concentrations of 2,630 mg/L and 2,700 mg/L, respectively. However, Red Hill had a silica con- centration of 83 mg/L compared to only 51 mg/L for Monroe Hot Springs. Farther south on the Sevier fault, Johnson Warm Springs (fig. 17), whose water chemistry is much different than the other two springs, has dis- solved solids of only 428 mg/L. 0 l I Temperature 20 _ probe not in — W . I contact with _ ater 6V6 formation 2 4° 43 m Lu 5 so _ Location (0—36-15) 20 bbd - 2 Depth 152 m E 80 _ Diameter 40 cm _ I. Altitude 1605: m Principal hot-water E aquifer, 70 to 110 m ”DJ 100 '- .. Maximum temperature 120 F 107.8°C at depths 85-95 m ‘ Bottom hole temperature 140 - 103.7°C ‘ l l l AL I l 75 80 85 90 95 100 105 1 10 TEMPERATURE, IN °C FIGURE 13.—Temperature profile of the Christensen Brothers ther- mal irrigation well near Newcastle, Utah. H24 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS R15 W 113°30’ "113‘37’30'", R 16 W 3’33“. EWNA'HON ~—*——- 60 ~— LINE OF EQUAL WATER TEMPERATURE, 1976-— Dashed where approximately located. Interval 10 degrees Celsius, except for 15 degree line . m7 ‘\ ‘atadapthto? 10019.3] e1 37°37‘30" / Well am] temperature in °C \\\ x ,‘ \ . J1 \_ ' \ A/ ,7 5/ 9| 1 X iraased' " ‘on wmuresmeasum é { /‘ rifleme‘n brothers we! w/ica-ss—‘aés ate-chm! A \ ‘K . v ’Base flor'n ‘UIS. Gedlugiw survey Newcas‘tte i1:24£00.1922 re :1 2 3 mam-as '11 it A 1L 9 Fr a _ m n 2 Mums 1171mm aka-mama fat 3 l'depfih to? 3150!) an in We Newcasele urea. RECONNAISSANCE OF THE HY DROTHERMAL RESOURCES OF UTAH H25 113°37’30" R 18 W H 15 W 113°30’ T 35' EXPIANA‘I‘ION S ‘ — 5 — LINE OF EQUAL CONDUCTWE HEAT FLOW—Dashed where ‘approxfinately locawd. Unis are u <21 urn-2 sec. -1 / j o 79 Well and beat—flow waits / .r \l 1 ‘1 ~ \ ‘ \ ”Em,“ , a 1| / 3 xx 1 S 37°37rw' ; Base from us. .SemogicalSuwey Mmeastle 1223;090, 1972 ' r! 2 MILES FIGURE m5.—Dmmw«dfm flow £10m the W Wmnqui’fef :in ‘fihe Newcastle «apes. H26 113°35’30” GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS “3032, The estimated reservoir temperature for the 37°40’30” H18 R15W \W Newcastle W Christensen brothers well V / / 1 Base from US. Geological Survey Adapted from Denton (1976) Newcastle 1:24.000, 1972 0 ‘| 2 KILOMETEHS |_____lfi_l 0 l MlLE EXPLANATION —-— 17 — - - LINE OF EQUAL HELlUM CON— CENTRATION—Dashed where approx— imately located. Units are parts per billion in excess of ambient concen- trations in air 37°37 ’ 30” FIGURE 16.—Helium concentrations in the Newcastle area at a depth of 0.6 m below land surface. Monroe-Red Hill Hot Springs hydrothermal system is 100°—160°C, as determined from data in table 6. The maximum depth of circulation to the hydrothermal res- ervoir, in order to produce these temperatures, is esti- mated to be about 2—4 km on the basis of a regional conductive heat flow of 2 HFU, thermal conductivity of the rock of about 5 x 10‘3 cal/cm/s°C, and an ambient land-surface temperature of 12°C. The calculation as- sumes the absence of a shallow magmatic heat source. The hydrothermal system probably is recharged on the Sevier Plateau, and thermal water circulates upward through a permeable fracture zone associated with the Sevier fault. In addition to the spring flow and evapotranspiration, discharge of mixed water may in- clude an undetermined amount of lateral subsurface flow from the fault zone. Figure 17 shows a few data points for heat flow and the area of probable above- normal regional heat flow confined to the vicinity of the fault. The estimated heat discharge by spring flow and evapotranspiration is 3 X 1013 cal/yr. This estimate is based on a waterflow from the hydrothermal reservoir of a minimum of 0.2 X 106m3/yr (50 percent of spring and evapotranspiration discharge) and an estimated maximum reservoir temperature of 160°C. Lateral sub- surface flow from the fault zone would discharge ad- ditional heat from the system. Joseph Hot Springs is 8 km southwest and across a low unnamed mountain ridge from Monroe Hot Springs (fig. 17). The general geologic and topographic settings are similar to those at Monroe Hot Springs. The springs flow from a travertine bench at the western foot of the H 4‘/z W HAW 112°15’ H4W 112°10’ RBW 112°05’ EXPLANATlON Area of probable high heat flow '3 Well. Number is heat T flow in g cal/cm2 sec 25 0 1 2 3 KILOMETERS S l_l_'_L___J 0 1 MILE 38°35’ l | l l FIGURE 17 .—-Estimated heat flow and measured spring temperatures in the Monroe-Joseph area. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH H27 TABLE 11.—Measured flow and temperature of Monroe and Red Hill Hot Springs [locations of sites shown on figure 18) Sites Date Total RedHill Tunnel 1 2 3 4 5 6 7 flow (Us) [/3 °C L/s °C L/s “C L/s °C L/s ”C L/s “C L/s ”C Us °C Us “C 3—16—76 ______ 5.7 -__- 0.6 ____ 0.3 ____ 0.8 ____ 0.4 ____ 0.1 ____ 0.1 ___- 4.1 -__- 0.2 ____ 12.3 4—13—76 ______ 5.7 75 1.0 45 .2 40 .4 56 4 50 .1 40 .1 ____ 4.4 47 .0 ____ 12.3 5—11—76 ______ 5.6 75 .2 43 .1 38 .4 66 3 48 2 38 .1 ____ 4.1 48 0 __-_ 11.0 6— 7-76 ______ 6.2 75 .5 45 .1 40 .4 65 4 50 2 40 .0 _-__ 4.7 48 ___- ____ 12.5 7—21—76 ______ 6.6 75 <0.1 44.5 .1 39.5 .4 65.5 4 50 2 40.5 .0 ____ 4.7 48 .0 ____ 12.4 8—24—76 ______ 6.6 75 < .1 45 .1 40 .4 65 4 50 2 40 .1 -__- 4.6 48 .0 _-__ 12.4 10— 5—76 ______ 13.6 75 < .1 45 .1 40 .4 65.5 4 48 2 40 .0 _-__ 4.7 48.5 .0 ____ 19.4 12—27—76 ______ 15.2 74 < .1 36 .5 40 .5 60 5 45 1 32 .0 ____ 4.0 45 .0 -___ 20.8 2— 1—77 ______ 15.2 72 < .1 37 .4 39 .4 60 4 46 1 30 ____ ____ 4.6 35 .0 ____ 21.1 3—26—77 ______ 11.2 73 < .1 36 .4 39 .3 60 4 41 1 30 __-_ ____ 4.2 44 .0 ____ 16.6 ridge. The permeable zone through which hot water flows upward is the Dry Wash Fault zone. The narrow flood plain of the Sevier River is west of the fault. Table 12 lists the periodic measurements of all flow from Joseph Hot Springs. The flow ranges from 1.3 to 2.7 US or an annual flow of about 0.07 x 106m3/yr. During the year that measurements weremade, flow was generally declining for unknown reasons, but no pattern was ob- served in any relationship between flow and tempera- ture as was observed at the Monroe-Red Hill Hot Springs complex. The maximum observed temperature of the springs was 65°C. In addition to the spring flow, a small amount of water, about 0.01 X 106m3/yr, is esti- mated to be discharged by evapotranspiration of phreatophytes on the travertine bench. Chemical analyses of samples from Joseph Hot Springs (Mundorff, 1970, p. 16) show that the highest silica concentration was 85 mg/L. That sample had a dissolved-solids concentration of 5,150 mg/L. The domi- nant ions were sodium and chloride. The estimated res- ervoir temperature (table 6) and depth of circulation of thermal waters are about the same as for the Monroe- Red Hill Hot Springs system. Recharge for the hydrothermal system may originate as precipitation in the nearby highlands or come from saturated alluvium underlying the Sevier River flood plain. During upward circulation from the hydrother- mal reservoir through the permeable fault zone, the thermal water is estimated to mix with nonthermal water in the proportion 35—65 percent, respectively. TABLE 12.—Measured flow and temperature of Joseph Hot Springs [Sites 1— 12 are progressively farther north, with 1 the southernmost and 12 the northernmost] Sites Date 1 2 3 4 5 6 Us C Us °C L/s ”C L/s °C L/s °C L/s °C 3—16—76 ____________ 0.8 __-- 0.1 ____ <0.1 ____ 0.1 _--- 0.3 ____ 0.5 ____ 4—14—76 ____________ .4 58 < .1 47 .1 50 .1 65 .3 65 .5 61 5—11—76 ____________ .2 47 .1 47 .1 49 .2 65 .3 64 .6 60 6— 7—76 ____________ .2 56 .1 48 .1 49 .2 64 .2 64 .7 60 7—21-76 ____________ .2 56 .1 47 .1 50 .2 64 .3 64.5 .6 60 8—26—76 ____________ .2 56 .1 47.5 .1 50 .2 64.5 .3 64.5 .7 60 10—26—76 ____________ .2 56 .1 48 .1 49.5 .2 64 .3 64.5 .7 60 12—27—76 ____________ .2 52 < .1 48 < .1 48 .1 64 .3 55 .4 62 2— 1—77 ____________ .2 45 < .1 46 < .1 48 .1 62 .4 57 .4 59 3—26~77 ____________ .2 44 < .1 45 < .1 49 .1 61.5 .2 61 .2 60 Sites Date 7 8 9 10 11 12 Total I flow L/s °c Us 0 : Us to Us 0 Us °c L/s °c (Us) 3—16-76 ____________ 0.1 ____ <0.1 ___- E 0.1 ___- 0.4 -___ 0.2 ___- 0.2 ____ 2.7 4—14—76 ____________ .2 52 .1 61 g .2 59 .4 42 .2 46 .2 37 2.7 5—11—76 ____________ .2 47 .1 59.5 :1 .2 58 .3 43 .2 59.5 .1 38.5 2.6 6— 7~76 ____________ .2 48 < .1 61 g .2 59 .2 43 .2 59 .1 38.5 2.4 7—21—76 ____________ .2 47.5 < .1 61 g .2 59 .2 42.5 .2 59.5 .1 39 2.4 8—26—76 ____________ .2 48 < .1 60.5 a, .2 60 .2 42.5 .2 60 .1 38.5 2.4 10—26—76 ____________ .2 47.5 < .1 61 i .2 59 .2 43 .2 59.5 .1 39 2.5 12—27—76 ____________ < .1 47 < .1 60 1 < .1 58 .5 43 .2 60 .1 38 1.8 2— 1—77 ____________ < .1 45 < .1 59 I .1 58 .3 41 .2 59 .1 38 1.8 3—26—77 ____________ < .1 48 < .1 58.5 l < .1 50 .3 44 .2 45 .1 33 1.3 H28 Therefore, only about one-third'of the earlier reported rates of spring and evapotranspiration discharge would have circulated through the hydrothermal system. An additional unknown amount of thermal water is being discharged into the alluvium from the fault zone. The heat-flow values on figure 17, ranging from 8 to 18 HFU, probably are the result of high temperature gradients in the sha110w ground-water system. Beneath the alluvial aquifers containing this shallow lateral flow, conductive heat flow toward the land surface prob- ably is much less, such as is illustrated by the tempera- ture profile in the Christensen Brothers well near Newcastle (fig. 13). The estimated heat discharge by spring flow and evapotranspiration is 0.3 X 1013 cal/yr on the basis of a cooling of thermal water from a temperature as high as GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS 170°C to an ambient land-surface temperature of 12°C. Lateral subsurface flow of mixed water from the fault zone to alluvium would discharge additional heat from the system. CRATER HOT SPRINGS Crater Hot Springs, also known as Baker Hot Springs and Abraham Hot Springs, is about 30 km northwest of Delta, Utah (fig. 1), at (C-14-8)IOS. The springs and Crater Bench, a nearby, young basalt flow, are on the northwestern part of the broad, nearly flat alluvial floor of the Sevier Desert. The springs flow from a low travertine and alluvial mound a few hundred meters east of the lava flow. The altitude of the land surface at the spring mound is about 1,410 m above sea level. H 3W 112°[B' 38°38’ - School T D 25 S l Hot Spring EXPLANATION Spring-deposited travertine; hydrothermally-altered rock Alluvial-apron deposits; mostly sand and gravel Volcanic rock of Miocene age. Includes some land- slide deposits _ Contact Sevier fault 0" Spring and identification number in table 11 0 200 400 METERS 0 1000 FEET Basefrom aerial photographs: U.S. Geological Survey, 1953 FIGURE 18.—Surficial geology at Monroe and Red Hill Hot Springs. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH Crater Bench is generally oval in shape, 16 km by 10 km. The high point on the bench is at the top of Fumarole Butte (fig. 19), the neck of the volcanic con- duit, having an altitude of 1,609 m. The butte takes its name from gas vents that were reported to have been active during historic time. The altitude of the sur- rounding area of the bench generally ranges from 1,460 to 1,520 m above sea level; therefore, the upper surface of the lava flow generally ranges from about 50 to 100 In higher than the springs. At the northeast end of the bench, rhyolite, possibly Pliocene in age, has been mapped by Galyardt and Rush (1979). The remainder of the bench is probably either Pleistocene (<2—3 million years old) or late Pliocene in age. The bench is cut by a series of northeast-trending normal faults. Crater Bench lies along the south edge of an east-west line of three calderas (fig. 19) described by Shawe (1972). The last eruption of the calderas is reported to consist of late Tertiary and Quaternary(?) basalt and rhyolite. Silicic-rock ages as young as 3.41-02 million years were reported by Shawe. Although the relation of a controlling fault of Crater Hot Springs to arcuate ring faults associated with magma-chamber roof collapse is unknown, the controlling faults and ring faults are proximal in location and age. As part of the present study, the surface lithology of the Crater Hot Springs area was mapped (fig. 20). The vertical flow of hot water to the land surface is assumed to be through a fault-controlled permeable zone that generally underlies hot springs in the Basin and Range province. However, no fault cutting the spring mound could be located during geologic field mapping. A grav- ity map of the area (Smith, 1974) (fig. 20) shows a high- gravity anomaly beneath the spring area that extends northwestward. The contour pattern is interpreted as possibly being caused by a shallow body of volcanic rock or hot-water deposits of relatively high density. The location of Smith’s gravity anomaly suggests that any controlling fault for Crater Hot Springs may extend beneath Crater Bench and the spring area, as shown in figure 20. Figure 21 is a generalized cross section of Crater Bench, based on Schlumberger resistivity soundings provided by A. A. R. Zohdy (written commun., 1975). The basalt flow is pictured as resting on top of alluvial valley fill; the depth to basement rock is about 1.1 km. An inventory was made of about 40 spring orifices (fig. 22). (See table 17.) The estimated spring flow was 90 US during February 1976. Some additional water seeps to and ponds on the land surface. No accurate measure- ment of this seepage was possible because of lack of observable flow in channels. In the table, this seepage is estimated as half the observed flow or a maximum of 45 US, for a total of about 140 US. This total flow rate H29 probably is slightly too high because some water was "counted” twice at orifice pools R1, R2, and R3 and perhaps elsewhere (fig. 22). Each of these three pools has two orifices —one yielding water, one receiving wa- ter. Four sites of which flows were relatively large and accurate measurements could be made were selected for periodic discharge measurements. Nearly all flow from the springs is included in the periodic measurements (table 13). For example, on February 26, 1976, the flow in the four channels was 87 US, compared to a total flow of 90 US as noted earlier during the same month. Flow was generally declining during this period, but no rela- tion between flow rate and water temperature was ob- served. During the period of measurement, the flow rate averaged 69 US. The largest single flow was from the main-drain orifice on the northeastern part of the mound (fig. 22). Flow was about 80 percent of the total channelized flow. The maximum observed temperature of Crater Hot Springs was 87°C (table 6). According to Mundorff (1970, p. 14), the dominant ions are calcium, sulfate, and chloride. A water sample collected in 1958 had a silica concentration of 28 mg/L and a dissolved-solids concentration of only 1,440 mg/ L. However, these concentrations represent flow from an orifice having a temperature of only 43°C, much less than the maximum. As a result, it is not known whether the sample analysis represents the hottest water. The estimated temperature of the hydrothermal res- ervoir in table 6 is 110°——140°C. The computed depth of water circulation to the hydrothermal reservoir is 1.3- 1.7 km, or only about 200—600 In into the bedrock under- lying the alluvial valley fill. This estimate is based on a regional heat flow of 2 HFU, a mean thermal conductiv- ity of 2.7 x 10‘3cal/cm/s°C, an ambient land-surface temperaure of 14°C, and the estimated reservoir tem- peratures. The hydrothermal system may be recharged by perco- lation of ground water from saturated alluvium or by infiltration of precipitation in the mountains, perhaps the mountains of the calderas to the north. Upflow from the reservoir presumably is along a fault. During the upflow, the thermal water mixes with nonthermal water in the proportion: 50 percent thermal water, 50 percent nonthermal water, as computed by a graphic method of Truesdell and Fournier (1977). The mixed water flows upward to Crater Hot Springs where it flows onto the surface and supports vegetation or evaporates. Additional mixed water seeps into the shallow alluvium in the general spring area and supports phreatophytes (fig. 23); their water consumption is summarized in table 14 using the same general procedure described for Thermo Hot Springs. The total evapotranspiration of H30 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS 113°15' “3.00. TOOELE CO 110° CALDERA 4? I “ 39°45' — ' / :- ' "’0 - ED :9 SPOR a MOUNTAIN 9 DISTRICT 39°30’ 0 10 20 KILOMETERS 0 5 10 MILES FIGURE 19.—Thomas, Keg, and Desert calderas, near Crater Hot Springs. RECONNAISSAN CE OF THE HYDROTHERMAL RESOURCES OF UTAH 112°30' DESERT CALDERA Crater Hot Springs Area of report UTAH ¢°°° 3 Adapted from Shawe, 1972 39°45’ ‘ H31 EXPLANATION Alkali rhyoiite I TERTIARY AND QUATERNARYC’) E — Water-laid tuft — Miocen¢(?), Pliocene, and younger L I _l Rhyolite of Key Mountains Intrusive quartz latite, quartz monzonite, and breed: Oligocenef?) l Quartz monzonite of Desert Mountain / \’\ :L} /_ Rhyolitic welded l TERTIARY sh-flow tutf Quartz-latitic welded ash flow tuff Latitic. andesitic. and basaltic — flows and agglomerates .1 Lower(?) Tertiary Marine sedimentary rocks _ PRECAMBRIAN AND PALEOZOIC Contact mill: £th Dotted where .concealed Low—angle fault Sawteeth on upper plate Ring fracture Dashed where inferred 2300 Aeromagnetic contour Interval 100 gammas. Flown at 9,000 feet barometric elevation; flight line spacing approximately 1 mile O Ore deposit near Thomas Caldera Be, beryllium F, fluorspar U, uranium Mn, manganese —1__. Attitude of flat silica laminae in geodes in 9°30' rhyoiite of the Keg Mountain FIGURE 19.—Continued. H32 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS R 8 W 112°42’30" Density: 2.67 g cm—3 \ 4 \ 3 Of Old Ob \ I 0c 0 \ \\ \ _ z\ ‘ \ ‘ \ Qt 9 ‘ \ ‘ \ 1o ‘“ Of Q at Of @Omc \ Crater Hot prings T 0| Of \ 14 S Ob Q: 16 15 {.9 / 39°39 — ‘“ Qc I‘ s T A61 ”3 Old 0f 0 Ob on \ [6’7 22 21 0| \\ \z \75“? 6% 05 0c Of OI Os 1 W 0 .5 1 KILOMETER Adapted from Smith (1974) I | 1 - . l | 1 Basefrom aerial photographs. 0 V2 1 MILE U.S. Geological Survey,1971 FIGURE 20.—Reconnaissance geologic map of the Crater Hot Springs area and a simple Bouguer gravity map of the Crater Hot Springs area. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH CORRELATION OF MAP UNITS Omc Om - Holocene Os Old Of Oc QUATERNARY OI ~ Pleistocene Ot Ob DESCRIPTION OF MAP UNITS HYDROTHERMALLY—RELATED DEPOSITS ch SPRING-MOUND CREST DEPOSITS—Mostiy red silt and sand deposits rich in manganese oxide. Where deposits are dry, soil is fluffy. Travertine and travertine debris locally present Om SPRING-MOUND DEPOSITS—Mostly light-gray silt and clay, commonly transported by wind and trapped by moist ground. Minor amounts of travertine and travertine debris on east slope of mound which has an average slope of about 20 m/km VALLEY—FLOOR DEPOSITS Os SAND DUNES—Mostly small dunes of windblown, light-yellowish-brown sand partly stabilized by greasewood. Direction of sand movement is northeast DISSECTED DEPOSITS OF LAKE BONNE- VILLE—Light—yellowish-gray silt and clay. Dissected by local runoff and axial drainage southward on the valley floor Ol LAKE BONNEVILLE SEDIMENTS—Undissected, light-yellowish-gray silt and clay underlying a valley floor having an average slope southwestward of less than 1 m/km. Also present on Crater Bench in relatively low areas ALLUVlAL—APRON DEPOSITS Of ALLUVIAL FAN DEPOSlTS——-Mostly basalt boulders and gravel in a matrix of light-yellowish-gray silt and sand below the mouth of small canyons cut into basalt flows of Crater Bench QC COLLUVIUM—Mixture of mostly light-yellowish-gray silt and clay with lesser amounts of slope-wash debris derived from local upslope materials. Underlies alluvial apron of intermediate slope Qt BASALT TALUS—Angular basalt blocks eroded from Crater Bench and underlie irregular, steep slopes on the flank of the Bench CONSOLIDATED ROCKS Ob FUMAROLE BUTTE LAVA FLOWS—Black vesicular to massive basalt forming Crater Bench. Mostly large, angular bloch. Bench extends 30 m to 60 m higher than adjacent valley floor Old Contact —-I-—— Fault, bar and ball on downthrown side — - — Lineament of unknown origin. May be fault or fracture 0w Hot spring — — 76'2— Gravity contours in milligals FIGURE 20.—Continued. H33 mixed water is estimated to be 5 X 106m3/yr; this sum is equal to a continuous flow of 160 US. Additional water of unknown volume flows laterally from the area to be discharged elsewhere. On the basis of the evapotranspi- ration estimate above, the estimate that 50 percent of the mixed water is thermal water, a reservoir tempera- ture of 140°C and an ambient land-surface temperature of 14°C, the convective heat discharge is 36 x 1013 cal/ yr. Heat-flow estimates made for four water wells near Crater Hot Springs are shown on figure 24. The highest heat flow was 2.8 HFU, only slightly higher than the regional heat flow. NAVAJO LAKE KGRA Navajo Lake KGRA is on the Markagunt Plateau, about 30 km southeast of Cedar City in southwestern Utah (fig. 1). The Markagunt Plateau, part of the Col- orado Plateaus, is a horst bounded by two major fault systems—the Hurricane fault zone on the northwest and the Sevier fault on the southeast. The width of the plateau near the KGRA is about 60 km. The Navajo Lake KGRA has an area of only 10 km2 (table 2). It was designated solely because of overlapping lease applica- tion on Federal land, as is required by law. Most of the rocks exposed at the surface are Tertiary limestones (Wasatch Formation) and volcanic rocks in- cluding highly permeable Quaternary basaltic flows. Beneath these units is a thick sequence of mostly Mesozoic and older sedimentary rock, of which sandstone is the most common (Hintze, 1963). Silicic lavas were erupted, but they are generally older than the basaltic lavas. Smith and Shaw (1975, p. 82) list the largely unvegetated Markagunt basalt field as less than 10,000 years old. No hot springs have been found in the Navajo Lake area. A well drilled at (C-36-7)33, about 10 km north- east of the KGRA had a reported temperature of drilling mud returning in the bore to land surface of 43°C with a drilling depth of 1,862 m. This mud temperature could be produced with a heat flow of less than 2 HFU. As a result, the only possible indicator of geothermal poten- tial is the very young basaltic flows. MEADOW AND HAT'I‘ON HOT SPRINGS Meadow and Hatton Hot Springs are 18 km southwest of Fillmore (fig. 1) and a few kilometers north of the Cove Fort-Sulphurdale KGRA (fig. 1) in southwestern Utah. The springs are at (C-22-6)27ddS and (C-22— 6)35ddS, respectively, on a low alluvial spring mount in Pavant Valley, a few kilometers west of the Pavant H34 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Location of section shown in figure 24 0 5 10 KILOMETERS | l | Adapted from A, A. R, Zohdy l l (USGS. Denver, wntten 0 5 MILES communication, 1975) Crater Bench West East A I 1600 " ~1609 rn (Fumarole Butte) r 1600 EXPLANATION Crater Hot 1 Basalt, averages. /Spn‘ngs about 90 m thick 1400 __ / _ 1400 2 lnterbedded basalt (‘1 and alluvium or (// weathered basalt, 7‘ ‘ averages about 130 m thick 1200 — ~—- 1200 9 3 Alluvium, about 3 1,000 m thick Inferred fault 1000 1000 Altitude scale. in meters DATUM IS SEA LEVEL FIGURE 21.——Generalized cross section of Crater Bench. Range. Underlying the alluvium of the spring area, volcanic rocks may be as shallow as 30 m from the land surface. The Pavant Range has Paleozoic and Mesozoic quartzite and other sedimentary rocks thrust westward over Navajo Sandstone of Triassic (?) and Jurassic age (Hintze, 1963). North, west, and south of the springs, upper Tertiary and Quaternary basalt and basaltic andesite flows form low hills. A basaltic flow about 13 kilometers north of Hatton Hot Spring has been dated as less than 10,000 years old by Smith and Shaw (1975, p. 82, Ice Springs field). Lithologic units are shown in figure 25. The most prominent geologic and topographic feature in the area is the spring-deposited travertine ridge at Hatton Hot Spring. The ridge is about 2 km long and rises perhaps 20 m above the general land surface. The shape and orientation of the ridge suggests fault control for the deposit-producing springs. Extending outward from the ridge is the very low alluvial spring mound (fig. 25) that is marked by many north-trending lineaments that may be faults. Travertine deposits at Meadow Hot Spring are small in volume, encircling the spring pool at the gen- eral land surface. TABLE 13.—Measured flow and temperature of Crater Hot Springs [Figure 22 shows locations of Spring E2, Spring R2, and the main—drain orifice. The Southwest ditch was measured south of Springs Dl—DIO (table 1 7) and about 3 m northeast of concrete pools] Spring E2 Spring R2 Main—drain orifice Southwest ditch Date Flow Temperature Flow Temperature Flow Temperature Flow Temperature (L/s) (°C) (L/s) (°C) (Us) (“0 (Us) (“C) 2—26—76 ____________ 4.8 65 3.1 70 72.5 52 7.4 __ 4— 8—76 ____________ 3.7 __ 4.5 __ 73.6 __ 8.4 __ 4.5 66 .3 74 52.1 59 6.5 74 3.6 66 .6 76 52.9 60 6.4 78 __ .0 __ 60.8 __ 6.5 __ 66 .0 __ 52.3 61 5.9 70 65 1.1 69.5 51.8 65 6.0 70 64.5 1.4 69 51.5 55 6.0 68 68 3.9 67 50.5 55 6.0 68 64 1.9 70 46.2 54 5.2 72 65 1.4 69.5 58.1 54 6.0 70 RECONNAISSAN CE OF THE HYDROTHERMAL RESOURCES OF UTAH 112°44’ ' EXPLANATION .NI Spring and number listed in table —» Ditch and direction of flow a Large main pool 39°37! Jaifl Section comer in T 14 S, 1514 R 8 W, Salt Lake Base and Meridian R8W DI—10 Spring— 0 mound 10 11 00% crest 15 14 ¢ 500 METERS _l_’__L__+__1_l 0 500 1000 FEET FIGURE 22.—Spring orifices and pools on the mound of Crater Hot Springs. The observed discharge from both Meadow and Hat- ton Hot Springs was less than 1 Us in the summer of 1976. Mundorff (1970, p. 40) reports a flow of about 4 Us from Meadow Hot Spring and an absence of flow at Hatton Hot Spring during several recent years. The temperatures at these two springs, and those at other nearby sites, are shown in figure 26. The highest tem- perature, 67°C, was measured in a 27-m well a few tens of meters north of Hatton Hot Spring. The 67°C temper- ature was measured from a depth of 5 m to total depth. H35 Hatton Hot Spring had a temperature of 36°C at that time; Meadow Hot Spring pool was 30°C. Agricultural wells, 3 km and more to the east of the spring mound, not shown on figure 26, generally have discharge tem- peratures of 13°C. Mundorff (1970, p. 16) has published several chemical analyses for both Meadow and Hatton Hot Springs. He reports water temperatures for Meadow Hot Spring ranging from 29°C to 41°C. A sample collected in 1967 had a silica concentration of 47 mg/L and dissolved solids of 4,900 mg/ L. The principal ions were sodium and chloride. The Hatton Hot Spring sample was collected at a temperature of 38°C, had a silica concentration of 44 mg/L, had dissolved solids of 4,670 mg/L, and had ion concentrations very similar to those of Meadow Hot Spring. The temperature of the hydrothermal reservoir, es- timated with geothermometers, is possibly in the range of 70°C to 120°C (table 6). The maximum depth of circu- lation required to produce the estimated reservoir tem- peratures with normal regional heat flow is about 2—3 km; this estimate is based on a mean thermal conductiv- ity of the underlying rock and thin alluvium of 5 X 10‘3 cal/cm/s°C and an ambient land-surface temperature of 12°C. The calculations assume the absence of a shallow magma heat source. Recharge to the hydrothermal system probably re- sults from percolation downward from saturated a1- luvium of the area or from infiltration of precipitation in the Pavant Range. During upward flow, the thermal water may be mixing with nonthermal water in the proportion of about 40 percent thermal water and 60 percent nonthermal water, as estimated by the graphic method of Truesdale and Fournier (1977). The mixed water is discharged by the springs, by evapotranspira- tion of pheatophytes on the spring mound, and by sub- surface flow from the mound area principally to the north and west. The observed spring flow in 1976 was small, but the evapotranspiration was large. Much of TABLE 14.-——Evapotranspiration of ground water from Crater Hot Springs hydrothermal system [1975 conditions. The combined and nonhydrothermal system discharge rates are based on research by Lee (1912), White (1932), Young and Blaney (1942), Houston (1950), Robinson (1965), and Hart and Price (1972) in other areas] Average annual evapotranspiration rates (approximate) E32233? Depth Combined Nonhydrothermal Net annual net to water discharge system hydrothermal discharge Phreatophyte area table rate discharge rate discharge Area (rounded, (See fig. 23) (m) (m) (m) (m) (X 103m”) x 103m“) Mostly greasewood ______________ 3— 15 0.07 0.06 0.01 2,200 20 Mostly bare soil, saltgrass, and picklewood ________________ 0—3 .3 .2 .1 4,000 400 Mostly bare soil of spring mound __________________ 0.4— 1.5 .1 .06 .04 200 10 Meadow ________________________ .5—2 .4 .06 .34 470 160 Wet meadow ____________________ < 1 .8 .06 .74 890 660 Mostly tules and very wet meadow ______________ <.5 1.2 .1 1.1 3,400 3,700 Total (rounded) ________________ _ - _ _ _ _ _ _ _ - _ - 11,000 5,000 (160 US) H36 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS R 8 W 112°43' a o 1‘ \ \ (J . ‘\ / \\// \\ \l § // ///"I/// CRATER 39°36’ — Base from aerial photographs: 0 .5 ‘l KILOMEI’ER US Geological IL I I L I Survey, 1971 0 1/z 1 MILE FIGURE 23.—Phreatophyte distribution in the Crater Hot Springs area. RECONNAISSAN CE OF THE HYDROTHERMAL RESOURCES OF UTAH DESCRIPTION OF MAP UNITS Mostly bare soil of lowlands. Minor amounts of saltgrass and pickleweed locally. Discharging water probably not from Crater Hot Springs Mostly greasewood, locally some rabbitbnish, saltbush, pickleweed, and big sage Mostly bare soil, saltgrass, and pickleweed on lowlands. Soils are commonly moist and soft in the winter and _ spring and firm during the summer and fall Mostly bare soil of spring mound; saltgrass and rabbitbrush grow locally Meadow, saltgrass, and pickleweed on slopes of spring mound and areas north of the mound. Some cottonwood and willow of north edge of spring mound. Soils firm most of the time ~ ’ (I \‘_ Wet meadow with pickleweed and saltgrass. Soils are wet ‘ and soft during winter and spring and dryer during summer and fall _ \\ \\ s Mostly tufts and very wet meadow on slopes of spring mound and nearby lowlands; locally some willow growth. Soils are usually soft at all times. Some lowlands may be covered with standing water during most winters and springs Crater Bench; no phreatophytes or evapotranspiration of ground water 0* Hot spring FIGURE 23.——Continued. R9W 112°45’H8W R7W I T l I T 13 S _ + + —l Crater T Fumarole - /HotSpnngsA' Bune l4 s . 2.8 l— + _ T 2 8 EXPLANATION o<1 15 0 Well. Number is estimated 5 heat flow in p cal cm —2 sec—1 39 30 ‘0 5 10 KlLOMETERS ‘ 0 5 MILES Sugarville D l l I l FIGURE 24.——Heat flow in the Crater Hot Springs area. the mound area commonly is saturated to land surface during the winter and spring; at other times, the water table is at a depth commonly no greater than 2—3 m. The spring mound has an area of approximately 32 km”. The net evapotranspiration rate over the mound is esti- mated to average on the order of 0.3 m; therefore, the estimated evapotranspiration of mixed water is about 10 x 106m3/yr. The rate of underflow to the west and north is unknown, but is estimated to be much less than H37 the rate of evapotranspiration. Only 40 percent of the total discharge of mixed water is convective flow from the hydrothermal reservoir or at least 4 X 106m3/ yr. The heat convectively discharged by this flow of water is at least 43 X 1013 cal/yr; this discharge figure is based on the cooling of the convective flow of thermal water from a possible reservoir temperature of 120°C to an ambient land-surface temperature of 12°C. Figure 27 shows a melting pattern of freshly fallen snow near Hatton Hot Spring. According to White (1969), very high he at flow ratesare required to produce such melting. Subsurface temperatures as high as 70°C can be expected at depths of as little as 10 m under these snowmelt areas. VICINITY OF SALT LAKE CITY Geothermal resources in the Salt Lake City area probably have no potential for electric power genera- tion, but they are discussed here because of their poten- tial value for uses, such as space heating, associated with urban development. Two areas of hydrothermal potential have been mapped by Marine and Price (1964). Their map has been modified as a result of ad- ditional data that became available after its compila- tion in 1959. The modifications involved the enlarge- ment of the hydrothermal areas, as shown on figure 28. The map is based on discharge temperatures of wells generally drilled to depths of 200 m or less. The northern and larger of two areas extends gen- erally westward from the faults along the Wasatch Range front at least to the Great Salt Lake. The western and northern boundaries of the warm-water body are generally unknown; however, the map shows approxi- mate southern and eastern limits. The warmest water in the northern area issues from Becks (B—1—1)14dch and Wasatch Hot Springs (B—1—1)25dbS, 56°C and 42°C, respectively. Both issue from Paleozoic limestone at the Warm Springs fault. The distribution of water temperatures suggests that most of the warm water originates from faults in the eastern part of the warm- water area, then migrates westward in the alluvium. The hydrothermal area at the south end of Jordan Valley has similar temperatures. Crystal Hot Springs (C—4— 1)1 1 and 12bS (fig. 28) has a reported temperature of 58°C. It, like Becks and Wasatch Hot Springs, prob- ably flows from a permeable fault zone. The heat in the water is probably the result of deep circulation. The warm-water areas (fig. 28) may be enlarged to a greater extent as more data become available. The southern area might be extended farther northward into the area northwest of Sandy, and the northern area may be enlarged westward and possibly northward. H38 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS 112°30' H 6 W H 5 W 9 I 1 E \10 I 11 12 Tabernacle /' \l ‘\ '1 Crater . ' 1 / (/‘1‘\ / ! ,’ White Mountain Q'd \ I 38°52' T228 T238 Crater 9 / 1o 11 T2” 12 Base from aerial photographs: US. Geological Survey, 1953 0 1 KILOMETER 0 1 MILE FIGURE 25.—-Reconnaissance geologic map of the Meadow and Hatton Hot Springs area. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH CORREIATION 0F MAP UNITS O sd Old El- Holocene 0 sh OI Omc1 Omcz O QUATERNARY mx Pleistocene 0 mt Ob szf TERTIARY Ob szf M DESCRIPTION OF MAP UNITS HYDROTHERMALLY RELATED DEPOSITS SPRING-MOUND TRAVERTINE—Highly porous to dense and banded calcium carbonate deposited during the cooling and evaporation of water discharged from the Hatton Hot Spring system. Commonly light tan, yellow or greeth gray. The mound of its crest extends 10 m above the adjacent valley floor SPRING-MOUND TALUS—Mostly angular fragments of travert'ine forming the flank of the spring mound, resulting from the disintegration of th, described above. Mostly sand and gravel size; some larger blocks SPRING-MOUND COLLUVIUM—Tan silty fine- to medium-grained sand. Dominantly travertine fragments carried by infrequent runoff from the mound. Minor salt crust present in some damp low-altitude areas. Surface generally is firm. Material saturated at shallow depths with ground water, having temperatures as high as 67°C SPRING-MOUND COLLUVIUM—Mostly tan to brown sandy silt and silty sand derived characteristically by disintegration and erosion of spring mound travertine. Areas to the north and west have ground-water saturation generally towithin 1 m of the land surface resulting in a surface salt crust or grassy wet areas. Locally, some areas have dry, firm surface VALLEY—FLOOR DEPOSITS LAKE BONNEVILLE SEDIMENTS—Mostly medi- um-brown sandy silt; relatively undissected. Land surface, usually near horizontal and flat, usually damp and hard; salt crust and marsh areas common near springs and seeps. Depth to ground-water saturation generally less than 1 m DISSECTED LAKE BONNEVILLE SEDI- MENTS—(See above description of Q1.) Dissected by infrequent runoff. Locally includes blow sand, sand dunes, and playa silt and clay. Depth to ground-water saturation variable and dependent on topographic position SAND BAR—Tan,medium- to coarse-sand and fine gravel. Mostly quartz, travertine, and volcanic-rock fragments. Bars extend about 1 m above the general land surface. Formed in Lake Bonneville at an altitude of about 1455 m above mean sea level SAND DUNES—Mostly fine- to coarse-grained, silty quartz sand transported mostly by southwesterly winds. Includes some dissected Lake Bonneville sediments (Qld) CONSOLIDATED ROCKS QUATERNARY BASALT—Dark gray, commonly vesicular LATE TERTIARY BASALT AND BASALTIC ANDESITE Contact Lineament on aerial photographs; may be a fault Ridge line on spring mound Spring or seep FIGURE 25.—Continued. H39 Discovery of warm water at shallow depths elsewhere in the valley is unlikely. Becks and Wasatch Hot Springs yield sodium chloride-type waters with fairly high concentrations of dissolved solids, 13,000—14,000 mg/L and 6,000-13,000 mg/L, respectively. Crystal Hot Springs yields water with lower dissolved-solids concentrations, in the range of 1,300—1,7 00 mg/L. Water flowing from faults to the alluvium near the springs probably has ion concen- trations similar to the springs. As this water mixes with nonthermal water in the alluvium, the mix will have chemical characteristics intermediate between water types. In the northern area, the water in the alluvium will, therefore, have dissolved solids that are highly concentrated but less concentrated than the spring flow. Water from Becks, Wasatch, and Crystal Hot Springs or from wells nearby may be useful for space heating, but development of wells in the vicinity of the springs may stop the spring flow. GREAT SALT LAKE DESERT Thick beds of high-porosity clay underlie the Great Salt Lake Desert of northwestern Utah (fig. 1). Such beds have an insulating quality, impeding the conduc- tive flow of heat to land surface. As a result, geothermal gradients must be high to discharge the heat flowing upward in the earth’s crust. For example, if a regional heat flow of 2 HFU and a thermal conductivity of porous clay of 2 X 10‘3 cal/cm/s°C are assumed to be reasonable values, the computation of the geothermal gradient would be: I: HFU x 102 K where I is the geothermal gradient in °C/km, HFU is heat-flow units in ucaUcmz/s, and K is thermal conductivity. The calculation becomes = 2><102 I 2 = 100°C/km. If surface ambient temperatures on the desert are about 10°C, subsurface temperature at a depth of 1 km would be 110°C if the described bed of clay were also of that minimum thickness. The implication is that areas of thick clay accumulation, like the Great Salt Lake Desert, may have low-temperature geothermal poten- tial for space heating without a near-surface source of heat or a permeable zone in which to circulate upward- flowing hot water from great depths. The temperature- H40 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS gradient data on the Bonneville Salt Flats in the west- ern part of the Great Salt Lake Desert, of Turk (1973) and Whelan and Petersen (1974), support a conclusion that a geothermal potential may exist in the area. OTHER AREAS The few geothermal areas described in this report were selected because they appear to have the highest development potential. However, other areas in west- ern Utah may have important geothermal potential. Additional sources of geothermal data on these areas are the thermal-spring report by Mundorfl' (1970), computer-stored temperature data of the Water Re- sources Division of the US. Geological Survey, and other data in the files of the US. Geological Survey. (See table 18.) Table 18 lists temperatures and heat-flow data for nearly 100 wells and springs. It is not designed to be a comprehensive listing of such data for Utah, but rather a limited listing for widely spaced data points selected to provide information on wells and springs having mostly above-ambient temperatures. SUMMARY AND CONCLUSIONS 1. Several publications have summarized data on hot and warm springs in Utah (Stearns, Stearns, and War- ing, 1937; Waring, 1965; and Mundorff, 1970). As a result, no attempt was made to include in this report comprehensive tables of descriptions and data for ther- 112030' RBW RSW 38°55’ — Heat-flow estimates based on subsurface temperature meas- urements EXPLANATION // Area of probable high // conductive heat flow 03 Well. Number is heat flow / in 14 cal cm—2 sec—1 i T: 13° \ 0. Spring and temperature \ in °C 0t Springs A, T=30° | I @3 ; 03:13" White . ,_\\ MountaIn// \ / v.T= 14° \ \ BT=15.5° \ \ 5 T=22° 0 1 2 KILOMETERS T F—L____'__J 22 0 iMlLE o3 o3.7 T=13°‘~ 111011051 Hot Springs / —‘ FIGURE 26,—Estimated heat flow and water temperatures in the Meadow and Hatton Hot Springs area. RECONNAISSANCE OF THE HYDROTHERMAL RESOURCES OF UTAH TO MEADOW R 6 W 111°3n; 38°51’ T22 S T238 EXPLANATION m Travert ine E Alluvium Area of rapid snowmelt — Contact W Hatton Hot Springs l Base from aerial photographs: ’ US, Geological Survey, 1976 T0 HA ”ON 0 500 METERS f_L[_l__lL__L'_l 0 500 TWO 1500 FEET FIGURE 27 .—Areas of rapid snowmelt near Hatton Hot Springs, March 1976. mal springs. However, data for selected hydrothermal systems are presented and summarized in table 15. 2. Geologic factors are more favorable for geothermal resources in the Basin and Range province than on the Colorado Plateaus or in the Middle Rocky Mountains. The structure of the Basin and Range province is a result of sizable east-west crustal extension and crustal thinning during the last 17 million years. Igneous rocks H41 less than 6 million years old crop out at many places, mostly in the southwestern part of Utah. Basalts as young as 10,000 years have been mapped. 3. The principal source of geothermal fluids is water stored in the hydrothermal reservoir and water enter- ing the system as recharge from precipitation. Because of the semiarid climate of most of the area, most geo- thermal development for generation of electricity will remove fluids from storage at a higher rate than natural replenishment. 4. Conductive heat flows to the land surface at‘a generally high rate in the Basin and Range province of Utah are probably due to crustal thinning and possibly due to intrusion of young magmas into the earth’s crust at shallow depths. The rates of conductive heat flow are commonly in the range of 1.5—2.5 HFU and probably average about 2.0 HFU. This is about normal for the Basin and Range province and considerably higher than the average of about 1.6 HFU for the entire earth. 5. High-temperature convection systems may be lo- cated by searching for high-silica volcanic and intrusive rocks that are of Quaternary age. Mapping of faults, hydrothermally altered rock, and thermal-spring de- posits, along with the drilling of temperature-gradient holes, would be desirable components of an exploratory program. 6. Drilling for hot, high-silica, buried bodies of rock would best be pursued in the areas of recent volcanic activity. , 7. Some geothermal systems may be related to cal- deras because of their potential for eruption of large volumes of silicic rock. An example is the potential relation of Crater Hot Springs and Crater Bench to Thomas, Keg, and Desert calderas. 8. The southwestern part of Utah probably has the most promising geothermal potential, judged on the basis of spring temperatures, silica concentrations, and deposits such as siliceous sinter and sulfur. TABLE 15.—Summary of data for selected hydrothermal systems in Utah Estimated water discharge from the hydrothermal reservoir Estimated . Temperatures . (X 10"“WVT) ifiifiiiiiii (°C> 5:32:35 Eiif§fiitfi饧§i Minimum dischar e Spring water in reservoir below Hydrothermal Evapo- Spring Ground-water total convective ( X 10' Estimated flow mixed land surface system transpiration flow outflow flow cal/yr) reservoir maximum water (km) Roosevelt ________________ Minor Minor Principal Unknown Unknown 260— 290 85 - _ _ _ 6— 7 Cove Fort- Sulphurdale ________________ do-___ None __-_do--__ ____do____ ___-do--__ 200: ____ ____ 5 Theme __________________ 0.6 0.02 Probalfily 0.6 15 140— 200 82.5 40 3—4 sma New Castle ______________ Minor None 0.4 .4 7 140— 170 ‘ 107.8 60 3—4 Monroe-Red Hill __________ .02 .2 Unknown .2 3 100— 160 75 50 2—4 Joseph __________________ Minor .02 -___do__-- .02 .3 100—170 65 35 2—4 Crater __________________ 2.5 (2) ____d0____ 2.5 36 110—140 87 50 1.3—1.7 Meadow-Hatton __________ 4 Minor ___-do____ 4 43 70-120 167 40 _ _ _ _ 1Measured in well. aSupports phreatophytes and evaporates. Included in the evapotranspiration estimate. H42 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS 1 4;“); Ed“ *5 + + + + + + /\\¢\\§l4=//:= §j§§:4\\;= + + + + + + 3:4:\\3°o:=:’/::¢=:Z’:=, + + + + + + + “V” itiilrfgiuf237bf/I’. + + + + + + + 40°50' — v3: ‘3‘4’=§f~‘m=’=f,’_=*;‘ 4 [WV lip/=3 I¢;\\= , 9,4559% 11: «“5 + + + + + + + +._ wvvv «u‘lr sifiilyfxlfiéf ‘J”“‘\ + + + + + + + + WWWi’WS’” ‘ 4" ‘J fi’éi“l\i= /’ \\ ‘ + + + + + 1:33:33: "’ “3’37”? -:;=”c: ‘57,, § 5 550 Becks Hot Springs ’ /Vvvvvvvvv 'L’;¢y,£z¢,/;:4’:=fi‘¢ _ ¢¢II¢§ + + + + ,\ $333333: 10’ 4“”\\\‘,“‘1::’=”IvZZZZZIZZSV ”j‘ + + + + + + + + g viliiilill’Y/Zv ”IQ/13”,,“*J‘ZSVZZZZZZXZZSZZX 5 + + + + + .+ + + q, [/vvv:::::::::::::: :174/5 "It. ,,¢ .. :vv Vvvvv := 4 Wasatch Hot Spnngs + + (évvvv vvvvvvvvvvvvvvv 1:;:;1*‘3031 : 23‘ + + + + + + ,zvv Wiiiiiii‘v’iélii‘v’tixu 4* = 1,7,: 4: W + + + + + low vvvvvvvvvvvvvvvvvzvvvv ” ‘7, 4’5 J. a 35 + + + “' v3 V3133:12:33:333333v333339\,‘l.:I"; WV 12 o + + + /Jv WWWVWWWWWWWWWW 4“ ., WW g vvv va vv v vvvvvvvvvvvvvvvvvvvvvv 4‘ I § ‘ vvvv + + + 0 é:VV5353:3liéitééiiétiixétéiiiiii‘y{In VCCZCV + + + /vvv vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv 4" vvvvv + + vv VVVVvVVVVVVVVVVVVVVVVVVVVVVVVVVVV n \ VVVVVVV V “VWV"WWWW“‘V’ZZXVZ‘V’X‘V’ZL’Q’WZZZZV w X‘Jl’itiiv + + + de/31 :1::zs:x:::::zzvvvvzvvvvvvvvxl’A/vvvv *“' vvvvvvv SALT LAKE CITY 4» 8 VV V VV VV VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV ‘ g a VVVVVVV vvvvvvvvvvvv vvv vvvvvvvv ‘14 vvvvvvv @ 25.5 + + Z + + + - “W szzzzzx- ‘vtzzzzzzzv + g + + + + + + D Mag“ szxzzzm Wm + + + + + + + + wv fivw + + + + + + + + ‘v’ 27 . 3:: + + + + + + + + WV + + + + + + + 010 + + + + + + + + M + + + + + + + + I: may + + + + + + + + + ° 13 ° 12 + + + + + + + + + + + + + + + + + g + + + + + + + + ; + + + + + + + + +I~ + + + + + + + + + + + g + + + + + + + + + + o + + + + + 5 + + + +§+ + + + + '- + + + + + + + + + + < + + + + + + + + + + 3 '+ + + + + + + + + F“ +3 + + + + :1: + + + + + 508° + + + E + + + + + + + + + + + 5 + + + + + + + + c + + + + + + + + +‘ + + 0+ + + + + + + + , + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + o 12 Riverton D + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + 40°30' - + + + + + + + + + + 4'- + + + + + + + + + + + + + + + + + + + + + - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + L + + + + + + + + 0 1 2 3 4 5 10 KILOMETERS + + + + PI II II I II I I I TRAVERSE + + LE“ 0 1 2 3 4 5 MILES + + + + \IN— + + + + uTA“ J + + + + Modified from Marine and Price (1964) FIGURE 28.—Areas of warm ground water in the Jordan Valley. RECONNAISSAN CE OF THE HYDRO’I‘HERMAL RESOURCES OF UTAH 9. Deep exploratory drilling near Roosevelt Hot Springs has demonstrated that this KGRA has high potential for electric power generation. Reservoir tem- peratures are at least 260°C (table 15), and well testing demonstrates high reservoir permeability. The meat source may be related to Pleistocene rhyolites as young as 490,000 years. 10. The Cove Fort-Sulphurdale area may have reser- voir temperatures as high as 200°C. No thermal water is known to discharge at land surface in the area, but sulfur deposits, altered ground, and gaseous emissions indicate past hydrothermal activity. Quaternary basalt flows are abundant in the area. The area extending northward 60 km to Neels, including Roosevelt Hot Springs and the Cove Fort-Sulphurdale KGRA’s, prob- ably has the best potential for geothermal development in Utah. 11. Thermo Hot Springs discharge from a hydro- thermal system having an estimated reservoir temper- ature between 140°C and 200°C. Estimated hot-water circulation through the hydrothermal reservoir is at a rate of 18 US. 12. The Newcastle area has many thermal water wells but no thermal springs. The estimated reservoir temperature for the hydrothermal system is between 140°C and 170°C. An irrigation well has pumped boiling water at a rate of 108 US. Thermal water is discharged from a range-front fault from which it flows northward into an alluvial aquifer. The thermal water discharges its heat mostly by conduction to the land surface. 13. The Monroe-Joseph KGRA contains two hydro- thermal systems, one at Joseph Hot Springs and the other at the Monroe-Red Hill Hot Springs complex. Res- ervoir temperatures appear to be at least 100°C but may be as high as 160°C—170°C. 14. Crater, Meadow, and Hatton Hot Springs (table 15) which discharge from hydrothermal reservoirs hav- ing estimated temperatures less than 150°C, could be EXPLANATION Alluvium + + Consolidated rocks vvvvvv Areas where ground-water temperature is commonly between 16°C and 21°C ‘ = =\\3 ,9 Area where ground-water temperature is commonly above , - 21°C Contact dashed where location is indefinite Well; number is temperature of water in °C FIGURE 28.~Continued. H43 considered for space and process heating. Despite their probable low reservoir temperatures, these prospects have the largest estimates of heat and water discharge of those systems listed in table 15. 15. Areas that may have hydrothermal potential but are inadequately defined are all in the Escalante Desert northwest of Beryl and Lund and east of Table Butte. I REFERENCES CITED Batty, J. C., Grenney, W. J., Kaliser, Bruce, Pate, A. J., and Riley, J. P., 1975, Geothermal energy and water resources in Utah, in Impacts of energy development on Utah water resources: Proce- dures Third Annual Conference Utah Section, American Water Resources Association, p. 223—241. Brown, F. H., 1977, Attempt at paleomagnetic dating of opal, Roosevelt Hot Springs KGRA: University of Utah, Department of Geology and Geophysics, Technical Report 77-1, 13 p. Crosby, G. W., 1973, Regional structure in southwestern Utah, in Geology of the Milford area 1973: Utah Geological Association Publication 3, p. 27-32. Denton, E. H., 1976, Helium sniffer field test: Newcastle, Utah, 10—26 March 1976: U.S. Geological Survey Open-File Report 76—42 1, 4 p. Erickson, M. P., 1973, Volcanic rocks of the Milford area, Beaver County, Utah, in Geology of the Milford area 1973: Utah Geologi- cal Association Publication 3, p. 13—21. Fenneman, N. M., 1931, Physiography of western United States: New York, McGraw-Hill Book Co., 534 p. Fournier, R. 0., White, D. E., and Truesdell, A. H., 1974, Geochemical indicators of subsurface temperature—Part 1, Basic assump- tions: U.S. Geological Survey Journal of Research, v. 2, no. 3, p. 259—262. Fournier, R. 0., and Rowe, J. J ., 1966, Estimation of underground temperatures from silica content of water from hot springs and wet steam wells: American Journal of Science, v. 264, no. 11, p. 685—697. Fournier, R. 0., and Truesdell, A. H., 1973, An empirical Na-K—Ca geothermometer for natural waters: Geochimica et Cos- mochimica Acta, v. 37, p. 1255—1275. Galyardt, G. L., and Rush, F. E., 1979, Geology of the Crater Hot Springs KGRA and vicinity, Juab and Millard Counties, Utah: U.S. Geological Survey Open-File Report 79-1158, 1 sheet, scale 1:24,000. Gilbert, G. K, 1890, Lake Bonneville: U.S. Geological Survey Mono- graph 1, p. 332—335. Godwin, L. H., Haigker, L. B., Rioux, R. L., White, D. E., Muffler, L. J. P., and Wayland, R. G., 1971, Classification of public lands valuable for geothermal steam and associated geothermal re- sources: U.S. Geological Survey Circular 647, 18 p. Harr, R. D., and Price, K. R., 1972, Evapotranspiration from a greasewood-cheatgrass community: Water Resources Research, v. 8, no. 5, p. 1199—1203. Heylmun, E. B., 1966, Geothermal power potential in Utah: Utah Geological and Mineralogical Survey, Special Studies 14, 28 p. Hintze, L. F., 1963, Geologic map of southwestern Utah: Provo, Utah, Brigham Young University map, scale 1:250,000. Houston, C. E., 1950, Consumptive use of irrigation water by crops in Nevada: Nevada University Bulletin 185, 27 p. Lachenbruch, A. H., and Sass, J. H., 1977, Heat flow in the United States and the thermal regime of the crust, in Heacock, J. G., ed., H44 The earth’s crust: Geophysical Monograph Series, v. 20, Ameri- can Geophysical Union, Washington, D.C., p. 626—675. Lee, C. H., 1912, An intensive study of the water resources of a part of Owens Valley, California: U.S. Geological Survey Water-Supply Paper 294, 135 p. Lee, W. T., 1908, Water resources of Beaver Valley, Utah: U.S. Geological Survey Water-Supply Paper 217, 57 p. Liese, H. C., 1957, Geology of the northern Mineral Range, Millard and Beaver Counties, Utah: Salt Lake City, Utah, University of Utah, unpublished MS. Thesis, 88 p. Lovering, T. S., and Goode, H. D., 1963, Measuring geothermal gra— dients in drill holes less than 60 feet deep, East Tintic District, Utah: U.S. Geological Survey Bulletin 1172, 48 p. Marine, I. W., and Price, D., 1964, Geology and ground-water re- sources of the Jordan Valley, Utah: Utah Geological and Mineralogical Survey Water-Resources Bulletin 7, 68 p. Milligan, J. H., Marselli, R. E., and Bagley, J. M., 1966, Mineralized springs in Utah and their effect on manageable water supplies: Utah State University, Utah Water Research Laboratory Report WG23-6, 50 p. Mower, R. W., and Cordova, R. M., 1974, Water resources of the Mil- ford area, Utah, with emphasis on ground water: Utah Depart- ment of Natural Resources Technical Publication 43, 106 p. Mundorff, J. D., 1970, Major thermal springs of Utah: Utah Geologi- cal and Mineralogical Survey Water-Resources Bulletin 13, 60 p. Olmsted, F. H., Glancy, P. A., Hamill, J. R., Rush, F. E., and Van Denburgh, A. S., 1975, Preliminary hydrogeologic appraisal of selected hydrothermal systems in northern and central Nevada: U.S. Geological Survey Open-File Report 75-56, 267 p. Parry, W. T., Berson, N. L.,and Miller, C. D., 1976, Geochemistry and hydrothermal alteration at selected Utah hot springs: Utah Uni- versity Department of Geology and Geophysics Final Report, v. 3, 131 p. Petersen, C. A., 1975, Geology of the Roosevelt Hot Springs area, Beaver County, Utah: Utah Geological and Mineralogical Sur- vey, Utah Geology, v. 2, no. 2, p. 109—116. Renner, J. L., White, D. E., and Williams, D. L., 1975, Hydrothermal convection systems, in Assessment of geothermal resources of the United States—1975: U.S. Geological Survey Circular 726, p. 5—83. Robinson, T. W., 1965, Water use studies utilizing evapotranspiration tanks, in Water resources of the Humboldt River near Win- nemucca, Nevada: U.S. Geological Survey Water—Supply Paper 1795, p. 83—104. Rowley, P. D., 1978, Geologic map of the Thermo 15-minute quad- rangle, Beaver and Iron Counties, Utah: U.S. Geological Survey Geologic Quadrangle Map GQ-1493, 1 sheet. Rowley, P. D., Anderson, J. J., and Williams, P. G., 1975, A summary of Tertiary volcanic stratigraphy of the southwestern high plateaus and adjacent Great Basin, Utah: U.S. Geological Survey Bulletin 1405-B, 20 p. Rush, F. E., 1977, Subsurface-temperature data for some wells in western Utah: U.S. Geological Survey Open—File Report 77-132, 36 p. Sass, J. H., Lachenbruch, A. H., Munroe, R. J., Greene, G. W., and Moses, T. H., Jr., 1971, Heat-flow in the western United States: Journal of Geophysical Research, v. 76, no. 26, p. 6376—6413. 1976, A new heat-flow contour map of conterminous United States: U.S. Geological Survey Open-File Report 76-7 56, 24 p. Sass, J. H., and Munroe, R. J ., 1974, Basic heat-flow data from the United States: U.S. Geological Survey Open-File Report 74-9, 426 p. Schmoker, J. W., 1972, Analysis of gravity and aeromagnetic data, GEOHYDROLOGY 0F GEO’I‘HERMAL SYSTEMS San Francisco Mountains and vicinity, southwestern Utah: Utah Geological and Mineralogical Survey Bulletin 98, 24 p. Schubert, Gerald, and Anderson, 0. L., 1974, The earth’s thermal gradient: Physics Today, v. 27, no. 3, p. 28—34. Shawe, D. R., 1972, Reconnaissance geology and mineral potential of Thomas, Keg, and Desert calderas, central Juab County, Utah: U.S. Geological Survey Professional Paper 800-B, p. B67—B77. Smith, R. L., and Shaw, H. R., 1975, Igneous—related geothermal sys- tems, in Assessment of geothermal resources of the United States—197 5: U.S. Geological Survey Circular 726, p. 58—83. Smith, T. B., 1974, Gravity study of the Fumerole Butte area, Juab and Millard Counties: Salt Lake City, Utah, University of Utah, unpub. M.S. thesis, 54 p. Stearns, N. D., Stearns, H. T., and Waring, G. A., 1937, Thermal springs in the United States: U.S. Geological Survey Water- Supply Paper 679—B, p. B59—B206. Stephens, J. C., 1974, Hydrologic reconnaissance of the Wah Wah Valley drainage basin, Millard and Beaver Counties, Utah: Utah Department of Natural Resources Technical Publication no. 47, 53 p. Stewart, J. H., 1971, Basin and Range structure: A system of horsts and grabens produced by deep-seated extension: Geological Soci- ety of America Bulletin, v. 82, no. 4, p. 1019—1044. Stewart, J. H., Moore, W. J ., and Zietz, Isidore, 1977, East—west pat- terns of Cenozoic igneous rocks, aeromagnetic anomalies, and mineral deposits, Nevada and Utah: Geological Society of America Bulletin, v. 88, no. 1, p. 67—77. Swanberg, C. A., 1974, The application of the Na-K-Ca geothermome- ter to thermal areas of Utah and the Imperial Valley, California: Geothermics, v. 3, no. 2, p. 53—59. Truesdell, A. H., and Fournier, R. 0., 1977, Procedure for estimating the temperature of a hot-water component in a mixed water by using a plot of dissolved silica versus enthalpy: U.S. Geological Survey Journal of Research, v. 5, no. 1, p. 49—52. Turk, L. J ., 1973, Hydrogeology of the Bonneville Salt Flats, Utah: Utah Geological and Mineralogical Survey, Water-Resources Bulletin 19, 81 p. U.S. Weather Bureau (no date), Normal annual and May-September precipitation for the state of Utah: Map of Utah, scale 1:500,000, 1 sheet. Waring, G. A., [revised by R. R. Blankenship and Ray Bentall], 1965, Thermal springs of the United States and other countries of the world—A summary: U.S. Geological Survey Professional Paper 492, 383 p. Whelan, J. A., 1970, Radioactive and isotopic age determinations of Utah rocks: Utah Geological and Mineralogical Survey Bulletin 81, 75 p. Whelan, J. A., and Petersen, C. A., 1974, Bonneville Salt Flats—A possible geothermal area?: Utah Geological and Mineralogical Survey, Utah Geology, v. 1, no. 1, p. 71—82. White, D. E., 1969, Rapid heat-flow surveying of geothermal areas, utilizing individual snowfalls as calorimeters: Journal of Geophysical Research, v. 74, no. 22, p. 5191—5201. White, D. E., and Williams, D. L., eds., 1975, Assessment of geother- mal resources of the United States—1975: U.S. Geological Sur- vey Circular 726, 155 p. White, W. N., 1932, A method of estimating ground-water supplies based on discharge by plants and evaporation from soil: U.S. Geological Survey Water-Supply Paper 659—A, 105 p. Young, A. A., and Blaney, H. F., 1942, Use of water by native vegeta— tion: California Department of Public Works, Division of Water Resources Bulletin 50, 154 p. TABLES 16, 17, AND 18 H46 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 16.—Inventory of Thermo Hot Springs1 [Flow rate: Estimated total visible flow of the orifices. Elevation: Reference to adjacent land surface (LS). Specific conductance is in umhos/cm at 25°C. + indicates very small flows. See fig. 10 for map no.] Flow Principal orifice Number Eleva- Tempera- Specific Map of Length Width Depth tion ture conduct- no. orifices L/s Direction (m) (m) (m) (m) (°C) ance Remarks West spring mound Springs in south meadow area: 1 1 + Southwest 0.7 0.4 0.1 At LS 39 2,100 Southemmost spring on mound. 2 1 0.01 Southwest .3 .3 1.0 At LS 70 ____ Located 3 m north-northeast of No. 1 along mound ax1s. 3 4 + Southwest .3 .3 .3 At LS 66 ____ 20 111 north of No. 2. Sinter deposits nearby. 4 4 + Southwest .3 .3 1.0 At LS ____ ____ North end of meadow at shoulder of mound. Sinter and east deposits nearby. Springs flowing east in mostly grassy channels: 5 2 + East .4 .3 .6 At LS 56 ____ 15 m north of No. 4. Located slightly east of mound axis. Deposits of salt on surface. 6 2 + East .4 .4 _ _ __ At LS 54 _ _ - _ Springs are oriented east-west. Only east spring sup- ports grass; on east flank of mound. 7 4 + East .3 .3 1.7 +0.3 72 ____ Mound is sinter, 1.5 m in diameter. Other three orifices are smaller and 3-6 m northeast. 8 5 + East .3 .2 .6 At LS --__ ____ Water flows to and ponds on lowlands near mound. Springs flowing in poorly developed channels: 9 5 + ____ .3 .3 1.5 At LS 61 _-__ At southwest edge of north-south elongated grassy area, 11 m north of No. 8. 10 4 + East .4 .3 .6 At LS 70 ____ Channel extends part way down flank. Springs flowing east in grassy channels to lowlands: 11 9 .1 East .5 .3 .6 —.7 65 1,900 Spring area has three distinct arms at mound axis. 12 3 .1 East .2 .2 .3 —.8 74 ___- Principal spring is middle spring of group. Spring flowing west: 13 5 .3 West .3 .3 __-_ —.8 67 ____ On mound axis near south edge of trail crossing mound. Flow ponds on lowlands. Springs north of trail crossing mound: 14 1 + East .3 .2 .1 At LS 66 _-_- At east margin of mound. 15 1 3 West .8 .7 .8 — .8 82.5 -_ -_ Located slightly west of mound axis in southern part of a 20-m diameter tule area. 16 2 2 East .2 .2 ___- —.5 54 1,650 At east mar 'n of mound. l7 1 + East .2 .05 ____ At LS 61 ___- On east éian of mound. Farthest north of springs on moun . Summary of springs on west mound: Number of orifices: 54 Maximum temperature: 825°C Total visible flow: 1—2 L/s Specific conductance range: 1,650—2,100 umhos/cm East spring mound Springs in south area: 1 + East .3 .2 .3 At LS 64 ____ Half way up east flank of mound. Wet grassy area 10 m north, but no orifice. 2 2 + East .2 .2 ____ —.4 57 ____ Half way up east flank of mound. 3 1 ____ ____ .2 .2 .6 At LS 61 _-__ Sli htly higher on mound flank than No. 2. Sinter eposits nearby. 4 1 .1 East .3 .3 .6 At LS 67.5 __- At west edge of tules area. 5 1 + East .4 .3 .3 —.4 56 1,700 On mound axis. Flow is to large tules area. Spring in tules area ( 30 m X 150 m): . 6 4 + East ____ ____ ____ ____ __ __ ____ Halfway up east flank ofmound. Three grassy chan- nels carry minor flow to lowlands. Springs north of tales area: 7 1 .2 East 2 1 >4 — .7 70 2,000 Two-thirds down east flank of mound and 20 111 north of tules area. 8 1 + East 1 1 .2 —.7 36 ____ At southwest comer of man-made reservoir. 9 0 + East ____ ____ ____ At LS ____ -___ Seep; 150 m northeast of reservoir. Sinter deposits nearby on lowlands. Spring at north end of mound: 10 3 Northeast __-- --__ 2 At LS 68 --__ Three-fourths down north end of mound. Summary of springs an east mound: Number of orifices: 15 Total visible flow: 0.5—1.0 L/s Maximum temperature: 70°C IInventory made in April and May 1976. RECONNAISSAN CE OF THE HYDROTHERMAL RESOURCES OF UTAH H47 TABLE 17.—-Inventory of Crater Hot Springs, February 1976 [Directions, distances, dimensions, and flow rates are estimated. See figure 28 for locations 003 158%}- springs. Size: Small, length and width less than 0.2 by 0.2 m. Flow: Small, less than about Maximum Remarks Location Size temper- Pool or (le .h X width X Flow sture orifice epth; in m) (L/s) (°C) Main pools: South (S) Southwest flank of spring—mound crest __32 X 7 X 0.3 0 28 Many small orifices, mostl in southeast comer of pool; undraine . North (N) Northwest flank of spring—mound crest --30 X 4 X 1 < .1 58 Drains in ditch to northeast. Along southwest ditch: D1 ______ 10 m southwest of main pool—south ______ 0.3 x 0.3 X 0.2 1—2 63 D2 ______ 5 m southwest of D1 ____________________ Small Small 74 D3 ______ 15 m southwest of D2 ____________________ 0.2 X 0.2 X 0.1 Small __ D4 ______ 2 m southwest of D3 ____________________ -___ ? 80 Group of three orifices, 3—4 m west of ditch. D5 ______ 9 m southwest of D4 ____________________ 1.5 X 1 X 1 ____ 77 Group of three orifices, 1 m west of ditch. D6 ______ 3 1n southwest of D5 near bush 2 m high__ __-_ ____ __ 1 m southeast of ditch. D7 ______ 1 m southwest of D6 ____________________ ____ ____ __ D8 ______ 9 m southwest of D7 ____________________ ____ Small 69 11 m northwest of swamp. D9 ______ 9 m southwest of D8 ____________________ Small ? __ 5 orifices in ditch and many to southeast. D10 ______ Southwest of fence ______________________ 0.5 x 0.5 x 0.5 ____ 84 Twogrifices; temperature is of northeast or1 ce. East of main pools: E1 ________ 20 111 east of main pool—south ____________ 6 x 5 X 0.5 0 38 No visible orifices. E2 ________ 12 in northeast of E1 ____________________ 7 x 4 X 0.2 0.3—0.6 82 Ditch extending east southeast. At end of ditcéi, flow = 5 Us and temperature = 65° . E3 ________ 100 m east of E2 ________________________ 1 X 0.5 X 0. .3 64 Similar orifice 2 m northeast. E4 ________ 140 m east southeast of main pool—north__1.5 X 17 X 0.5 .3 87 Trenched east; flow dissipates. E5 ________ 90 in east ofE4 ________________________ 1.5 X 1.5 X 0.2 0 __ Recycling pools east of main pools: R1 ________ 20 in southwest of E5 ____________________ 1 X 1 X 0.6 Small 71 Swamp on south; brush on north. R2 ________ 25 m east of R1 ________________________ 5 X 0.7 X 1 3 70 Source orifice in west end of pool. R3 ________ 15 m southeast of R2 ____________________ ____ 5 29 Source is surface flow from swamp. Several other recycling orifices to east. Northeast of main pools: N1 ______ 30 in east of main pool—north ____________ 5 X 5 X 0.5 0 25 N2 ______ 70 m northeast of main pool-north ______ 3 X 3 X 1 .1 59 N3 ______ 30: m west of N2 ______________________ 1 X 1 X 0.2 Small 77 N4 ______ 151- m west of N3 ______________________ ____ .3 __ Several orifices; along ditch draining north- east from main pool~north. Main drain; spring: __________ 500 m east of main pool—north __________3 X 4 X 1 72 52 Drains in ditch to northeast and then north toward set of buildings. Specific conduc- tance was 5,800 ,umhos. Summa : Observe flow: 190: US Flow from seeps and the like (approximate): 245: US Total (rounded): 140: US Observed orifices and pools: 40: 5 Maximum observed temperature: 87°C lMay include some recycled water. zEstimated as half the observed flow. H48 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS TABLE 18.—Selected subsurface temperature and heat-flow data not summarized on maps [Tem rature description: M, maximum in well; probably bottom-hole temlperature in most wells; D , discharge temperature of well or spri . Temperature gradient: Estimated, computed from list temperature, and-surface ambient temperature, and well depth. stimated heat flow: HFU, heat-flow units, in ucal/cm’/s‘ base on average temperature gradient and generalized thermal conductivity by lithology] Temperature Average Estimated Ap roximate Total —————_— Temperature conductive Remarks wel or spring depth Degrees Descrip- gradient heat flow location (In) “C tion ”C/km (HFU) Beaver area (C—28—7)15bb ______________ 300 10 D Very small <1 Irrigation well. (C—28—7)3lad ______________ 34 11.2 D Very small <1 Unused. (C—29—7)15cd ______________ 46 11.6 D Very small <1 Public-sulpply well. (0- 29—8)9ba ______________ 46 18 D 140 3.6 Stock we . (C—29—8)3 lad ______________ 94 12.1 D Very small <1 Irrigation well. (C—29—8)35ab ______________ 157 18.5 D 45 1.3 Do. (0— 29—8)36ac ______________ 1 10 22 D 96 3 Do. (C-30—7)5cd ______________ 245 22 D 43 1.3 Do. East of Cedar City (C-36—7)33 ________________ 2,016 ____ __-- ___- _-__ Mud temperatures were reported as high as 43°C. Eastern Utah (D—5422)22ac ______________ 1,311 46 D 27 1.6 (D— 11-24)8ca ______________ 2,002 25 D 7 <1 (D—12—21)19de __________ ___- 19.5 D -___ ____ Sulphur (SIC) Spring. (D—22—6)4ca ______________ 492 26.5 D 33 1.3 Milford-Minersville area (C—27—10)6da ______________ 30 13.5 D 49 1.5 Stock well. (C—27— 10)31dc ____________ 213 27 D 70 2 Irrigation well. (C—28— 10)18ac ____________ 138 21 D 65 2 Do. (C—28— 10)14bb ____________ 78 20.5 D 1 10 3.3 Stock well. (C—28— 10)31dd ____________ 59 13.5 D 25 <1 Irrigation well. (C—28—11)10ac ____________ 69 16.5 D 65 1.5 Stock well. (C-28—11)23cb ____________ 29 14 D 69 2.1 Irrigation well. (C—29—10)27bb ____________ 45 12.5 D Very small <1 Stock well. (C—29— 1 1)27ad ____________ 36 14.5 D 69 2.1 Irrigation well. (C—30—9)7ad ______________ 22 33.5 D 980 29 Near warm springs, east of Minersville. Reflects convective flow. (C—30— 10)19ab ____________ 89 21 D 100 3 Irrigation well. Northwestern Utah (A—12—1)16dd ______________ 74 22 D 150 4 (B—1—1)31d ________________ 183 28.5 M 72 <2 Near Salt Lake City airport. (B—1—9)24cd ______________ 79 24 D 160 4 (B—1—18)29cc ______________ 50 28 D 350 7 (B— 1— 18)31cb ______________ 70 24 D 200 4 (B—4—3)19ca 146 24 D 96 3 (B—5—1)30ad _____________ 274 55 D 160 5 (B—5— 13)31ac ______________ 61 22 D 180 3.6 Temperature data from Stephens (1974, p. 44). (B—6—3)19aa ______________ 67 19 D 130 4 (B—6—5)21aaS ______________ --__ 21 D ____ ____ (B—7—5)15ch ______________ ___- 25 D _--- ____ (B—7—5)22cdS ______________ A--- 22 D ____ ____ (8—8—5)5cdS ______________ ___- 22 D _--_ _-__ (B—10—6)9bbS ______________ ___- 22 D ___- ____ (8—10-15)6cdS ____________ -___ 20 D ____ ____ Warm Spring No. 2. fB—ll—ll)6dbS ____________ __-_ 19 D ---_ .1.. Black Butte Spring. (8—12—5)22daS ____________ ____ 20 D ____ ___- (B—12—6)33dbS ____________ --__ 20.5 D -___ ____ (B—13—12)30caS ____________ __-_ 25 D -___ ____ (B-lB—l3)27ddS __________ ___- 21 D __-- _--_ (B—13—13)34ch ____________ -n- 21 D _-__ --__ (B—l3—13)35bbS .......... ____ 23 D ____ ____ (B—l3—l4)21ddS .......... -n- 19.5 D V--- ____ (8-13—14)24ch ............ ____ 23 D __-- __-_ (B-13—16)23ccS ............ ____ 21 D ____ ____ Head Spring. RECONNAISSANCE OF THE HY DROTHERMAL RESOURCES OF UTAH H49 TABLE 18.—Selected subsurface temperature and heat—flow data not summarized on maps-—Continued Temperature Average Estimated Ap roximate Total Temperature conductive Remarks wel or spring depth Degrees Descrip- gradient heat flow location (In) “C tion “0/ km (HFU) Northwestern Utah—Continued (B—14—9)4bb ______________ 107 22 D 110' 2.8 Stock well. (B— 14— 9)9bb ______________ 1 10 21 D 100 2.5 Irrigation well. (B—14—10)33ch ____________ -___ 43 D ____ -___ Coyote Sprin . (B—15—9)28cb ______________ 122 24 D 110 2.9 Irrigation we 1. (B—15—9)30ab ______________ 124 21.5 D 90 2.8 (C— 1—8)6ab ________________ 20 26.5 D 790 20 (C- 1—17)34ba ______________ 1,298 67 M 44 <2 (0— 1—19)1bb ______________ 50 24 D 260 5 (C- 1— 19)3dc ______________ 53 24 D 240 5 Temperature data from Stephens (1974, p. 45). (C— 1— 19)9db‘ ______________ 27 24 D 480 10 Do. (C—1—19)10ba ______________ 33 31 D 610 12 (C—1—19)34cd ______________ 351 32 D 60 1.2 Do. (0— 2— 6)23cb ______________ 64 20 D 140 4 (C— 2~ 19)24c ______________ 499 88 M 160 4 Temperature date from Turk (1973, p. 9). (C—4— 19)7S ________________ ____ 29 D ____ _-__ Blue Lake Spring. (C—5—1)23bd ______________ 32 21 D 310 9 27 22 D 400 12 30 24 D 430. 13 30 23 D 390 12 45 35 D 540 16 32 46 D 1,100 33 Near Saratoga Springs. 152 50 D 120 3.7 (C—6—1)18dc ______________ 85 27 D 180 5 (C—10—2)15 and 22' ________ ____ -___ M 104 ____ Temperature gradient is average for 3 1 values (Lovering and Goode, 1963, table 1 1). (C—1'0-2)15 and 22 ________ __-_ ____ M 140 ___- Do. (C—10—2)15dd ______________ -___ 54 ____ ____ ____ Mine effluent. (D—2—5)32bb ______________ 53 21 D 190 6 (D— 7—3)28bd ______________ 103 32 D 200 6 (D—8—2)28cc ______________ 84 33 D 260 8 (D—8—2)23dc ______________ 174 59 D 280 8 Southwestern Utah (C—12—5)31 ________________ 114 29.5. D 135 4 Heylmun (1966, p. 21). (C— 18—4)31’db ______________ 159‘ 51.5 D’ 250 8 (C—41—15)3230 ____________ 183 43.5 D 130 3.8 Near Thermo Hot Springs (C— 29- 1 1)4ba ______________ 1 7 13 M Very small <2 Martin well; (C— 29— 1 1) 17aa ____________ 18 12.5 M Very small < 1 Windmill. (C—39—11)19db _______ _ 21 15.5 M 170 4 Do. (0— 30— 13)34bb ____________ 20 13 M Very small < 1 Do. (C—31—12)30cd ____________ ____ 17.5 D __-_ -___ May indicate significant heat flow. (C—31-13)1"Saa ____________ 28 15 D 100 <2 (C—31—13)23bb ____________ ____ 12.5 D ____ <1 Windmill. (C— 32— 12)6cb ______________ 21 9 M Very small <1' Do. The Geothermal Hydrology of Warner Valley, Oregon: A Reconnaissance Study By EDWARD A. SAMMEL and ROBERT W. CRAIG GEOHYDRO'LOGY OF GEOTH’ERMAL SYSTEMS GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—1 A description and evaluation of a moderate-temperature hydrothermal convection system UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1981 UNITED STATES DEPARTMENT OF THE INTERIOR JAMES G. WATT, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Catanging in Puincation Data SammeI, Edward A. The geothermal hydroiogy of Warner VaIIey, Oregon. (GeohydroIogy of geothermaI systems) (GeologicaI Survey Professiona] Paper 1044-1) Biinography: p. I45-I47. Supt. of Docs. no.: I 19.2:0r3/5 1. GeothermaI resources--0regon--warner VaIIey. 2. water, Underground--0regon--Warner VaIIey. 1. Craig, Robert w. II. Title. III. Series. IV. Series: GeologicaI Survey ProfessionaI Paper 1044-I. GBII99.7.07SZ4 553.7 81-607830 AACR2 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Page Quality of water—Continued ________________________________ Page Abstract __________________________________________________ I 1 Isotopes in ground water ________________________________ I 21 Introduction ______________________________________________ 1 Distributions of temperatures and heat flow __________________ 23 Objectives and scope of study ____________________________ 1 Temperature of ground water ____________________________ 23 Acknowledgments ______________________________________ 1 Heat flow ____________________________________________ 25 Numbering system for wells and springs __________________ 3 Temperatures in the geothermal reservoir: estimates from mixing Units of measurement __________________________________ 3 models and chemical geothermometers __________________ 26 Physical description of the area ______________________________ 3 Crump geyser area ____________________________________ 29 Climate ______________________________________________ 4 Fisher Hot Spring area ________________________________ 33 Geologic setting ________________________________________ 5 Other thermal waters __________________________________ 37 Surface water ______________________________________________ 9 Conceptual models of the geothermal system __________________ 41 Ground water ______________________________________________ 9 Heat—flow constraints on the depth of fluid circulation ______ 41 Quality of water ___________________________________________ 15 Constraints based on chemical data and mixing models __-_ 42 Quality of surface water ________________________________ 15 Reservoir characteristics ________________________________ 43 Quality of ground water ________________________________ 15 '. Referean ________________________________________________ 45 ILLUSTRATIONS PLATE 1. Map showing locations of springs, wells, and heat-flow holes, and Stifl‘ diagrams for water sampled for chemical analysis» In pocket Page FIGURE 1. Index map of the Warner Valley area, Oregon ______________________________________________________________________ 12 2. Subdivision of township and range sections for well-numbering system ________________________________________________ 3 3. Geologic map of the Warner Valley area ____________________________________________________________________________ 7 4. Piper diagram of chemical relations in water from wells and springs: percentage reacting values __________________________ 17 5. Graph of chloride concentration versus fluoride concentration in ground water __________________________________________ 19 6. Graph of chloride concentration versus temperature in ground water __________________________________________________ 19 7. Graph of chloride concentration versus silica concentration in ground water and hypothetical reservoir water ______________ 19 8. Graph of the weight ratio (Cl/Si02) versus chloride concentration in ground water ______________________________________ 20 9. Graph of the weight ratio (Cl/H003) versus chloride concentration in ground water ______________________________________ 20 10. Graph of 8180 versus 8D in ground water as departures, in parts per thousand, from standard mean ocean water __________ 22 11. Frequency distribution of temperatures (<40°C) measured in springs and wells __________________________________________ 24 . 12. Profiles of temperature measured in heatflow holes __________________________________________________________________ 25 13. Profiles of temperature measured in water wells ____________________________________________________________________ 25 14. Map of the Crump geyser area, showing temperatures measured and estimated at a depth of 2 meters below land surface --__ 27 15. Graph showing correlation between temperatures estimated by the silica and Na—K-Ca geothermometers __________________ 28 16. Graph of silica concentration versus enthalpy in waters having surface temperatures >30°C ______________________________ 31 17. Graph of chloride concentration versus enthalpy in waters having surface temperatures >30°C ____________________________ 32 18. Graph of boron concentration versus enthalpy in waters having surface temperatures >30°C ______________________________ 35 19. Graph of chloride concentration versus 8D in ground water, showing a single—stage boiling curve between 220°C and 96°C __ 36 20. Graph of chloride concentration versus 8130 in ground water, showing a single-stage boiling curve between 220°C and 96°C __ 36 21. Hypothetical profile of normal temperature versus depth in Warner Valley ____________________________________________ 37 TABLES Page TABLE 1. Annual precipitation and temperature at three US. Weather Service stations in and near Warner Valley for the period 1963— 77 ____________________________________________________________________________________________________ I4 2. Average annual streamflow entering Warner Valley in Deep Creek, Honey Creek, and Twentymile Creek for the period of record prior to 1976 __________________________________________________________________________________________ 9 3. Data from wells and heat-flow holes ______________________________________________________________________________ 10 CONTENTS III 1V CONTENTS Page TABLE 4. Data from selected springs ______________________________________________________________________________________ 112 5. Chemical analyses of surface water ______________________________________________________________________________ 15 6. Chemical analyses of water from wells and springs ________________________________________________________________ 16 7. Molal ratios of selected dissolved constituents in ground water ______________________________________________________ 18 8. Isotopes of oxygen ("‘0) and hydrogen (deuterium and tritium) in ground water. ____________________________________ 22 9. Calculated temperatures of equilibration for waters having surface temperatures of 20°C or more ____________________ 29 CONVERSION OF UNITS OF MEASUREMENT Used in Multiply by To obtain this report inch (in) foot (ft) mile (mi) square foot (ftz) acre square mile (miz) acre-foot (acre-ft) cubic mile (mi3) gallon per minute (gal min“) gallon per minute (gal min“) kilogram per second (kg s“) foot per day (ft d“) cubic foot per day per foot (fta d—l ft") degree Celsius (°C) degree Celsius per kilometer (°C km") watt per meter-kelvin (W(m‘k)“) milliwatt per square meter (mW m‘z) watt per square meter (W m‘z) joule (J) joule per second (J s“) joule per cubic centimeter per degree Cel- sius (J cm‘3 °C") micromho per centimeter (”.th cm“) Length 25.40 0.3048 1.609 Area 0.0929 0.4048 2.590 Volume 1.233 x 10‘3 4.168 Flow 0.06308 3.785 1.000 Permeability 0.305 Transmissivity 0.0929 Temperature 1.8 (and add 32) Thermal gradient 0.05486 Thermal conductivity 2.390 , Heat flow 0.0239 23.90 Thermal energy 0.239 1.000 Specific heat (volumetric) 0.239 Electrical conductance 1.000 millimeter (mm) meter (m) kilometer (km) square meter (m2) hectare (ha) square kilometer (km?) cubic hectometer (hm3) cubic kilometer (km3) liter per second (L s“) liter per minute (L min“) liter per second (L s“) for pure water at 3.89°C meter per day (m d") meter squared per day (m2 d“) degree Fahrenheit (°F) degree Fahrenheit per hundred feet (°F/ 100 ft) millicalorie per centimeter per second per degree Celsius (mcal cm“s-"°C“) microcalorie per square centimeter per sec- ond (acal cm‘zs“) microcalorie per square centimeter per sec- ond 1 pvcal cm‘2 s‘1 = 1 heat—flow unit (HFU) calorie (cal) watt (W) calorie per cubic centimeter per degree Celsius (cal cm‘3 °C‘1) microsiemens per centimeter (”S cm”) THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON: A RECONNAISSANCE STUDY By EDWARD A. SAMMEL and ROBERT W. CRAIG ABSTRACT Warner Valley and its southern extension, Coleman Valley, are two of several high-desert valleys in the Basin and Range province of south-central Oregon that contain thermal waters. At least 20 ther- mal Springs, defined as having temperatures of 20°C or more, issue from Tertiary basaltic flows and tuffs in and near the valleys. Many shallow wells also produce thermal waters. The highest measured temperature is 127°C, reported from a well known as Crump geyser, at a depth of 200 meters. The hottest spring, located near Crump geyser, has a surface temperature of 78°C. The occurrence of these thermal waters is closely related to faults and fault intersections in the graben and horst structure of the valleys. Chemical analyses show that the thermal waters are of two types: sodium chloride and sodium bicarbonate waters. The warmer waters are likely to have higher concentrations of sodium and chloride, as well as sulfate, silica, and dissolved solids, than the cooler waters. Chemical indicators show that the geothermal system is a hot-water rather than a vapor-dominated system. Conductive heat flow in areas of the valley unaffected by hy- drothermal convection is probably about 75 milliwatts per square meter. The normal thermal gradient in valley-fill deposits in these areas may be about 40°C per kilometer. Extensive areas underlain by thermal ground water occur near Crump geyser and Fisher Hot Spring. These two areas, located along the western and eastern boundary faults, respectively, are believed to be zones in which hot water, derived from geothermal reservoirs, spreads from the fault zones and mixes with local ground water. Thermal gradients in the valley-fill deposits are extremely high in these areas. Geothermometers and mixing models indicate that temperatures of equilibration are at least 170°C for the thermal components of the hotter waters. The thermal waters probably originate as local meteoric water which circulates to great depths in the fault zones. The depth of circulation may be as great as 4 kilometers, on the basis of the thermal gradients (about 40°C per kilometer) estimated for the valley-fill deposits and bedrock. The size and location of geothermal reservoirs are unknown. If the mixing models are valid, thermal waters of the Crump geyser area and the Fisher Hot Spring area could be derived from a common reservoir. If so, a probable maximum size for such a reservoir is about 38 cubic kilometers at a depth near 4 kilometers. Total heat stored in the reservoir,ab0ve a base temperature of 10°C, could be as much as 1.6 x 1019 joules. A probable minimum value for stored heat, esti- mated on the basis of the assumption that the thermal fluids at depth are entirely restricted to the major boundary faults, may be about 5 X 1016 joules. The reservoir at a depth of 4 kilometers constitutes a resource under present economic conditions only to the extent that hot water can be withdrawn from upflow conduits at shallower depths. Because of conductive cooling, water temperatures in the upflow conduits will have lower temperatures than water in the reservoir. Recoverable heat stored in mixing zones in the valley-fill deposits at depths less than 300 meters may total at least 5 X 1016 joules, however, and may represent an economic resource at temperatures greater than 100°C. INTRODUCTION OBJECTIVES AND SCOPE or STUDY The objectives of the study can be summarized as follows: 1. To determine the large-scale characteristics of the hydrologic regime with emphasis on the occur- rence and nature of the ground water. 2. To map the geology as it relates to the hydrologic regime. 3. To determine the chemical characteristics of the nonthermal waters of the area and their relation to the geothermal waters. 4. To interpret, if possible, the geologic and hydrologic data in terms of the nature and extent of the geothermal systems and the characteristics of the geothermal reservoirs. The investigations described in this report were made during the fall of 1974 and the summer of 1976. The area of study comprises about 875 square miles (mi2) in Warner Valley, Coleman Valley, and adjacent highlands in south-central Oregon (fig. 1). Within this area, nearly all wells and springs were inventoried, all known sources of geothermal water were investigated, and most were sampled for chemical and isotope analysis. The geologic map by Walker and Repenning (1965) provides the framework for the stratigraphy and lithology described in this report. Detailed structural mapping in much of the area was added in 1976 by N. E. Voegtly. ACKNOWLEDGMENTS The authors are grateful for assistance received from many individuals and organizations. Information and Il 12 42°30’ THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON 119°3o’ 120°00’ 119°45’ / / I I x/ / I I COYOTE HILLS ) (" S ‘\./”'\‘ _ T I F’" k x Q HART/MOUN TA IN ’ b I o I § NA TIONAL )AN TEL OPE REFUGE I I /II/ / I, v / A0 I 42°15’ Fisher Hot Spring W / Crump Geyser I I _ Valley bottom I'EIIC I an ‘7‘0 Lek . /’ ake ‘ _______ _ (30 In ”’Wv Ummproved road i/ ’ 0,) [IA // — Improved road 1’ 99,0 Cree (Adel —--— Generalized topographic boundaries Study area 42°00’ SCALE 0 5MILES O 5 KILOMETERS FIGURE 1,—Index map of the Warner Valley area, Oregon. PHYSICAL DESCRIPTION OF THE AREA 13 facilities provided by the Lakeview District Office, U.S. Bureau of Land Management was especially helpful. Thanks are also due to the U.S. Soil Conservation Ser— vice for the use of aerial photographs, to the Geother— mal Division of Gulf Oil Co. for permitting us to drill on land leased by them, and to Jack Stooksberry, Lakeview well driller, for much helpful information. Finally, we are grateful to scores of landowners who granted access to wells and springs, provided informa- tion, and cooperated in many other ways with the in- vestigation. NUMBERING SYSTEM FOR WELLS AND SPRINGS In this report, wells and springs are numbered ac- cording to a system based on their locations within townships, ranges, and sections referred to the Willamette Baseline and Meridian. Thus, the number 36/24—4 baaa specifies a well in Township 36 South, Range 24 East, and Section 4. The letters which follow indicate respectively the 1/4, 1/16, 1/64, and 1/256 quadrants within the section according to the scheme illustrated in figure 2. A number following the letters indicates that more than one well was inventoried in the same 1/256 sec- tion quadrant; such wells are numbered serially in chronological order of inventory. Each well number specifies a location within a quadrant that is 330 feet on a side. WELL 36/24—4 baaa / b l. a // — a _ // // C l d // ////// //// / / // \\\ \ \ \\ 1” \ >/ \\\ \VV/ \ \\\b l\a \ —-‘b \ —— a c l d l 4 __ c __ d _ Section UNITS 0F MEASUREMENT The units of measurement in this report are those used in the field or obtained from the reports of others. For the most part, these units are in the foot-pound system, and this system has been retained for the con— venience of most readers. Temperatures were meas- ured in degrees Celsius (°C) and are consistently re- ported as such. Derived units are generally calculated and reported in the units obtained in the field. In the sections of the report entitled “Distributions of Tem- perature and Heat Flow” and “Conceptual Models of the Geothermal System,” the units used for meas- urements of thermal gradients and heat flow are in the International System in order to avoid conversion of data obtained from others and to facilitate comparison with most published data. Units of mass and length used in these sections of the report are also given in the International System for the sake of consistency. The table, “Conversion of Units of Measurement,” at the beginning of this report, will enable units to be con- verted from one system to the other. PHYSICAL DESCRIPTION OF THE AREA Warner Valley is a narrow graben about 60 miles long and 4 to 5 miles wide. It is alined approximately north—south, paralleling the Warner Range which lies some 15 miles to the west. The valley lies in a transi- tion zone between the Great Basin and the Columbia FIGURE 2.—Subdivision of township and range sections for well-numbering system. 14 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON Plateaus physiographic provinces. It is unique among valleys in south-central Oregon in having an abundant supply of surface water, most of which is derived from precipitation on the mountains. At the south end of Warner Valley, and separated from it by an upraised fault block, is Coleman Valley, about 10 miles long and 3 miles wide, which extends into Nevada. Unlike Warner Valley, Coleman Valley is dry most of the year and contains playa lake beds. In this respect, Coleman Valley is more typical of the high-desert valleys that occur to the south in Nevada and to the east in Oregon. The upraised fault blocks that define the valley have precipitous slopes along most of the valley length. Maximum relief between the valley floor and the fault—block summits is about 3,560 feet and occurs be- tween Warner Peak on Hart Mountain and the adja- cent Valley floor in a distance ofonly about 3 miles. The sunken blocks that underlie the floor of Warner Valley are covered by alluvial and lacustrine deposits that are 800 feet deep in some places and may be much deeper in at least one area. Eleven shallow lakes and many ephemeral ponds occur in the valley, fed by three pe— rennial streams and many springs. Warner Valley and its tributary, Coleman Valley, form a closed basin with respect to surface water. It is probable, however, that ground water flows from the basin at its northern end. Irrigated crops, principally alfalfa and some cereal grains, flourish in Warner Valley, with surface water furnishing most of the water supply in the south half of the valley and ground water supplying the north half. Thousands of cattle forage in the highlands adjacent to the valley during the summer months and are brought into the valley each fall for sale or winter feeding. The economy of the valley is almost entirely dependent on cattle ranching. The lakes and pastures of Warner Valley provide a natural sanctuary and feeding ground for water fowl. Deer are prevalent in the higher country, and herds of pronghorn anteope occupy the barren high-desert country east of Hart Mountain. The area was probably a favored hunting ground for prehistoric Indians, and their pictographs are found on several large boulders on the east side of the valley. At least 20 geothermal wells and springs occur in Warner and Coleman Valleys. This number is based on the assumption that any water having a temperature of 20°C or more has a geothermal component. With the exception of a warm spring on Hart Mountain and a slightly warm well in the valley northwest of Hart Mountain, all of the known thermal occurrences are located south of Hart Lake. The present study has been largely confined to the areas containing the known thermal waters and the adjacent highlands at the south end of the basin. A high valley, 10 miles west of the main area on the lower slopes of the Warner Range, was also studied. This area, Big Valley, contains the headwaters of Deep Creek as well as several thermal springs. CLIMATE The climate of Warner Valley is relatively mild be- cause of its protected location. Average annual tem- perature in the valley at Adel for a recent period of 15 years is 94°C, as compared with 8°C at Lakeview, about 30 miles west of Adel, and 64°C on Hart Mountain (table 1). Average annual precipitation at Adel is about 9.9 inches as compared with 16.5 inches at Lakeview. Much of the precipitation brought into the region by westerly winds is apparently intercepted by the Warner Range and fails to reach the Warner Valley area. The valley would be considered semiarid were it not for the three perennial streams that flow into it from the Warner Range. Rates of evapotranspiration in Warner Valley have not been determined. A crude estimate of total evapo- transpiration may be obtained, however, from esti— mates of precipitation and streamflow. It seems reasonable to assume that precipitation at Adel is rep- resentative of precipitation on the valley floor south of the 42°30’ parallel. Over this area of approximately 390 mi2, therefore, the average amount of water de- rived annually from precipitation is calculated to be about 207,000 acre-feet. Total streamflow entering the valley may add about 167,000 acre-feet. (See section entitled “Surface Water.”) There is no surface-water outflow, and if it is assumed that ground-water outflow from the valley is equal to ground-water inflow, then total evapotranspiration is about 1.5 feet per year (374,000 acre-feet). The estimate derived in the manner outlined above may be too high because ground-water outflow from the valley probably exceeds ground-water inflow. This may occur because of high water tables induced by sur- face inflow and irrigation. Estimates of transmissivity in the surficial deposits underlying the valley suggest, however, that ground-water outflow is a small fraction of the water lost through evapotranspiration. (See sec- tion entitled “Ground Water.”) The estimate of evapo- transpiration appears reasonable in the light of the TABLE 1.—Annual precipitation and temperature at three US. Weather Service stations in and near Warner Valley for the period 1963—77 {Data from the National Climatic Center, Asheville, NC] Approximate Station Average annual Average annual altitude precipitation temperature (ft) (in) (DC) Adel ____________ 4,560 9.9 9.4 Hart Mountain 5,616 11.8 6.4 Lakeview ________ 4,740 16.5 8.0 PHYSICAL DESCRIPTION OF THE AREA 15 available water supply, the many lakes and ponds, and the relatively high water table. The amount of direct evaporation from lakes in the valley is not known. On the basis of a comparison with three lakes in the surrounding region that have depths similar to the larger lakes of Warner Valley, it seems likely that evaporation from lake surfaces is on the order of 3.3 feet per year. The lakes used for this com- parison are Abert, 3.5 feet; Summer, 3.5 feet; and Har- ney, 3.3 feet (Langbein, 1961). Because of the large fluctuations in lake size from year to year, no estimate has been made of the total evaporative loss from lake surfaces in Warner Valley. In the area south of parallel 42°30’, the maximum surface area of lakes is about 30 mi2 and evaporation from the lake surfaces is probably less than 20 percent of the total evapotranspiration. Near the north end of the valley, ponded water covers a large fraction of the area, and evaporation from lake surfaces may be more than half the total evapotranspi- ration. GEOLOGIC SETTING Warner Valley is in the Basin and Range physio~ graphic province near its northern boundary with the Columbia Plateaus province. The Basin and Range province is characterized by extension in a thin conti- nental crust, evidence for which in the Warner Valley area consists of typical graben structures having nor- mal boundary faults, seismic activity, and volcanism. The crustal extension in the area began at least 17 million years ago and has continued through Pleis- tocene time and probably Holocene time (Christiansen and McKee, 1978). Associated with the zone of crustal extension in Ore- gon is a belt of bimodal volcanism that originated in southeastern Oregon about 13 million years ago, pro- gressed toward the northwest, and culminated in ex- tensive silicic volcanism at Newberry Volcano, near Bend, Oreg. (MacLeod and others, 1975). The progres- sion of silicic volcanism across southeastern Oregon occurred at a rate of about 71/2 inches per year during the period 5 to 10 million years ago and decreased to about 21/2 inches per year during subsequent times. Precise dates are not available for silicic volcanic rocks in Warner Valley, but extrapolation of ages from sur- rounding areas suggests an age of about 9 million years for these rocks. _ The regional structural pattern in the vicinity of Warner Valley is similar to that described by Donath (1962) in the Summer Lake-Abert Rim area, a horst and graben region located about 25 miles west of Warner Valley. In both areas, a rhombic pattern of faults has occurred in a crustal stress field Whose prin- cipal stress axis was north-south and whose minimuni stress axis was east-west. The resulting faults have major strike-slip components. Superimposed on the rhombic pattern of strike-slip faults are normal faults related to the regional crustal extension. The northern boundary of the zone in which these patterns appear is the Brother’s fault zone, a complex transform fault zone located about 10 miles north of Warner Valley on the northern margin of the Basin and Range province. Two sets of high-angle normal faults occur in Warner Valley, striking about N. 35° W. and N. 20° E. These faults probably originated during Pleistocene time or even earlier, and continued to be active in Holocene time. Several thousand feet of dip-slip dis- placement has occurred in northeast-trending faults (Walker and Swanson, 1968; Larson, 1965). No surface movement along fault planes in Warner Valley has been observed in historic times. The faults blocks that form the valley walls have been tilted and fractured with only minor warping of the strata. Some, such as the massive block of Hart Mountain, have apparently been raised above previous levels (Larson, 1965). At the south end of Warner Val- ley and in Coleman Valley, fracturing and tilting have occurred to a high degree, forming a confused pattern of angled blocks. It is likely that the blocks dropped beneath the valley floor form similar patterns. Warner Valley is near the center of an active earth- quake region in southeast Oregon. A map showing epicenters of historic earthquakes since 1841 (Couch and Lowell, 1971) reveals that the northern extension of the Basin and Range province in Oregon contains at least six areas in which significant numbers of earth- quakes have occurred. The earthquakes probably originate in the earth’s crust at an average depth between 12 and 15 miles. The crust in this area may have a thickness between 20 and 25 miles. The earthquakes are the result of crustal extension in a stress field whose minimum compressive stress is oriented east-west. The underlying cause ap- pears to be related to the movement of crustal plates. An earthquake swarm occurred in Warner Valley in May and June 1968, causing moderate damage to many buildings in the Adel area. The largest quake, with an estimated magnitude of 5.1 (Richter scale), had an epicenter located between Crump and Hart Lakes (Couch and Johnson, 1968). No detectable motion oc- curred along surficial faults in the area during the quakes, and there were no documented changes in thermal springs and wells, although there are conflict- ing reports of a significant change in Crump geyser. A comprehensive analysis of 169 seismic events in the swarm has been made by Schaff (1976). He reports that nearly all the epicenters fall within a rectangle about 15 kilometers long (north-south) and 6 kilome- ters wide, centered about 3 kilometers northwest of 16 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON Adel. Only one shock registered clearly enough at a sufficient number of stations to reliably indicate a focal mechanism. This shock, the largest of the swarm, oc- curred on a plane having a dip of 80° E., and a strike of N. 4° E. The motion on this plane was oblique slip, with about equal components of left-lateral strike-slip and reverse fault motion (Schaff, 1976, p. 26). The reverse fault motion implies that the valley block moved upward with respect to the mountain block, and that the maximum stress was compressional rather than tensional. This motion is contrary to the motion as determined by Couch and Johnson (1968), who based their conclusions on data from 11 shocks less well-documented than the one used by Schaff. Other earthquake swarms in the region, such as the one in 1973 at Denio, Nev., about 65 miles east of Warner Valley, have demonstrated tensional horizon— tal stress as would be expected for the Basin and Range province (Richins, 1974). ’ Apart from the question of the focal mechanism for the Warner Valley earthquakes, the data of Schaff suggest strongly that the swarm was associated with p movements along the western boundary fault and that seismic focal points occurred at depths between 2 and 13 kilometers. The inference is clear that a fault ex- tends to the depths determined for the earthquake hypocenters, and it may also be inferred that such a fault could serve as a conduit for the circulation of thermal waters. Geologic units in the south half of Warner Valley as modified from the map of Walker and Repenning (1965) are shown in figure 3. Several of the mapped units have been combined for the sake of simplicity. Faults shown on this map were located in 1976 by N. E. Voegtly. The oldest rocks in the area, exposed on the west face of Hart Mountain, are gently dipping, commonly al- tered tuffs, tuffaceous sediments, andesite flows, and flow breccias of late Oligocene to middle-Miocene age (Walker and Swanson, 1968). The unit is designated Tvp in figure 3. Thin, discontinuous basalt flows ex- posed at the mouth of Deep Creek may be about the same age. Rocks shown as be are andesitic and basaltic flows and flow breccias of middle to late Miocene age that correlate in part with the Steens Basalt (Steens Mountain Basalt of Fuller, 1931) which was potassium-argon dated 14.5 to 14.7 my. on a flow high in the unit at Steens Mountain (Evernden and others, 1964). They are also correlative with the Steens Mountain Volcanics (Steens Mountain Andesitic Series of Fuller, 1931)~and with the andesite flows overlying the Steens Basalt, which may correlate with the Owyhee Basalt (Renick, 1930). The unit be shown in figure 3 includes, at places, the subdivided units Taf and be mapped by Walker and Repenning (1965). The Steens Basalt is particularly well exposed at Deep Creek Falls, west of Adel, where massive basalt flows and interbedded tuff and scoria crop out in the canyon. The upper part of the unit is highly feldspathic; diktytaxitic and subophitic textures are common. Where the rocks are porphyritic, plagioclase (An 50-70) is the dominant phenocryst. Olivine, com- monly altered, occurs in small to moderate amounts. Flows in the lower section contain calcite (aragonite) and zeolites. Intrusive rhyolitic and dacitic flows and breccias of Miocene to Pliocene age form small plugs and domes at a few places. Within the area shown in figure 3, these intrusives occur only on Hart Mountain. Overlying the basaltic and andesitic flows in the upraised blocks are Tertiary tuffaceous rocks (Tts). These are mostly fine-grained tuffaceous sedimentary rocks and tuffs representing flood plain and shallow lake deposits. They grade upward into pumice lapilli tuffs. Elsewhere, these deposits contain middle to late Miocene fossils correlative to faunas in the Mascall Formation (Downs, 1956) and at Beatty Butte (Wal- lace, 1946). The unit is correlative in part with the Mascall Formation (Merriam, 1901), the Virgin Valley Beds, (Merriam, 1910), and the upper part of the Cedarville Series (Russell, 1928). These rocks form the steep upper part of the valley walls adjacent to Deep Creek and are believed to underlie the alluvial deposits on the valley floor in at least the southern part of the valley. An erosional surface with hundreds of feet of relief developed on the tuffaceous sediments (Walker and Swanson, 1968). Overlying the erosional surface are Tertiary basalt flows (Tb). These flows cover the sur- face of nearly all uplifted fault blocks, including Hart Mountain. The basalt is highly feldspathic, contains slightly altered olivine, and exhibits diktytaxitic tex- ture (Walker and Repenning, 1965). The unit is cor- relative in part with the widespread grey open- textured olivine basalt (Pliocene) of Wells and Peck (1961) in south-central Oregon. Constructional volcanic features that are younger than the Tertiary basalt flows occur at two places in the area shown in figure 3: in the canyon of Twelvemile Creek and on the northeast flank of Hart Mountain. These rocks (Qva) form large complex exogenous domes and related flows and flow breccias of rhyodaci- tic composition that are highly modified by erosion. These features may have been emplaced as early as the Miocene or as late as the early Pleistocene. The remainder of the rock units exposed in the study area are sedimentary volcanic deposits and lacustrine and alluvial sediments dating from late Pliocene to Pleistocene and Holocene. Some of the oldest of these 42° 15' Base from U. 8. Geological Survey, 196 PHYSICAL DESCRIPTION OF THE AREA CORRELATION OF MAP UNITS Holocene and Pleistocene Pleistocene and ' Pliocene — AND > 4 Early Pleistocene(?) TERTIARY Qva . . 7 L « Pliocene, and Miocene _ Pliocene and Miocene Miocene and Oligocene d DESCRIPTION OF MAP UNITS t“: Alluvial deposits. Includes playa deposits, Qp, of Walker and Repenning, 1965. ' A Landslide debris. ": Lacustrine, fluviatile, and aeolian sedimentary ‘ rocks; includes interstratified tuft“, ashy diato- mite, and small masses of hot-spring sinter. Also includes ediment or fluvial gravel, th, of Walker and epenning, 1965. 4 A Rhyodacitic domes and related flows and flow breccias of rhyodacitic composition. %% Basalt flows; feldspathic olivine basalt. Tuffaceous sedimentary rocks and tuffs ‘ m Basalt and andesite flows and flow breccias, * variable in texture and mineral composition. Includes subdivided units Taf and TM of Walker and Repenning, 1965. Tuffaceous sedimentary rocks and tuffs. Contact J——--- Fault-Dashed where approximately located; dotted where inferred. Bar and ball on down- thrown side. 0 5 10 KILOMETERS '__;_'_r_L,_ifil__fi_T—J O 5 MILES FIGURE 3.—Geologic map of the Warner Valley area (modified from Walker and Repenning, 1965). }QUATERNARY QUATERNARY -TERTIARY 18 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON rocks, designated Qts, consist of lacustrine, fluviatile, and eolian sediments, with interstratified tuff, ashy diatomite, and unconsolidated clay, silt, sand, and gravel. This unit is correlative in part with Pliocene and Pleistocene waterlaid volcanic deposits and Pleis- tocene terrace deposits of Wells and Peck (1961). Pedi- ment or fluvial gravel (included with Qts in fig. 3) that covers the lower slopes of the valley walls at places may represent former pluvial lake levels. During Pleistocene time, a single large lake occupied the graben. According to Weide (1974), the water reached a maximum depth of about 320 feet prior to 17,000 HP. The lake had a surface area of about 500 mi2 and drained to the north into the Harney Basin. Beach features are not well defined at the highest levels, but conspicuous shorelines exist on the fault scarps at several lower levels. According to Weide, the shorelines dip toward the south and have warped at an average rate of 2.4 feet per thousand years over the past 17,000 years. A large delta was constructed at the mouth of Deep Creek during the period of high lake levels. The sur- face of this delta is now at least 200 feet above the valley floor. As lake levels dropped, the delta was dis- sected by Deep Creek, leaving steep-sided remnants along a former stream course. Deep Creek now flows south of the delta in order to reach the valley floor. Two US. Geological Survey drill holes in the valley floor, MCI and MC2 (pl. 1), penetrated lacustrine sed- iments to depths of 839 and 658 feet, respectively, be- fore reaching basaltic bedrock. An exploratory hole drilled by the San Juan Oil Co. 1 mile east of Adel reportedly encountered basalt at a depth close to 800 feet. The basalt is probably part of a sequence of tuf- faceous sedimentary rocks, tuffs, and interbedded basaltic and andesitic flows similar to those exposed in an upraised block about 3 miles east'of the study area (rock unit st of Walker and Repenning, 1965). This unit may be correlative in part with the Yonna Forma- tion (Newcomb, 1958) that is prominently exposed in several valleys west of the area, and with numerous similar formations occurring elsewhere in eastern Oregon and Idaho. Northward from Plush, the valley floor contains swampy areas and playa lakes, surrounded by windblown sand. Similar conditions are found in Cole- man Valley, near the Nevada State Line. Evaporite deposits, including potash, borates, and ulexite, occur in the playas in noncommercial quantities. Talus deposits and landslide debris cover the base of most steep fault scarps (QTls, fig. 3). Below Hart Mountain, for example, a large landslide forms a series of hummocky hills that extend several miles onto the valley floor. The debris consists of unsorted mixtures of basaltic and tuffaceous material ranging in grain size from large boulders to sand and silt. Geophysical investigations conducted by the US. Geological Survey in Warner Valley and Coleman Val- ley include a truck-borne magnetometer survey, an aeromagnetic survey, gravity measurements, and par- tial coverage by the audio-magnetotelluric method (AMT) (Gregory and Martinez, 1975; Plouff and Con- radi, 1975). Reports on the findings provide estimates of thickness of the valley—fill deposits at several places and show the locations of major boundary faults. The AMT data define a shallow zone of low-resistivity rocks in one part of the valley near Crump geyser. (See sec— tion entitled “Distributions of Temperature and Heat Flow.”) The magnetometer and gravity surveys generally coincide in depicting the north-trending faults that bound the east and west sides of the valleys. Both the magnetic and gravitational discontinuities are more pronounced on the west side of the valley than on the east side, suggesting that the western boundary fault zone is better defined than the eastern at most places. Several of the northwest-trending faults that cut the upraised blocks on both sides of the valley are indi- cated by trends in the data. The cross trends are espe- cially notable in the vicinity of Crump geyser and Fisher Hot Spring, and the intersecting faults mapped in the upraised blocks may extend ,into the valley—fill deposits. At other places along the main boundary faults, such as the mouth of Deep Creek canyon, step faulting with a lateral spacing of approximately 34 miles is indicated by the magnetometer data. A major feature of the gravity map is a large region of low gravity centered in the valley about 3 miles south of Adel. A maximum thickness of alluvium of about 2,700 feet can be calculated from the residual anomaly if the average density of the alluvium is as- sumed to be 2.0 grams per cubic centimeter (g cm‘3) and the density of underlying Tertiary rocks is 2.5 g cm'3. However, gravity data from elsewhere in the val- ley indicate that the rocks most likely to underlie the valley floor in the southern half of Warner. Valley are tuffaceous sedimentary rocks and tuffs which may have densities not much greater than those of the deeper valley-fill deposits. Thus, the depth of alluvium may be much less than the 2,700 feet calculated on the basis of a maximum density contrast between alluvium and an underlying basaltic bedrock. Depths of al- luvium estimated from some of the more reliable truck-borne magnetometer data range from 200 feet in Coleman Valley and the northwest edge of Greaser Reservoir to 500 feet at a location 1 mile south of Fisher Hot Spring. Near the southern boundary of Warner Valley in Township 40 8., Range 24 E., the GROUND WATER 19 estimated depth to bedrock is 400 feet. The aeromagnetic map shows a major magnetic low that coincides with the large gravity low south of Adel and which is consistent with the presence of a thick slab of relatively nonmagnetic sediments in this part of the Valley. A large magnetic high is centered on the boundary between sections 9 and 16, Township 40 8., Range 24 E. at the extreme south end of Warner Val- ley. The cause of the magnetic anomaly at this location is not known. The gravity and magnetic surveys are helpful in out— lining the large-scale structural features in the area, but they are not sufficiently detailed to delineate small structural features and they do not provide clues to the depth and extent of deep reservoirs that could supply the geothermal waters. SURFACE WATER Warner Valley is a closed basin with no surface drainage outlet. Two perennial streams, Twentymile Creek and Deep Creek, enter the valley near its south- ern end, and a third, Honey Creek, enters near Plush. All three streams have their sources in the Warner Range. During the winter and early spring, many ephemeral streams flow into the valley from canyons in the valley walls. Surface flow in the southern part of the valley is generally toward the north with a gradient of approx- imately 1.5 x 10““1 (0.8 feet per mile). To the north of Hart Lake, the gradient is virtually zero and the area is an evaporative sump. Streamflow entering the valley in Twentymile, Honey, and Deep Creeks is gaged, and records are available for periods of years ranging from 44 to 53. Data for the period prior to 1976 are given in table 2. The amount of water entering the valley in ungaged streams has been measured during only one period of TABLE 2.——Auerage annual stream/low entering Warner Valley in Deep Creek, Honey Creek, and Twentymile Creek for the period of record prior to 1976 [Data from 0.5. Geological Survey Water-Data Report OR-75-1 (1976)] Latitude. Years of record Average discharge Station longitude prior to 1976 (fti‘s' ‘) (acre-ft yr") Deep Creek, 42°11’21" 47 131 94,910 5 miles west 120°00’02” of Adel. Honey Creek, 42°25’30” 49 30.4 22,020 1 mile north- west of Plush. 119°55’20” Twentymile 42°04’20” 40 53.2 38,540 Creek, 1.5 119°57’42” miles down- stream from Twelvemile Creek. time, so far as is known to the authors. For the years 1909—15, average ungaged flow is estimated to have been about 7.3 percent of the gaged streamflow (Whis- tler and Lewis, 1916). The annual percentages ranged from 2.4 percent to 14.1 percent, with the lower per- centages occurring in years of lower flow. If the same conditions hold for the period of record shown in table 2, the average ungaged streamflow may be about 11,350 acre-foot per year (acre-ft yr“). Most of this flow occurs during the 4-month period January through April. GROUND WATER The characteristics of ground water in the study area are known from data on 66 wells and heat-flow holes and 57 springs. The data from these wells and springs are given in tables 3 and 4 and locations are shown in plate 1. Most of the wells in Warner Valley penetrate allu- vial or lacustrine deposits to depths less than 300 feet. Many of the wells situated near the valley walls also penetrate small thicknesses of volcanic breccia, cin- ders, and one or more thin basalt flows. Only one well is known to penetrate a fairly thick sequence of basal- tic rocks in a raised fault block. Water levels at most places are within 30 feet of the land surface except in the former delta of Deep Creek, near Adel, where water levels are as much as 70 to 90 feet below land surface. Wells having depths greater than about 250‘ feet generally have hydraulic heads above land surface as the result of pressures transmitted from aquifers at higher levels in the upraised fault blocks. No measurements of static heads in these artesian wells are available, but heads meas- ured in three heat-flow holes, MCI, M02, and 0K1, at depths ranging from 350 to 500 feet, were 1 to 3 feet above land surface. Artesian heads are believed to be generally low throughout the valley. Springs are widely scattered in the valley and the adjacent highlands. They range in type from simple gravity springs fed by local precipitation to thermal- artesian springs fed, in part, by deep geothermal reser- voirs. Nearly half the springs inventoried for the study are cold gravity springs that issue on the surfaces of fault blocks that flank the valley. The remainder, in- cluding most thermal springs, occur on or near the val— ley floor, most of them at the base of talus slopes or fault scarps. Many have no well-defined openings, but are merely seeps that, in some cases, extend for hun- dreds of feet along the bases of the fault scarps. A few springs occur in the alluvium at some distance from the valley walls where there is no visible reason for their occurrence. Thermal springs in the study area cannot be distin- guished from nonthermal springs on the basis of 110 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON TABLE 3,—Datafrom wells and heat-flow holes Sequence No.—Identifyin number or letter used in tables and illustrations. Location—Based on townsiip, range, and section. See figure 2 for key. Completion—~OH, open hole; P, rforated (depth interval in feet). Altitude—Land cslurfacle at wel , in feet above National Geodetic Vertical Datum of 1929, estimated from US. Geological Survey topographic qua ran e ma 5. Depth to water~—-Feet elow and surface; F, flowing at land surface. Yield—«Gallons per minute (gal min“). Temperature—Maximum water temperature in degrees Celsius after pumping; in unused wells where temperature profile was obtained, temperature is maximum measured at depth given (feet). S ecific conductance—Micromho per centimeter at 25°C. se—D, domestic; I, irrigation; S, stock; T, USGS test hole; U, unused. Driller’s log—O, owner; D, driller’s records; SE, Oregon State Engineer’s records; T, Trau er, 1950; US, US. Geological Survey. Other data——C, chemical analysis, table 6; I, isotope analysis, table 8; T, temperature pro 1e, USGS, figures 13 and 14. Se- Depth Specific uence Longitude Date Inside Comple- to Temp» conduct- Date Driller’s Other ll‘lo. Owner Location Latitude completed De th diameter tion Altitude water Yield erature ance measured Use log data (deg) (min) ( t) (in) (ft) (ft) ° Wells _--, Flynn Bros. 35/24-33dbcc 113 3330 -_-_ 190 14 OH 4510 34 300 13(184) -_n 12-09-74 U D T(12-09-74) . 5 __-. Martin Anderson 35/25-31bdda 119 49.55 1915 375 2 OH 4465 F 5-10 15 320 Owners D&S .___ ..._ 42 29.75 report .___ KileyRanch 35/25-32ccca 119 $3.70 .___ __._ 1.5 he. 4480 F 2-5 12 160 09-28-74 S .0. .___ 42 .45 .___ Phillip Lynch 36/24-laccd 1112) 33% 1959 450 -.__ OH 4470 F 11 k" .__. report S .___ .1.. 2 Cook Laird 36/24—4baaa 119 . 33:70 08-05—58 190 13 P90611014) 4505 33 2500 14 308 08-00-74 I SE C(04—27-76) 42 9.00 & .__. PhillipLynch 36/24—10ddn 113 5%.35 .-._ .___ __.- .___ 4470 F 4 .__. -___ -_1. S _.._ .___ 2 . 5 .-__ Laird&SonsRch. 36/244483de 119 59.80 ~1914 400 2 OH 4470 ~1“ .__. 15 400 09-28-74 D -__. *Flows much 42 2 '10 ofyear. ____ Laird&SonsRch. 36/24-14aadb2 119 2571.80 ~1950 300 -___ OH 4470 __-- _-__ 13 680 -___ -___ -___ -___ 42 .10 .__- Fitzgerald Reh. 36/24-17dm 1:9 543(5) 1960 135 16 OH 4485 17.5 2500 Cold .__. .___ I O/T ___. 2 26. __-_ CharlesChalstand 36/24—22ba 119 52.90 1940 35 9 OH 4470 3.3 -___ -_-_ .__. 08-16-48 _._. T Data from 42 26.10 Trawler (1950). Well not located in field. Phillip Lynch 36/24-27bbcc 119 53.00 260* 6 OH 4490 10 12(143) 100474 U “1004.7“ 42 25-25 *Fillad into 143 . .___ Joseph Flynn 36/24-2803bd 119 52.32 1958 447 15 OH 4495 16 _..- 13(447) ..-_ 10-04—74 U .___ T(10—04—74) 2 2 . -___ Plush School Dist. 36/24-23ccbd 119 52.115 08-23-69 65 6 P2313113) 4500 24.5 15 13 300 09-27-74 D SE -___ 42 2 . 5 -___ Frank Riverman 36/24—280006 119 54%: 03-00-76 80 6 P(52/72) 4500 29.5 20 .___ -_-_ 09-27-74 D SE 1... 2 24. .___ Joseph Flynn 36/24-28cdcd 119 512.22 ~1955 48 6 OH 4490 13 .__- 13 240 09-27-74 D SE .___ 42 2 . Phillip Lynch 36/24-29adab 119 54.30 10-00-68 100 6 OH 4515 6 15 12 430 1023-68 D SE 1m, pm,“ 42 25.25 tates from water ---_ Taylor Rch.lnc. 36/24-33abbc 11; 32:2 11-00-60 262 14 OH 4490 30 ~2000 15 320 02-0060 1 SE -___ _-__ Tom Sullavan 36/24-33bbbb 119 31:21 -__- 15-20 ____ OH 4490 2-3 5-10 12 340 09-28-74 D -___ -___ 42 .55 4 PhillipLynch 36/25-7cddb 119 49.6: 1958 650 6 .___ 4470 F 5.5 20.5 275 09-28-74 S -___ C,I(04-28-76) 42 27.5 __-- McKeeRanch 36/25-80ddd 11% figs .___ ~200 10 OH 4535 _.-. .___ 12 150 .___ D&S .___ .___ .__- Fitzgerald 37/24-4hbbb 119 532(0) ____ ~100 -___ OH 4475 ._-_ _._. _.__ _-_. .___ D .___ _.._ 42 2 . -___ Vern Pura 37/24~9cdcc 119 33.33 12-00-68 170 6 flag?) 4480 9 25 14 320 12-30—68 S SE _--_ 42 . 8: 6 Glen Perky 37/25-20ddda 11% 43:2 1962 160 8 OH 4500 F 5 20 188 11-01-74 S D C(04—21-76) 2 . 8 Glen Perky 37/25-33bdcd 119 47.40 ~1958 81 8 -___ 4480 60 200 14(82) 219 04-21-76 S _.__ 4 _ 42 19.20 121 210* *11-01-74 wfw‘i’fitand fllal'lefil—rept. owsprlng 1974. C,I(04—21-76) T022676) 12 Charles Crump 38/24—34ccad1 119 52.90 1955 121 4 4515 1-2 124* 1390 12—07-74 3 0 [(031576) 42 ”-60 *Muimum temperature at depth. 13 Charles Crump 3024—3406462 119 52.90 1959 1664 10 OH 4515 4.5 127* 1310 12-07-74 U C “031576) 42 13.45 maximum temperature at depth. 15 GlenPerky 38/25-16aaab 11; 46.70 -___ 80-90 8 OH 4485 15 .___ 18 1260 me; D&S -.._ C,I(03-12- 6) 16.95 ‘ -___ GlenPerky 38/25-168834‘1 119 42.33 ~1958 180 8 OH 4490 ---- 200 18 650 -___ D853 D -0- 2 . -___ Henry O'Keefe 39/24—3cccc 11; 53.05 08-21-74 294 6 OH 4500 ___. .___ 60 _.__ ____ U 0 _.._ 12.65 -__- Henry O’Keefe 39/24—160ddc 119 53% 03-06-64 162 6 P0061131) 4600 80 20 14 300 03-05-64 D SE .___ 42 . & 20 Henry O’Keefe 39/24—16dabc 1:: 53.2? 07-06-67 360 16 ”208128) 4500 F 400 17 1130 07-07-67 S SE C,I(03—16-76) 1 . .___ HenryO’Keefe 39/24-16dcda 113 53.45 03-24-70 240 6 P05881198) 4590 85 10 17 470 08-24—70 D SE ..._ 1 .90 8:0 .___ GlennCleland 39/24—2laadd 119 53 20 10-16-66 100 6 P(90/100) 4580 69 16 12 180 11-19—74 D SE .___ GROUND WATER 111 TABLE 3.—Data from wells and heat-flow holes—Continued _ Depth Specific . Y lsueance longitude Date Inside Comple— to . Temp- conduct- Date Drillers Other 0‘10. Owner Location Latitude completed De th diameter tion Altitude water Yield eravture ance measured Use log data deg) (min) (lg) (in) (ft) (ft) (C) _-__ Raymond Poore 39/24—21abbb 113 53.35 04-03-68 158 6 'OH 4590 88 15 16(7) 300 04-03-68 SE ---- 1 . 5 ____ OregonState 39/24-2lacbb 119 53.65 11-10-62 110 8 P(80/90) 4590 76 20 cold 230 11-02-62 D SE _-__ Hw .Dept. 42 10.60 -_-_ J.N. riener 39/24-21bdba 1:3 53.23 ---- 65 6 OH 4535 18 ---- cold 180 11-18-74 D ____ -_-_ 1 . ---_ Adel School District 39/24—21bdab 119 53.22 1959 85 6 OH 4540 32.5 --__ --__ _-__ 11-18—74 D&U -__- _--_ 2 1 . ---_ Adel School District 39/24-2lcadc 113 53;; 06-07-68 123 --__ OH 4560 ---_ _--- _-_- 190 -__- D825 SE _--_ 1 . 21 Adel School District 39/24-21cdbb 1:9 530(5) 1961 100 6 OH 4550 55.5 ---_ 13 232 11-18—74 D _.__ C(03-29-76) .1 ___- HenryO’Keefe 39/24-23bccb 113 53.23 08-00-74 310 6 ---_ 4485 ---_ --__ ---_ _--- --__ U 0 ---_ --_- HenryO’Keefe 39/24-23bdda 113 55:30 ---- 300 6 ___. 4485 -___ -___ ____ _-__ __-_ U 0 ___- .50 ____ HenryO’Keefe 39/24-230ccd 113 53.38 ---- _--- 2.5 OH 4485 F ~2 11 200 12-14-74 U ._-_ ---- 22 HenryO‘Keefe 39/24—26baad 119 338(5) 1969 375 ---_ __-_ 4485 F ~50 13.5 190 1969 S O C,I(04—20~76) 2 . ___- MC Ranch 39/24—2Saadc 113 33.22 1959 250 10 OH 4500 --_- 200 Warm 200 ---- S ___. ___- ____ MC Ranch 39/24288bbb 119 563556955 -_-_ ---- 6 OH 4510 --__ -_-_ Cold 175 ---_ D -__. --_- 42 . --__ Bill Lane 39/24—28bbac 119 33.3; ~1968 ~200 8 OH 4545 52 _--. Cold 850 11-18-74 D -_-- ___. ---- MCRanch 39/24—28bbcd 119 33%?) 04-27-61 106 6 P(50/106) 4530 32 60 13 280 11-18—74 D SE ___. 2 . 23 Lonny Schadler 39/24-29cdcc 113 55.18 -_-- __-- 6 ---- 4500 ---_ ___- 33 366 ___- S ---_ C,I(03-28—76) 09.1 24 JohnLane 39/24-29dbab 119 53.28 07-21-69 252 8 P(85/234) 4520 25 _--_ 17 354 07-01-69 S SE C(05-05—76) 42 0 . 26 MCRanch 39/24-32baad 119 33.33 -_-_ 302 14 OH 4485 F 10—15 14 349 11-17-74 S T C,I(03-28—76) DonRobinson 39/24—33abcb1 119 53:65 <1925 110 3 OH 4490 1-2 12 330 11-1074 D .Reptiwm, 42 08-95 level rises infall.de« clines belowLSD inDec., since 1968 earthquake. 27 TerryCahill 39/24—34abbb 11: 33.5506 07-21-69 231 6 P(60/2111) 4490 F 6-8 9.5 220 11-16—74 D825 SE C(04<22-76) 28 MC Ranch 39/24—34cbaa 119 52:80 <1925 156 4 OH 4490 3-5 12 267 10-28-40 S ---_ C(03-29-76) 42 08.60 10‘K 10" 266‘ *C(08-09—48), Trauger (1950). -_-_ Mrs.O’Sullivan 39/24—35bbab 119 3332 ---_ 24“ 4 ---- 4490 9.5 -_-_ ---_ ___- 11-19—74 U -_-- *Siltedin 42 . 30 BLM (Sandy Well) 39/25288de 113 33% 11-12—70 430 8 OH 4805 330 _--- 25.5 242 09—30—70 S SE C,I(04—25-76) __-_ WarnerValley 39/25-30cb 119 50:00 1934 690 6 ---- 4485 5 ---- -.-. ---- 10-28-74 S&D T Stock Co. 42 09.45 ”1‘52?“ 32 JamesG.Dyke 40/23-24ddbb 1:3 31.33 ~1962 160 6 OH 4560 21 21 721 12—07-74 D C.I(04-27-76) -___ WarnerValley 40/24-3bc 119 53:05 ~1948 566 10 _-_- 4490 F 10 11 ---_ 08-22-48 * T w 8ka Co. 42 07.85 Des‘myed HughCahill 40/24-15cdbc 119 56.40 ~1950 ~100 4 OH 4505 8.5 51500 ~15 400 11-04-74 D .Yieldmuld 42 05.70 ‘ notbesuar tained. Value pro- mm 1g . 35 USBLM 40/25-4dcba 119 39.23 05—01-61 150 6 P(50/150) 4550 68.5 40 12 445 11-19-74 S SE C,I(05—05-76) 42 . ---_ HughCahill 41/24-ldbad 113 gig ~1964 405 13 OH 4525 36 _--- 17(405) ---_ 12-06-74 U --_- T(12»06-74) --__ Lonny Schadler 47/19-25adcd 119 47:60 ---_ 162 6 ___- 4740 3 ---_ 15(162) ———- 12-10-74 U -_-_ _ (Nevada? 41 58.00 311113 $5,: onmap 43 LonnySchadler 47/20<3labbc 119 41.65 _--- 400- 10 ---- 4750 F ---- 27 207 11-21-74 S ---- C(04-29-76) (Nevada? 41 57.50 600 mambo“ onmap Heat-flow holes 19 MCRanch(MC-l) 39/24—llddbb 119 51.00 07-2576 885 6 * 4480 F 10 23.5(790) 344 07-27-76 T SE/US C,I(06-02-76) 42 1130 T(11-06—76) "Casing groutedto total depth. ---- HenryO’Keefe (OK-1) 39/24—26bbac 119 51.70 08-06-76 552 6 " 4480 2 ___- 17.5(533) _--_ 09-27~76 T SE/US Casing 42 09.90 3,0“de total depth. 39, MC Ranch(MC-2) 40/24—17cacd 119 55.40 08-16-76 696 6 *‘ 4490 2 ---- 24(696) 710(#39) 08-16-76 T SE/US *39(342’-357’) 4 , 42 05.75 40(492'506’) 41 41(606'696’) C,I(#39,0& 28-76) C,I(#40, 08- 26-76 C,1(#41,08- 25—76) T(11-06—76) ”Casing groutedto total depth. 112 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON TABLE 4,—Data from selected springs Sequence No.—Identifyin number used in tables and illustrations. Location—Based on towns ip, range, and section. See figure 2 for ke Altitude—Land surface at spring, in feet above National Geodetic maps. Yield—Gallons per minute (gal min"). Specific Conductance—Micromho per centimeter at 25°C. Use—D, domestic; S, stock; U, unused. Other data—C, chemical analysis in table 6; I, isotope analysis in table 8. Vertical Datum of 1929, estimated from Geological Survey topographic quadrangle Temp- Specific Se uence Owner Spring name Location Longitude Altitude Probable water-bearing Occurence Yield erature conduct- Date Use Other data So. Latitude (ft) materials (“0) ance . _ . _ Kiley Ranch Anderson Spr. 35/25-32dcdd 1133310 4525 Valley-fill deposits gravity 5- 10 16 100 09-28-74 S , - ,1 .25 45 US Fish 8; Valet Spr. 35/26-28bccc 11940.65 5880 Contact-tuffaceous sedi- ____ "A 14 448 07-15-80 D&S C(07-15-80) Wildlife 42 30.55 ments, volcanic dome 1 US Fish & Antelope Hot Spr.35/26-32baba 119 41.75 6000 Volcanic dome thermal- 10-12 40 876 ”,7 D&S C(08-10-48) Wildlife 42 30.00 artesian 1(03-00-77) _ _ _ . McKee Ranch a" 36/25-8cd(a&d) 113 $.35 4520 Lake terrace sediments A," O 8 160 09328-74 S 2 large seeps. .60 5 US Fish & _ . ., 36/25-9cdcd 119 47.45 4980- Tertiary basaltic flows gravity 10-15 4 77 09-28-74 U C,I(05-03-76) Linear seeps Wildlife 42 27.50 5000 indeep canyon. "A US Fish& ,._. 36/25-30daaa 119 48.85 4825 Tertiarybasaltic flows gravity .W- 6 230 H” S t"- Wildlife 42 25.30 _ _ . _ US Fish 8: , , , , 36/25-30daba 1 1949.10 4660 Contact-tuffaceous sedi- gravity 10* 1 1 245 09-20-48’k S *Trauger (1950). Wildlife 42 .5.20 merits, basaltic flows _ _ . . Ancore . . W 37/24-23bacd 1 19 51.55 4480 Tertiary baalts Thermal- 10 21 320 09-29-74 U Artesian head ~ ‘4 it above 4221.15 artesian LSDincasing. _ _ . . US BLM Lynch Spr. 37/24-30cdcb 113 53.30 5740 Tertiary basalts gravity 3 10 120 1 1-03-74 S , __ _ 1 . 5 . _ V _ Mrs, Calder- _ _ _ . 37/25-28bcbd 119 47 .40 4490 Volcanic and lacustrine thermal- 3-5 20 200 11-0 1-74 S _ _ . _ wood 42 20.20 sediments artesian an Glen Perky “.1 37/25-28bdca 1194735 4500 Volcanic and lacustrine Thermal- 1-3 24 200 11-01-74 S _.__ 42 20.10 sediments ariesian 7 Dixon , , , , 37/25-30dccb 1 19 49.50 4480 Contact-Tertiary basalts, gravity 27 3 70 10-04-74 S C(04-20-76) 42 19.65 valley-fill deposits 1-5 "H Dixon .v- 37/25-30dccc 1 1949.50 4490 Contact-Tertiary basalts, gravity 1 11 430 04-20-76 S "A 42 19.55 valley-fill deposits ,1. . US Fish & .d- 37/35-33ba 11947.35 4500 Volcanic and lacustrine gravity 20 ____ ".1 09-24-48 ,_,- Trauger(1950). Not Wildlife 42 19.50 sediments visited. -H , John Lane _,__ 38/24—3bdad 119 52.50 4485 Landslide debris and Ter- thermal- 2-4 20 210 09-30-74 "A 42 18.40 tiary basalt flows gravity ".7 John Lane ,_-, 38/24—3dcdc 119 52.35 4485 Landslide debris and Ter- thermal- <1 24 200 09-30-74 S 1-1, 42 17.85 tiary basalt flows artesian US BLM Coxes Spr. 38/24-6dbad 1 19 55.80 6090 Tertiary basalts gravity 5- 10 6 1 10 1 1-03-74 S C,I(04—18-76) 42 18.15 ”a John Lane __,_ 38/24—1080db 119 52.35 4490 Valley-fill deposits (Ter- thermal— 5-10 25 220 09-30-74 S -M, 42 17.50 tiary basalts?) gravity 10 John Lane _ _ . _ 38/24— 10dadd 1 19 52.00 4500 Tertiary basalts thermal- 10- 15 29 162 09-30-74 S Almost continuous seeps 42 17.25 gravity and flows for V4 mile. C,l(04-18-76). A” John Lane .___ 38/24-14b(b—c) 119 51.75 4480 Valley-fill deposits (Ter— gravity 0 _.__ _..- 09-30-74 S Water level ~ IR below 42 16.70 tiary basalts?) (inter- LSD 09-30-74. mittent) u" _.__ ____ 38/24-22bd(a-d) 119 52.23 4475 Tertiary basalts gravity 4—5 24 275 12-14-79 U Line ofseeps at lake level. 42 15. 1 1 Charles Crump , , , _ 38/24-27cddb 119 52.60 4500(?) Tertiary basalts thermal- 5 40 935 09-09-48 S Trauger (1 950). C(09- 1 6- 42 14.45 artesian 48) 44 MC Ranch _ _ _ . (38/2434ch 1 19 52. 70 4490 Lacustrine sediments thermal- 5- 10 70-80 1490 12-07-74 S Numerous widespread 42 13.90 (Tertiary basalts) artesian seeps. C,I(08-03-72) 14 Glen Perky Fisher Hot 38/25- lobbbb 1 19 46.55 45 10 Tertiary basalts thermal 15-20 68 5 13 09-22-74 D See 'I‘raug‘er (1950), Bowen Spr. 42 17.85 and Peterson (1970), Mariner and others (1974,1975).C(08-02r 72) 1(03-00-77) 16 US BLM 0-- 38/25-2labdc 1194700 4490 Valley-fill deposits thermal- 1-3 38 440 11-01-74 S Developed, 2.5 itcasing in 42 15.90 artesian large seep area. C,I(03- 17 Terry Cahill . _ _ . 38/25-30dddd 1142) 42.32 4478 Valley-fill deposits artesian 1- 10 14 366 05-05-76 S C(05-05-76) 1 . 46 Henry O'Keefe _ _ _ _ 39/24—3dcbc 1 19 52.45 4485 Valley-fill deposits thermal- 2-3 58 1400 07-19-80 U Large area of sinter depos- 42 12.75 artesian its. At least six other seeps and spring 35 to 49"C. C(07-l9—80) 18 Charles CrumpCrump Spr. 39/24—4badc 1 19 53.80 5960 Probable contact Tertiarygravity 10 6 95 03- 16-76 S C,I(03- 1 6-76) 42 13.45 tuffs over basalt flows ..__ US BLM a" 39/24—18cddd 113 53.55 5400 Tuffaceous sediments gravity 0.2 13 200 11-04-74 S .1“ 1 . 0 25 MC Ranch .___ 39/24—31babd 119 56.20 5150 Contact, tuffaceous sedi- gravity 10 13.5 212 11-20-74 D C(03-29—76) 42 09.00 ments, older basalt flows physiographic setting or type of occurrence. Both thermal and nonthermal springs may be gravity fed or have artesian flow. Both occur as seeps on the valley floor and as well-defined artesian flows issuing high on fault blocks. A noteworthy example of the latter is An- telope Hot Spring, which issues from volcanic rocks on Hart Mountain at an altitude of about 6,000 feet. Thermal waters have been defined, in this area, as GROUND WATER 113 TABLE 4. —Data from selected springs—Continued Temp- Specific Sefiuence Owner Spring name Location Longitude AltitudeProbable water-bearing Occurence Yield erature conduct- Date Use Other data 0. Latitude (f t) materials oC) ance _ _ _ . Robinson _ _ . _ 40/22- 15bacd 120 06.95 5590 Rhyodacitic tuff gravity 10- 1 5 11 240 12- 1 1-74 S One of numerous springs Ranch 42 06.20 in draw. _ _ v _ Robinson . _ . . 40/22-20dccc* 120 09.10 5620 Rhyodacitic tuff gravity 15-20 7 70 1 2-11-74 3 One outlet at base of mas- Ranch 42 04.70 sive outcrop at change in slope.* Not on location map. 3 1 Robinson _ _ _ _ 40/22.22babc 120 07.05 5550 Rhyodacite flows thermal _ _ _ _ 23 1 5 1 _ , i . S C,l(04-28—76) Line of seeps Ranch 42 05.45 at base ofdomal mass. _ _ _ _ Robinson _ _ . _ 40/22-28baca" 120 08.10 5590 Rhyodacitic tuff gravity 1-2 6 80 1 2- 1 1-74 D&U Shallow, du out, near Ranch 42 04.50 stream.* ot on location map. _ _ _ _ Robinson _ _ _ 1 40/22-29acdd* 120 08.85 5660 Rhyodacitic tuff gravity 20—25 12 140 12- 1 1-74 S&D Shallow, dug out and Ranch 42 04.28 walled up.* Not on loca- tion map. _,__ HenryO’Keefe _.__ 40/23-36bca 11;) 33-53 4595 'I‘uffaceous sediments gravity 15-30 34 430 1207-74 _.._ __._ 33 MC Ranch Foskett Spr. 40/24-25caaa 11332.33 4480 Tertiary basalts tiligermal- 15 21 270 1 1-17-74 . _-. C,I(04-05-76) . a man 34 Hugh Cahill Hallinan Spr. (?)*40/24-29bcbb 1:9 32.23 4530 Tertiary basalts thermal 3-5 66 945 1 1-04—74 S *Té‘au7g6er (1950). C,l(04- 2 . 5- ) _ . .1 Hugh Cahill Hallinan Spr. (71‘40/24—30aaac 113 32.22 4530 Tertiary basalts thermal 3-5 45 670 1 1-04—74 S *Trauger (1950). __._ ___- i." 40/24-36acba 119 50:15 4480 Playadeposits (Tertiary artesian 3-5 16 150 11-17-74 S _.._ 42 03.55 basalts?) __ __ US BLM _ -__ 40/25—19cbac 1:3 31-5133 4475 Playa deposits artesian < 1 8 650 1 1-17-74 S _ _ ._ 36 US BLM Chukar Spr. 40/25-23be 11% 32.22 4820 Tertiary basalt gravity 5 7 130 04 18-7 6 S C(04- 1 8-76) _ _ _ _ US BLM - _ _ _ 40/25-23c-d l 19 44:35 5600 Tuffaceous sediments and gravity 2-5 7 2 10 1 1-21-74 S Springs come from many 42 04.80 (Max) basalt (Tertiary) soulrl'ces along canyon wa s. _ . - . US ELM Burro Spr. 40/25-34bacd 1 19 45.75 4740 “iffaceous sediments gravity 2 11 420 1 1-21-74 S Much calcium deposits on 42 03.60 nearby rocks. 37 US BLM Rosebriar Spr. 40/2536dcac 1 19 43.05 6330 Tuffaceous sediments gravity <1 8.5 1 30 1 1-23-74 S Water level ~ 3 it below 42 03.00 LSD. C.I(04-08—76) , _-- US BLM Y Spr. 41-23-10bdac 119 59.;5 5200 Tuffaoeous sediments gravity 1 12 160 1 1-18—74 S “i, 42 01. 0 38 HughCahill .i__ 41/24—1dcda 113 32.18 4510 Landslide debris gravity 1 11 400 11-17-74 S C(04—07-76) . .1 ___ _ US BLM Sp§arpoint 41/24—10bdaa 1112) (5)2132 5460 Tertiary basalts gravity 3-5 9 260 11~17-74 S __ __ pr. 1. ___- Hugh Cahill .__- 41/24—12cdbd1 11330.;8 4750 Tertiary basalts artesian(?) 1-2 6 220 11-20~74 S .-__ 1. ___. Hugh Cahill ._._ 41/24—12cdbd2 11230.38 4730 Tertiary basalts gravity ,___ 10 250 s... S ___. ___. US BLM ___- 41/24-13abcd 11338.3(; 4755 Tertiary basalts gravity 1 14 220 11-2074 S Slight smell ofHZS. -1-_ Hugh Cahill __ _ _ 41/24—13bacb 113 30.33 4970 Tertiary basalts gravity 3-5 13 170 1 1-20—74 _ .-_ Slight smell of H28. .-_- Lolnny Schad- .. __ 41/25-7bdab 11239.2(; 4490 Valley-fill deposits artasian 2—3 15 310 1 1-17-74 S _ -_ - er . ___- US BLM Jackass Spr. 41/25-15adca 11945.10 5330 Tufl‘aceous sediments gravity 10 12 170 11-20-74 S ___- 42 00.80 (Tertiary basalts?) _ _ _ _ US BLM Hill Camp Spr. 4 l/26-8aaaa 119 40.35 5960 Tufl'aceous sediments gravity 5—3 8 150 11-23-74 S , . . - 42 02.00 (Tertiary basalts?) . _ . . US BLM Tim Spr. 4 1-26— 18bbdd 1 19 42.45 6480 Contact, Tertiary basalts, gravity 0 4 160 11-23-74 S Water level ~ 3 It below . 42 00.95 tuffaceous sediments LSD. 42 US BLM Gravely Spr. 47/ 18- 13dadd 119 54.35 5505 Tertiary volcanic rocks gravity 0 10 194 04-07-76 . ___ C,I(04—07-76)* Not on loca- (Nevada)* 41 59.55 tion map. those having temperatures greater than 20°C, or about 11°C above mean annual air temperature in the valley. This definition is almost entirely arbitrary, and several springs having temperatures lower than 20°C may con- tain conductively cooled thermal water. Some springs having temperatures greater than 20°C may be con- ductively heated at shallow depths. However, most springs with temperatures of 20°C or greater are con- firmed by chemical and isotopic evidence to have significant fractions of thermal water. The nonthermal and thermal springs range from fresh meteoric water to relatively old water derived almost directly from geothermal reservoirs. Most of the thermal waters are mixtures of these two extremes. The thermal springs especially seem to represent a fortuitous combination of circumstances involving permeable fault or fracture zones, the intersection of these zones with permeable lithologic horizons, and the mixing of deep thermal waters with shallow ground water in these zones. Most springs have flow rates that are estimated to be less than 10 gallons per minute (gal min“). Several springs in Big Valley and in the canyon of Twentymile Creek have flow rates as high as 20 to 40 gal min“, but these springs are exceptional. The total flow of all 114 » THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON springs inventoried is probably in the range 300 to 400 gal min‘1 (485 to 645 acre-ft yr‘l). Spring discharge, therefore, is less than 1 percent of the water that enters in surface streams. _ Little is known of the ground-water regime in basal- tic rocks of the fault blocks adjacent to the valley and beneath the valley floor. The BLM well (sequence number 30) located in section 28, Township 39 8., Range 25 E., at an altitude of 4,805 feet derives water from basaltic rocks and volcanic sediments in a raised block. The water level in this well, in September 1970, was at an altitude of approximately 4,475 feet, which is very nearly the altitude of shallow ground water in the valley about 11/2 miles to the west. Water levels in highlands west of the valley are prob- ably higher than those to the east as the result of recharge from precipitation and snowmelt on the Warner Range. Most surface-water inflow to the valley enters from the west side, and this fact implies a pre- dominance of ground-water inflow from this side. The distribution of springs in the highlands is not helpful in estimates of ground-water altitudes and flow rates because springs occur on both sides of the valley at equally high altitudes and are likely to represent only shallow perched water tables in the nearly horizontal basaltic strata. The hydraulic conductivity and transmissivity of the rocks are not known anywhere in the area. A few data are available from which the specific capacities of wells may be calculated and transmissivities estimated. (Specific capacity equals yield divided by the draw- down of the water level during pumping.) In the sedimentary deposits, specific capacities in 11 wells ranged from approximately 1 to 60 gal min‘1 per foot of drawdown. Applying methods described by Theis (1963) and Hurr (1966), the range of transmissivities estimated from the specific capacities is several hundred to 13,000 cubic feet per day per foot (equiva- lent to ft2d‘1). Most of the wells are believed to pene- trate less than half the aquifer thickness, however, and many of the wells are fairly old. Thus, the observed specific capacities are probably not reliable indicators of the transmissivity of the aquifer. Lithologic logs from many wells consistently describe the sediments as predominantly fine grained and organic rich, and the average transmissivity is likely to be in the lower part of the range given above—perhaps 4,000 to 5,000 ftzd‘l. Transmissivities in the volcanic rocks are even more difficult to estimate than those in the valley-fill depos- its. Data on discharge and drawdown are available from drillers’ logs or ownerS’ reports for eight wells in the valley that penetrate at least small thicknesses of basaltic flows and volcanic sediments beneath the val- ley fill. The wells range from large-diameter irrigation wells to small stock-watering wells. Their depths are mostly less than 250 feet. Specific capacities range from about 3 to several hundred gal min‘1 per foot of drawdown, indicating that apparent transmissivities are in the range 800 to 55,000 ft2 d“. These wells pene- trate no more than a few tens of feet of volcanic rock, however, and the water-bearing formations appear to be mostly volcanic sediments and breccias. Data from these wells probably are not representative of condi- tions in the deeper basaltic aquifers. The volcanic formations in the area appear to be similar in most respects to formations near Klamath Falls, where data are available from about 140 wells completed in basaltic flow rocks and volcanic sedi- ments (Sammel, 1980). Assuming that the hydraulic characteristics of the Warner Valley rocks are similar to those in the Klamath Falls area, average transmissivity in tuffs and tuffaceous sediments is es- timated to be about 8,000 ft2 (1‘1 and in the basaltic flow rocks, about 10,000 ft2 d'l. Flow paths of ground water in the area have not been accurately determined. Probable patterns of flow can be deduced, however, from measurements of static heads in wells and from some of the chemical data. Within the valley-fill deposits, water probably moves upward from the bedrock and laterally from the valley walls as the result of the presumably higher poten- tiometric heads in the adjacent highlands. Artesian heads observed in several deep wells and highly permeable zones penetrated by wells in bedrock aqui- fers near the valley walls indicate that such flow must occur. The artesian heads are generally no more than 10 or 15 feet above normal hydrostatic heads, however, and the amount of flow may be rather small. Chemical analyses of water from one drill hole, MC2, located near the south end of Warner Valley, show that water having the characteristics generally found in bedrock aquifers moves upward more than 200 feet into the unconsolidated deposits at this site (table 3). A significant amount of ground-water recharge may oc- cur, therefore, at the base of the valley-fill deposits if conditions at MC2 are prevalent. Throughout most of the thickness of the valley-fill deposits, ground-water flow is toward the north with a gradient that is probably similar to the surface-water gradient of about 0.8 feet per mile. The water table forms a generally uniform surface which is interrupted only by a ground-water mound within the former delta of Deep Creek, near Adel. Un- confined ground water in the deltaic deposits has static heads as much as 70 feet above levels in adjacent valley-fill deposits, and hydraulic gradients directed away from the delta are as high as 40 feet per mile. QUALITY OF WATER At the north end of Warner Valley, much of the ground water is discharged by evapotranspiration. Ir- rigation from large wells in this area contributes to the evaporative discharge, but most of the water probably evaporates from lake and swamp surfaces. Ground water probably flows out of Warner Valley toward the topographically low area surrounding Har- ney Lake, about 10 miles to the north. Much of the flow may occur in a thick section of tuffaceous sediments and welded tuffs that cover the older Tertiary volcanic rocks north of Warner Valley (Walker and Repenning, 1965; Greene and others, 1972). A crude estimate of ground-water discharge can be made on the basis of an assumed hydraulic gradient and an estimate of transmissivity in the region north of Warner Valley. The average hydraulic gradient be- tween the north end of Warner Valley and Harney Lake is on the order of 0.007. Transmissivity in the welded tuffs and tuffaceous sediments is estimated to be in the range 5,000 to 8,000 ft2 d”‘. If most of the flow is assumed to occur through an 8-mile wide section of rock below the 4,500-foot altitude contour north of Bluejoint Lake, the total annual ground-water outflow is calculated to be in the range 12,000 to 20,000 acre-ft. This estimate may contain large errors, but it seems certain that ground-water discharge is a small fraction of the total recharge from streamflow and precipita- tion. QUALITY OF WATER QUALITY OF SURFACE WATER The chemical quality of water in the perennial streams that enter the valley is excellent, as would be 115 expected of water that originates as precipitation on the Warner Range only a few tens of miles west of the valley. The data in table 5 and the Stiff diagrams in plate 1 show that the waters in Deep Creek, Twen- tymile Creek, and Honey Creek are mildly bicarbonate waters in which calcium concentrations slightly exceed concentrations of sodium and magnesium. The concen- tration of dissolved solids is less than 200 milligrams per liter (mg L‘l). As the water moves downgradient (northward) in Deep Creek and into the chain of lakes that form the channel in much of the area north of Adel, dissolved solids increase and sodium becomes the dominant cat- ion. Some of the added constituents, particularly solids increase and sodium becomes the dominant ca- tion. Some of the added constituents, particularly sodium and nitrogen, are probably derived from fer- tilizers used in the valley. Much of the increase in dis- solved solids, however, is due to natural constituents dissolved by ground water from the valley-fill deposits. Finally, evaporation from lakes and the shallow water table and the addition of water from thermal springs add to the concentrations of dissolved solids in the sur- face waters. QUALITY OF GROUND WATER Shallow ground water in the Warner Valley area is generally of good to excellent quality. Concentrations of dissolved solids are mostly less than 300 mg L“1 (table 6). Waters having temperatures less than 10°C are of excellent quality and are chemically similar to stream waters entering the valley. The coldest waters are found in gravity springs located in the highlands. They are mildly calcium bicarbonate to sodium bicarbonate-type waters. TABLE 5.—Chemical analyses of surface water [Concentrations in milligrams per liter unless otherwise indicated] s . Dis- CFfieC :3:- As B F M sollvaed conduc- Sources - e n soi s tance Date of 59'1- ture Ca Mg Na K HCO CO SO Cl F 510. ( L" L“ —1 _. No. Name Location (0’) a a l 1 ug ) (Mg ) (MEL ) (”EL ) (”‘th egg-711) sampled data 1 .- 2 ~ _, 21 23 1370 82 1820 511 206 504 __ 20 __ ._ 500 __ 3951 iv _ 09-06-12 A ls Bl“eJ°'“‘Lak° Xfizfl‘figvfifiuh 5.0 16 7.6 29 31.7 139 0 12 10 0.3 30 38 410 140 20 180 230 7.2 03-12-76 B 285 CrumpLake 38/24—22cdaa .. 13 5.6 18 (incl.K) 95 0 8.6 4.2 A 3.00) 0.0 __ _. a101 __ 1 0907-12 C 215 CrumpLake 38/24-22 16.7 34 14 54 6.3 246 6 26 20 39 __ ._ __ __ 3322 __ 8.4 04-27-61 0 27:3 CrumpLake 38/24—22 89 -- -» 9° —— A— -— ~ 30 »— — -- 7' _, “500 752 -- 090%1 C 2dS CrumpLake 38/24—22 156 —~ -- 16 -» -~ -- —» 5-0 -- 23 -— —_ 1 -. 3120 162 -. 04-19-62 0 255 Crum Lake 38/24-22 6.0 7.8 3.6 6.8 2.6 58 0 2.1 1.1 0.1 27 0 0 230 10 89 105 7.7 03-10—76 B 3138 Deep reek 39/23-24acdb .- -— u -— -— 46 0 9-2 1.0 ,, —— A- -- -- -- _. 79 6.9 03.31.51 0 31,3 DeepCreek 39/2345!) 18.3 79 3.8 7.1 2.8 62 0 , 1.2 -, 31 1. _, __ __ 390 105 7_1 08-21‘56 C 328 DeEPCreek 39/23-15‘3 ._ 19 19 102 15 238 60 24 37 __ 28 __ ,1 600 ._ 461 __ -_ 09-06-12 A 48 Flagstafl'Lake’ AtClifTHouse nearpump f %m:£%nl¥itx1820 t 5 33 3,; 23 (. 511K) 2:;(15 8 1‘1; 3.9 0.4 1113 29 323 20 10 3231 305 6.9 03-09-76 B 533 1.1mm _ -— . mc. .. __ .0 __ .. 230 ._ -. 09-05-12 C 51,3 Hmmfi 33311231)“ 15 52 30 92 16 518 43 94 62 2.1 35 __ __ __ _, 3737 1190 8.8 04.27.61 c 5cS HartLake 36/24—14 11 14 6.5 9.6 2.9 99 0 1.6 1.3 0.1 34 0 10 60 10 117 170 7.8 03-10-76 B 633 HoneyCreek 36/24—29bbab 22 19 8.1 12 3.4 126 0 _. 2.2 __ 22 n __ __ ,- 3135 204 7_9 08-21-56 0 6bS HoneyCi-eek 36/24-ZQa 3 11 4.9 17 2.8 87 0 7.7 4.5 0.3 26 3 60 200 10 125 165 7.1 03-10—76 B ’I‘wentymileCreek 40/23-25bcad Sources of data: A. Van Winkle (1914). B. E. A. Sammel and N. E. Voegtly. C. Phillips and Van Denburgh (1971). ‘ Sequence numbers of surface waters are identified with an "S" symbol. ’Not In area covered bythis re rt. Data has been included for com arison. 3Estunated by summation of C a x 0.4917 and other ions listed ( em, 1970). 116 THE GEOTHERMAL HYDROLOGY 0F WARNER VALLEY, OREGON TABLE 6.—Chemical analyses of waterfrom wells and springs [Concentrations in milligrams per liter unless otherwise indicated] A. E.A. Sammel and N. E. Voegtly. B. R. H. Mariner, and others (1974). C. R. H. Mariner, and others (1975). D. Trauger(1950). E. E.A. ammel and R.W.Craig. ‘Estirnated by summation ofI'ICOa X 0.4917 and other ions listed (Hem, 1970). Within the valley, most nonthermal waters have temperatures in the range 10°C to 16°C. Their contact with the basin sediments is reflected in concentrations of major ions that are high relative to the cold spring waters from the highlands. Sodium and bicarbonate are the predominant ions. The chemical contents of the colder waters of the Warner Valley area correspond to those found in wa- ters throughout the Lava Plateau region, as documented by Klein and Koenig (1977). According to these authors, cold spring waters having temperatures less than the mean annual air temperature are charac- terized by sodium, calcium, and magnesium concen-' trations which are generally less than 10 mg L‘1 each. Sulfate and chloride concentrations are generally less than 2 mg L". Bicarbonate concentrations range from 41 to 66 mg L”. Silica concentrations are generally less than 30 mg L“, but range to more than 40 mg L”. The chemical similarity between the colder waters of Warner Valley and the regional pattern is probably the result of two factors: the region’s widespread volcanic rocks which provide a generally similar chemical basis for water-rock interactions, and the relatively short res- idence times of the waters in the rocks which minimizes the effect of the water-rock reactions. In the discussion that follows and throughout the remainder of this report, the sequence number as- analysis is used interchangeably to refer to both the _. sample and to the spring or well from which it came. Locations and other data for these springs and wells signed to each ground-water sample taken for chemical ‘ Tem- . S cific S Seq 0 “er few Ca Mg Na K Hco. con so.1 c1 F Sio. As B Fe Li Mn Dissolved gamma pH Date ”f,“ No: orvriame Location (81;;3 (#ng) (”‘ng ‘) (“31‘7” (”31‘8” (”5147!) SOhds (#th/Cm’ '1 sampled D81 Waters having surface temperatures 220°C 12 CharlesCrump 38/24-34ccad1 93 12 0.1 270 12 137 0 190 240 5.3 180 1200 15000 10 __ 0 1070 1500 7.9 03-15—76 A 13 CharlesCrump 38/24-34ccad2 93 14 0.8 280 12 140 0 200 250 4.9 190 1200 14000 10 __ 220 1060 1480 7.6 03-15-76 A 44 MC Ranch (spg) 38/24-34cdbd 78 16 0.2 280 11 153 <1 200 240 4.9 180 H 13600 ,_ 400 __ 11020 1490 7.26 08-03-72 B 14 Fisher Hot Spring 38/25-10bbbb 68 8.4 1.0 92 7.9 105 1 59 56 3.5 77 100 2200 <20 40 <20 '360 513 7.93 08-02-72 B,C 34 Hugh Cahill (spg) 40/24-29bcbb 66 2.1 0.4 200 4.0 209 0 120 85 12 98 390 2400 100 160 0 659 945 7.2 04-05-76 A 46 Hen O’Keefe (Sp ) 39/24-03dcbc 58 12 0.5 270 12 158 0 180 230 6.5 180 1 100 25000 <10 320 40 '995 1409 7.9 07-19-80 E 1 Us. ish& Wildli e 35/26-32baba 40 10 2.5 191 13 376 __ 57 64 3.6 168 q 1500 20 ,_ __ 695 876 8.3 08-10-48 D (Antelope Hot Spring) 11 Charles Crump(spg) 38/24-27cddb 40 18 2.0 175 8.7 130 ._ 116 150 1.9 125 __ 7300 40 __ __ 662 935 8.7 09-16-48 D 16 U.S.BLM (Sp ) 38/25218de 38 5.2 0.3 88 5.6 108 0 48 42 2.2 58 150 1600 0 __ 0 297 440 7.7 03-17-76 A 23 Lonn Schad er 39/24—29cdcc 33 5.1 2.5 54 20 1 17 0 37 25 0.5 97 3 1 660 100 . _ 0 300 366 7.7 03-28-76 A 10 John ane(spg) 38/24—10dadd 29 10 3.6 19 5.6 93 0 3.8 2.8 0.4 39 12 70 0 L 0 132 162 7.4 04-18-76 A 7 Dixon (spg) 37/25-30dccb 27 8.1 3.2 61 9.2 104 0 35 35 0.6 52 95 1100 30 H 0 268 370 7.3 04—20-76 A 43 Lonny Schadler 4730—31‘abbc 27 2.7 0.4 36 9.9 87 0 13 7.9 0.4 78 6 150 20 L 0 189 207 8.2 04-29-76 A ( eva a) 30 U.S.BLM 39/25-283dad 25.5 9.5 2.2 32 10 92 0 15 15 0.3 58 19 680 70 _ _ 10 185 242 7.2 04-25-76 A (Sand Well) 31 DonRo inson(spg) 40/22-22babc 23 4.3 0.4 30 0.5 87 0 2.1 1.4 0.3 31 3 110 0 _ 0 120 151 7.5 04-28-76 A 41 MC-2 (606'-696') 40/24-17cacd 22.5 3.0 0.4 78 8.9 157 0 23 24 1.9 90 100 1000 310 10 10 318 399 6.8 08-25-76 A 32 JamesD ke 40/23-24ddbb 21 7.7 3.2 140 12 209 0 85 54 0 66 230 370 0 ,v 0 477 721 7.6 04-27-76 A 33 Foskett pring 40/24—25caaa 21 7.6 319 42 5.7 111 0 19 13 0.7 62 85 340 10 _ 0 207 270 7.7 04-05-76 A 40 MC-2 (492-506,) 40/24—17cacd 21 2.7 0.4 77 8.4 1 18 0 15 22 2.0 82 100 1 100 330 10 10 298 383 7.2 08-26-76 A 4 Philli L nch 36/25-7cddh 20.5 3.5 2.1 59 4.3 160 0 2.5 8.3 0.8 68 12 10 20 , . 70 206 284 7.8 04-28-76 A 6 Glen er y 37/25-20ddda 20 15 6.3 14 4.4 101 0 5.2 3.1 0.1 37 3 40 0 _. 0 125 188 7.4 04—21-76 A Waters having surface temperatures <20°C 15 Glen Perk 38/25—168aab 18 61 37 170 26 627 O 140 27 0.7 63 48 4800 0 0 10 830 1260 7.6 03-17-76 A 20 Hen 0’ eefe 39/24-16dabc 17 1 10 66 45 12 267 0 350 23 0.2 56 12 460 10 0 10 850 1130 7.3 03-16-76 A 24 John ane 39/24-29dbab 17 19 9.5 35 12 148 0 33 13 0.1 52 1 1 800 300 . _ 30 238 354 7.4 05-05-76 A 39 MC-2 (342'-357') 40/24-17cacd 16 1 1 5.4 140 7.6 397 0 2.8 34 0.7 57 8 1200 170 0 240 452 7 10 7.2 08-28-76 A 19 MC-l 39/24-1 lddbb 15 2.5 0.6 70 7.7 156 0 14 18 1.7 59 200 1600 60 0 40 245 344 7.5 08-02-76 A 2 Cook Laird 36/24-4baaa 14 11 7.3 41 7.0 136 0 17 14 0.4 53 18 4 0 ._ 0 274 308 7.7 04-27-76 A 8 Glen Perky 37/25—33bdcd 14 14 6.1 20 6.0 114 0 6.7 5.4 0.1 44 5 70 30 __ 20 150 219 718 01-21-76 A 17 Te Cahill (spg) 38/25-30dddd 14 717 4.2 64 7.9 139 0 26 25 0.5 56 50 980 240 _ - 0 254 366 8.0 05-05-76 A 26 Mflanch 39/24-32baad 14 1 1 7.9 44 7.9 174 0 1.2 21 0.4 56 31 380 190 , V 340 230 349 7.7 03-28-76 A 22 Henry O‘Keefe 39/24-26baad 13.5 3.1 2.0 33 6.5 111 0 2.1 2.1 0.3 53 19 250 80 _, 50 154 190 7.4 04-20-76 A 25 MC Ranch (spg) 39/24-31babd 13.5 18 6.8 12 7.3 117 0 3.8 3.0 0.2 63 1 30 0 ,_ 0 177 212 7.7 03-29-76 A 21 Adel School 39/24—21cdbb 13 15 8.9 15 5.3 116 0 14 4.8 0.3 53 2 20 20 .1 0 171 232 7.1 03-29-76 A 28 MC Ranch 39/24-34cbaa 12 515 3.8 46 4.7 151 0 1.1 9.0 0.4 55 75 250 150 M 140 200 267 7.9 03-29-76 A 10 5.9 5.5 46 5.6 156 __ .7 10 0.8 65 .. 40 350 _. __ 217 266 8.5 08-09-48 D 35 U.S.BLM 40/25-4dcba 12 40 17 26 5.4 218 0 22 16 0.1 64 8 340 130 A- 220 300 445 7.2 05-05-76 A (MC—Well) 45 U.S.Fish& Wildlife 35/26—28bccc 12 12 4.1 75 11 170 0 30 29 1.4 66 1 650 70 30 6 ’313 448 7.3 07-15-80 E (Valet Spring) 38 Hugh Cahill (spg) 41/24-1dcda 11 2.3 14 39 5.6 160 0 33 28 0.5 54 12 160 10 .1 20 280 400 7.4 04-07-76 A 42 Gravely Spring 47/18-1332dd 10 12 3.8 19 2.3 59 0 17 15 0.1 28 0 180 60 L 10 127 194 7.0 04-07-76 A (Neva ) 27 Te Cahill 39/24—34abbb 9 5 5.7 4.3 39 4.6 137 0 1.5 5.8 0.2 51 95 160 330 W 260 177 220 718 04-22-76 A 37 U883LM 40/25-36dcac 8 5 9.8 5.5 10 1.3 0 7.2 5.0 0.2 31 1 50 110 __ 0 115 130 6.8 04-08-76 A (Rosebriar Spg) 36 U.S.BLM 40/25-23bbbd 7 8.5 4.0 11 4.7 66 0 5.6 3.9 0.1 22 1 80 340 0 122 130 6.8 04—18-76 A (Chukar Spg) 9 U.S.BLM 38/24-6dbad 6 9.8 4.4 517 0.7 56 0 1.9 1.0 0.1 27 0 30 10 __ 0 91 110 6.6 04-18-76 A (Coxes Spg) 18 Charles Cnimp(spg) 39/24-4badc 6 7.4 2.0 4.8 4.2 41 0 3.0 0.9 0.1 43 0 90 0 L 0 92 95 6.9 03-16-76 A 5 U.S.Fish & Wildlife 36/25-9cdcd 4 6.4 2.6 6.1 0.9 43 0 2.0 1.1 0.1 25 1 250 230 ,, 0 60 77 7.2 05-03-76 A (spg) Sources of data: QUALITY OF WATER 117 may be found by referring to the sequence numbers in tables 3 and 4. Thermal waters in the study area range from sodium bicarbonate waters at low temperatures to sodium chloride waters at higher temperatures. The hotter waters contain the highest measured concentrations of sodium, chloride, sulfate, and silica. Three samples of this type of water were obtained in the Crump geyser area, located on the west side of the valley north of Adel. Samples 12, from a 121-foot well, and 44, from a nearby spring, have temperatures of 93°C and 78°C respectively. The waters have virtually identical chem- ical compositions and are clearly derived from the same parent water (pl. 1 and fig. 4). The third sample, number 13, is from a 1,684-foot well located only 140 feet from well 12. Most of the constituents in sample 13 have slightly higher concentrations relative to sample 12. Boiling observed at the water surface in well 13 may have concentrated the dissolved constituents. 6L «’0 0 Spring 920°C 0 Spring <20°C + Well >20°C + Well <20°C FIGURE 4.—Piper diagram of chemical relations in water from wells and springs: percentage reacting values. Sample 11, from a spring located approximately a mile north of Crump geyser, is also a sodium chloride water. Chemically it is a dilute version of the Crump geyser waters and is clearly related. In the remaining thermal waters, with temperatures in the range of 20°C to 68°C, sodium and chloride gen- erally increase with respect to calcium and bicarbonate as temperatures increase (table 7). The increasing sodium and chloride concentrations probably represent increased proportions of thermal water. The two hottest waters in the latter group are Fisher Hot Spring (sample 14), located about 4 miles north of Crump geyser on the east side of the valley, which has a surface temperature of 68°C, and a large seep at the south end of Warner Valley (sample 34), with a tem- perature of 66°C. Cooler waters of this type are scat- tered throughout Warner Valley and Coleman Valley. The relationship between these waters and the waters of the Crump geyser area is discussed in subsequent sections of this report. Concentrations of fluoride are high in most of the thermal waters, ranging from 1.9 to 12 mg L~1 in wa- ters warmer than 40°C. The fluoride concentrations vary greatly and are not closely correlated with con- centrations of more stable ions such as chloride (fig. 5). The high concentration of fluoride in sample 34, 12 mg L“, may be related to the unusually low concentration of calcium in this water, 2.1 mg L", which is the lowest concentration observed in a ground-water sample in Warner Valley. The calcium concentration is approxi- mately the amount required for equilibration with the mineral fluorite (Can) which is commonly found in volcanic terranes (Hem, 1970). This relationship is not typical of thermal waters in Warner Valley however. Sample 20, obtained at 17°C from a 360-foot well located about 1/2 mile south of Crump geyser, may have an admixture of thermal water but is anomalous with respect to its high concentrations of calcium, mag- nesium, and sulfate (fig. 4). The specific conductance, 1,130 micromhos per centimeter ()1.th cm“), is excep- tionally high for a low-temperature water. Silica and chloride concentrations are useful in some areas for determining possible mixing trends in ther- mal waters. (See Sammel, 1980.) Chloride concen- trations acquired at high temperatures in geothermal reservoirs tend to remain constant as the waters as- cend to the surface unless the water has boiled or is diluted by shallow ground water. If the cooling waters do not become saturated with respect to amorphous silica, the silica concentrations also may remain rather stable. Thus, if variations in these two constituents can be attributed solely to mixing, they may be used to determine the amount of mixing that has occurred. A graph of chloride concentration versus tempera- ture in the thermal waters (fig. 6) shows that chloride 118 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON TABLE 7.—Molal ratios of selected dissolved constituents inground water Setitlience Temp- 0. Owner or name ”(23m VCa/Na Na/K Ca/Mg Cl/Ca Cl/Mg Cl/Na Cl/SO, Cl/B Cl/(HC03+CO;.) Waters having Surface temperatures 220°C 12 Charles Crump 93 1.5 38.3 72.8 31.6 2552 0.58 3.69 4.88 4.39 13 Charles Crump 88 . 1.5 39.7 10.6 20.2 326 .59 3.61 5.45 3.53 44 MC Ranch (spg) 78 1.6 43.3 48.5 17.0 823 .56 3.25 5.50 2.70 14 Fisher Hot Spring 68 3.6 19.8 5.18 8.86 45.9 .40 2.71 7.76 1.00 34 Hugh Cahill (spg) 66 .8 85.0 3.42 56 194 .28 2.00 10.8 .71 46 Hen O’Keefe (spg) 58 1.5 38.3 14.6 21.7 315 .55 3.46 2.81 2.51 1 Ante ope Hot Spring 40 1.9 25.0 2.43 7.24 17.6 .22 3.04 15.2 .29 11 Charles Crump (spg) 40 2.8 34.2 5.46 19.4 51.4 .56 3.50 6.27 1.99 16 U S BLM (spg) 38 3.0 26.7 10.6 9.13 105 .31 2.45 8.00 .68 23 Lonn Schadler 33 4.8 4.6 1.23 5.54 6.86 .30 1.90 11.5 .38 10 John ane (spg) 29 19.1 5.8 1.69 0.32 .55 .10 2.11 12.2 .05 7 Dixon (spg) 27 5.4 11.3 1.53 4.88 7.91 .37 2.84 9.70 .58 43 Lonny Schadler 27 * 5.2 6.2 4.08 3.31 14.1 14 1.68 16.1 .16 30 USBLM (Sandy Well) 25.5 11.1 5.4 2.62 1.84 4.83 30 2.84 6.73 .28 31 Don Robinson (spg) 23 7.9 102.1 6.53 .37 2.44 03 1.85 3.88 .03 41 MC-2 22.5 2.5 14.9 4.55 9.04 41.1 20 2.83 ~10 26 32 JamesD ke 21 2.3 19.8 1.72 7.93 1.16 25 1.72 44.5 .45 33 Foskett pring 21 7.5 12.5 1.18 1.93 2.37 20 1.94 11.7 .20 40 MC-2 21 2.5 15.6 4.09 9.21 37.7 19 3.97 ~5 .23 4 Philli L nch 20.5 3.6 2.3 1.01 2.68 2.71 .09 9.26 253 .09 6 Glen er y 20 31.8 5.4 1.45 .24 .34 .14 1.74 23.6 .05 Waters having surface temperatures <20°C 15 Glen Perky 18 5.3 11. 1.00 0.57 0.57 0.10 0.62 1.72 0.08 20 ‘ Henry O’Keefe 17 26.8 6.4 9.76 2.87 2.81 3.35 .23 15.2 .15 24 JohnLane 17 14.3 5.0 1.21 .81 .98 .24 1.16 4.96 .15 39 MC-2 16 2.7 31.3 1.24 3.49 4.32 .16 34.9 ~10 .15 19 MC-l 15 2.6 15.5 2.54 8.42 2.14 .17 3.56 3.43 .20 2 Cook Laird 14 9.3 10.0 .91 1.49 1.36 .22 2.36 1067 .18 8 Glen Perky 14 21.5 5.7 1.39 .44 . .62 . .18 2.32 23.5 .08 17 Te Cahill (spg) 14 5.0 13.8 1.11 3.84 4.26 25 2.72 7.78 .32 26 MET-finch 14 8.7 9.5 .85 2.21 1.87 .31 50.3 16.9 .21 22 Henry O’Keefe 13.5 6.1 8.6 .94 .78 .73 .04 2.77 2.56 .03 25 ' MC Ranch 13.5 40.6 2.8 1.61 .19 .31- .16 2.30 30.5 .04 21 Adel School 13 29.7 4.8 1.02 .37 .38 .21 1.00 73.1 .07 28 MC Ranch 12 5.9 16.6 .88 1.89 1.67 .13 2.29 11.0 .10 35 U S BLM (MC well) 12 27.9 8.2 1.43 .47 .67 .40 2.24 14.4 .13 45 Valet Spring 12 5.3 11.6 1.78 2.73 4.85 .25 2.62 13.6 .29 38 Hugh Cahil (spg) 11 14.1 11.8 .99 1.44 1.43 .47 2.52 53.3 .30 42 Gravely Spring 10 20.9 14.1 1.91 1.41 2.77 5.13 2.51 25.4 .44 27 Te Cahill 9.5 7.0 14.4 .81 1.17 .95 .10 10.8 11.0 .07 37 U S LM (Rosebriar Spg) 8.5 35.9 13. 1 1.08 .58 .63 .32 1.97 30.5 .14 36 USBLM (Chukar Spg) 7 30.4 4.0 1.29 .52 .68 .23 1.96 14.9 .10 9 USBLM (CoxesSpg) 6 63.1 13.8 1.35 .12 .16 .11 1.49 10.2 .03 18 Charles Crum (s g) 6 65.1 1.9 2.25 .14 .31 .12 .84 3.05 .04 5 U S Fish & Wi dli?e (spg) 4 47.6 11.5 1.50 .20 .29 .12 1.53 1.34 .04 generally increases with increasing temperature. Be- cause chloride concentrations are fairly uniform in the colder waters (represented by the triangular point in fig. 6) and in the hotter waters (samples 12, 13 and 44), mixing of waters from these two groups should produce a linear trend for waters of intermediate composition. The scatter in the data suggests, however, that factors other than simple mixing of hot and cold waters are important in determining the temperatures of many of the low-temperature waters. Conductive heat loss may explain some of the temperature differences, and this mechanism is invoked in a subsequent section of the report (“Temperatures in the Geothermal Reservoir”) in order to account for the observed temperatures of Fisher Hot Spring and several other waters. Silica concentrations in the thermal waters gen- erally increase with increasing chloride concentration, and a graph of silica versus chloride (fig. 7) suggests that both of these constituents might be used with some confidence in mixing models. A dilution trend is clearly indicated by line A in figure 7. Trend A con- tains all but two of the thermal waters having temper- atures greater than 30°C. The exceptions are samples 1 and 23. Sample 1 (Antelope Hot Spring) has an excep- tionally large silica concentration in relation to its temperature and chloride concentration. This spring occurs at a much higher altitude than other thermal springs (6,000 feet versus 4,500 feet for springs on the valley floor), and its relationship to other thermal wa- ters is uncertain. Sample 23 was obtained from a rarely used well of unknown depth located near the west val- ley wall south of Adel. The silica concentration is anomalously high in relation to the dissolved-solids concentration and concentrations of other major ions. Exceptions to the trend of line A at temperatures below 30°C are samples from wells 4, 43, and MC2 (samples 40 and 41). These waters, with samples 1 and 23, may define a separate mixing trend, line B in figure QUALITY OF WATER 7, or, more probably, may have increased their silica concentrations by dissolving amorphous silica from the valley-fill deposits or bedrock. The low-concentration end point in figure 7, shown by the open triangle, represents the average concen- 14 I I | I I I: E ' I nd h d _ ._ Basedorisiicaa clorie __ -' 12 .34 relationships shown else- a; where; included for E reference only a) A Average of 14 writers 2 1° _ with temperatures <17°C _ < - O a: - Sprlng,r20 C ‘3 + Well,r20°C .1 :' E 8 — _ E i 2 ms 2 6— — E /,+12 2 13 3 d‘ 44. i g 4 —- 9 fig / __ u 14- .1 mud) / LU ed / E Sufi / a: .16 /P‘5/ g 1’ “Hm / ~11 - u. / / 31 ’4 23 .33! .7 o .6 3° IE? I I | | O 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGBAMS PER LITER FIGURE 5.—Graph of chloride concentration versus fluoride concentra- tion in ground water. The assumed mining trend is based on silica and chloride relationships shown elsewhere and is included for' reference only. 119 trations of silica and chloride in 14 low-temperature waters from Warner Valley. The average concen- trations of silica, chloride, and dissolved solids in these waters fit well in mixing models described in a sequent section of this report. Hence, these average concen- trations are assumed to be close to actual concen- trations in the recharge waters of the area. The 14 waters selected do not include 6 dilute cold waters from springs located on raised fault blocks adjacent to the valley. Although the waters of these springs may be typical of upland recharge sources, their chemical compositions generally do not fit the geothermal mix- ing models as well as do the compositions of waters that have circulated more deeply and have acquired slightly higher temperatures and concentrations of dissolved solids. The implications of this fact are dis- cussed further in the section entitled “Temperatures in the Geothermal Reservoir.” A further evaluation of the usefulness and reliability of the silica and chloride concentrations as geothermal indicators may be made with the help of graphs such as those shown in figures 8 and 9. In these graphs, concen- trations of chloride, silica, and bicarbonate are plotted as ratios versus the chloride concentration. As com- pared with simple two-parameter graphs, such as figures 5, 6, and 7, ratio plots may convey additional information insofar as the direction and magnitude of changes in the ratio of the two parameters are significant. As in figures 5, 6, and 7, separate data points are given for each thermal water (220°C) and a single point represents a probable average low- temperature recharge water. In both graphs, a dashed a; 100 I I I I I 2 12+ M3 m d u 80 — .44 T :0 Lu g '1‘ .34 3 6° ‘ ~46 ‘ D 5 LI; 4o _ '16 .1 .11 ._ a: 3 1043 +23 2 {($157 A Average of 14 waters with m 20 :2 .40 Q32 temperatures <17°c _ g: 4 33 . Spring,z20°C E o Well,r20°C '- 0 I I I I I 0 50 100 150 200 250 300 250 I I I I I ~ I: 200 13 - E g P. ‘2 53M '2 -' .1 % 44 a: S (‘D/ l- m 150 — '- z 0: Lu U s / .11 g g A Average of 14 waters with 0 o :3 100 _ 23*/ 34./ temperatures <17 C _ < -‘ 41+ G Hypothetical reservoirs for the 2 2‘ +4?“ .14/ thermal component of spring :' 5 +33 +32 14 and well 12 m E 50 _ 33° ”.16 /Possible mixing trend _ ‘10 / Based onwell 12-,® \31 based on spring 1 I Spring,:20°c 9 Well,r20°C o I l l I | 0 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PEH LITEH FIGURE 6.—Graph of chloride concentration versus temperature in ground water. CHLORIDE CONCENTRATION, IN MILLIGRAMS PER LITER FIGURE 7.—Graph of chloride concentration versus silica concentration in ground water and hypothetical reservoir water. 120 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON line indicates the curve that would be followed if the average cold recharge water were mixed in all pro- portions with a hot water represented either by sample 12 (fig. 9), or by a hypothetical reservoir water (fig. 8). The concentrations of silica and chloride in the reser- voir water were estimated by means of silica-enthalpy and chloride-enthalpy diagrams discussed in the sec— tion of this report entitled “Temperatures in the Geothermal Reservoir.” The graph of figure 8 shows that the ratio of chloride to silica is much lower in the presumed recharge water than in the hotter geothermal waters. This is the result of the relatively high solubility and the general avail— ability of amorphous silica at shallow depths in vol- canic terranes. Chloride, on the other hand, is not read- ily available to the shallow ground waters of Warner Valley, and concentrations greater than about 20 mg L‘1 probably indicate the presence of a thermal-water component. In the thermal-reservoir water, the high ratio of chloride to silica reflects the limiting of silica concen- trations by the temperature-dependent solubility of quartz (Fournier, 1973), whereas the chloride concen- tration is limited only by the availability of chloride- bearing minerals and residence time. The high concen- trations of chloride in the hotter waters indicate that "6 l l I l l 1.4 — _ 13 1‘2: 2 0 //'4e 44 1.2 — .11 / ._ a: // in; 1 0 _ / —I o 9‘ 134 I- 0 8 _ +32/ < - __ E .154: I '7 (D I: 0.6 — / . II _ g 0 Spring :20 C / + Well 220°C 0_4 _ / .1 A Average of 14 waterg with _ temperatures <17 C 40 41 G) Hypothetical resecvoir water 23 (see ”Temperatures intne 0-2 _‘\33 30 geothermal reservoir”) _ 1/1/43 —— Hypothetical mixing curve (see text explanation) t—G k 10 o 31 I I I I I O 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER LITER FIGURE 8.—Graph of the weight ratio (Cl/SiOz) versus chloride concen- tration in ground water. chloride is available to the thermal waters and that the residence times are fairly long. The reasonably close fit of the intermediate data points to the hypothetical mixing curve in figure 8 in- dicates that simple mixing of reservoir water with re- charge water could account for the observed silica and chloride concentrations in most of the moderate- temperature waters. The fit would be slightly better if the water of well 12 were used as the high-temperature end point, but the small difference may be due to ana- lytical imprecision and is probably not significant. At the steeply-sloped low—chloride end of the curve, some scatter océurs which is partly masked by the relative independence of the chloride-silica ratio from the chloride concentration. Only the data point for spring 1 (Antelope Hot Spring) is clearly anomalous, however, and there is little suggestion of the double mixing trends that appear when the silica and chloride concen- trations are plotted as in figure 7. The ratio of equivalents chloride to equivalents (H003 + CO3) can be useful in distinguishing waters of differing thermal aquifers (Fournier and Truesdell, 1970; White, 1970). Concentrations of (HCO3 + CO3) in thermal waters may be established by temperature- dependent reactions involving the partial pressure of 002 gas and sodium or potassium silicates, all of which commonly occur in hydrothermal environments. To the extent that these constituents are uniform within an aquifer and vary between aquifers, the equivalent Cl/HCOs + 00,) ratio will be distinctive for a given aquifer and may be retained for some time in the near-surface environment (Fournier and Truesdell, 1970). Mixing of thermal water with shallow water 2.2 I I I I I 2.0 _ 0 Spring =20°C _ 1.8 — + Well 220°C 12 +13 o" A Average of 14 waters with temperatures are /+ u 1.6 - ___ Hypothetical mixing curve (see text // .42344 - g 1_4 _ for explanation) / — U / . _ / _ E 1.2 ~11/ / g 1.0 I- / _ :l: 0.8 - _ ‘2 // "u 0.6 — / _ g 6 10 31 4 / ~ - ’ ’ . / 34 _ 0 4 .7/16/ , 33 +32 0‘2 7 1513427340 .1 _ +\ 0 0 +4 43 i I I I I 0 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER LITER FIGURE 9.——Graph of the weight ratio (Cl/£1003) versus chloride concen- tration in ground water. QUALITY OF WATER 121 having a different Cl/HCO3 + 003) ratio will obviously result in ratios intermediate between the two. For the Warner Valley waters, the atomic ratio Cl/ HC03, is used in figure 9 in place of the equivalent Cl/HC03 + CO3) ratio because only two of the waters contain 003 and the concentrations of this ion are 1 mg L“ or less. The graphical relations are virtually iden— tical to those that would be shown by the equivalent Cl/(HCO3 + CO3) ratio. The change in the Cl/HCOs ratio shown in figure 9 between the high-chloride (hot) waters and the low- chloride (cooler) waters clearly indicates that the two extreme groups of waters represent differing chemical regimes. As is true for the chloride-silica relation (fig. 8) the most reasonable explanation for the distribution of intermediate data points appears to be mixing of the end-member waters. The hypothetical mixing curve based on sample 12 as the high-chloride end point probably accounts fairly well for the composition of spring 11, a spring of moderate temperature located 1 mile north of Crump geyser. The scatter of data points near the low-chloride end of the curve may be the normal effect of a hetero- geneous environment such as Warner Valley. It seems equally probable, however, that some of the departures from values predicted by the mixing curve represent differing mixing trends. It may be significant, for example, that three warm springs located near the north end of Crump Lake (samples 7, 14, and 16) have uniformly higher bicarbonate concentrations than pre- dicted, whereas a well and a spring located in the south end of Warner Valley (samples 32 and 34) have lower than expected bicarbonate concentrations. A sixth sample (1), from Antelope Hot Spring, has a position with respect to the mixing curve that is similar to its position in figure 9 and which seems clearly anomal- ous. In the absence of additional high-chloride waters that might help to establish possible origins in sepa- rate reservoirs, no firm conclusions regarding precise mixing trends can be reached on the basis of figures 8 and 9. Additional evidence is discussed in the section of the report entitled "Temperatures in the Geothermal Reservoir.” Several general conclusions about the origin and na- ture of the thermal waters in Warner Valley may be drawn from the chemical data. It seems certain, for example, that the thermal waters originate in a hot- water system rather than a vapor-dominated system. The occurrence of high concentrations of alkali chlorides, silica, boron, and arsenic in waters of near- neutral pH support this conclusion (White and others, 1971). Ratios of major ions (shown in table 7) also sup- port the concept of a hot-water system. Fairly high ratios of Ca/Mg, Cl/Mg, and Cl/(HCO3 + 003) and low ratios of Ca/Na in the hotter waters are typical of hy- drothermal systems having moderately high reservoir temperatures (White, 1970; Mahon, 1970; Ellis, 1970). Adiabatic cooling of the hot waters as they ascend to the surface would not greatly change the relations among major ions such as silica, chloride, sodium, and sulfate. Extensive sinter deposits observed in two thermal discharge areas north of Adel (pl. 1) indicate that the thermal waters have had temperatures that may have been as high as 180°C (White and others, 1971). Field observations indicate that silica is still being deposited in these areas. Finally, all thermal waters in Warner Valley are assumed to be mixtures of hot waters and cooler re- charge waters. This conclusion is not well substan— tiated by the data presented thus far, but it is sup- ported by the graphs of temperature versus chloride, chloride versus silica, and chloride versus bicarbonate (figs. 6—9). ISOTOPES IN GROUND WATER Twenty-nine samples of water from the Warner Val- ley area were analyzed for stable isotopes of oxygen (“‘0) and hydrogen (deuterium). Samples were taken from a wide range of waters in the study area in order to provide information on background concentrations as well as concentrations in the thermal waters. Re- sults of the analyses are shown in table 8. The amounts of 18O and deuterium, D, in precipita- tion decrease with altitude and other factors in a man- ner that is approximated by the expression, SD = 86180 + 10, where 8D and 6180 are ratios of these isotopes to ‘H and 160, respectively, expressed in parts per thousand (per mil) as departures from the ratios in standard mean ocean water (SMOW) (Craig, 1961).1 The “meteoric line” resulting from the above expres- sion is shown in figure 10, along with data from the springs and wells in the Warner Valley area. Samples numbered 5, 9, and 37 in figure 10 are from cold springs at high altitudes. The data points are roughly parallel to, but offset from, the meteoric line. The position of sample 18, from a cold, high-altitude spring, and sample 19, a cold water from a depth of approximately 400 feet in a GS heat-flow hole (MCI), are anomalous but are within the limits of analytical error. The hottest waters, samples 12, 13, and 44, show an enrichment in 18O of at least 2 per mil relative to the colder waters. The increase is presumably due to isotope exchange reactions between water and rock ‘For a more complete description of the application of isotope analyses to geothermal investigations see, for example, Coplen (1976). I22 [Concentrations of “‘0 and deuterium in parts per thousan THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON ‘80 l I I l ’V I I Spring 220°C 0 Spring <20°C 15+ -90 __ + WeII220°c _ + Well <20°C :1 Standard error 2 51.0: (10.2%.) o: 1" -100 _ 80: {(21340) _ Z —. 15 and E 35 I 2 sampL/ 5 110 '- _ _ 3 Lu a 0 no 12++13 32+ +2 '16 -120 '1 ° ggfi+41334 _ 30 _ l I l | l 130-17 —16 -15 —14 —13 I\’ ~9 8 OXYGEN"? m PERMIL FIGURE 10.—Graph of 8‘90 versus 8D in ground water as departures, in parts per thousand, from standard mean ocean water (Craig, 1961). TABLE 8.—Isotopes of oxy en (‘80) and hydrcsgen ( deuterium and tritium) in ground water as departures from stan ard mean ocean water (Craig, 1961). Concentrations of tritium in tritium units (TU). One TU = 3H/H = 104'] Se§uence Date mO Deuterium Tritium Temperature 0. Name Sampled (0/00) (0/00) (TU) (°C) 1 Antelope Hot Spring 03-00—77 — 16.45 - 123.45 - __ _ 40 4 Phillip Lynch 04-28-76 — 15.10 — 1 16.85 _ __- 20.5 5 US Fish & Wildlife (spg) 05-03-76 — 14.95 —111.90 ___- 4 8 Glen Perky 04-21-76 —16.15 —119.90 ___- 14 9 Coxes Spring 04-18-76 —15.50 ~115.55 91.5:49 6 10 John Lane (spg) 04-18-76 —15.85 —119.05 ___- 29 12 Charles Crump 03-15-76 — 13.60 — 116.55 52:03 93 13 Charles Crump 03-15-76 —13.55 — 116.75 ___- 93 14 Fisher Hot Spring. 03-00-77 — 14.80 — 117.10 _ __ _ 68 15 Glen Perky 03-17-76 —9.00 —86.90 ___- 18 16 US BLM (spg) 03-17-76 —14.90 —118.25 _-__ 38 18 Charles Crump (spg) 03-16-76 — 16.05 — 117.65 ---_ 6 19 MC-1 08-02-76 —15.55 —112.15 ___ 15 20 Henry O’Keefe 03—16-76 — 14.65 — 112.80 2 9t0.2 17 22 Henry O’Keefe 04-20-76 -—15.10 —-114.45 ___- 13.5 23 Lonn Schadler 03-28-76 —15.15 — 118.65 -_-_ 33 26 MC nch 03-28-76 —15.35 —118.60 __ 14 30 US BLM (Sandy Well) 04-25-76 —15.35 —119.10 0.15:0 2 25.5 31 Don Robinson (spg) 04-28-76 — 15.35 — 115.30 -_ __ 23 32 James D ke 04-27-76 —15.25 —118.55 ___- 21 33 Foskett pring 04-05-76 — 16.05 — 120.85 _ _ _ _ 21 34 Hugh Cahill (spg) 04-05-76 — 15.00 - 119.30 _ _-_ 66 35 US BLM (MC-well) 05-05-76 —15.15 —116.45 __-- 12 37 Rosebriar Spring 04-08-76 - 15.40 — 1 14.50 ___ _ 8.5 39 M -2 08-28-76 — 16.25 — 125.05 ___- 16 40 MC—2 08-26-76 ~ 15.45 — 120.55 ___- 21 41 MC-2 08-25-76 — 15.20 — 119.60 051102 22.5 42 Gravely Spring 04-07-76 — 15.10 — 1 14.45 _ _ _ _ 10 44 MC Ranch (spg) 08-03-72 — 13.28 — 1 15.5 ___ _ 78 DISTRIBUTIONS OF TEMPERATURES AND HEAT FLOW 123 oxygen in a reservoir that is deeper and hotter than the shallow aquifers. The amount of 18O enrichment in these waters may be compared to increases of 1.4 per mil in hot waters at Klamath Falls, Oreg. (Sammel, 1980), and as much as 3—5 per mil in a high- temperature system such as Steamboat Springs, Nev. (White, 1968). Sample 15, with a temperature of 18°C, was obtained in a 90-foot well located on the east side of Warner Valley and is clearly anomalous in relation to other waters in the area. Its position on the graph coincides, however, with a trend for warm waters of the Klamath Hills area, near Klamath Falls, in which chemical and isotopic compositions are, for the most part, similar to those of sample 15 (Sammel, 1980). Low-temperature evaporation has been invoked in order to explain the characteristics of the Klamath Hills waters. This explanation does not satisfactorily account for the chemical nature of sample 15. Concen- trations of dissolved solids and of the major ions, Ca, Mg, Na, K, HC03, in sample 15 are from four to six times the mean concentrations of these ions in 18 sam- ples of water from Warner Valley having temperatures between 10°C and 20°C. This apparent uniformity of change suggests evaporation as a mechanism. How- ever, concentrations of silica and chloride in sample 15 are low, respectively 1.1 and 2 times concentrations in the 18 other waters. As silica and chloride are expected to be conserved during evaporative concentration, this mechanism does not satisfactorily account for their low values. An alternative explanation is that the high concentrations of constituents were acquired by solu- tion of minerals in an atypical zone of the aquifer. This explanation could account for the chloride concentra- tion which would not necessarily be greatly increased by solution of the valley-fill deposits, but the low silica concentration would still be difficult to explain. The question remains unresolved, but the explanation, if obtained, may also account for the reported chemistry of sample 20, from a 109-foot well near Adel, which has the same temperature and nearly the same dissolved solids concentration as sample 15, although greatly dif- fering ratios of major ions. Sample 20 also shows an enrichment in 18O which places it on the trend for sam- ple 15 (fig. 10). Except for samples 15 and 20, the warmer waters of Warner Valley show an enrichment of 180 generally commensurate with their temperatures. Small dis— crepancies, such as those between springs 14, 16, and 34, can probably be readily accounted for by differing mixing proportions of thermal waters with cold waters. Springs 1, 10, 31, and 33, however, with temperatures ranging from 21°C to 40°C, fall relatively close to the meteoric line, suggesting either that the isotopes are not in equilibrium or that the waters have obtained some heat by conduction. The latter possibility is most likely for spring 33, which is located in Coleman Valley in an area of probable high heat flow. Sample 39, ob- tained at a depth of about 350 feet in heat-flow hole MC2, is a cold water that has probably been enriched in 18O, as well as in major ions, by solution of minerals in the valley-fill deposits. The exceptionally low deuterium concentration probably implies that the water originated at high altitude. Concentrations of the radioactive isotope tritium (3H) were determined in five samples (table 8). Tritium is formed naturally by cosmic radiation in the upper atmosphere, and decays, with a half-life of 12.3 years, to helium-3. Concentrations are reported in tritium units (TU) which equal a 3H/H ratio of 10—18, about 3 picocuries per liter. Most tritium now in the hydro- sphere is the result of thermonuclear testing in the period 1953 to 1963. Tritium levels in precipitation in Warner Valley prior to 1953 were probably about 9 TU (T. A. Wyerman, written commun., 1978). According to estimates by Wyerman, levels peaked in 1966 at about 1,000 TU and have declined since then to about 20 TU in 1976. Samples of tritium collected at one point in time (synoptically), as is the case for the Warner Valley samples, cannot be used to determine exact ages of the waters, but they do provide a basis for estimating rela- tive times of residence in the aquifers. The relatively large amount of tritium in sample 9, its low tempera- ture (6°C), and the location of this spring on the valley wall indicate a short residence time, probably on the order of a few years. The other four samples, numbers 12, 20, 30, and 41, have relatively small amounts of tritium which indicate long residence times, on the order of tens of years to hundreds or thousands of years, depending on whether the aquifer is assumed to be recharged by piston flow or complete mixing (Pear- son and Truesdell, 1978). The slightly higher level of tritium in sample 12, from the hot well at Crump geyser, suggests that the water is mixed with a larger proportion of young recharge water than are the sam- ples from deeper wells in basaltic rocks or valley-fill deposits (samples 20, 30, and 41). These relations are used in a subsequent section of this report in one of the mixing models for the waters of the Crump geyser area. (See section entitled “Temperatures in the Geothermal Reservoir.”) DISTRIBUTIONS OF TEMPERATURE AND HEAT FLOW TEMPERATURE OF GROUND WATER Temperatures were measured in springs and wells at 92 sites in the study area. The observed temperatures 124 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON are shown on the location map (pl. 1) and the frequency distribution of temperature is shown in figure 11 for waters having temperatures less than 40°C. Seven samples of hot water, ranging in temperature from 40°C to 93°C are omitted. A majority of the waters sampled had temperatures in the range 10°C to 15°C, or about 1 to 6 degrees above the mean annual air tem- perature at Adel. The temperature distribution shown in figure 11 is biased by the large number of cold springs sampled in the highlands adjacent to the basin and by the large number of cold wells located near the mouth of Deep Creek. It is clear, however, that most shallow ground water in the basin either has not been heated by deep circulation or has not retained a significant amount of crustal heat. The distribution in figure 11 is highly skewed and would appear even more so if the seven hot-water samples were added to the graph. The gradational change from cold meteoric waters to geothermal waters in figure 11 provides no basis for distinguishing between these groups. The chemical na— ture of the water also shows a gradational change (fig. 4) and indicates that many waters having tempera- tures greater than about 16°C have a component de- rived from hot waters of deep circulation. The tempera- ture of water probably provides an approximate indica- tion of the amount of mixing that has occurred, but, as is shown elsewhere in this report, temperature is not a completely reliable indicator. The areal pattern of temperature distribution in ground water (pl. 1) does not clearly define areas of geothermal anomalies, but the data are sufficient to 28 show significant differences between areas. In the area north of Hart Lake, the ground waters sampled within the basin boundaries show only a normal increase in temperature with depth and there is no clear evidence of a geothermal resource. Outside the valley walls, An- telope Hot Spring on Hart Mountain is the only known geothermal occurrence. South of Hart Lake, warm waters occur at shallow depths in several areas: at a fault intersection north of Crump Lake, near an inter- section of faults on the east side of the valley (Fisher Hot Spring area), between fault intersections on the west side of the Valley (Crump geyser area), and at the base of a tilted fault block at the south end of Warner Valley. Several warm springs and wells are scattered at additional isolated locations in the southern part of Warner Valley, Coleman Valley, and Big Valley. The factor common to nearly all these occurrences, close proximity to major faults and, in most cases, to fault intersections, is noteworthy. Concealed faults undoubtedly occur at many places beneath the valley floor, and it seems probable that undiscovered geother- mal waters occur in association with these faults. Subsurface resistivity measurements made by audio-magnetotelluric methods show that a large area near Crump geyser is underlain by rocks having anomalously low resistivities (Gregory and Martinez, 1975). Apparent resistivities as low as 2.5 ohm meters occur at apparent depths between 500 and 1,000 feet and resistivities of 4 ohm meters or less were measured beneath an area covering about a square mile. These data are probably best explained by the assumption that a reservoir of hot water having high electrical 24- -| N O) O l I FREQUENCY G I 1' 4t06 7t09 37m 39 TEMPERATURE, IN DEGREES CELSIUS FIGURE 11.—Frequency of temperatures (<40°C) measured in springs and wells. DISTRIBUTIONS OF TEMPERATURES AND HEAT FLOW 125 conductivity occurs at the indicated depths. Resistivity measurements are not available in the same detail for other sites, and the areal extent of shal- low warm waters elsewhere in the valley cannot be estimated by this method. Temperature profiles and heat~flow data, described below, suggest that at least one other extensive area of the basin contains thermal waters. HEAT FLOW Temperature profiles were obtained in six unused water wells and three test holes (table 3 and figs. 12 and 13). Locations of these wells and test holes are shown in plate 1. The three test holes, MC1, MC2, and 0K1, were completed as heat-flow holes in 1976. Ten additional heat-flow holes were drilled in 1979 by the Geological Survey. Temperature and heat-flow meas- urements obtained in the 10 holes by means of an in- situ probe (Sass and others, 1979) are available only in preliminary form. Data and interpretations from these sites will be released in the near future. 0 0 100 -— —50 200- 300 —— —100 (I) E E a 40° - a 2 E _. z I —. E 500 — — 150 E "‘" a. ‘3 Lu «:1 600 — —200 700 — 800— _ —250 5 10 15 20 25 TEMPERATURE, IN DEGREES CELSIUS FIGURE 12.—Profiles of temperature measured in heat-flow holes. Preliminary interpretation of the thermal data in- dicates that, in the areas north of Hart Lake and in the center of Warner Valley south of Adel, heat flow is about 75 milliwatts per square meter (mW m‘z), which is in the lower part of the normal range for this region of the Basin and Range province (Charles Mase, writ- ten commun., 1980). On the basis of temperature pro- files from holes M01 and 0K1, the average normal thermal gradient in the valley-fill deposits may be in the range 40°C km“1 to 45°C km“. Thus, if tempera- tures near land surface are about 10°C, temperatures in parts of the valley unaffected by recharge or thermal convection should be at least 16°C at 500-foot depths and at least 22°C at 1,000 feet. The minimum depth of circulation required to attain temperatures of 20°C or more would be 730 feet. Thermal gradients and heat flows are apparently higher than normal at several of the 10 recent test sites. These sites include a location 1 mile west of Hart Lake, another about 5 miles north of Crump geyser, and three sites near Fisher Hot Spring. Thermal gra- dients at these sites ranged from 113°C km‘1 to 600°C km‘l, and apparent heat flows ranged from 125 to 689 mW m'2, the highest value occurring at a site a few hundred feet from Fisher Hot Spring (Charles Mase, written commun., 1980). No measurements were ob— tained at the extreme south end of the valley near the site of spring 34, but temperatures reportedly meas- ured in the tilted fault block at this location by a pri- vate exploration company suggest that heat flow may be above normal. 0 o [\G/en‘ Perky 37/25-33 bdcd 1 00 PM [4% \nny Schadler (Nevada) — 36/24-27 bbcc 47/19—25 adcd - 50 m 1. Jack and Joe Flynn atI ‘dj 200 35/24-33 dbcc '_ “- LU 5 i =5 -. '— I f; 300 E D “ — 100 ‘5‘ Hugh Cahi/l 400 4 ”24—7 dbad \lae Flynn 36/24-28 cabd l— 5 I 1 | 1 a 150 009 11 13 15 17 19 TEMPERATURE, IN DEGREES CELSIUS FIGURE 13.—Profi1es of temperature in water wells. 126 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON The high thermal gradients and heat flows measured in the areas described above occur as the result of cir- culation of hot water that is believed to rise in the nearby faults and spread laterally through shallow basaltic rocks and the valley-fill deposits. The extent and depth of the spreading zone in the Crump geyser areas were estimated, on- the basis of AMT meas- urements, to be at least 1 mi2 and 1,000 feet, re- spectively. The heat-flow data obtained near Fisher Hot Spring suggest a zone of comparable size at this location. Elsewhere in the valley, the volumes of shal- low hot water are unknown. The subsurface temperature distribution in the Crump geyser area was investigated further by means of measurements obtained in holes driven or augered to depths of 2 meters. The area covered by these small-diameter holes is shown in figure 14, where gen- eralized lines of equal temperature have been drawn on the data obtained at 2-meter depths. Temperatures obtained at shallower depths showed a large amount of scatter resulting from surface effects and insolation, but the measurements at 2 meters were fairly consis- tent. Many of the holes did not reach the full depth of 2 meters, and for these holes, the temperatures given in figure 14 were extrapolated from shallower meas- urements. The number of data points shown in figure 14 is in- sufficient to define the temperature distribution fully, but the general characteristics of the shallow thermal anomaly in the vicinity of Crump geyser are probably indicated fairly well. The thermal disturbance result- ing from the geologically recent imposition of the two hot wells (wells 12 and 13) appears as a minor pertur- bation in the temperature field. The cause is probably the spreading of hot water that erupts and leaks from the wells. A much larger anomaly appears in the area of siliceous-sinter mounds and hot springs that oc- cupies about 41/2 acres to the east of the wells. The spread of hot water from these sources, and possibly from others that may be concealed, raises tempera- tures to more than 60°C over a large area. The mounds of siliceous sinter that dot the area stand 4 or 5 feet above the surrounding terrane. Springs and seeps occur both at the edges of these mounds and in isolated locations. Strata of sinter were encountered in the temperature holes at depths of about 3 feet at several places where no springs exist at present. The mounds and strata suggest that hydro- thermal activity was formerly greater than at present and that the hydraulic heads in the springs were higher than existing heads. The shallow thermal anomalies in the Crump geyser area appear to be caused by the lateral spread of hot water from seeps, springs, and wells. The large ther- mal anomaly inferred from geophysical measurements is believed to be caused by spreading, at greater depths, of hot water from two or more intersecting faults. Two faults have been mapped as striking N. 48° W. and intersecting the north-trending boundary fault at points located a few hundred yards north and south of the geyser area. A large area of low heat flow and low thermal gra- dients occurs in and adjacent to the former delta of Deep Creek near Adel (Charles Mase, written com- mun., 1980). In this area, ground-water inflow from the canyon of Deep Creek, a probable fault zone, lowers temperatures of ground water in the adjacent valley- fill deposits. Reports of temperatures encountered in the San Juan Oil Co. test well, located 1 mile east of Adel, indicate that this cold-water inflow persists to great depths in the Tertiary rocks beneath the valley- fill deposits. Profiles of temperature measured in six water wells in Warner Valley (fig. 13) show good agreement with data from the heat-flow holes. The correlation is sur- prising because thermal gradients derived from meas- urements in water wells are often unreliable owing to convection in the well or outside the casing. Several of the Warner Valley profiles show large effects of convec- tive flow, most notably in well 36/24—280abd (Flynn), where the thermal gradient is reversed in the upper 150 feet of the well. However, the relatively uniform gradients observed in the lower sections of most pro- files generally confirm the data from the heat-flow holes and suggest that thermal gradients are low at the north end of the valley and are higher than normal at the south end and in Coleman Valley. Gradients ob- served in the water wells range from 36°C km‘1 to 58°C km‘1 in three wells north of Hart Lake, and from 82°C km‘1 to 86°C km‘1 in two wells in Coleman Valley. A sixth well, 37/25—33bdcd, located about 1% miles north of Fisher Hot Spring, had an extremely high gradient in the upper 50 feet which is probably evidence of the convective flow of hot water at shallow depths in this area. TEMPERATURES IN THE GEOTHERMAL RESERVOIR: ESTIMATES FROM MIXING MODELS AND CHEMICAL GEOTHERMOMETERS Temperatures of equilibration were calculated for all waters having surface temperatures of 20°C or more (table 9). The geothermometers used for these calcula- tions are the silica geothermometer of Fournier and Rowe (1966) and Fournier (1973) and the Na-K-Ca geothermometer of Fournier and Truesdell (197 3). The silica geothermometer based on quartz equilibrium is assumed to apply if the apparent equilibration temper- ature is greater than 110°C. Below 110°C, equilibra- tion is assumed to occur with chalcedony. This distinc- tion is made arbitrarily for the Warner Valley waters on the basis of data from Iceland, where chalcedony 127 TEMPERATURES IN THE GEOTHERMAL RESERVOIR All mazE «an _o_u< O... Al - 89 Q 33 E53 .SQSm “ES 323 .239: N go 5.3% m am 33558 EB @2535 gwhafifi Marga .33 “Show 95:0 2: no nuznléa museum 83 003 :25 cm. 9: . 5:5 0 5. 008:» 83 _ wmfi .6930 o - $9 83 - R3)» 89., , x. Lt .V ‘4 . 88 $.0an O . 38 00V 06%. . 83 ES 550". 3.3 . mm 93 8.25% 0v n6 - o 89 Ami 3 $9 3302 2: cm o “Snow one. 323:— .EouoE a .3 596 “a nous—£30 .5 69.5»qu 3.5.2353 .250 .3 0:3 Imbml 25:35 uEnnBcao. E :35 E Coo: >o=a>v :22: .«c 025 \.:.J_H4 ® 250E SEE «I,» \x 33.13 3 3:53.: din—00 means“. 5 0.53.395: “can? 9:305 503 EEuuFH O 03.23 «a V3535 £3200 avenue“. 5 SEEKEB 3.55 9:255. 65...: 3555. DJ anon—ow 3.5» 2535» .«o no“ 3 .322: E .595 Z .3355an 5 :39? £306 .22: N 0. 3.3092»: mo=~a>du30E a («o 592.. a E 33:53 .8 “.9532: Jim—co 3932. E 9:523an 9:32: .29: 232353 30:55 . ZO~H20°C 13+ #12 — . (c) NaAK-Ca geothermometer ’ '3‘“ corrected for Mg 46- : 160 (Fourruer and Potter, 1978) _ E (Chalc) Assumed equilibvation : with chalcedony Lu n- 3 140 — _ E :7: if .1 I: g 120 I——- 23(0) 14“) __ E a (Club); 2 2 MM . +41(c) g 3 100 _ (Chalc) 40(c) + (Chalc) +43“)— c: r: 4(0) 32( ) (Chem (Chalc) “J z +(Chalc) (Ch‘l E " so _ -33(c " °’ +30m .16(c) _ o ‘°""°’ 7(a) (Chalc) (Chalc) g '(Chalc) < _ 5( ) ,10(c) __ g 60 (Chafc) (Chalc) : .31 U: (Chalc) 4o — — l | I I I l l I 40 60 80 100 120 140 160 180 200 Na—K—Ca GEOTHERMOMETER TEMPERATURE, IN DEGREES CELSIUS FIGURE 15. —Gr' ,‘n showing correlation between temperatures estimated '5! the silica and Na—K-Ca geothermometers. TEMPERATURES IN THE GEOTHIERMAL RESERVOIR Magnesium concentrations are low and values of R are correspondingly low for well 12 and springs 44 and 46 in the Crump geyser area; the Na-K-Ca geother- mometer consequently needs little or no correction for these waters. Other waters, such as springs 1, 11, and 14, have larger R values, but are highly mixed waters in which the magnesium concentrations may have been acquired during or after mixing. No correction is used for springs 1 and 11, but the indicated tempera- ture for spring 14 has been corrected in order to bring the Na-K-Ca temperature into agreement with the quartz equilibration temperature of 123°C. This tem- perature fits well in the mixing models discussed be- low. In the remainder of this section, a detailed examina— tion is made of the chemical relations in 10 of the hot- ter waters in order to justify the conclusions regarding the geothermal reservoir. The final conclusions depend largely on the relations between waters from the Crump geyser area and those from the Fisher Hot Spring area. Other thermal waters are fitted into the conceptual model with varying degrees of confidence. I29 CRUMP GEYSER AREA Samples 12, 13, and 44 were obtained in two boiling wells and a spring, respectively, which are located near the western boundary fault about 3 miles north of Adel. Measured surface temperatures are, re- spectively, 93°C, 93°C, and 78°C. The temperature at a depth of 660 feet in well 13 is reported to be 127°C and the maximum temperature in well 12 was 124°C. Tem- peratures in the spring (44) have previously been ob— served to be somewhat higher than the temperature reported here. Chemical analyses of major constituents (table 6) show that the three waters are Virtually iden- tical. The quartz conductive geothermometer indicates that the temperature of equilibration is at least 173°C for the three waters, whereas the Na-K-Ca geother- mometer indicates a temperature close to 150°C. If the quartz conductive geothermometer is accepted as the more reliable of the three silica geothermometers, and if the chloride concentrations and mixing models (dis- cussed below) are accepted as probably correct, the TABLE 9,—Calculated temperatures of equilibration for waters having surface temperatures of 20°C or more Calculated reservoir temperature (“C) Sefiuence Measured Qtz. tz. Chalced. Na/K+B VCa/Naz 0. Owner or name mmfieéature (adiabJ‘ (cond.)‘ (cond.)‘ M correcgted" R‘ 6‘“O(SO4)5 12 Charles Crump 93 163 173 151 152 152 0.9 208 13 Charles Crump 93 165 177 155 149 142 6.1 _-_- 44 MC Ranch (spg) 78 163 173 151 144 144 1.5 202 14 Fisher Hot Spring 68 121 123 95 169 123 11.7 139 34 Hugh Cahill (spg) 66 132 136 109 125 105 13.7 179 46 Henry O’Keefe (s g) 58 163 173 151 152 150 4.3 219 1 US Fish and Wildlife (Antelope Hot Spring) 40 159 169 146 168 87 19.8 125 11 Charles Crump (spg) 40 144 150 125 147 112 12.8 200 16 US BLM (spg) 38 109 110 81 158 148 5.8 ---- 23 Lonny Schadler 33 131 136 109 253 105 21.2 ___- 10 John Lane (spg) 29 93 91 60 777 766 31.6 ---- 7 Dixon (spg) 27 104 104 74 193 59 29.2 ____ 43 Lonny Schadler 27 122 124 96 230 177 7.8 ____ 30 US BLM (Sandy Well) 25.5 109 109 80 233 100 19.9 --__ 31 Don Robinson (spg) 23 84 81 50 729 (3) 12.6 -___ 41 M02 22.5 128 132 104 191 153 8.0 ____ 32 James D ke 21 114 115 87 177 62 27.6 -_-_ 33 Foskett pring 21 112 112 83 792 48 37.9 -___ 40 MC-2 21 124 127 99 189 149 8.6 ____ 4 Philli L nch 20.5 116 117 88 161 36 37.8 ____ 6 Glen er y 20 91 88 58 758 (3) 37.6 __-_ lFourniel' (1977). ”B = *3 except where noted (Fournier and Truesdell, 1974). ”Fournier and Potter (1978). _ M8 Mg+Ca+K sMcKenzie and Truesdell (1977). “Nehring and others (1979). ' 78 = 4/3 (Fournier and Truesdell, 1974). ‘R x 100; units of concentration are equivalents (Foumier and Potter, 1978). 8UFormula not applicable: - estimated Na-K-Ca temperature less than 70°C (Fournier and Potter, 1978). I30 lower temperatures indicated by the Na-K-Ca geother- mometer can be explained as the result of water-rock reactions that occur as the waters ascend from the res- ervoir and cool. The likelihood of this explanation is supported by data from two of the samples, well 12 and spring 44, in which a difference of 8°C between the two Na-K-Ca temperatures results from minor chemical changes that occur in the spring as the water rises and cools 15°C. The fairly high reservoir temperature indicated by the silica geothermometer is supported by the observed shift in the 8180 ratio (fig. 7) of about 2 per mil relative to values in cold regional waters. The amount of in- crease suggests equilibration at fairly high tempera- tures and is compatible with the estimated tempera- ture of 175°C. High rates of chloride to calcium, magnesium, sodium, sulfate, and total carbonate species are also compatible with relatively long residence times at high temperatures (table 7). The ratio of sodium to potas- sium is fairly large, however, indicating that reservoir temperatures are probably not extremely high (White, 1970). The close similarity in ionic concentrations and ratios among the three waters clearly points toward common origins and subsurface movements. The lower temperature of the spring water is probably explained by the additional conductive cooling of this water. If these waters are assumed to rise and spread from the adjacent boundary fault, it might be expected that some mixing occurs with local ground water. The tritium concentration in sample 12 (table 8) indicates that such mixing does occur. The amount of mixing can be estimated by assuming, on the basis of cold-water sample 9 in table 8, that the tritium concentration of the recharge water is about 90 tritium units (TU). If the tritium concentration in the geothermal reservoir is assumed to be zero, the fraction of cold water in sample 12 is calculated to be 0.06. The uncertainty in this fraction is large because of uncertainty as to the exact nature of the cold recharge water. The results of other mixing models discussed below show, however, that the fraction is probably close to the value indi- cated by the tritium model. Possible modes of origin for the thermal waters of the Crump geyser area are suggested by the silica- enthalpy relations shown in figure 16 (Fournier, 1977). The three simplest of the possible modes are described below as they apply to sample 12, the water least af- fected by near-surface boiling or conductive cooling. The extension of the mixing line through sample 12 in figure 16 intersects the lowest possible enthalpy of steam separation for the altitude of Warner Valley at 96 calories per gram (cal g“). A vertical line from this THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON point (M 2) intersects the curve for maximum steam loss due to adiabatic cooling (Fournier, 1977) and then may be directed horizontally to intersect the curve for quartz equilibration at M1 (enthalpy = 166 cal g"). Point M1 indicates the minimum enthalpy and silica concentration of a possible reservoir water supplying . water M2, which is assumed to mix with a typical cold recharge water (point A) to form the water of sample 12. The horizontal boiling path from point M1 to the maximum steam-loss curve represents the increase in quartz concentration due to boiling. The vertical path from the maximum steam-loss curve to point M2 repre- sents the energy loss (cooling) due to boiling. The en- thalpy estimated for the reservoir is not greatly sensi- tive to the choice of the steam-separation point (M2). An alternative mode of origin for sample 12 is one in which the water cools conductively from the enthalpy indicated by the quartz conductive geothermometer (near point P in the diagram) and then mixes from point M2. A third possible origin would produce the shallow thermal water by direct mixing of cold water with water from a reservoir having an enthalpy of about 245 cal g‘1 (point M3). The waters of well 13 and spring 44 are believed to originate in the same way as the water of well 12. The displacement of sample 13 relative to sample 12 is readily explained by the surface boiling that is ob- served in the large-diameter well of sample 13. The position of sample 44 in the diagram appears to be the result of additional conductive cooling as the hot water flows to the spring at shallow depths. The third hypothesis, mixing from a high-enthalpy reservoir, is believed to be unlikely on the basis of in- dications from mixing models. The fraction of cold water calculated from the tritium mixing model is 0.06, whereas the fraction indicated by mixing from a high-enthalpy reservoir in figure 16 is 0.66. A cold- water fraction of 60 percent or more would probably introduce a significant concentration of magnesium into the mixture, whereas observed concentrations are less than 1 mg L“, thus lending weight to the evidence favoring only minor amounts of mixing. Either of the first two models, adiabatic cooling with steam separa- tion or conductive cooling from near point P is possible on the basis of the silica-enthalpy relations, and each postulates a reservoir temperature of at least 165°C. However, the lack of any evidence of steam separation or of vapor transport and the high probability of con- ductive cooling at the observed flow rates favors the conductive-cooling model. A second approach to mixing models utilizes the chloride-enthalpy relation and assumes that most chloride in thermal. waters is released by water-rock reactions at high temperatures in the geothermal res- I31 TEMPERATURES IN THEGEOTHERMAL RESERVOIR 8m .OoomA $536988 manta wig: £353 5 315:0 3pm.; 5555538 85m mo namuoléfi 859m 25:545. awn. 95:534.: 2. .zo_._.<¢hzwuzou < D 002v 355.252 5;, £28: 3 3 322E 4 cask—33 O 9...? £55 0 _ _ _ _ _ _ _ _ _ 53 95.3 2:33“. I I. EB: uEfiE £5.88 .Ill 8N 0mm con was new saluowa m 'AJ1VH1N3 132 ervoir (Fournier and others, 1979; Truesdell and Four- nier, 1976; Fournier, 1977). The diagram of figure 17 shows possible equilibration points, conductive cooling paths, and mixing paths for the 10 samples having temperatures greater than 30°C. In figure 17, the high-chloride, low-enthalpy waters of samples 12, 13, and 44 are shown in relation to two possible points of origin, M1 and M4. Point M1 corres- ponds closely to the similar point in the silica-enthalpy diagram from which boiling might occur. Point M4 is an equilibration point determined by the silica geo- thermometer from which conductive cooling could occur. The line connecting points M1 and M2 in figure 17 represents an adiabatic cooling curve constructed by extending a line from M2, the point of steam separa- tion, to a point of zero chloride concentration and an enthalpy of 639 cal g‘l representing the composition of pure steam (Truesdell and Fournier, 1976). The mixing fractions of cold water calculated for well 12 from the chloride-enthalpy and silica-enthalpy graphs are identical (0.05), and are close to the value 400 | I | I I Enthalpy=639 cal gm" at Cl=o °S”’i"g"39 C _ 350 —- - Well >30 C (Truesdell and \ AA ' f 14 t ‘th Fournier 1976) \ verage 0 we egs m ’ temperatures <17 C \ 0 Average of 5 low-silica waters ©Possible equilibration point \ Olntermediate stage 300 ‘ Possible conductive \ cooling path ___ Possible mixing trend 5 \ _._ Possible boiling path o: u 250 — \ \ — c a \ a 01 Q _ / \ _ / __ g 200 \ / _I < \ / P (#48X#12) U ®N2 . @M4 2 /. M‘l/ . ; -_ (#16) A ' >_ 150 © // . (#11) / — n. ' NL®/ I \ / -‘ (#14) / / ‘ / E , < . © / / /)( / 2 : IE (#23) 3 ' (#34? / . // : . 2 <9 = 9 ’ " \ / = \ L|.l 100 _ . . -/ / ; / .1 ' _ - : I(#l)//: / - zO/Rl ~ DMZ . . // ; / '/ . '13 . . Q , / /, / . _ 1 / : /’ /’ : //’ : 044 Z ./ // : // z ; : 14; /"3:/ /<>/ 546 50 — 2 . 2/ // /// : 2 i in //61/ // 611 2w/916//// / / // ring” 0 BA 1 I I I I 0 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER KILOGRAM FIGURE 17.—Graph of chloride concentration versus enthalpy in waters having surface temperatures >30°C. THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON obtained from the tritium mixing model (0.06). Al- though the coincidence inspires some confidence in the mixing models and in the probability of an origin in a reservoir at about 170°C, the data do not permit a clear choice between the alternatives of conductive or adiabatic cooling. These alternatives are discussed further at the end of this section of the report. Corroborative evidence for the mixing trend between M2 and cold water is provided by the data from spring 11. This spring, with a temperature of 40°C and a flow of about 5 gallons per minute (gal min“), is located along the west boundary fault about a mile north of well 12. The water has a chemical composition similar to those of the Crump geyser waters (fig. 4), but roughly half the dissolved-solids concentration. It seems reasonable to assume that the water in spring 11 has cooled conductively, perhaps 15°C or 20°C, as spring 44 is believed to do. This assumption places spring 11 on the mixing trend for well 12 in both the chloride and silica diagrams. On the basis of the mixing line from point M2, cold-water fractions in spring 11 are calculated to be 0.48, 0.44, and 0.52, re- spectively, for the constituents silica, chloride, and dis- solved solids. The dissolved solids concentration at M2 was estimated from ratios of silica and chloride to dis- solved solids in sample 12. Thus, the mixing model shows excellent agreement for these constituents and indicates that the thermal component of spring 11 could have originated in water represented by point M2. There remains the possibility that the water of spring 11 has attained its present state entirely by conductive cooling from its indicated quartz conductive geothermometer temperature of about 150°C. This pos- sibility cannot be entirely ruled out, but the small con- centration of dissolved solids (662 mg L“) in the water is incompatible with equilibration at this temperature and indicates that mixing has occurred. Ratios of major ions in spring 11 also support the concept of low- temperature mixing. Ratios of calcium, magnesium, and bicarbonate to chloride and sodium are increased by the mixing, while ratios of sodium to potassium, chloride to sodium, and chloride to sulfate remain simi- lar to those in the higher temperature water. The mix- ing model explains these observations better than the conductive-cooling hypothesis. A third spring which is included with the Crump geyser waters on the basis of its chemical composition is spring 46, located about one mile south of Crump geyser and about 0.6 miles east of the boundary fault. This spring, with a temperature of 58°C and a flow rate of about 3 gal min—1, issues at the end of a penin- sulalike sinter terrace about 0.6 miles long extending southeast from the fault zone. At least six additional TEMPERATURES IN THE GEOTHERMAL RESERVOIR springs and seeps having low flows and temperatures ranging from 35°C to 49°C also issue from the terrace. The chemical composition of spring 46 is similar in all important respects to the compositions of wells 12 and 13 and spring 44; the silica concentration is identi- cal to that of well 12 and the chloride concentration is only slightly lower than that of well 12 (230 mg L’1 versus 240 mg L") (table 6). The virtual coincidence of ionic concentrations in the Crump geyser water and spring 46 compels the belief that the two waters originate in the same reservoir and probably ascend in the same fault. It is significant, therefore, that the water of spring 46 undergoes little change in ionic concentrations (with the possible ex- ception of an increase in boron) as it flows 0.6 miles from the fault zone to its outlet while cooling about 35°C. The lack of water-rock interaction during this passage implies that the water moves in a conduit that is probably armored with silica and that has remained open during the time required to precipitate the large amount of silica laid down in building the sinter ter— race. Further important implications are that the hot waters of the Crump geyser area can presumably move upward from the reservoir in similar armored conduits while changing little in composition as they cool con- ductively. The indirect evidence afforded by spring 46 thus tends to increase confidence in the belief that the Crump geyser waters closely approximate the chemical composition of the geothermal reservoir. Waters from the two wells and three springs in the Crump geyser area appear to define a lower limiting trend on the silica-enthalpy and chloride-enthalpy dia- grams (figs. 16 and 17). Spring 1, which falls below this trend on the silica graph, is an isolated spring at high altitude whose relation to thermal waters in Warner Valley is unclear. This spring is discussed below. All other thermal waters fall on or above the trend estab- lished by 12, 13, and 44 if about 20°C of conductive cooling is accepted for spring 11, and 30°C or more of cooling has occurred in spring 46. FISHER HOT SPRING AREA An upper limiting trend in both the silica and chloride diagrams is determined by the mixing lines through spring 14, Fisher Hot Spring. This spring has a surface temperature of 68°C and an estimated flow of 15—20 gal min—1. The calculated temperature of equilibration is 123°C for both the quartz conductive and the magnesium-corrected Na-K-Ca geother- mometers. The concentration of dissolved solids (360 mg L“) and the position of this water on the chloride and silica diagrams (figs. 16 and 17) and the trilinear diagram (fig. 4) suggest that this is a mixed water. Ratios of major ionic constituents (table 7) are similar to those in 133 spring 11 and show changes relative to the hotter wa- ters that are consistent with low-temperature mixing. For these reasons, the line showing a direct conductive cooling path in figure 17 is not favored as an explana- tion for the origin of this water. If the water of Fisher Hot Spring is assumed to be highly mixed, as its chemical composition suggests, several possible mixing relations can be postulated from figures 16 and 17. In figure 16, thermal water at N1, having an equilibration temperature of 143°C may mix with cold water represented by point A. The frac- tion of cold water calculated for this mixture is 0.58, and is the same for the corresponding mixture on the chloride—enthalpy graph. The temperature of equili- bration indicated by point N1 is 20°C higher than the temperature estimated by the quartz and Mg-corrected Na-K-Ca geothermometers but is close to the tempera- ture of 139°C indicated by the sulfate-oxygen geother- mometers. The latter geothermometer may not be reli- able if the water is mixed, however, and low- temperature sulfate is introduced in the mixture (Nehring and Mariner, 1979). A second possible mixing path utilizes average val- ues of silica in five low-silica cool waters (point B, figs. 16 and 17). The five waters (5, 6, 36, 37, and 42) include four high-altitude cold springs on the eastern fault blocks and a well located about 3 miles north of Fisher Hot Spring. These waters differ somewhat from other cool waters in the area, and they could represent a mixing water that occurs on the east side of the valley. The equilibration temperature for quartz at which this trend ends in figure 16 (point N2) coincides with the uncorrected Na-K-Ca temperature (about 168°C). Other explanations for the origin of spring 14 are possible, including various combinations of conductive cooling and mixing from high- or low-temperature res- ervoirs. One alternative hypothesis, given below, seems more probable than others when all the availa— ble evidence is considered. Abnormally high thermal gradients and heat flow in the area near Fisher Hot Spring have been described in the section of the report entitled “Distributions of Temperature and Heat Flow.” The presence of large volumes of warm ground water in the area makes it probable, therefore, that the spring water has acquired heat by conduction and by mixing with ground water having temperatures as high as 30°C to 40°C. The ef- fects of these processes would probably establish a mix- ing line for spring 14 on the silica-enthalpy diagram originating at some position above point A and trend- ing roughly parallel to the line through spring 34. The mixing line would probably include sample 16, a highly mixed water from a spring located about 2 miles south of Fisher Hot Spring. Spring 16 has a surface 134 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON temperature of 38°C and a flow of 1—2 gal min“. The reservoir equilibration point could range from near M1 to P, depending on the point of origin and the dis- placement of spring 14 due to conductive heating. The indicated temperature of equilibration is in the range of 165°C to 175°C and the fraction of high—temperature water in the mixture is about 0.10. A similar trend would appear on the chloride-enthalpy diagram (fig. 17). General confirmation of the mixing relations for Fisher Hot Spring and other thermal waters is pro- vided by the boron-enthalpy diagram (fig. 18), which shows a configuration similar to that of the chloride- enthalpy relation. However, plots of the relation be- tween boron and other major ions (not presented in this report) show much more scatter than similar plots using chloride. In particular, samples 1, 14, and 34 have exceptionally low boron-chloride ratios that have the effect of displacing these samples on the boron- enthalpy graph (fig. 18) from the positions they occupy on the chloride-enthalpy graph (fig. 17). If boron con- centrations in these samples are low as the result of near-surface reactions with clay minerals, for example, adjustments for these losses applied to the boron con- centrations in figure 18 for samples 1, 14, and 34 would bring them into conformity with the relations on the chloride-enthalpy graph. The boron-enthalpy graph thus provides a partly independent confirmation of the chloride relations and supports the concept of high- enthalpy, low-concentration mixed waters for Fisher Hot Spring and spring 16.2 In spring 16, thermal water apparently has mixed with local water to produce ratios of major ions that are similar in all important respects to those in spring 14 (table 7). Most of these ratios contrast strongly with those in three shallow low-temperature wells in the area (6, 8, 15). The relatively high temperature of spring 16 probably reflects the high ambient tempera- tures in the area rather than a large component of thermal water, however. The fraction of thermal water may be only about 0.05, as determined by the probable mixing line with background water of elevated tem- perature and dissolved solids concentration. The mix- ing water could readily be derived from waters like those in wells 6, 8, and 15 by conductive heating and reequilibration, with consequent decreases in calcium, magnesium, bicarbonate, and sulfate relative to sodium and chloride. Relations among major ionic constituents of the thermal waters and mixing calculations based on these 2The anomalously high concentration of boron in spring 46 may be the result of water- rock interaction as the spring water flows about 0.6 mi thmugh the siliceous sinter terrace to its outlet. Chemical analysis of the sinter shows that it contains 700 parts per million by weight of boron. relations tend to reinforce the hypothesis of low- temperature mixing for the waters of Fisher Hot Spring and spring 16. These springs consistently fall on or near the trends of chloride versus silica and chloride versus bicarbonate in figures 7, 8, and 9. They also have consistent ratios to the positions of spring 11 and the hotter waters in these diagrams. Mixing trends are believed to provide good evidence that waters in positions along well-defined trends probably reflect similar origins and are mixed with chemically similar waters. From these arguments, and as a check on the rela- tionship between the silica-enthalpy and chloride- enthalpy graphs, it may be assumed that, if the water of spring 14 is derived from a reservoir between points P and M1, having a temperature of 170°C, a chloride concentration of about 215 mg L‘1 (fig. 17) and a silica concentration of about 175 mg L‘1 (fig. 16) the position of this reservoir water should fall along the mixing trend in the silica-chloride graph (fig. 7). The point labeled P’ in figure 7 is consistent with this hypothesis and with the positions of the Crump geyser samples. ' The graphs based on chloride concentration, silica con- centration, and enthalpy thus demonstrate fairly con- sistent relationships among waters in the two major thermal areas of Warner Valley. The mixing hypothesis described above indicates that the temperature of equilibration for the thermal component of Fisher Hot Spring is at least 170°C. The apparent equilibration temperature of 123°C indicated by the quartz geothermometer (table 9) is, of course, in agreement with the proposed mixing formula derived from the silica-enthalpy graph. The identical tempera- ture obtained from the magnesium-corrected Na-K-Ca geothermometer could be fortuitous, and the uncor- rected temperature of 169°C may be correct. The addi- tion of magnesium during mixing at shallow depths would not have changed the ratios used in the Na-K-Ca geothermometer. To summarize the conclusions for the Fisher Hot Spring and Crump geyser areas, the reservoir waters indicated by mixing models for the two areas have similar, although not identical, temperatures and con- centrations of dissolved constituents. Apparent equilibration temperatures range from a mimimum of 165°C for Fisher Hot Spring to 173°C for Crump geyser. Chloride concentrations may be in the range 215 to 240 mg L‘1 and silica concentrations may be in the range 165 to 190 mg L“. On the basis of ratios to silica and chloride, dissolved solids concentrations could range from 900 to 1,150 mg L“. The apparent differences between the reservoirs supplying the two areas may be commensurate with the uncertainties inherent in the mixing models and hence should not be taken as con- 135 TEMPERATURES IN THE GEOTHERMAL RESERVOIR .OoomA $3.338»: «aim mega: @863 E 3355 99:3 :oflgnwonoo :83 we anwuwlwfi 559m E<¢wo.:v_ cun— ms<=o_._.:2 z_ .zo_._.' 445! 30 / zj/ ’ X .31 26 I; _ _ o 1“ £41 Bolling from 220°C to 96°C (single stage) «a I379+19‘° (From TruesdelI, Nathenson, and Rye; 1977) .10 ‘16 .33 _ WK +39 18 8 .1 -17 4 r r 1 l 0 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER KILOGRAM FIGURE 20.—Graph of chloride concentration versus 6‘80 in ground water, showing a single-stage boiling curve between 220°C and 96°C. -105 I I I I I ° Spring, 220°C ° Spring,<20°C + Well, I20°C -110 — + We||,<20°C ‘ _, (9 Possible reservoir water ._ 05 +19 5 22 +20 Standard error in SD [(11705) O: 3 E 115 M” ' $35,322 .44 M2 _ E - 12+ 9 o c ‘+ .y ______________ I} (96 C) E °r?“': _ _ / D 26++33§=fi¢ ~ ‘ / — 010 ‘5 32‘ \. 34‘“ ~ \ _. / a: —120 *8 “ ‘ ~~~~~~~~ / ”,2 33. No \\\\\ ,‘AVP (180°C) 3 9?) (220°C) ”a." to '1 Boiling from 220°C to96°C (single-stage) -125 — +39 (From Truesdeu, Nathenson, and Rye, 1977) — 430 r I I I I 0 50 100 150 200 250 300 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER KILOGHAM FIGURE 19.~Graph of chloride concentration versus 8D in ground water, showing a single-stage boiling curve between 220°C and 96°C. TEMPERATURES IN THE GEOTHIERMAL RESERVOIR 137 rather large; despite this fact, the correlation of figures 19 and 20 with the silica-enthalpy and chloride- enthalpy graphs is fairly good. OTHER THERMAL WATERS Spring 34 is located at the base of a tilted fault block at the south end of Warner Valley. The rate of flow is difficult to determine but is estimated to be in the range of 3 to 5 gallons per minute. The surface temper- ature is 66°C and the apparent temperatures of equilibration are 109°C and 105°C respectively, accord- ing to the chalcedony and the magnesium-corrected Na-K-Ca geothermometers. Spring 34 is probably a mixed water having fractions of cold water of 0.68 and 0.66 calculated from the silica-enthalpy and chloride-enthalpy graphs, re- spectively. The graphs of isotope concentration versus chloride concentration (figs. 19 and 20) also suggest that the water has formed by direct mixing of local ground water with reservoir water. Fractions of cold water calculated by assuming mixing from point P in figures 20 and 21 range from 0.62 and 0.66, depending on the point chosen for the cold-water component. It should be noted, however, that quantitative determi- nations from figures 19 and 20 are highly dependent on the chloride concentrations and hence add little inde- pendent support to the determinations based on the chloride-enthalpy relation. The interpretations given above suggest that the geothermal regime at the south end of Warner Valley in the vicinity of spring 34 is similar to the regime postulated for the areas north of Adel. Several aspects of the chemistry of spring 34 cloud this picture, how- ever. Spring 34 has high concentrations of sulfate, bicar- bonate, and fluoride and low concentrations of calcium and boron relative to other thermal waters. The pH is fairly low (7.2) and the deuterium concentration is significantly depleted relative to concentrations in more typical thermal waters. Finally, the concentra- tion of dissolved solids in this water is too high to fit any of the mixing relations utilized above. 0 K I | r I I l I l \Vallev-flll deposits (800 ft =243 m)"‘ V\ " \\ \ Basalt, tufts, and tuffncaous sediments E 2000 \ \ Thermal conductivity ‘ 1.90 W (m—k)“ 1000 *— \\ Thermal gradlont = 39.4°C km‘1 — \ — 4000 \ \\ Hest-How= 75mW m-2 \ m \ — 6000 3: Reported bottom-hole Lu 2000 “ temperature In San Juan _ E l— 0" Company teat hole: \ Lu g 94°C a: 2127 m(6980 m \ u. z \\ — 8000 E . \ :E E \ ’a‘. s \ g Q 3000 — \\ ‘— 10,000 \ \ \\ \ — 12,000 \ 4000 _ 170°C at 4054 mg _ (13,300 ft) 0 x Thermal conductivity - 1.88 W (m—Id'1 — 14,000 Thermal grndlont = 40°C km-1 1 l 1 I mm 5000 l l | | 0 20 40 60 80 100 120 140 160 180 TEMPERATURE, IN DEGREES CELSIUS FIGURE 21.—Hypothetica.l profile of normal temperature versus depth in Warner Valley. 138 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON Ratios of several major ions, Ca/Mg, Cl/Mg, and Cl/Na (table 7), indicate a fairly close relationship be- tween spring 34 and the hotter thermal waters. The silica-chloride relation in figure 7 also confirms the probable origin of the thermal component of spring 34 in hot water similar to that of Fisher Hot Spring and Crump geyser. Ratios that are sensitive to low- temperature equilibration, such as Ca/Mg and Cl/(HCOR + 003), are closer to those observed in two samples from a deep test hole located near the south end of the valley (samples 40 and 41). The samples from depths below 492 feet in this test hole have higher concentrations of bicarbonate, sulfate, and fluoride than shallower ground waters and they may represent a type of water that occurs in the Tertiary volcanic sediments that underlie the valley floor. The positions, in figure 4, of samples from wells in bedrock or deep wells in the valley-fill deposits tend to support this con- clusion. (See samples 4, 19, 23, 32, 40, 41, and 43.) Ground water of this type is depleted in deuterium and probably contains little or no tritium; it may represent old water recharged during pluvial times from high altitudes. On the basis of the chemical composition and mixing models, therefore, it is believed that the thermal wa- ters of spring 34 and other warm springs at the south end of the valley originally equilibrate at a tempera- ture of 170°—180°C. Mixing with local ground water adds high concentrations of fluoride, sulfate, bicarbo- nate, and dissolved solids while ratios of chloride to silica remain on the mixing trend observed for more typical thermal waters. The fraction of cold water in the mixture may range from 0.65 to 0.71 or more, de- pending on the selection of a low-temperature point for the local ground water. The sulfate-oxygen geother- mometer (table 9) apparently confirms a high- temperature equilibration for this water (179°C), but the reliability of this geothermometer is questionable when it is applied to a mixture presumably containing large amounts of low-temperature sulfate. Another thermal water, that of Antelope Hot Spring on Hart Mountain (1), does not fit any of the models thus far described. The sample is a low-enthalpy, high-silica, sodium bicarbonate water that is not clearly related to other thermal waters of the area. Antelope Hot Spring appears far from the mixfng trends established by other thermal waters in figures 7, 8 and 9, although the separate trend that could be deduced from the graph of figure 7 is not substantiated by the graphs of figures 8 and 9. This water contains low concentrations of deuterium and oxygen-18 which place it outside the field established by other thermal waters on the graphs of isotope-chloride relations (figs. 19 and 20). The quartz and Na-K-Ca geothermometers are in close agreement at a temperature of about 168°C, but the high concentration of magnesium (2.5 mg L“) suggests that the corrected Na-K-Ca tempera- ture of 87°C may be applicable. A comparison of the silica-enthalpy and chloride- enthalpy graphs (figs. 16 and 17) suggests that the water has acquired excess silica relative to the chloride concentration. Computer calculations of phase equilib- ria (Kharaka and Barnes, 1973) lend support to this assumption by showing that the water is supersatu- rated with respect to amorphous silica at its measured temperature. Molal ratios of VC—a/Na, Na/K, and Cl/SO, are simi- lar to ratios in the waters of Crump geyser and Fisher Hot Spring, but other ratios are highly dissimilar (ta- ble 7). Most ratios suggest that the water is highly mixed and has reequilibrated with shallow ground water at low temperature. Several hypotheses provide partial explanations for the origin of Antelope Hot Spring. First, the water may have cooled conductively from the reservoir indicated by the quartz and Na-K-Ca geothermometers at about 170°C. If so, the low chloride concentration in relation to silica and dissolved solids concentrations must be accounted for in some other way. A second explanation is based on the single-stage boiling and mixing rela— tionship determined by the mixing line through spring 1 on the silica-enthalpy graph (fig. 16). This process would require that the thermal component of the mix— ture originate in a reservoir at about 215°C. The data do not clearly rule out such a high-temperature reser- voir, and this possibility is discussed further in the summary of this section of the report. It may be significant, however, that the mixing line from R1 in figure 16 closely follows the equilibrium curve for amorphous silica solubility and that the spring water is supersaturated with respect to amorphous silica. The most probable explanation for the characteris- tics of Antelope Hot Spring involves conductive cool- ing, mixing, and the solution of silica during the cool- ing and mixing process. Many combinations of these processes would fit the available data and none of them is susceptible to convincing proof. The position of sample 1 on the chloride-enthalpy diagram (fig. 17) is probably the most reliable indicator of the relation of this Water to other thermal waters in the area. On this diagram, sample 1 falls above the mixing line for the Crump geyser waters. In view of the isolated location of spring 1, its high altitude, and the absence of any other thermal water on Hart Mountain, it seems probable that this water cools conductively from a position even farther above the Crump geyser mixing trend and that this water is not necessarily derived from the reservoir that supplies the Crump TEMPERATURES IN THE GEOTHERMAL RESERVOIR geyser water. Furthermore, the water that recharges Antelope Hot Spring is probably derived from the same high-altitude sources that recharge the shallow aqui- fers on Hart Mountain. The result is seen in the un— usually low concentrations of oxygen-18 and deuterium in the spring water (figs. 10, 19, and 20). The exceptionally high silica concentration in spring 1 is probably derived from Tertiary tuffaceous sedimentary rocks that underlie and are incorporated in the volcanic dome of Hart Mountain (fig. 3). Cor- roborative evidence for this view is provided by the chemical composition of Valet Spring, a cold—water spring located about 1 mile north of Antelope Hot Spring on Hart Mountain. The silica concentration in Valet Spring is 66 mg L“, the highest concentration measured in a cold spring, and the water also contains higher concentrations of sodium, bicarbonate, sulfate, and dissolved solids than other cold springs in or near the valley. The water of Antelope Hot Spring probably rises in the permeable zone of a major fault intersection (fig. 3), cools conductively, and mixes with local ground water. The reservoir supplying the spring has a temperature of at least 87°C, as indicated by the magnesium- corrected Na—K-Ca geothermometer (table 9), but the temperature could be considerably higher if possible mixing relations on the chloride-enthalpy graph are considered. The equilibration temperature for Antelope Hot Spring probably cannot be estimated on the basis of available data. Despite agreement between the quartz conductive and uncorrected Na—K-Ca geothermome- ters, neither of these is believed to be reliable in this instance. The quartz geothermometer is almost cer- tainly erroneous if the chloride concentration is, as as- sumed, a valid indicator of mixing. The Na-K-Ca geothermometer is probably erroneous if, as assumed, the thermal water originates from and mixes with low-temperature ground water from Hart Mountain. When the Na-K-Ca geothermometer is applied to the water of spring 45, for example, calculated equilibra- tion temperatures are 191°C and 63°C, respectively, for the uncorrected and magnesium-corrected geother- mometers. These temperatures do not differ greatly from those indicated for spring 1, suggesting that even the magnesium-corrected value may not be a reliable indicator of equilibration temperature in the hot spring. The last of the nine samples having temperatures greater than 30°C is number 23, obtained from a well of unknown depth located along the west boundary fault about 2 miles south of Adel. The well is known to be old and is used only rarely for stock watering. As a conse- quence, there is some doubt about the reliability of the 139 chemical data. The data from well 23 demonstrate many of the com- plexities inherent in the determination of chemical re- lations in low-temperature waters. Sample 23 falls near the trends for Fisher Hot Spring and spring 34 in figure 17 (chloride—enthalpy); it falls near the trend for Crump geyser waters in figures 5 (chloride-fluoride) and 16 (silica-enthalpy); and it is on the trend with Antelope Hot Spring and low-temperature waters from deep wells in figure 7 (silica-chloride). Ratios of major ions in this water (table 7) are similarly varied. The chloride concentration should be one of the more reliable indicators for well 23, and in relation to this constituent, both temperature and silica concentration are higher than expected. The water shares significant characteristics, however, with other deep well waters from the southern part of the valley and also with water from the thermal spring 34. On the basis of the ratios Cl/Na, Cl/SO4, Cl/(HC03 + CO3), Ca/Mg, and Na/K (table 7), and the chloride- enthalpy and SD-chloride diagrams (figs. 17 and 19), the water is assumed to be a combination of typical high-temperature water mixed with a large fraction of shallow ground water having higher than normal tem- perature and silica concentration. The relations are probably similar to those in spring 34 with additional complexities of unknown origin. The most probable interpretations of the chemical relations and geothermometers in the waters of Warner Valley lead to the conclusion that the thermal components of these waters equilibrate in similar res- ervoirs at temperatures of at least 170°C and attain their surface temperatures and compositions largely by conductive cooling and mixing. Adiabatic cooling with steam separation is believed to be minimal for even the hotter waters. At the al- titude of Crump geyser (4,500 feet), water in an uncon- fined aquifer could boil only at depths less than about 230 feet if the temperature were as high as 170°C. Data from wells 12 and 13 at this location indicate that tem- peratures were about 124°C and 127°C at depths of 120 and 660 feet, respectively, suggesting that the shallow water is far below temperatures required for boiling under normal hydrostatic heads. If the mixing rela- tions deduced for these wells are valid, they imply that a reasonably good hydraulic connection exists between the reservoir water and the shallow waters, and this suggests that the reservoir waters could not boil at great depths. The implausibility of extensive boiling thus favors the conductive cooling hypothesis for the Crump geyser waters and the concept of deep mixing for the remaining thermal areas. The silica-enthalpy and chloride-enthalpy relations for the Warner Valley thermal waters can be inter- 140 THE GEO’I‘HERMAL HYDROLOGY OF WARNER VALLEY, OREGON preted to imply the existence of a deep parent reservoir from which the 170°C waters of shallower reservoirs are derived. In figures 16 and 17, mixing lines for two springs, Antelope Hot Spring and Fisher Hot Spring, can be projected to equilibration points that represent such a deep, hot reservoir. One line through Fisher Hot Spring (14) projects to a reservoir at Q on the chloride- enthalpy graph, and a line through Antelope Hot Spring (1) projects to a similar point (Q1) on the silica- enthalpy graph. Although highly speculative, point Q on the chloride-enthalpy graph could be a point of origin for all thermal waters in the area. According to this hypothesis, reservoir water at Q cools adiabatically to M2 and mixes with cold water to form the waters of the Crump geyser area (12, 13, and 44). At intermediate points, M1 and R, the water mixes directly with cooler water to form the waters of spring 34 and spring 1. Water from Q also mixes directly with waters of lower temperature to produce the mixture at N2, from which waters of the Fisher Hot Spring area are derived by further mixing. (The point N2 was established as the end point of a possible mixing trend for Fisher Hot Spring in fig. 16). Inspection of the silica-enthalpy graph (fig. 16) shows that if the water of Antelope Hot Spring is de- rived from a high-temperature reservoir, the reservoir water could cool adiabatically from Q1 to R1 and then mix with recharge water, A, to produce the observed silica concentration in the spring. The equilibration point equivalent to R1 on the chloride-enthalpy graph could be anywhere along the 96 cal gm‘1 steam separa- tion line. If it is placed, as shown in figure 17, at a point where the mixing fractions are identical to those for Rl-l-A on the silica diagram, it falls conveniently near the extension of a boiling curve through N2. The reser- voir equilibration point equivalent to Q1 would then presumably occur at the 218 cal gm‘l enthalpy level along the boiling curve as shown. The hypothesis outlined above requires the existence of two intermediate reservoirs beneath Warner Valley, one having high silica and low chloride concentrations and the other having the reverse. An occurrence of this kind seems implausible. Furthermore, the relations proposed for Antelope Hot Spring are not substantiated by the isotope-chloride graphs (figs. 20 and 21) or by calculated mixing fractions. The unreliability of indi- cations from Antelope Hot Spring has been discussed above, and the evidence from this spring should prob- ably be given very little weight in conclusions concern- ing a high-temperature reservoir. The mixing ‘line through Fisher Hot Spring to point Q also cannot be substantiated by the isotope-chloride graphs. If the most probable of several possible mixing lines for Fisher Hot Spring is accepted, the equilibration point falls close to M1 and the evidence for a reservoir at Q vanishes. Additional evidence supporting the concept of a high-temperature reservoir are the equilibration tem- peratures estimated by means of the sulfate-oxygen geothermometer (McKenzie and Truesdell, 1977) for springs 11, 44, and 46 and well 12 in the Crump geyser area (table 9). Apparent temperatures of sulfate- oxygen equilibration range from 200°C to 219°C in these waters for the conductive heat-loss model. The consistency of the estimated temperatures is remarka- ble in View of the inclusion of spring 11, a highly mixed water having a measured temperature of 40°C. The maximum temperature, 219°C, is consistent with the indications of the chloride-enthalpy graph (fig. 17). The most damaging argument against the evidence provided by the silica-enthalpy and chloride-enthalpy diagrams is the extremely low probability that water from a high-temperature reservoir has cooled adiabati- cally in accordance with the boiling curves suggested in figures 16 and 17. There is no evidence, for example, that the waters represented by possible equilibration points N2, M1, or P are derived from or are heated by steam that would have separated from a deep reser- voir. There are no fumaroles in the Crump geyser area that would indicate steam separation, although this lack does not preclude the condensation of steam in the shallow ground water. However, ratios of chloride to boron, which would be expected to indicate deep boil- ing, show evidence only of normal heating of meteoric water. Other constituents, such as fluoride, arsenic, and lithium, similarly give no indications of extensive boiling. There would be no need to invoke adiabatic cooling, however, if fluid from a deep high-temperature reser- voir were assumed to ascend and cool conductively and then reequilibrate at shallower depths in a reservoir represented by points in the vicinity of M4 in figure 17. All thermal waters in the area could be derived from a point near M4 by the same means proposed above for an origin at point Q. The silica and N a-K-Ca geother- mometers thus would represent the reequilibration temperatures, whereas the slow—responding sulfate- oxygen would continue to indicate the temperature of the deeper reservoir. Many uncertainties still surround the application of the sulfate-oxygen geothermometer. A discussion by Nehring and Mariner (1979) shows, for example, that thermal waters may dissolve “fossil” sulfate from min- erals formed at higher temperatures, thus giving higher sulfate geothermometer temperatures than chemical geothermometers. This explanation was ad- vanced to account for discrepancies in geothermome- CONCEPTUAL MODELS OF THE GEOTHERMAL SYSTEM 141 ters in the Klamath Falls area (Sammel, 1980). On the other hand, Nehring and Mariner present evidence to show that some high-temperature waters precipitate silica at concentrations greater than about 200 mg L", thereby causing the discrepancies between the silica and sulfate geothermometers. A third possible hypothesis, therefore, is that the precipitation of silica has resulted in erroneously low temperatures from the quartz geothermometer. In this case, the reservoir would be deeper than estimated and the thermal fluids would be assumed to cool conduc— tively to temperatures at which the mixing relations proposed previously would apply. Some of the mixing proportions would be incorrect in this case but the gen- eral relationships would still be applicable. The significance of these several hypotheses for strategies of exploration and development is discussed in the final section of this report. In view of the uncer- tainties, however, the occurrence of a deep high- temperature reservoir in Warner Valley is speculative and has an essentially unknown probability. CONCEPTUAL MODELS OF THE GEOTHERMAL SYSTEM HEAT-FLOW CONSTRAINTS ON THE DEPTH OF FLUID CIRCULATION Thermal waters of the Warner Valley area are be- lieved to originate in a deeply-circulating hot-water system. The chemical compositions of the thermal wa— ters show no evidence that the waters have been in contact with magma or have been heated by steam ris- ing from a magma chamber. The chemical evidence is thus in accord with the data of MacLeod, Walker, and McKee (1975), who conclude that silicic domal masses or magma chambers in this region are probably 8 or 9 million years old and would have cooled to ambient temperatures by the present time. Geothermometers and mixing models indicate that temperatures of equilibration for the thermal compo- nents of the waters are at least 170°C prior to mixing. The depth at which this temperature would occur in a steady-state conductive regime beneath Warner Valley is estimated to be at least 13,000 feet. The assumptions used for this estimate are depicted in figure 21 and are described as follows: average heat flow in the valley is 75 milliwatts per square meter (mW m‘2); thermal conductivity in the valley-fill deposits is 1.88 watts per meter kelvin (W (m'K)“); the average depth of the valley fill in the area north of Adel is 800 feet, based on depth to bedrock in the US. Geological Survey test hole, MCI, and on a similar depth reported in a San Juan Oil Co. test hole near Adel; thermal conductivity in the underlying Tertiary basalt and tuffaceous sedi- ments is 1.9 W (m-K)’1, which is lower than the con- ductivity in a dense basalt but which may be a reason- able value for the thin flows, tuffs, and tuffaceous sed— iments that underlie the valley. The depth of 13,000 feet (4 km) is calculated with the further assumption that the average thermal conduc- tivity in the bedrock remains constant with depth. Characteristics of the bedrock reported in logs of the San Juan Oil Company well show no overall change to a depth of 7,500 feet, but if deeper rocks were to have higher thermal conductivities as the result of greater densities, the depth to the 17 0°C horizon would be cor— respondingly increased. Calculated gradients of about 39°C km‘1 in the bedrock and 40°C km‘1 in the valley- fill deposits produce a temperature-depth relation that is in close agreement with a reported bottom-hole tem- perature of 94°C at a depth of 6,980 feet in the San Juan Oil Co. test hole (figure 21). The requirement that thermal waters circulate to a depth of at least 4 km in order to attain the estimated reservoir temperature implies that the flow probably occurs in major boundary faults. Evidence that major faults are active and are possible conduits for ground- water flow to such depths is provided by the Adel earthquake swarm which had focal centers at depths ranging from 2 km to nearly 13 km and averaging 9 km (Schaff, 1976). It is clear from geometric considerations that if thermal water is confined to a nearly vertical fault plane, it will not capture the vertical conductive heat flux in the crust as efficiently as if the water were spread laterally in a horizontal aquifer. Numerical models developed by M. L. Sorey for the Grass Valley, Nev., geothermal system show, for example, that for some fault-controlled hydrothermal systems in the Basin and Range province, rates of heat absorption by meteoric water circulating down a fault are insufficient to raise fluid temperatures to known or inferred reser- voir temperatures within reasonable depths of circula- tion (Welch, and others, 1981). In addition to vertical flow in the fault, lateral circulation is needed in order to capture a sufficient amount of heat. In the absence of lateral circulation, the meteoric recharge water would cool the adjacent rocks. In order to estimate whether or not lateral circula- tion might be required in the Warner Valley fault- system, flow rates were estimated for all known ther- mal springs and wells in Warner Valley and the flow rates of the thermal components, presumably derived from the 170°C reservoir, were calculated on the basis of proportions derived from mixing models. The sum of the flow rates of thermal springs and wells in Warner Valley is approximately 7.5 kilograms per second (kg s“) and the total rate of flow from the thermal reser- 142 THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON voir is approximately 2 kg s“). Total thermal dis- charge at the land surface above a base temperature of 10°C is calculated to be 1.8 X 106 watts (W). If the estimated mass flow from the reservoir is as- sumed to be replaced by recharge water circulating down the faults, calculations based on numerical models show that, at a depth of 4 km, this circulation would lower the rock temperature about 20°C below that which would exist in the absence of fluid circula- tion (Welch and others, 1981). The assumed boundary conditions include a 70°C dip of the fault planes on either side of the valley and circulation to 4 km. The dip of the faults is taken as the smaller of two esti- mates: 70°C estimated by Peterson (1959), and 80°C estimated by Schaff (1976) on the basis of seismic data from the Warner Valley earthquake swarm of 1968. The smaller estimate is slightly more favorable for ef- fective heat transfer. Half of the total flow is assumed to occur on each side of the valley, and the horizontal length of fault through which flow occurs is assumed to be 4 km on each side. If downflow occurs over greater lengths of the fault or if fluid circulation has not con- tinued for sufficient time to establish a thermal steady state, resulting temperatures at depth in the fault con- duit would be correspondingly higher. This analysis indicates that if the total throughflow of the thermal water is approximately 2 kg s“, the hydrothermal system in Warner Valley could consist largely of water circulating within the boundary faults to a depth of little more than 4 km. The storage of both water and heat in the system would be extremely small in this case. The existence of a more extensive reser- voir is not ruled out, however, and this possibility is discussed in the final section of this report. Numerical simulations by Sorey (1978) further sug- gest that thermal waters having flow rates similar to those estimated for the Warner Valley springs would cool appreciably by conduction prior to reaching the surface. Conductive cooling at a flow rate of 2 kg s“ might lower the temperature as much as 75°C from a reservoir temperature of 170°C. These results are qual- itatively confirmed by calculations, reported by Trues- dell, Nathenson, and Rye (1977), which indicate that flows less than 1.7 kg S‘1 at 200°C would cool to below boiling temperature in a vertical ascent of 4 km. Thermal water arising from such a depth would have to flow at rates greater than 15 kg s‘1 in order to main- tain reservoir temperatures. CONSTRAINTS BASED ON CHEMICAL DATA AND MIXING MODELS The results of the numerical models must be recon- ciled with chemical data which are partly ambiguous (for example, the geothermometers) but which have a generally satisfying consistency (for example, the isotope data and the mixing calculations). Indications based on ratios of major ions, isotope concentrations, and mixing relations can probably be taken at face value as evidence of a high-temperature reservoir. Silica—concentrations in the hotter waters are consis- tent with this evidence and indicate that the tempera- ture is at least 170°C. Two factors may mitigate the supposed effects of conductive cooling. The flow rate from the reservoir could be as much as double the estimated rate, and much of the flow could be dispersed Within the valley- fill deposits. The large volumes of heated ground water known to exist in the Crump geyser and Fisher Hot Spring areas are evidence of this dispersion. Higher flow rates would decrease the temperature loss in the upflowing thermal water and the higher ambient tem- peratures in the mixing zones would further increase the final temperature of the mixing water. The differing influences of these factors are probably involved to some extent in the temperature differences observed between the Crump geyser area and the Fisher Hot Spring area. It is likely, for example, that the flow system is not symmetric as postulated in the heat-loss model described above. Discharge on the west side of the valley could be as much as twice the dis- charge on the east side, thereby accounting in part for the temperature differences between the two areas. On the west side of the valley, thermal water from the reservoir probably ascends and cools conductively in several conduits within the boundary fault. The flow rate may be in the range 1 to 3 kg 5". Near the base of the valley-fill deposits, the continuity of the channels is broken and the water spreads and mixes with local ground water. Clear evidence for the hot saline water in this zone is seen in the resistivity data described in a previous section of the report. The water may reequilibrate to some extent in this mixing zone and attain the cation ratios that, in the Na-K-Ca geothermometer, indicate a temperature of 152°C. Silica concentrations and dissolved solids con- centrations remain high, however, in the slowly cir- culating hot water. Vertical permeabilities in this zone are probably low and one half or more of the hot water spreads laterally through the valley-fill deposits. The remainder ascends to the surface through the conduits that supply the thermal springs in this area. As previ- ous sections of this report have shown, some springs contain highly mixed water (samples 7, 10, and 11, for example) while other springs and two wells (springs 44 and 46 and wells 12 and 13) tap only slightly mixed water that retains most of the characteristics acquired in the thermal reservoir. In the Fisher Hot Spring area, the cooling and mix— ing relations are not entirely clear, but much of the CONCEPTUAL MODELS OF THE GEOTHERMAL SYSTEM 143 evidence points toward a reservoir temperature of 170°C or more. The reservoir could be the one that supplies the Crump geyser waters or it could be hydraulically unconnected and have slightly lower concentrations of chloride and slightly higher silica concentrations than the Crump geyser reservoir. The resolution of mixing diagrams for these waters is not sufficiently precise to permit distinctions of this sort to be made. It seems fairly certain, however, that the reser- voir water mixes directly with local warm waters to pro- duce the water of Fisher Hot Spring and spring 16. There is at present no evidence to indicate a large mixing zone in the deeper valley-fill deposits beneath Fisher Hot Spring similar to the zone inferred beneath Crump geyser. Heat-flow holes give clear evidence, however, of a zone of lateral spreading that extends to depths of at least several hundred feet. The areal ex- tent of this zone has not been well defined but is at least several square miles. Conditions similar to those in the Fisher Hot Spring area are also believed to exist at the south end of Warner Valley, where thermal waters having temper- atures of 170°C or more appear to mix directly with local warm waters. The area may differ from the northern areas in that much of the warm water may exist in a fractured fault block that dips into the valley and forms its southern boundary. Little else can be deduced from the data available at this time, and ad- ditional data collected in the area by a private explora- tion company may refine or alter these conclusions in the future. RESERVOIR CHARACTERISTICS Discussions in this report referring to the geother- mal reservoir in Warner Valley have thus far assumed only that circultion occurs along fault planes to depths sufficient to attain indicated reservoir temperatures. Available data neither confirm nor deny the possibility that circulation also occurs in a direction perpendicu- lar to the fault plane in a laterally extensive volume of rock. Some evidence supports this possibility. The great enrichment of oxygen-18 in the reservoir waters rela- tive to recharge waters could require that the thermal waters percolate through large volumes of permeable rock outside the fault zone, especially if leaching has depleted the oxygen-18 in the rocks of the fault con- duits. The similarity of thermal waters on opposite sides of the valley (if the mixing models are valid) may also support the concept of a laterally extensive reser- voir. If a more extensive reservoir exists, the narrow range of reservoir temperatures indicated by the geothermometers and mixing models for waters on both sides of the valley may imply that the reservoir rocks have a fairly small vertical thickness. The known stratigraphy of the area also implies that permeable strata at great depths are likely to be thin, perhaps confined to alternating interfiow zones in basaltic rocks. If rock temperatures are assumed to range from 170°C at the top of the reservoir to 180°C at the bottom, this temperature difference would probably occur over a vertical distance of no more than a few hundred me- ters, assuming that vertical convective flow in the res— ervoir is minimal. For the calculations that follow, a minimum reservoir thickness of 200 meters (650 ft) and a probable maximum thickness of 500 meters (1,600 ft) are assumed. It is convenient to assume that the reservoir is con- fined to the rocks that lie between the main boundary faults, although it may be equally probable that a major reason for the occurrence of the thermal waters at this location is a fortuitous juxtaposition of perme- able rocks on both sides of the two faults. The first possibility seems much the more likely of the two, however, and is assumed to be the case in the following discussion. It may also be assumed that the reservoir supplying Crump geyser and Fisher Hot Spring is confined to a 9-mile length of the valley between Pelican Lake and Hart Lake. This assumption is based on the lack of thermal manifestations to the north and south of this area and on data from the Geological Survey test holes MCI and 0K1 and the San Juan Oil Co. well east of Adel, all of which show less than normal temperatures at their respective total depths of 885, 522, and 7,500 feet. By projecting an assumed 70° dip of the boundary faults in the area underlain by the reservoir, the aver- age width of the reservoir between the faults at a depth of 13,000 feet is determined to be 3.2 miles and the area of the reservoir is, therefore, about 29 mi2 (75 km"). For assumed thicknesses of 650 feet (200 m) and 1,600 feet (500 m), the reservoir volume is calculated to be 3.6 mi3 (15 kmi‘) and 9 mi3 (38 km"), respectively. Utilizing the International System of Units for ease in comparison of results, a porosity of 0.10 is assumed in the reservoir rocks and volumetric specific heat is assumed to be 2.7 joules per cubic centimeter per de- gree Celsius (J cm‘3 °C'1). Volumetric heat above a base temperature of 10°C stored in the reservoir is then calculated to be 6.5 X 10‘8 J for the 200 m thickness and 1.6 X 1019 J for the 500-meter thickness. Applying the 25 percent recovery factor used by Nathenson and Muffler (1975), recoverable thermal energy is esti- mated to be 1.6 x 1018 J and 4 X 1018 J for the two thicknesses of reservoir assumed. I44 Several additional calculations can be made to serve as rough checks on the heat and mass balances implied by the preceeding analysis. For example, assuming that normal heat flow through the valley-fill deposits is 0.075 W m”, the total conductive heat flow in the area underlain by the reservoir (117 kmz) is calculated to be 8.8 X 106 W. The calculated convective heat discharge at or near land surface, 1.8 X 106 W, is, therefore, about one-fifth of the conductive heat flow. The difference be- tween the values is sufficiently large to ensure an adequate conductive heat-flow base from which to de- rive the convective discharge. The convective dis- charge is large enough, however, to account for a prob- able small deficit in the measured heat flows in the central and northern parts of the valley. It may also account, in part, for the lower than normal thermal gradients observed in the San Juan Oil Company well and the Geological Survey test holes MCI and 0K1. A second calculation may be made in order to esti- mate mass flow rates and apparent permeabilities in the reservoir rocks. For this calculation, it is assumed that mass discharge from the thermal reservoir is dou- ble the estimated flow at the surface, or 4 kilograms per second (kg s“). Corrected for density at 170°C, this becomes 4.7 liters per second (L s"). The hydraulic gradient in the reservoir is assumed to be 0.005, com- parable to the near-surface gradient estimated for re- gional ground-water flow. For the width and range of thicknesses assumed for the reservoir, the apparent permeability is calculated to be in the range 0.10 to 0.15 feet per day (ft (1“), which, corrected to surface temperature, is about 2 X 10’2 to 5 X 10‘2 ft d". It could also be assumed that as little as one-fourth or as much as three-fourths of the total flow in the reservoir results in upward flow in the conduits, but in all cases, the calculated permeabilities remain in a range that appears reasonable when compared with published data from other areas. (See, for example, 6 X 10’3 ft d“1 for welded tuffs in Nevada (Davis and De Wiest, 1966, p. 339), 6 X 10‘3 ft (1‘1 for welded tuffs and tuffaceous sediments in Long Valley, Calif. (Sorey and others, 1978), and 2.5 X 10‘4 ft d—1 for moderately dense Hawaiian basalts (Davis, 1969, p. 65). All of the parameters assumed for the above calculations have extremely large uncertainties, and the fair agreement with a few published data is not intended to imply any degree of reliability in the results. The calculations do no more than indicate that the assumptions used may be reasonable ones. At the south end of Warner Valley in the large tilted fault block and the adjacent valley-fill deposits, the reservoir may have an area of about 8 mi2 (21 km2). On the basis of the chemical data, the reservoir may be assumed to occur at the same depth as the northern THE GEOTHERMAL HYDROLOGY OF WARNER VALLEY, OREGON reservoir. For the same range of reservoir thicknesses the stored heat is calculated to be about 3.2 X 1018 J and the recoverable energy is about 8 X 1017 J. The combined recoverable thermal energy in the Crump geyser, Fisher Hot Spring, and south Warner areas is estimated to be about 5 X 1018 J. The estimate is based, as described above, on a single reservoir sup- plying Crump geyser and Fisher Hot Spring and a sep- arate reservoir supplying springs at the south end of the valley. The estimate is regarded as a probable maximum. The probable minimum heat content in the Warner Valley systems is calculated as heat stored only in the major faults and associated fractured rocks in the fault zones. The total heat stored in faults adjacent to Crump geyser, Fisher Hot Spring, and spring 34 at the south end of Warner Valley could be as small as 5 X 1016 J. This calculation includes a total fault length of 25 kilometers, a depth interval of 500 meters and a fault- zone thickness of 10 meters; porosity and heat capacity are the same as assumed for the reservoir calculations above, and the heat stored in the upflow conduits is ignored. The recovery factor for such a system is un- known but could be assumed to be comparable to that calculated for the extended reservoir systems. Thus, recoverable thermal energy for the fault-restricted sys- tem may be about 1 X 1016 J. Additional geothermal resources occur in Coleman Valley and Big Valley, where warm springs or wells indicate probable low-temperature systems. The areas covered by this report may also contain undiscovered resources. The additional geothermal energy in these areas is expected to be small. Unless the geothermal heat in a reservoir can be tapped by wells within economic drilling limits, the thermal energy represents an accessible resource base rather than a resource, to use the terminology em- ployed by Muffler and Guffanti (1979). The recov- erability of the deep reservoir water in Warner Valley is uncertain. However, the hot water that can be ob- tained from shallow wells penetrating the fault zones or the valley-fill deposits represents a low-t0- moderate-temperature resource which could be readily utilized at the present time. Much of the thermal energy stored in the Crump geyser and Fisher Hot Spring areas is available to wells less than 300 meters deep within the valley-fill deposits. In each of these areas, the volume of sedimen— tary deposits containing water at temperatures of 100°C or more may be about one-half km3. If porosity and volumetric specific heat in the valley-fill deposits are conservatively assumed to be the same as in the bedrock, the total thermal energy above a base tem- perature of 10°C is calculated to be about 1 X 1017 J for REFERENCE each area. Assuming 25 percent recovery, the total re- coverable energy in the sedimentary deposits of the upper valley may be about 5 X 1016 J. In the absence of comparable data from the south end of Warner Valley, no estimate has been made for this area. Thermal energy may be recoverable from the deep reservoirs in Warner Valley by drilling into the fault zones in the Crump geyser and Fisher Hot Spring areas. Wells tapping the fault zones should encounter water at temperatures intermediate between the res- ervoir temperature and the boiling temperature at land surface (approximately 96°C). Estimates based on the numerical model results of Sorey (1978) indicate that thermal water ascending and conductively cooling in a fault plane should retain a temperature of at least 155°C at a depth of 2 kilometers for the flow rates and boundary conditions applicable to the Crump geyser area. At a depth of 1 kilometer, the temperature should be at least 140°C, and at 0.6 kilometer, about 130°C. If the flow is confined to a cylindrical conduit, the temperatures at any depth would be considerably higher than those estimated for the fault-plane model. Estimates of the temperature decrease caused by con- ductive cooling in cylindrical conduits and fault planes based on analytical solutions of Nathenson, Urban, and Diment (1979) and Sorey (1978) indicate that thermal upflow in the principal geothermal areas of Warner Valley is probably restricted to lengths of the faults that may approach the dimensions of a cylindri- cal conduit. The analytical results show, for example, that reservoir water having a flow rate of 2 kg s‘1 would cool to about 96°C at land surface in either a cylinder of small dimension or a fault plane less than 400 min length. Greater lengths of fault surface would permit cooling to much lower temperatures than are actually observed in the Crump geyser area, while flow rates much greater than those estimated would appar- ently result in higher temperatures than those ob- served near land surface and would imply that consid- erable boiling must occur at shallow depths. The implications of these conclusions for geothermal exploration and development are clear. Water can probably be obtained at temperatures of 100°C or more by wells penetrating the unconsolidated deposits to depths of 200 to 300 meters within a kilometer or two of the Crump geyser and Fisher Hot Spring areas. Water approaching the estimated reservoir tempera- ture (at least 170°C) is probably available only to wells that tap the fault conduits close to the hot springs. As the foregoing calculations show, water should be available at high temperatures in the upflow conduits within acceptable drilling depths. If a permeable con- duit were to be tapped by a pumped well, the decrease, in pressure caused by pumping would be transmitted to I 145 the reservoir and the flow rate would increase. The result would be an increase in the discharge tempera— ture. Although discharge temperatures would eventu- ally decline somewhat under long-term pumping as the rocks adjacent to the conduits reequilibrated, the final steady-state temperature would be greater than the temperature initially encountered at any given depth. If the sulfate-oxygen geothermometer correctly in- dicates the presence of a reservoir having a tempera- ture of 220°C, the two most probable conceptual models for the system are (1) a deep high-temperature reser- voir that supplies a shallower reservoir in which the thermal waters reequilibrate, and (2) a deep high- temperature reservoir from which the thermal waters ascend to the surface and cool conductively, some of them mixing with cooler waters at fairly shallow depths. In terms of present-day exploration and develop- ment, case 1 would probably not differ in any practical way from the situation discussed above which assumes the presence of only one reservoir having a tempera- ture of 170°C to 180°C. In either situation, drilling costs would probably preclude drilling into the reser- voir and development would occur by tapping thermal. conduits above the 170°C reservoir. Case 2 would differ from case 1 in making possible the withdrawal of hot water directly from the deep reservoir by tapping the upflow conduits. Thus, even though conductive cooling would still occur, temperatures obtained at given depths would be higher than those obtainable in the absence of a 220°C reservoir. Temperatures attainable by tapping the fault con- duits have not been estimated for any of the cases de- scribed above. Factors limiting the temperature would probably be the permeability, volume, and configura- tion of the reservoir or reservoirs and the permeability of the upflow conduits. Aquifer tests in exploration wells tapping the conduits could provide much-needed evidence concerning these factors. The hydraulic system in Warner Valley is probably a relatively simple one, and stresses imposed on the sys- tem during an aquifer test should make it possible to obtain reliable estimates of the temperature, size, and hydraulic characteristics of the geothermal reservoir. These estimates, combined with additional chemical and isotopic analyses, should permit the prediction of long-term responses in the system and lead to an esti- mate of the economic potential of the Warner Valley geothermal resource. REFERENCES Arnorsson, Stefan, 1975, Application of the silica geothermometer in low temperature hydrothermal areas in Iceland: American Journal of Science, v. 275, p. 763-784. I46 Bowen, R. G., and Peterson, N. V., 1970, Thermal springs and wells in Oregon: Oregon Department of Geology and Mineral Indus— tries Miscellaneous Paper 14, 1 p. Christiansen, R. L., and McKee, E. H., 1978, Late Cenozoic volcanic and tectonic evolution of the Great Basin and Columbia Inter— montane region, in Smith, R. B., and Eaton, G. P., eds., Cenozoic tectonics and regional geophysics of the Western Cordillera: Geological Society of America Memoir 152. Coplen, T. B., 1976, Cooperative geochemical resource assessment of the Mesa geothermal system: Final report to the US. Depart- ment of the Interior, Bureau of Reclamation, IGPP-UCR-76-1, Institute of Geophysics and Planetary Physics, University of California at Riverside, 97 p. Couch, R., and Johnson, 8., 1968, The Warner Valley earthquake sequence, May and June, 1968: The Ore Bin, v. 30, no. 10, p. 191—204. Couch, R. W., and Lowell, R. P., 1971, Earthquakes and seismic energy release in Oregon: The Ore Bin, v. 33, no. 4, p. 61—84. Craig, H., 1961, Standard for reporting concentrations of deuterium and oxygen-18 in natural waters: Science, v. 133, p. 1833. Davis, S. N., 1969, Porosity and permeability of natural materials, in DeWiest, R. J. M., ed., Flow through porous media: New York, Academic Press, p. 54—89. Davis, 3. N., and DeWiest, R. J. M., 1966, Hydrogeology: New York, John Wiley, 463 p. Donath, F. A., 1962, Analysis of basin-range structure, South- Central Oregon: Geological Society of America Bulletin, v. 73, p. 1—16. Downs, Theodore, 1956, The Mascall fauna from the Miocene of Ore- gon: California University, Department of Geological Science Bulletin, v. 31, no. 5, p. 199—354. Ellis, A. J., 1970, Quantitative interpretation of chemical charac- teristics of hydrothermal systems: U.N. Symposium on the De- velopment and Utilization of Geothermal Resources, Pisa, 1970, in Geothermics, Special Issue 2, p. 516—528. Evernden, J. F., Savage, D. E., Curtis, G. H., and James, G. T., 1964, Potassium-argon dates and the Cenozoic mammalian chronology of North America: American Journal of Science, v. 262, p. 145— 198. Fournier, R. 0., 1973, Silica in thermal waters: laboratory and field investigations: International Symposium on Hydrogeochemistry and Biogeochemistry, Proceedings, Tokyo, 1970, Washington, D.C., The Clarke Company, p. 122—139. 1977, Chemical geothermometers and mixing models for geothermal systems: Geothermics, v. 5, p. 41—50. Fournier, R. 0., and Potter, R. W., 1978, A magnesium correction for the Na-K-Ca chemical geothermometer: US. Geological Survey Open-File Report 78—486, 24 p. . Fournier, R. 0., and Rowe, J. J ., 1966, Estimation of underground temperatures from the silica content of water from hot springs and wet steam wells: American Journal of Science, v. 264, no. 9, p. 685—697. Fournier, R. 0., and Truesdell, A. H., 1970, Chemical indicators of subsurface temperature applied to hot spring waters of Yel- lowstone National Park, Wyoming, U.S.A., in U.N. Symposium on Development and Utilization of Geothermal Resources, Pisa, 1970: Geothermics, Special Issue 2, p. 529—535. 1973, An empirical Na-K-Ca geothermometer for natural wa- ters: Geochimica et Cosmochimica Acta, v. 37, p. 1255— 1275. 1974, Geochemical indicators of subsurface temperature—— Part II, Estimation of temperature and fraction of hot water mixed with cold water: US. Geological Survey Journal of Re- search, v. 2, no. 3, p. 263—270. Fournier, R. 0., Sorey, M. L., Mariner, R. H., and Truesdell, A. H., 1979, Chemical and isotopic prediction of aquifer temperatures THE GEO’I‘HERMAL HYDROLOGY OF WARNER VALLEY, OREGON in the geothermal system at Long Valley, California: Journal of Volcanology and Geothermal Research, v. 5, p. 17—34. Fuller, R. E., 1931, The geomorphology and volcanic sequence of Steens Mountain in southeastern Oregon: Washington Univer- sity Geology Publication, v. 3, no. 1, p. 1—130. Greene, R. C., Walker, G. W., and Corcoran, R. E., 1972, Geologic map of the Burns quadrangle, Oregon: US. Geological Survey Miscellaneous Geologic Investigations map 1—680, scale 1:250,000. Gregory, D. 1., and Martinez, R. J., 1975, Audio-magnetotelluric ap- parent resistivity maps, Warner Valley, Oregon: US. Geological Survey Open-File Report 75—652, 4 sheets. Hem, J. D., 1970, Study and interpretation of the chemical charac- teristics of natural water: US. Geological Survey Water-Supply Paper 1473, p. 176—178. Hurr, R. T., 1966, A new approach for estimating transmissibility from specific capacity: Water Resources Research, v. 2, no. 4, p. 657—664. Kharaka, Y. K., and Barnes, Ivan, 1973, SOLMNEQ: Solution- mineral equilibrium computations: US Geological Survey Computer Contribution, National Technical Information Service PB 215—899, 81 p. Klein, C. W., and Koenig, J. B., 1977 , Interpretation of analytical results, thermal and non-thermal waters, lava plateaus region of northeastern California and southern Oregon: Second workshop on sampling geothermal effluents, Environmental Protection Agency, Proceedings, p. 166- 173. Langbein, W. B., 1961, Salinity and hydrology of closed lakes: US. Geological Survey Professional Paper 412, 20 p. Larson, E. E., 1965, The structure, stratigraphy and paleomagnetics of the Plush area, southeastern Lake County, Oregon: Ph. D. dissertation, University of Colorado, 166 p. McKenzie, W. F., and Truesdell, A. H., 1977, Geothermal reservoir temperatures estimated from the oxygen isotope compositions of dissolved sulfate and water from hot springs and shallow drillholes: Geothermics, v. 5, p. 51—61. MacLeod, N. S., Walker, G. W., and McKee, E. H., 1975, Geothermal significance of eastward increase in age of upper Cenozoic rhyolitic domes in southeastern Oregon: Second United Nations Symposium on Development and Use of Geothermal Resources, San Francisco, 1975, Proc., p. 465—474. Mahon, W. A. J ., 197 0, Chemistry in the exploration and exploitation of hydrothermal systems, in U.N. Symposium on Development and Utilization of Geothermal Resources, Pisa, 1970: Geotherm- ics, Special Issue 2, p. 1310—1322. Mariner, R. H., Presser, T. S., Rapp, J. B., and Willey, L. M., 1975, The minor and trace elements, gas and isotope compositions of the principal hot springs of Nevada and Oregon: US. Geological Survey Open-File Report, 27 p. Mariner, R. H., Rapp, J. B., Willey, L. M., and Presser, T. S., 1974, The chemical composition and estimated minimum thermal res- ervoir temperatures of selected hot springs in Oregon: US. Geological Survey Open-File Report, 27 p. Merriam, J. C., 1901, A contribution to the geology of the John Day Basin (Oregon): California University, Department of Geology Bulletin, v. 2, no. 9, p. 269—314. 1910, Tertiary mammal beds of Virgin Valley and Thousand Creek in northwestern Nevada, pt. 1, Geologic history: Califor- nia University, Department of Geology Bulletin, v. 6, no. 2, p. 21—53. Muffler, L. J. F., and Gufi‘anti, Marianne, 1979, Introduction, in Muffler, L. J. P., ed., Assessment of geothermal resources of the United States—1978: US. Geological Survey Circular 790, p. 1—7. Nathenson, Manuel, and Muffler, L. J. P., 1975, Geothermal re— REFERENCE 147 sources in hydrothermal convection systems and conduction- dominated areas, in White, D. F., and Williams, D. L., eds., As- sessment of geothermal resources of the United States—«1975: US. Geological Survey Circular 726, p. 104—121. Nathenson, Manuel, Urban, T. C., and Diment, W. H., 197 9, Approx- imate solution for the temperature distribution caused by flow up a fault and its application to temperatures measured in a drillhole at Raft River Geothermal Area, Cassia County, Idaho: Geothermal Resources Council Transactions, v. 3, p. 447—480. Nehring, N. L., and Mariner, R. H., 1979, Sulfate-water isotopic equilibrium temperatures for thermal springs and wells of the Great Basin: Geothermal Resources Council, Transactions, v. 3, p. 485—488. . Nehring, N. L., Mariner, R. H., White, L. D., Huebner, M. A., Roberts, E. D., Haron, Karen, Bowen , P. A., and Tanner, Lane, 1979, Sulfate geothermometry of thermal waters in the western United States: US. Geological Survey Open-File Report 79— 1135, 11 p. Newcomb, R. C., 1958, Yonna formation of the Klamath River Basin, Oregon: Northwest Science, v. 32, no. 2, p. 41-48. Pearson, F. J ., and Truesdell, A. H., 1978, Tritium in the waters of Yellowstone National Park: Short papers of the 4th Interna- tional Conference, Geochronology, Cosmochronology, Isotope Geology, 1978: US. Geological Survey Open-File Report 78—701, p. 327—329. Peterson, N. V., 1959, Lake County’s new continuous geyser: The Ore Bin, v. 21, no. 9, p. 83—88. Phillips, K. N., and Van Denburgh, A. S., 1971, Hydrology and geochemistry of Abert, Summer, and Goose Lakes, and other closed-basin lakes in south-central Oregon: US. Geological Sur- vey Professional Paper 502—B, 86 p. Plouff, Donald, and Conradi, Arthur, Jr., 1975, Gravity and mag- netic profiles and maps, Crump Geyser Area, Oregon: US. Geological Survey Open-File Report 75—346, 2 p., 3 sheets. Renick, B. C., 1930, The petrology and geology of a portion of Malheur County, Oregon: Journal of Geology, v. 38, p. 387—496. Richins, W. D., 1974, Earthquake swarm near Denio, Nevada, Feb— ruary to April, 1973: MS. Thesis, University of Nevada, Reno. Russell, R. J ., 1928, Basin Range structure and stratigraphy of the Warner Range, northeastern California: California University Publications in Geological Sciences, v. 17, no. 11, p. 387—496. Sammel, E. A., 1980, Hydrogeologic appraisal of the Klamath Falls geothermal area, Oregon: US. Geological Survey Professional Paper 1044—G, 45 p. Sass, J. H., Kennelly, J. P., Jr. Wendt, W. E., Moses, T. H., and Ziagos, J. P., 1979, In-situ determination of heat flow in uncon- solidated sediments: US. Geological Survey Open-File Report 79—593, 73 p. Schaff, S. C., 1976, The 1968 Adel, Oregon, earthquake swarm: M. A. Thesis, University of Nevada, Reno, 63 p. Sorey, M. L., 1978, Numerical modeling of liquid geothermal sys- tems: US. Geological Survey Professional Paper 1044PD, 25 p. Sorey, M. L., Lewis, R. E., and Olmsted, F. H., 1978, The hydrother- mal system of Long Valley Caldera, California: US. Geological Survey Professional Paper 1044-A, 60 p. Theis, C. V., 1963, Estimating the transmissibility of a water-table aquifer from the specific capacity of a well, in Bentall, Ray, com- piler, Methods of determining permeability, transmissibility, and drawdown: US. Geological Survey Water—Supply Paper 1536—1, p. 332—336. Trauger, F. D., 1950, Factual ground-water data in Lake County, Oregon: US. Geological Survey Open-File Report in cooperation with Office of Oregon State Engineer, 287 p. Truesdell, A. H., and Fournier, R. 0., 1976, Calculations of deep temperatures in geothermal systems from the chemistry of boil- ing spring waters of mixed origin: Second United Nations Sym- posium on the Development and Use of Geothermal Resources, San Francisco, 1975, Proc.', p. 837—844. Truesdell, A. H., Nathenson, Manuel, and Rye, R. 0., 1977, The effects of subsurface boiling and dilution on the isotopic compo- sitions of Yellowstone thermal waters: Journal of Geophysical Research, v. 82, no. 26, p. 3694—3704. US. Geological Survey, 1976, Water resources data for Oregon, Water year 1975: US. Geological Survey Water-Data Report OR-75-1. Van Winkle, W., 1914, Quality of the surface waters of Oregon: US. Geological Survey Water-Supply Paper 363, 137 p. Walker, G. W., and Repenning, C. A., 1965, Reconnaissance geologic map of the Adel Quadrangle, Lake, Harney, and Malheur Coun- ties, Oregon: U.S. Geological Survey Miscellaneous Geologic In- vestigations map I—446, scale 1:250,000. Walker, G. W., and Swanson, D. A., 1968, Summary report on the geology and mineral resources of the Poker Jim Ridge and Fort Warner areas of the Hart Mountain National Antelope Refuge, Lake County, Oregon: US. Geological Survey Bulletin 1260—M, 22 p. Wallace, R. E., 1946, A. Miocene mammalian fauna from Beatty Buttes, Oregon: Carnegie Institution of Washington Publication 551, p. 113—134. Weide, D. L., 1974, Postglacial geomorphology and environments of the Warner Valley-Hart Mountain Area, Oregon: Ph. D. disser- tation, University of California, Los Angeles, Department of Ge- ography, 160 p. Welch, A. H., Sorey, M. L., and Olmsted, F. H., 1981, Hydrothermal system in southern Grass Valley, Pershing County, Nevada: US. Geological Survey Open-File Report (in preparation). Wells, F. G., and Peck, D. L., 1961, Geologic map of Oregon west of the 12lst meridian: US. Geological Survey Miscellaneous Geologic Investigations map I—325, scale 1:500,000. Whistler, J. T., and Lewis, J. H., 1916, Warner Valley and White River projects: U.S. Reclamation Service, 123 p. White, D. E., 1968, Hydrology, activity, and heat flow of the Steam- boat Springs thermal system, Washoe County, Nevada: US. Geological Survey Professional Paper 458—C, 109 p. 1970, Geochemistry applied to the discovery, evaluation, and exploitation of geothermal energy resources, in UN. Symposium on the Development and Utilization of Geothermal Resources, Pisa, 1970: Geothermics, Special Issue 2, p. 58—80. White, D. E., Muffler, L. J. P., and Truesdell, A. H., 1971, Vapor- dominated hydrothermal systems compared with hot-water sys— tems: Economic Geology, v. 66, no. 1, p. 75-97. GPO 587-0h1/2h PROFESSIONAL PAPER 1044-1 PLATE 1 119°37'30" 42°31' UNITED STATES DEPARTM NT OF THE INTERIOR GEOLOGICAL SURVEY 120°07'30" 5' R.23E. R.22E. 120°00' R.23E. R.24E. 4231',_ _ .. 1- _ y .1 ._ .. , ,p __ , .Wd .25E. 1" /1 R. 26 E. .‘ 1 111 e I , ‘2 '11,. Thunderb {d' V \\ . Ire z’ .11- ? ride bird 11-1 rm, EXPLANATION Nonflowing well. Nonflowing well, sampled for chemical or isotope analysis. Well which normally flows at land surface. Flowing well, sampled for chemical analysis. US. Geological Survey heat—flow hole. 21 OJ Perennial spring. Tail indicates direction of flow. ' , ' 1 I 1 " 1' ' W i ‘ ~V '1 I " 7 ' : ._ I ‘l . ‘ ’ r ' , 1 / ‘ ' _L , ‘\ ‘ .4611: .JJPerennialspring,sampledforchemicalanalysis. ) , . m . f' " _ 1 ‘ ~ . _ 7 / \ 1 ‘ _1 1; " 1 q \ 1 11/ , p ’ _. 1‘ 1‘ ‘ ‘ _ r‘ Reservoir ‘1 A Surface water site, sampled for chemical analysis. g , '1 ’l 7’ I ,, ‘ ~. , ’ , ' _, , ' I ‘ , " I i 7 1 ‘ . 1 ' ‘ ' ran-11..., \* Depth (ft) Log available 240L Temperature (°C) Specific conductance 470 O 17 C = reportedly cold (pmho/cm at 25°C) 4505 W = reportedly warm Altitude of water surface (ft above NGVD of 1929) Depth (ft)\ Log available ‘ , . Conductive temp. gradient 40.5 E 17.5 Maximum temperature (°C) 1' T " ' 1' ’ ‘ 1 ' ' ” fl ‘ . ' . . “ ‘ ‘ ' " * ‘ I 1 ; _ ' ’ - ~. .' . \ 1 1' j ‘ \ _ (°C/km) 4482 Altitude of water surface (ft above NGVD of 1929) Specific conductance 366 14 Temperature (°C) (umho/cm at 25°C) \ - Mair 1 A A Specific conductance 105 6 Temperature (°C) (pmho/cm at 25°C) SCALE FOR STIFF DIAGRAM Secluence No. S = surface-water sample 112’: Spring ernolte , > 1 \ , 1 . , Ann/hon): Springs 51"“8 1' \ _ hallcw » ‘ ‘ ' ~ 1 ' ' 1" 1 a ‘ Lake *1 * 11 2 1 1 2 M illiequivalents per liter School Section Lake Stiff diagram r"\ l \ \__/ Area of siliceous sinter rat-Ir Hiir’ / Dyl 1.1154 _ Parsrup > Springs Little, ' Parsnrp L: ‘/ T. 38 S. « hump . New}: , T. 39 S. Eurr‘l - rap: iug _ - Head- win-whole ‘ 4 1 1 .18611’3 12 451 {10L Q r > ‘1 \Iugllperl‘llm /5Ering \ ., 1 1 . . , Wakqneia ‘ v . 1 1 1w 1M1 I 1' Cabin ,. ' 3 T. 39 S. T. 40 S. Twenty-«ire W1 .1 e , 511311qu ' MW. 175461 Kyhuya Wareulvolu \ 1 11/11:; I NW1 \\-. T. 40 S. .1113 Hale RaseKvolr \ \ ‘ . - T.4‘l S. \’ . I hwy 1 1 l caceiejeu‘ard ‘11 512118 \foI‘Or/e éweyor ’.Spr1ng Antelope J. \\«”%l 1.91% .. 1 1‘ 5 ,1, 342', I ‘1 I 19 ,.' .7 , » 42000: R. 25 E. R. 26 E. 119°37'30" 42°00’ IR ‘3 E I I, " " 1 -. ' ‘ ~1 ‘ 1 1 .2 o I u . 2 . 120 O7 30 INTERIOR -— GEOLOGICAL SURVEY, RESTON, VA -— 1981 - W81674 Base from U.S.G.S. 7.5’ Advanced field completion sheets: Calderwood Reser- . SCALE 1552 500 voir, 1968, 20 ft; Blizzard Gap, 1968, 20 ft; Coleman Lake, 1968, 20 ft; May 1 y 0 1 2 3 4 5 MILES Lake, 1968, 20 ft; Collins Rim, 1968, 40 ft; Mud Lake Reservoir, 1968, 20 ft; I ,_‘ ,_{ ,_f H H ,_______.____, ;—__. I—I Adel, 1968, 20 ft; Sage HendButte, 1968, 20 ft; JacgbsDReserlvoir, 1967, 20 gt; E Crump Lake, 1967, 20 ft; Pri ay Reservoir 1967, 20 t; rake eak, 1968 20 t; 0 Hart Lake, 1967, 20 ft; Plush, 1967,10 ft; Rabbit Hills sw, 1967,20 ft; Flagstaff g 1H H ,f’H H0 1 2 3 “,‘__f’ ‘7‘_.__47 K'LOMETERS OREGON Lake,’ 1967, 20 ft; Warner Peak, 1967, 20 ft; Campbell Lake, 1967, 20 ft; Drake g Peak NE, 196$, 20 ft; Cooper Draw, 1968, 20 ft; Oregon. Compiled Menlo Park CONTOUR INTERVAL 40 FEET Base Map seem” (1'75X15'37) 33mm L APPROXIMATEMEAN DOTTED LINES REPRESENT 20-FOOT CONTOURS DECL'NAT'ON'm‘ NATIONAL GEODETIC VERTICAL DATUM OF 1929 AREA OF MAP MAP SHOWING LOCATIONS OF SPRINGS, WELLS, AND HEAT-F LOW HOLES, AND STIF F DIAGRAMS FOR WATER SAMPLED FOR C *MICAL ANALYSIS, WARNER VALLEY, OREGON 5 Hydrology and Geochemistry of Thermal < Ground Water in Southwestern Idaho and 7 North-Central Nevada ’ ‘ By H. W. YOUNG and R. E. LEWIS L’ GEOHYDROLOGY OF GEOTHERMAL SYSTEMS p 71’ GEOLOGICAL SURVEY PROFESSIONAL PAPER 1044—] UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1982 UNITED STATES DEPARTMENT OF THE INTERIOR JAMES G. WATT, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Young. H, We (Harold William), 1942- Hydrology and Geochemistry (Geohydrology of Geothermal Systems) (Geological Survey Professional Paper 1044-1) Bibliography: 20 p. I, Geothermal Resources—Idaho. 2. Geochemistry—Idaho l, Lewis, R. E. (Robert Edward). 1935- ll. Title. "L Series. W. Series: Geological Survey Professional Paper 1044—] GB] l99t7tl2Y68 553.7'09796‘2 81—6817 AACRZ For sale by the Branch of Distribution, US. Geological Survey, 604 South Pickett Street, Alexandria, VA 22304 1- A 5», CONTENTS Page Abstract ............................................................................ J1 Introduction ........................................................................ 1 Description of study area ........................................................ 1 Purpose and scope .............................................................. 2 Acknowledgments ............................................................... 2 Previous investigations ........................................................... 2 Well; and spring-numbering system ----------------------------------------------- 2 Geologic and hydrologic setting ....................................................... 4 Geochemical characteristics of the hydrothermal system ................................ 4 Chemical character .............................................................. 4 Chemical geothermometers ....................................................... 6 Isotopes ........................................................................ 3 Tritium .................................................................... ‘8 Deuterium and oxygen-18 .................................................... 10 Thermal ground-water occurrence and movement ...................................... 12 Temperature distribution and heat flow ............................................... 15 Conceptual model of the system ...................................................... 17 Summary --------------------------------------------------------------------------- 18 Selected references .................................................................. 19 ILLUSTRATIONS PLATE 1. Map showing generalized geology of southwestern Idaho and north-central Nevada -------------------------------------------------------------- In pocket 2. Map showing generalized potentiometric surface of thermal water and hydro- logic-data site locations in southwestern Idaho and north-central Nevada -- In pocket FIGURES 1—4. Diagrams showing: Page 1. Well- and spring-numbering system -------------------------------- J3 2. Comparison of reservoir temperatures estimated by the silica and sodi- um-potassium-calcium geothermometers -------------------------- 8 3. Tritium content of water assuming piston-flow and well-mixed ground- water systems in Yellowstone National Park ----------------------- 9 4. Relations between concentrations of deuterium and oxygen-18 -------- 10 TABLES Page TABLE 1. Chemical analyses of water from selected wells and Springs .................... J5 2. Estimated reservoir temperatures and free energy of formation for selected hot- water wells and hot springs ................................................ 7 3. Tritium in water from selected wells and springs .............................. 9 4. Stable-isotope analyses of water from selected wells and springs ................ 11 5. Record of wells ............................................................ 13 6. Estimated thermal-water discharge, on the basis 0f enthalpy ................... 16 III IV CONTENTS CONVERSION FACTORS The following conversion table is included for the convenience of those who prefer to use the SI (International System of Units) rather than the inch-pound system of units. Chemical data for concentrations are given in mg/L (milligrams per liter) or ug/L (micrograms per liter), which are, within the range of values presented, numerically equal to parts per million or parts per billion, respectively. Thermal parameters are reported in “working” units. Multiply inch-pound unit By To obtain SI unit Length inch (in.) 25.40 millimeter foot (ft) 0.3048 meter mile (mi) 1.609 kilometer Area acre 4047 square meter square mile (miz) 2.590 square kilometer Volume acre-foot (acre-ft) 1233 cubic meter cubic foot (fti‘) 0.02832 cubic meter cubic mile (mi3) 4.168 cubic kilometer Volume Per Unit ’l‘inre gallon per minute (gal/min) 0.06309 liter per second cubic foot per second (fta/s) .02832 cubic meter per second Multiply working unit By To obtain SI unit Energy and Power calorie (cal) 4.187 joule calorie per second (cal/s) 4.187 joule per second; watt Heat Flux Density microcalorie per square 4.187>< 10'2 watt per square meter centimeter nsecond (Meal/cm2 . s) heat flow unit (HFU) Thermal Conductivity calorie per centimeter 4.187 X 102 watt per meter odegree second - degree Celsius kelvin cal/(Cm . s.°C) Temperature Conversion of °C (degrees Celsius) to °F (degrees Fahrenheit) is based on the equation °F= (1.8) (°C) +32. Water temperatures are reported to the nearest one-half degree. rev HYDROLOGY AND GEOCHEMISTRY OF THERMAL GROUND WATER IN SOUTHWESTERN IDAHO AND NORTH-CENTRAL NEVADA By H. W. YOUNG and R. E. LEWIS ABSTRACT The study area occupies about 14,500 square miles in southwestern Idaho and north-central Nevada. Thermal ground water occurs under artesian conditions, in discontinuous or compartmented zones, in ig- neous or sedimentary rocks of Tertiary age. Ground-water movement is generally northward. Temperatures of the ground water range from about 30° to more than 80° Celsius. Chemical analyses of water from 12 wells and 9 springs indicate that nonthermal waters are a calcium bicarbonate type; thermal waters are a sodium carbonate or bicarbonate type. Chemical geothermometers indicate probable maximum reservoir temperatures are near 100° Celsius. Concentration of tritium in the thermal water is near zero. Depletion of stable isotopes in the hot waters relative to present-day meteoric waters indicates recharge to the system probably occurred when the climate averaged 3° to 5° Celsius colder than at present. Temperatures about 3.5° Celsius colder than at present occurred dur- ing periods of recorded Holocene glacial advances and indicate a resi- dence time of water in the system of at least several thousand years. Residence time calculated on the basis of reservoir volume and ther- mal-water discharge is 3,400 to 6,800 years for an effective reservoir porosity of 0.05 and 0.10, respectively. Preliminary analyses of car- bon-14 determinations indicate an age of the hot waters of about 18,000—25,000 years. Water is heated to temperatures between 80° and 100° Celsius by circulation in the Idavada Volcanics to depths of about 5,000 feet. Water of about 80° Celsius occurs in wells about 2,900 feet, which sug- gests transfer of heat convectively within the Idavada Volcanics. Heat flux into the system is about 8.7 ><107 calories per second. With about 4.3)(107 and 1.6><107 calories per second discharged by conduction and thermal springs, respectively, 25.0 cubic feet per second, or about 18,000 acre-feet per year of 50° Celsius water is required to transport 2.8><107 calories per second advectively in ground water moving northward out of the area. The proposed conceptual model for the area is one of an old system, where water has circulated for thousands, even tens of thousands, of years. Within constraints imposed by the model described, reservoir thermal energy for the geothermal system in southwestern Idaho and north-central Nevada is about 130><1018 calories. INTRODUCTION Thermal ground water occurs over a large area south of the Snake River in Owyhee and Twin Falls Counties in southwestern Idaho and northern Elko County in north- central Nevada (pl. 1). The area supports a rural popula- tion dependent primarily on irrigated agriculture and ranching. Wells drilled for irrigation generally yield ther- mal water, which ranges in temperature from 15° to more than 80°C. For use in this report, thermal waters that have temperatures between 40° and 81°C are re— ferred to as hot waters; waters that have temperatures between 20° and 30°C are referred to as warm waters. Cold, nonthermal water in the area is generally 12°C or less. Except for some space heating of homes locally, development of the thermal ground-water resource in the area has been largely for irrigation; consequently, ther- mal energy associated with the resource is unused. Although temperatures of the thermal water are only moderate, because of the very large reservoir volume, the available thermal energy in the geothermal system is enormous. The most recent assessment of the geother- mal resources of the United States (Brook and others, 1979) indicates that the geothermal energy contained in the hydrothermal systems in this area is the largest out- side of Yellowstone National Park and represents almost one-third of the energy in known systems above 90°C. This study is an effort to further evaluate this large resource and is part of a comprehensive program by the US. Geological Survey to better understand the nature and occurrence of the geothermal resources in the Nation. DESCRIPTION OF STUDY AREA The study area includes about 14,500 mi2 in Owyhee and western Twin Falls Counties in southwestern Idaho, from the Oregon border eastward to Twin Falls, and northern Elko County in north-central Nevada (pl. 1). Included in this study is the 1,100 mi2 Bruneau-Grand View area described by Young and Whitehead (1975) in northern Owyhee County. (See pl. 1 for boundaries of Bruneau—Grand View area.) Later references are made to a “Bruneau—Grand View area” by White and Wil- liams (1975), who stated that areal boundaries include 869 mi2 on the basis of “surface manifestations, geophysical data, well records, and geologic inference.” Brook and others (1979, p. 34) stated that the Bruneau-Grand View area consists of 573 mi2 on the basis of “twice the identified component.” The Bruneau and most of the Castle Creek KGRA’s (known geother- mal resource areas) are within the study area. Climate in the region is semiarid, characterized by J1 J2 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS hot, dry summers and cool winters. Topography is varied. The Jarbidge and Owyhee Mountains rise to altitudes of 10, 185 and 8,122 ft, respectively. Much of the area consists of extensive plateaus, which range in altitude from 3,000 to 7,000 ft. Lowlands include the Snake River valley, where altitudes range from 2,000 to 3,800 ft. Perennial and intermittent streams drain the area. Principal drainages include the Bruneau and Owyhee Rivers and Salmon Falls Creek. These streams head in the Owyhee and Jarbidge Mountains and flow north across the plateaus in deeply incised canyons to the Snake River. PURPOSE AND SCOPE The purposes of this report are to describe the hydrology and geochemistry, to better define the areal limits of thermal water in southwestern Idaho and north- central Nevada, and to develop a conceptual model of the system that contains the thermal water. Water samples from 12 wells and 9 springs were col- lected for chemical analyses. Constituents analyzed in- clude the common ions and silica, and the minor ele- ments arsenic, boron, lithium, and mercury. Additional samples for isotope analyses were collected from wells and springs that were representative of thermal and non- thermal waters in the area. Six wells and six springs were sampled for tritium and four wells and nine springs for deuterium and oxygen-18. In addition, water levels were measured or reported values were obtained for 148 wells, and a generalized potentiometric-surface map was con- structed. Where possible, water temperatures were measured or reported values were obtained. For all thermal-water wells and springs sampled, reservoir temperatures were estimated using the silica and Na-K- Ca (sodium-potassium-ca1cium) geothermometers. Ratios of selected chemical constituents, concentrations of deuterium and oxygen-18, and tritium in sampled waters were used to characterize and thereby distinguish water from different aquifers. ACKNOWLEDGMENTS Many farmers and landowners in Owyhee and Twin Falls Counties in southwestern Idaho and Elko County in north-central Nevada cooperated fully in this study. The authors express appreciation to these residents for allowing water-level measurements and water-quality samples to be obtained from their wells and for allowing access to their property. Several colleagues within the Geological Survey contributed significantly to this in- vestigation: A. H. Truesdell and N. L. Nehring provided sulfate-isotope analyses; discussions with R. H. Mariner aided in interpretation of geochemical data; T. A. Wyerman performed tritium-isotope analyses; and M. Nathenson assisted with calculations for converting hot- water heads to their cold-water equivalents. PREVIOUS INVESTIGATIONS Data from known hot springs in southwestern Idaho and north-central Nevada are included in Stearns, Stearns, and Waring (1937). Historical data from ther- mal springs and wells in southwestern Idaho were sum- marized by Ross (1971); however, few water-chemistry data were presented. Young and Mitchell (1973) gave chemical analyses from 16 hot-water wells and 6 hot springs in Owyhee and Twin Falls Counties in their evaluation of thermal water in Idaho. They recom- mended that additional studies be made in 23 areas of the State where aquifer temperatures greater than 140°C were indicated. 0n those recommendations, Young and Whitehead (1975) inventoried and sampled 94 wells and springs in the Bruneau-Grand View area. In addition to compilation of chemical data, their report contains (1) a description of the areal extent and chemical character of the thermal water, (2) estimates of reservoir tem- peratures using chemical geothermometers, (3) a description of geophysical data available for the area, (4) a description of geology, and (5) a description of the probable source of the thermal water. Rightmire, Young, and Whitehead (1976) collected and analyzed isotopic and geochemical data for the Bruneau—Grand View area to determine the possible source of recharge to the hydrothermal system. On the basis of an estimated sub- surface temperature of 145°C and reservoir volume of 810 mi3, Renner, White, and Williams (1975, p. 38—39) calculated that heat stored in the Bruneau—Grand View geothermal system was about 263x10“ cal. A more re- cent estimate by Brook and others (1979, table 6) places the stored heat at 107i26X1018 cal, on the basis of a mean reservoir temperature of 107i6°C and reservoir volume of 43921101 mi3. Young, Lewis, and Backsen (1979) estimated the discharge of thermal ground water and associated convective heat flux from wells and springs in the Bruneau—Grand View area in 1978 to be 50,500 acre-ft/yr and 4.97X107 cal/s, respectively. WELL- AND SPRING-NUMBERING SYSTEM The well- and spring-numbering system (fig. 1) used by the US. Geological Survey in Idaho indicates the location of wells or springs within the official rectangular subdivision of the public lands, with reference to the Boise base line and meridian. The first two segments of the number designate the township and range. The third segment gives the section number, followed by three let- J3 THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA 4S—1E—34BADI ng system. spring-numberi .—Well- and FIGURE 1 J4 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS ters and a numeral, which indicate the 1A: section (160- acre tract), the IA- 1/4 section (40-acre tract), the 14-1/4-14 section (lo-acre tract), and the serial number of the well within the tract, respectively. Quarter sections are let- tered A, B, C, and D in counterclockwise order from the northeast quarter of each section. Within the quarter sections, 40—acre and 10—acre tracts are lettered in the same manner. Well 4S—1E—34BAD1 is in the SE 1ANEWNWM: sec. 34, T. 4 S., R. 1 E., and is the first well inventoried in that tract. Springs are designated by the letter “S” following the last numeral; for example, 14S-14E—11CAB1S. Wells and springs in Nevada are located in the same manner with reference to the Mount Diablo base line and meridian. GEOLOGIC AND HYDROLOGIC SETTING The rocks of southwestern Idaho and north-central Nevada range in age from Precambrian to Holocene and represent a varied and complex geologic history. Tertiary lava flows ranging in composition from rhyolite to basalt form much of the uplands and plateaus. Tertiary and Quaternary sedimentary rocks cover most of the lower areas and are locally overlain by Quarternary basalt. The cores of the Owyhee and Jarbidge Mountains are composed of sedimentary and metamorphic rocks of Precambrian and Paleozoic age, intruded by granitic rocks of Mesozoic and Tertiary age. For purposes of this report, the geology of the area is generalized and divided into six units. (See pl. 1 for areal extent and description.) Each unit is delineated on the gasis of age, origin, and lithology. Three of the geologic units are related to occurrence of thermal ground water: sedimentary rocks of Quaternary and Tertiary age, basalt of Quaternary and Tertiary age, and silicic volcanic rocks of Tertiary age. The sedimentary unit, consisting chiefly of rocks of the Idaho Group, contains thermal ground water in the northern part of the study area. This sequence of fluvial and lacustrine deposits, ranging from clay to gravel, has a maximum thickness of about 2,000 ft near Oreana and Grand View, Idaho (Young and Whitehead, 1975). The basalt unit, which consists chiefly of Banbury Basalt, is exposed throughout the study area. This unit is also widespread in the subsurface and, along with rocks of the Idavada Volcanics, contains most of the thermal ground water. The thickness of the basalt unit varies and is unknown in many parts of the area. In the Bruneau—Grand View area, Young and Whitehead (1975) indicated that the unit ranges in thickness from less than 100 ft to nearly 1,200 ft. The silicic volcanic rock unit, which includes the Idavada Volcanics, is also exposed throughout the area south of the Snake River Plain and may underlie the en- tire study area. Thickness of the Idavada Volcanics is probably 2,000—3,000 ft; however, the total thickness of the silicic volcanic rock unit is largely unknown and may be as much as 7,000 ft (McIntyre, 1979). Low concentra- tion of magnesium in the thermal waters seems to favor silicic volcanic rocks as the principal reservoir rock in the regional thermal ground-water system. Due to head dif- ferentials, some thermal water probably moves upward into the overlying Banbury Basalt. A system of northwest-trending faults, particularly in the northern part of the area, has fractured and dis- placed the rock units. Most of the faulting probably oc- curred in early Pliocene time and progressively diminished through Pleistocene time. Thermal ground water in southwestern Idaho and north-central Nevada generally occurs under artesian conditions in both the volcanic and sedimentary rocks. The artesian head is variable and may range from a few feet above the water-bearing zone to several hundred feet above land surface. Generally, large well yields and high- temperature waters are from wells penetrating the volcanic-rock aquifers. These wells range from 600 to 3,600 ft in depth, and the water temperatures range from about 30°C to more than 80°C. Wells penetrating the sedimentary~rock aquifers yield smaller amounts of lower temperature water. Climate in most of the area is semiarid. Plateau and valley areas receive less than 15 in. of precipitation an- nually (Thomas and others, 1963). The highest parts of the Owyhee and J arbidge Mountains to the south receive as much as 50 in. annually and may be the recharge areas for the thermal aquifers. GEOCHEMICAL CHARACTERISTICS OF THE HYDROTHERMAL SYSTEM CHEMICAL CHARACTER Water samples for chemical analyses were obtained from 12 wells and 9 springs. Results of the analyses are included in table 1. Additional chemical analyses of thermal and nonthermal waters from the area are given in Rightmire, Young, and Whitehead (1976, table 2). Generally, the waters may be divided into three prin- cipal types on the basis of proportions of major dissolved constituents and concentration of dissolved solids: (1) Nonthermal water from springs, located mostly at high altitudes in the Owyhee and Jarbidge Mountains to the south, (2) hot water from springs and wells, and (3) warm water from springs and wells, probably representing a mixture of hot water and cooler, local ground waters. The nonthermal waters are generally calcium bicar- bonate in character. Only one spring water exceeded a dissolved-solids concentration of 120 mg/L. All other nonthermal waters sampled had a concentration of dis- J5 THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA .33: 3332: can .950» 535:3? .ma nan wusumuuasmu you»: Haw: Mo mnaumm ms» an wumconuaUAQ wan muncoA~mu mm uwusaduumwm avauwamxam Hmuokfl A: OS 03 m 3. SJ SN 3 m.m S 3 Nm 2 3 TN 2 «a E u. a.” o m 93 m.m 2; .. 2-34 -- BESTuEéS HZ cm c2 2 SJ 3. NE 2 D4 5 o: o o: m4 h. 3 I m; 3 c an o4” m.» SN ow 31%;. 235???sz H; 3 cm H 8; 2. fix 3 m.~ 2 9.2 o 2: ma h. S 2 «A mm o 02 can m; Num 1 EA -2 SN; SmuhTmmwlzz H on em a mo. 3. SN o2 u.m o.~ 5‘. «m m 3 H.~ In m» cm H; 2 o 2 mi“ ma m2 om m7m~$ 33:-3 $3 on E m 3. as E Z a.” 3 an as o o: Tm ~A E on 3 S m OS ”13 a; mmm m2 m72$ mmm 25: -mm -m: A; 8 o: 3 8. we. Eu : 3 1m 2 e: 3 cm m. 3 mm om HJ m4 o n W: Tm 3m com; meTw maumommumfi -mNH H; cm 3 a SJ o4 EN mm a; 3 mm mm o o2 a; >4 3 mm m.m wm o S 0.3 9m SM 82; Elfilw a; 883$ng HC 3 c3 ow S. 8. m: an E om «m o2 2. on HA mm 3 o: H; N; o m mg: fim So S E; IN maogméiém H] on 02 cm 8. E. mun Z 2 Ta m~ OS 3 on m.~ 2 mm 2: H; 1N o m oJm Tm m2 o2 EATo a; 31mm Imw L? H7 m 2: Z 8. 2. Zn 2 «a as 2 on p 3 N; m.m 3 mm H; 5w o 5 man fim EN cmo.m meTo 2c; :35 .mm .3 H. 3 33 v S. me. o: o: 2 3 3 OS 2. 2. Ta 3 3 OS A; ¢.~ o w 9; ma «.3 S: 2-2-w EQN 383$: -mm 1 3 Z 3. 3. a: 2 m. o; 1.. 3 o 3 ~6 N4 3 S a. 2 m cm 92 m.» 2; no: 3-2; ooo.~ :34; -5 .mm H. S .2 an 2. 3. 3n E 2 3 3 92 mm m: w. 2 mm a: H; H; o m T: «.0 c2 13 EATw SQN EmmeTE .wv Hf. m a: h 8. 2. N: E 54 Tu mm o: 3 S o.~ 2 mm mm 1v a.» o 2 0.3 mg mg 334 EATw :3; 323...: -m~ 1. 3 Sn . N S. 8 men 3 S 2 2. 93 on 2: NJ on mm o2 1v .1” o w 93 «4. 2% c2 fiATw 9:; :62 4a -3 as a saw H 8.0 .56 o: 3 2 «N 2 2: mm o: o.~ NM 3 OS 1o a; c h 0.3 «a 2% q mTfié 235:: .E mwuu30m Huiuwza V 3 3 1 SJ 3. mv~ 2 a. m4 F; can a o5 w. H. m m.m 3 mo 3 3m m6 m; a: «N EéNIw magafllmmmémq H; N m Av He. 90. 2. mm H. v. h.N HA 0 MA w.N m. Hm via ~.v m.H o m o.¢ A.w mm A: mhIhNIw m~uu OMA v mwlw IF «muumUmalmm Imm A. c om N S. S. S S . c.~ NA 2 o S F. q. 2 H; e.« 3 c an ma Hg 2: w MTN é VERSE: -mm N5 5 cm H 8.0 m.“ 3 on fie mg m.» 2 0 cu o.~ m5 3 ~.m «4 9m 0 om 0.: fir S S MTN A umdaoiém -mm mmuu=On Hmauusunoz ) .A d S a u 0.8 S w n m. w w um m. m. m w m w u m. u m. a m. w m m m a H d m n E m 1 1 x s o 0 3 s I n I I X x a 3 P 1 P 5 I u 1 3 a u a o o I o u o a s z 1 s r. o o 3 2 q a e r. a r. u a a P a m. I M x u m m u w. w mu. m w u u. u u m 1 s m w m H m m w 1 ; 1 o a fi A m ) a o 0 a n P P a u a W W _ 3 I. m q s 3 m. n 3 m.W n a 1 z a ) a a t, 3 u 2 ) m o s a a a u 1 ) ) ( ) n (d P S ) 3 a 9 m P s N ) u m o 0 PM U 1 V s I I ) ) S A 3 s o e 3 E a d o I a m b I. ) s I n s 0 d D O ) a ) o D. ( ) u. 3 s a m T. ST. 3 ( ( n ( s o z ( I . E D x x I H ( a J u a n I T b I ( ( ( s 0 ( d E 3 e P a 1 I ) ) / ) m u I c m a m ( q 9 3 n I M ED n n 1 n 2 I. D. O ( D I. E D n D W I. as u b 5 ( b e 3 s e l O o W cm H a I 0 ad n w w w I u ) w c u u u w u fu. w ( ( ( E 3 a c 1 J a ) a ) e s a e e s a a 9 J I 3 s D 1 d E a I. ( M ( s n o m I N u > m 5 / m m m ~=nnu mafia .v ukuuomwu mumv 0: .II “uumadxounmm m .couo: anus: umouxw .hwuwa uwm nESHU«HHHE :« munwauwumnuo Havafimsow m??% 33 «:2: 39833 Set .533 x: 333:: BSEBSI A mqmfib d < #. A _ ) ‘ y p > p n y t l . . I . by A A A u ‘ ‘ . _ ‘ r o v, > t » r p b x J I! A! ,‘r I‘ I a II ‘y“‘ l‘ ptl ‘ ‘ r J6 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS solved solids of 120 mg/L or less. Hot waters do not seem to be mixed and are, therefore, representative of the hot- test water from the deep convective system. These waters are generally sodium carbonate or bicarbonate in character, having concentrations of dissolved solids less than about 400 mg/L. Concentrations of chloride and fluoride in the 80°C water are about 15 mg/L; concentra- tion of boron is about 550 ug/L. Warm waters are mostly sodium bicarbonate in character, and the concentration of dissolved solids is generally less than about 300 mg/L. In the vicinity of Oreana and Grand View, however, warm waters sampled by Rightmire, Young, and Whitehead (1976) had con- centrations of dissolved solids as high as 1,160 mg/L. All the warm waters that had high concentrations of dis- solved solids were sampled from wells producing from sedimentary rocks of the Idaho Group. The relatively dilute nature, the wide range in con- centration of dissolved solids, and the variability in con- centrations of chloride and boron in all the thermal waters indicate a large, complex, interconnected hydrologic system. The vertical multiplicity of aquifers, along with areal compartmentalization created by faulting, present an almost infinite number of avenues for deep circulation of thermal water. The depths and duration of circulation, contact with different rock types, and mixing with shallow, local ground waters serve to produce the variations in temperature and water chemistry evident in the thermal waters. ACHEMICAL GEOTHERMOMETERS Reservoir temperatures estimated using the silica and Na-K-Ca geothermometers are valid only for hot-water systems and only if the following basic assumptions are met: (1) The chemical reactions at depth are tempera- ture dependent; (2) an adequate supply of chemical con- stituents used for the thermometry is present in the aquifer; (3) chemical equilibrium is established between the hot water and the specific aquifer minerals; (4) re- equilibration of the chemical composition of hot water as it rises to the surface is negligible; and (5) hot water rises rapidly to the surface with no dilution or mixing of hot and cold waters (White, 1970). Reservoir temperatures in the study area were es- timated using the silica geothermometer (Foumier and Rowe, 1966) and the Na-K-Ca geothermometer (Four- nier and Truesdell, 1973). Silica equilibrium with both quartz and chalcedony was considered assuming conduc- tive heat loss. No magnesium correction was applied because the Na-K-Ca geothermometer indicated temperatures of 74°C or less for samples containing ap- preciable (greater than 1 mg/L) magnesium. (Fournier and Potter (1979) suggested that a magnesium correc- tion should not be attempted if the Na-K-Ca geother- mometer indicates a temperature of less than 70°C.) Dissolved silica (SiOz) reported in chemical analyses is actually present as silicic acid (H4Si04) and various amounts of dissociated species (particularly H38i04‘). In alkaline waters, hydroxide (OH‘) reacts with the silicic acid to reduce the proportion of silicic acid to total dissolved silica: Silicic Dissociated acid Hydroxide silicic acid Water H4SiO4 + OH' = HasIO4_ + H20. The pH of thermal waters sampled ranged from 7.3 to 9.6. For waters that have pH values greater than about 8.2, the total concentration of dissolved silica measured in the laboratory (H48i04+H38i04‘) was reduced by the calculated concentration of H38i04‘ to obtain a better estimate of the temperature of the thermal reservoir (Brook and others, 1979). Estimated reservoir temperatures for 11 hot-water wells and 5 hot springs sampled in the study area are given in table 2. Locations of the sample sites are shown on plate 2. In general, most of the temperatures esti- mated using the Na-K-Ca and the H38i04‘ corrected quartz geothermometers are in good agreement. Tem- peratures estimated using the H38i04’ corrected chal- cedony geothermometers generally are less than tem- peratures estimated using the Na-K-Ca geothermome- ter and, occasionally, were less than the water tempera- ture measured at the surface. Figure 2 is a comparison of reservoir temperatures es- timated by the HaSiOi‘ corrected quartz and Na-K-Ca geothermometers. Waters that plot on or very near the equal-temperature line are likely to be unmixed waters or waters that have reequilibrated with the aquifer minerals after mixing (Foumier and others, 1979). If the calculated temperature is significantly higher than the measured temperature for these unmixed or re- equilibrated waters, some heat has probably been lost conductively as the water moves through the aquifer. Waters that plot significantly above the equal- temperature line have undergone evaporation or contain silica dissolved from amorphous aquifer material. The Na—K-Ca geothermometer uses ratios of constituents and is less sensitive to concentration changes that occur during evaporation than is the silica geothermometer. Waters that plot significantly below the equal- temperature line may have had (1) mixing with another type water without sufficient time to equilibrate with the surrounding rock (in mixed waters, silica estimated temperatures are generally decreased more than Na-K- a) THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J7 TABLE 2.—Estimated reservoir temperatures and free energy of formation for selected hot—water wells and hot springs Estimated reservoir temperatures, in degrees Celsius, based on geothermometers Silica- Water quartz Silica— temperature conduc- 1 chal- Free energy of formation 2 at the hive, Sodium- cedony Reference Spring or surface “3510,“ potassium- H35i0.' number well number (°C) corrected calcium corrected Aragonite Calcite chalcedony Quartz (fig. 2) IN— 3W—21ACDIS 56.0 96 90 66 0 7 0 8 0.2 0 6 l 15- 2W- 7CCBl 45.0 85 90 54 1 2 .1 7 2 2S— 2W—35ACBl 40.0 81 81 49 9 9 .1 7 3 4S— 1E-34BADl 76.5 82 80 51 2 3 -.4 1 4 SS- 4W- 8ADA1 20.0 95 72 64 —2 0 -1.9 6 1 2 5 55- 3E-263CBl 81.0 90 89 59 7 8 -.3 1 6 78- 51::- 7ABBl 39.5 114 106 85 1 2 .6 1 3 7 75- 6E- 9BADl 51.0 85 108 56 5 6 .1 6 8 8S—l4E—30ACDIS 70.5 96 92 66 2 3 —.1 4 9 9S—l3E-33CBD1 30.0 127 74 100 O 0 9 1.5 10 125- 7E-33CBC1S 71.5 69 73 37 3 4 -.5 — 1 11 145- 9E- ZBAAl 26.5 118 57 89 - l 0 8 1.4 12 165- 9E-24BBlS 54.5 140 62 113 - 4 - 3 9 1.4 13 47N-65E-17CBC1 36.0 63 63 31 - 4 - 4 —.1 5 14 46N-64E-23BBDlS 24.0 55 51 13 - 6 - 6 0 6 15 45N-64E-20AC81 54.0 106 96 76 - 1 0 .3 9 16 ‘Temperatures calculated using B = 4/3 (Fournier and Truesdell, 1973). 2Values are departure from theoretical equilibrium in kilocalories; (+) values indicate supersaturated, (—) values indicate unsaturated, Calculations from computer program SOLMNEQ (Kharaka and Barnes, 1973). Ca estimated temperatures), (2) precipitation of silica during cooling, Whereas Na-K-Ca proportions remain unchanged, or (3) precipitation of calcite or aragonite due to loss of carbon dioxide without an adjustment in the concentration of sodium and potassium through reaction with clays, zeolites, or other minerals (Fournier and others, 1979). As shown in figure 2, most sampled waters plot on or near the equal-temperature line and are probably un- mixed waters or waters that have had time to re- equilibrate with the surrounding waters after mixing. Estimated reservoir temperatures between Givens Hot Springs and Murphy (sample Nos. 1, 2, and 3) range from 81° to 96°C. Estimated reservoir temperatures in the area between Oreana and Grand View (sample Nos. 4 and 6) range from 80° to 90°C. Reservoir-temperature estimates for spring 128—7E—330BCIS (sample No. 11) are within 2° of the temperature of 715°C measured at the surface. Estimated reservoir temperatures for the two hot-water wells and one warm spring (sample Nos. 14, 15, and 16) in north-central Nevada were in good agreement at each location, although the estimated temperatures at the three sites ranged from 51° to 106°C. Reservoir temperatures near Buhl (sample No. 9) range from 92° to 96°C. Samples 5, 10, 12, and 13 (significantly above the equal-temperature line) may have undergone evapora- tion, or dissolution of vitreous material in the aquifer may have increased the dissolved silica concentration. The Na-K-Ca geothermometer temperatures are probably the better estimate because of the unsaturation of calcite or aragonite (table 2). Samples 7 and 8 (below the equal-temperature line) probably indicate a mixed water that has not reequilibrated chemically with the aquifer materials after mixing. Reservoir temperatures were estimated for samples 4, J8 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS l l I l l l I 140 — 013 _ Q; . (05‘ .63 0% a 01° 4‘ :I'.‘ 63‘ 'lll 120 — .12 <09 _ 5E "’ \J D 2 0V . 7 9g (00 < O .16 9 m :‘ Lu u: :2 100 — .— 2 0 o 5 ‘00 9 O LIJ I D 6 “ z 3:” A- . 2 . 8 E 8 4 E '5 80 — 3 _ Hg :2 EXPLANATION o: LIEJ 8 .11 .13 Therma| well or spring and '— reference number 0 14 Lu ti 60 — Numbers refer to wells and _ 5 ' l‘ ted ' t bl 2 r; 15. sprIngs IS In a e (1) Lu 40 I L I I l I I 40 60 80 100 120 140 160 180 200 ESTIMATED TEMPERATURE FROM SODIUM-POTASSIUM-CALCIUM, IN DEGREES CELSIUS FIGURE 2.—Comparison of reservoir temperatures estimated by the silica and sodium-potassium-calcium geothermometers. 6, 7, and 11, using the sulfate-water isotope geother- mometer described by McKenzie and Truesdell (1977). Assuming conductive heat loss, N. L. Nehring, US. Geological Survey, Menlo Park, Calif., determined res- ervoir temperatures of 103°, 95°, 106°, and 115°C for samples 4, 6, 7, and 11, respectively (table 2). Dissolved sulfate and water are probably in isotopic equilibrium in all reservoirs of significant size having temperatures above about 140°C (McKenzie and Truesdell, 1977). The reasonably close agreement of the sulfate-water isotope geothermometer temperatures ob- tained for three of the waters with those obtained using the silica and Na—K-Ca geothermometers may indicate that, given sufficient residence time, equilibrium could also be achieved in reservoirs at a temperature as low as 100°C. 0n the basis of available geothermometry, estimated reservoir temperatures in southwestern Idaho proba- bly range from about 80° to near 100°C. Reservoir temperatures in north-central Nevada probably range from 51° to 106°C. ISOTOPES Samples of thermal and nonthermal water from 7 wells and 5 springs were obtained for analysis of tritium, deuterium, and oxygen-18. Several conclusions may be drawn concerning age, origin, and mixing pat- terns of the geothermal water on the basis of these analyses. TRITIUM Tritium (3H) is a radioactive isotope of hydrogen and decays with a half-life of about 12.4 years. It is formed in the upper atmosphere during bombardment by sub- atomic particles from outer space and is introduced into the water cycle dissolved in rain and snow. Although concentrations of tritium in precipitation vary both seasonally and geographically, prior to extensive ther- monuclear testing between 1953—63, average tritium levels in precipitation were about 10 ‘18 TU (tritium units). (One TU equals a 3H/H ratio of about 1018, or about 3.2><10'18 picocuries per liter.) By 1963, worldwide THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J9 tritium levels in precipitation had increased several 0r- ders of magnitude and were reported to be 7,000 TU in the vicinity of Yellowstone National Park (Pearson and Truesdell, 1978). Since the ending of atmospheric nuclear tests, tritium levels in precipitation have de- clined and presently (1979) average about 50 TU. Tritium content in ground water is a function of tritium content in the recharge water and the residence time and nature of flow in the system. Two basic types of flow models were discussed in detail by Nir (1964): (1) the piston-flow model, which has parallel flow lines of constant and equal velocity, so that a water sample taken at some point would include only water originating at the point of recharge; and (2) the completely mixed reservoir model, where it is assumed that the recharge water is continually and instantly mixing throughout the entire system. Tritium content in water of various residence times, assuming piston-flow and well-mixed reservoir systems, in Yellowstone National Park was calculated by Pearson and Truesdell (1978) and is shown in figure 3. Curves similar to those in figure 3 probably would result for comparable ground-water systems in thermal areas of southwestern Idaho and north-central Nevada. Results of tritium analyses of water samples from selected wells and springs are shown in table 3. Samples 1000 I I IIIIIIII I ||||IIII 700 C 500 - 300 — Assuming a well-mixed system .- \IO 00 lllllll Illllll 50 Assuming piston flow MEAN RESIDENCE TIME, IN YEARS 8 I _. 01V 0 I Illll‘i l l I 1 I l l I l l l I l L J_L J l 1 L1 5 10 20 30 4050 100 200 300 500 1000 2000 TRITIUM UNITS FIGURE 3.—Tritium content of water assuming piston-flow and well- mixed ground—water systems in Yellowstone National Park (from Pearson and Truesdell, 1978). analyzed in the Reston, Va., laboratory of the U.S. Geological Survey were predistilled and enriched by electrolysis. Enriched aliquots from the 100-mL (mil- liliter) samples were counted by a liquid scintillation counter; enriched aliquots of 470-mL samples were counted by a gas proportional counter. In table 3 errors are given for one standard deviation and include those incurred during radioactive counting, as well as volume errors. For the 100-mL enrichment method of analysis, the minimum error is 0.6 TU; for the 470-mL enrichment method, the minimum error is 0.2 TU. All samples were corrected for tritium decay to the collection date using a half-life of 12.361 years. As shown in table 3, concentration of tritium in water sampled from wells and springs having temperatures between 51° and 81°C ranges from 0 to 1.4 TU. Concentration of tritium in water sampled from two TABLE 3.—Tritium in water from selected wells and springs Water temperature Well or spring at surface Tritium number (°C) (TU) NONTHERMAL llS— 5W- 2DABlS 7.5 284 138— 4E—12CDD1 8.0 267 l4s—l4E—11CABlS 12.0 295 46N—60E—l3ACClS 4.0 259 45N-55E—25DAAlS 6.5 2119 THERMAL lN— 3W—21ACDlS 56.0 1o 1 0 4s— lE—34BAD1 76.5 11.4 i 0. SS- 4W— 8ADAl ' 20.0 21.9 55— 3E—26BCBl 81.0 1o i 0 75— SE— 7ABBl 39.5 10.9 i 0 78— 6E— 9BADl 51.0 10 i 0 165— 9E—24BBlS 54.5 10.5 i 0 lSamples analyzed by T. A. Wyerman, U.S. Geo- logical Survey, Reston, Va. 2Samples analyzed in U.S. Geological Survey Central Laboratory, Arvada, Colo. J10 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS cold-water springs in the Jarbidge Mountains to the south was 59 and 119 TU. Tritium in water from a slightly thermal (20°C) well was 1.9 TU. Considering the standard deviation of the 100- and 470-mL aliquots, it is unlikely that any of the water sampled from the hot wells and springs contains significant amounts of post-1954 water. This conclusion is based on an estimate of pre- 1954 precipitation at 10 TU and about two half-lives decay. The actual choices concerning the origins of the hot water are either all pre-1954 precipitation or con- siderably older water that has small amounts of post- 1954 precipitation (T. A. Wyerman, written commun., 1979). Some degree of mixing, therefore, is not ab- solutely precluded from the tritium data. From the small amount of tritium in the hot water sampled, residence time in the system is at least 30 years. Regarding the high concentration of tritium in the cold spring water to the south, a residence time of as lit- tle as 2 years or as many as 20 years is indicated (fig. 3), assuming piston flow in the system; for a well-mixed reservoir, residence time up to 20 years may be possi- ble. If the warm water is unmixed and has a tritium concentration of 1.9 TU, using the piston—flow model, a reSidence time of about 30 years is indicated. DEUTERIUM AND OXYGEN-18 Concentration of the stable isotopes, deuterium (D) and oxygen-18 (180), in water from different sources is used to characterize and indicate the origin and mixing patterns of individual waters. Principal stable molecular species in water are H2160, H2170, H2180, and HDO. In ocean water, the proportions of these species are 106:2,0001420:316 (Craig, 1963); this composition is referred to as SMOW (standard mean ocean water). Atmospheric water derived from the ocean is depleted in the stable isotopes. The isotopic composition of precipitation depends on the fraction of water remaining in the air mass from which the rain or snow is derived, the first precipitation being the rich- est in heavy isotopes (Ellis and Mahon, 1977). Stable- isotope concentrations are generally expressed in delta units (5) and are reported in parts per mil (700), or parts per thousand. These units represent relative deviations in the heavy isotope fraction in water and are defined as Rsample _ R standard 5 2 ——-——— x 1,000, Rstandard where Rsam pl e=ratIo or Isotopic concentration (180/160, D/H) of the sample, and Rstandard=ratio of isotopic concentration of the standard SMOW, A worldwide study of freshwater samples by Craig (1963) showed that the isotopic compositions of cold meteoric waters were related by the equation 6D=86180+10. This straight line, commonly referred to as the meteoric water line, is shown in figure 4; the slope of the line may vary regionally. Depletion of stable isotopes in meteoric waters (reporting units become more negative) in general can be correlated with dis- tance from the ocean, latitude, and altitude (or tempera- ture). Waters affected by extensive nonequilibrium evaporation, as in inland basins, lie off the meteoric line. However, at ordinary air tempertures, evaporated sur- face waters are connected approximately to the original precipitation composition 61800, (SDo by a line expressing the equation 6D=5(¢3130—61800)+Do (Ellis and Mahon, 1977). During passage through an aquifer, thermal and non- thermal ground water retains the D composition charac- teristic of precipitation in the recharge area. A compara- ble hydrogen-isotope shift does not occur because rock _100 I I I I I I I I EXPLANATION Reference symbols refer to wells and springs listed in table 4 -110— a. Nonthermal well or spring _ and reference symbol A 30 Thermal well or spring \3 and reference symbol of .1 E 120 — — 0: [LI n. m l- E -130 o. _ _. E 2‘ 2 E ,._ —140 — ._ D I.” D so -150 _ Standard error _ 5'80... 10.2 °/,, 6 DI 21.0 “/0, —160 I I 1 I I I I -22 —20 -18 —16 -14 -12 80XYGEN-18, IN PARTS PER MlL l°/.,.,l FIGURE 4.—Relations between concentrations of deuterium and oxygen-18. i 14 .. «4 THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J11 minerals contain little hydrogen and because the water to rock ratios of geothermal systems are seldom so low that the hydrogen of rock minerals is a significant frac- tion of the total hydrogen (Truesdell and Hulston, 1980). However, hydrogen and oxygen isotope-exchange reac- tions between clay minerals and water have been ob- served by O’Neil and Kharaka (1976) in laboratory bomb runs at temperatures between 100° and 150°C. The 180 content in thermal water is usually enriched (becomes less negative) to varying degrees during circulation within the system, due to reaction with the more en- riched 18O of the confining rock. Isotope-exchange reac- tions between water and rock occur slowly at tempera- tures below 150°C (Hobba and others, 1979) and are generally detected only in systems having reservoir temperatures equal to or greater than about 100°C (Ellis and Mahon, 1977). For a more complete discussion of stable—isotope geochemistry, refer to Gat (1971) or Ellis and Mahon (1977). Water samples for analyses of 18O and D were col- lected from four wells and nine springs in and adjacent TABLE 4.—Stable-isotope analyses of water from selected wells and springs Hell or spring R:;;;:2ce number 1°C GDSMOW 6 I aosnow A‘ 50 (fig. 4) NONTHERMAL 46N-60E-13ACC15 4.0 —127 -16.7 +0.43 a 45N—55E-25DAAIS 6.5 -123 ~16.6 + .65 b 95- 2w—26ccc1s‘ 7.0 -122 -16.5 o c 115— 5w— ZDABIS 7.5 -118 -15.4 + .60 d as- 1E-2OCAA152 9.5 -120 -17.1 + .16 e 135- 4s—12c0n1‘ 8.0 -116 -13.8 +1.93 f l4S-l4E-llCABlS 12.0 -126 -17.1 o g THERMAL lN- 3W-21ACD15 56.0 -139 -16.3 +1.32 h 25- 2W-35ACBl 40.0 —145 -17.3 +1.53 i 4s- 15-26ABc1’ 27.0 -144 -1a.2 +1.05 j 45— 12-3413’A012 76.0 -145 -17.5 +1.80 k ss- 3E-26BC312 03.0 -l46 -17 5 +2.00 1 7s— 55- 771131312 39.0 -135 —17.6 + 55 m 7s- 65- 93A01° 50.5 -142 -1e.2 + 80 n 85—14E-30ACDlS 70.5 -137 -17.2 +1.18 0 95-135—33c301 30.0 -131 -16.2 +1.42 p 125- 7B—33CBC15 71.5 -130 -16.6 + 90 q 145- 9E- zsAAi‘ 26.5 —124 -16.4 + 42 r 165— 9E-24BBlS 54.5 -131 -17.3 + .32 s 'Crosthwaite, unpublished data (1977L 2Rightmire, Young, and Whitehead (1976L to the study area. Additional stable-isotope analyses are available for the Bruneau—Grand View area (Right- mire, Young, and Whitehead, 1976; and Crosthwaite, unpubl. data, 1977). All the recent (1979) and selected older data from the Bruneau—Grand View area are shown in table 4. A plot of stable-isotope data in standard 6 values (°/oo) relative to SMOW, for the above data is shown in figure 4. Cold waters, indicated by closed circles in figure 4, range in temperature from 4° to 12°C and probably represent local, present-day meteoric water. Tritium analyses of the cold waters indicate all are relatively young (less than 20 years old). Several of the cold-water samples plot off the meteoric water line and probably in- dicate evaporation prior to being recharged. Waters k and | in figure 4, having temperatures near 80°C at the surface, were sampled from wells in the Oreana and Grand View areas. On the basis of the silica and Na-K-Ca geothermometer results, these waters have not attained temperatures much higher than those measured at the surface and probably represent un- mixed hot water from the deep system. Waters that plot between the hot and cold waters in figure 4 range in temperature from 72° to 27°C and are probably mixtures of the two waters. The unmixed hot waters lack the conventional “ox- ygen shift” or enrichment of 180 common to many geothermal systems (Ellis and Mahon, 1977). Instead, hot waters sampled from southwestern Idaho and north- central Nevada exhibit a general depletion trend in both 180 and D relative to the local cold meteoric waters. Some discussion of this phenomenon is warranted at this point. Constraints to be applied to any interpretation of the isotopic composition of the waters shown in figure 4 re- quire that the hot water, due to a depletion of about —17 and —2 °/oo D and 180, respectively, relative to the present-day cold waters, might have originated at a higher altitude or might have occurred as precipitation in the past when the regional climate was colder. Com- parison of concentrations of stable isotopes with mean annual temperatures at springs in the Appalachian Mountains (Hobba and others, 1979) indicate a deple- tion rate for D of about —4 °/ 66 per °C, which is within the range of —3.2 to —5.6 °/oo per °C suggested by Dansgaard (1964). On the basis of these estimates of D depletion with temperature and the normal temperature lapse rate of about 2°C per 1,000 ft, meteoric water hav- ing a D concentration comparable to that of the hot waters could occur if climatic conditions were about 3° to 5°C cooler than at present or if precipitation fell at altitudes 1,500—2,700 ft higher than the cold springs sampled during this study. Rightmire, Young, and Whitehead (1976) suggested J12 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS that recharge to the hot-water system in the Bruneau-Grand View area might have originated from areas of high altitude to the south in the Bruneau River drainage, or might have occurred during a time when the regional climate was much colder. Other processes dur- ing circulation that could produce waters depleted in 18O and D are boiling and separation of steam; mixing of' an older, hotter water with younger, cooler, local water; or isotope-exchange reactions between clay minerals and water. Neither of the first two processes lends it- self to the conceptual model proposed for the system. Because it was previously suggested that hot waters in the Oreana and Grand View areas are not mixed and probably have not attained temperatures much in excess of 90°—100°C, any interpretation involving separation of steam or mixing of older, hotter water seems unwar- ranted. The possibility that the stable-isotope composition of the hot waters has been altered due to hydrogen and oxygen isotope-exchange reactions with clays also is dis- missed. Such waters probably would contain concentra- tions of dissolved solids in excess of those observed due to dissolution of the clay minerals, and their composition of stable isotopes would be enriched relative to SMOW. Stable-isotope analyses of cold-water samples a and b (fig. 4) were collected in 1979 from springs near the head of the Bruneau River drainage at altitudes near 8,300 and 7,000 ft, respectively. Enrichment of stable isotopes in the cold springs relative to the hot water and the absence of nearby higher topography that could yield precipitation lighter in isotopic composition, indicates that none of the hot water discharged from the geother- mal system is derived from present-day, local precipita- tion. Considering the previous statement, the most probable explanation for the isotopic composition of the hot waters is recharge from precipitation that fell some time in the past when ambient temperatures were 3°—5°C colder than at present. To obtain sufficient recharge depleted in the stable isotopes, colder climatic conditions must have persisted over long periods of time, rather than a few tens of years, and are probably cor— relative with one of the major ice advances. While working in the St. Elias Mountains in southern Yukon Territory and Alaska, Denton and Karlen (1973) identified two major intervals of worldwide Holocene glaciation that occurred 200—350 years (Little Ice Age) and 2,400—3,300 years ago. During these episodes, they suggested that average temperatures were about 3.5°C colder than at present. A third, post-Wisconsin advance occurred 4,900—5,800 years ago but was less intense than the younger intervals. Prior to the 4,900- to 5,800-year in- terval, glacial advances were associated with late Wisconsin Glaciation in North America and occurred at about 2,500-year intervals beginning about 7,800 years BP (before present) (Denton and Karlen, 1973). The 2,400- to 3,300-year dates suggested by Denton and Karlen (1973) are significantly close to the 2,000- to 3,500-year age given for post-hypsithermal glacier ad- vances at Mount Rainier, Wash., given by Crandell and Miller (1964), and for the La Sal Mountains of eastern Utah, given by Richmond (1962); these ages may bracket the time during which precipitation and recharge to the geothermal system occurred. Occurrence of a colder climate about 3,000 years ago is within the time frame suggested by estimates of residence time in the geother- mal system; however, this does not preclude the pos- sibility of precipitation and recharge nearly 8,000 years BP, or even earlier, during the Wisconsin Glaciation. Considering the available data, meteoric water was probably recharged to the geothermal system when am- bient temperatures were about 3.5°C colder than at pre- sent. The water could then circulate to some depth, be heated to temperatures near 100°C, and become enriched in 180 through water-rock reaction to ultimately resemble, isotopically, the unmixed hot waters shown in figure 4. THERMAL GROUND-WATER OCCURRENCE AND MOVEMENT Thermal ground water occurs principally in discon- tinuous, or at least, compartmented zones created by the complex pattern of faulting in the area. Between Grand View and Oreana, thermal water occurs generally below about 2,000 ft in the Banbury Basalt and Idavada Volcanics. This area is approximately coincident with the resistivity low delineated on the apparent-resistivity maps and skin-depth pseudosections drawn by Hoover and Tippens (1975) from audio-magnetotelluric data. Two wells that penetrate to about 2,900 ft in the Idavada Volcanics yield water having temperatures up to 81°C at discharge rates near 2,000 gal/min. Wells in this vicinity completed below about 2,000 ft in the Banbury Basalt yield water generally greater than about 65°C. Also, between Grand View and Oreana, and throughout the rest of the study area, thermal water occurs in one or more shallow aquifers; wells between about 250 and 1,900 ft, completed in sedimentary rocks of the Idaho Group, yield water between about 20° and 30°C at a dis- charge rate between about 5 and 1,000 gal/min. Artesian head in the deep aquifer is generally several hundred feet above the artesian head in the shallow aquifer(s), and some upward movement of water between the two aquifers probably occurs (Young and Whitehead, 1975). Depths of wells, water levels, and water temperatures are highly variable in the volcanic rocks. Water levels in these rocks range from several hundred feet above land surface (as artesian heads) to nearly 900 ft below land surface. Generally, artesian heads are above land surface in wells in the northern part of the study area adjacent to _ THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J13 v ~ TABLE 5.—Record of wells (Altitude: From 0.5. Geological Survey Water temperature: R - Reported topographic maps Remarks: DL - Driller's log] Water level: a — Recently pumped R - Reported Altitude of land Water level surface (feet above Reported Altitude of National depth of Feet potentiometric Geodetic well (feet above (+) surface (feet Vertical below or below Water above National Datum land land Date temperature Geodetic Vertical Well number of 1929) surface) surface measured (°C) Datum of 1929) Remarks 3N-6W-36ABBl 2,452 —- 9.30 2— 9-79 -- 2,433 DL 36BAC1 2,498 330 15.47 2—13—79 -— 2,483 3N—5W- SCBBI 2,350 -— 16.25 2- 9-79 -- 2,334 19DCD1 2,378 325 74.393 2—13-79 -- 2,304 25ADC1 2,292 386 37.27 2-15-79 16.5 R 2,255 DL 30ADA1 2,388 415 34.15 2-13-79 21.0 R 2,354 3N—4W-30CCD1 2,288 544 26.85 2-15-79 17.5 R 2,261 DL 3ZBBA1 2,302 250 36.65 2—15-79 -— 2,265 DL 2N—5W-13BCD1 2,468 190 17.92 2-15-79 18.5 R 2,450 DL 2N-4W- 3BBBl 2,305 960 59.60 3- 2-79 26.5 R 2.245 70331 2,356 1,100 61.88 3— 2-79 —- 2,294 26CCD1 2,455 450 80 R 3-28—75 -- 2,375 DL lN—4W—12DBBl 2,375 1,210 2.20 2- 2-79 35.5 2,373 DL 13BAC1 2,460 2,860 95.77 2- 1—79 21.5 R 2,364 DL 1N-3W— 6DCC1 2,270 900 +21.42 2—28—79 37.0 2,291 6DDC1 2,240 560 +22.62 2-28—79 45.0 R 2,263 8CDA1 2,248 512 +29.72 2-12-79 36.0 R 2,278 17CCD1 2,420 800 84.15 1-31—79 -- 2,336 DL 18ABD1 2,370 525 25.40 2- 1—79 18.5 R 2,345 DL ZOBCBl 2,455 400 113.58 1-31—79 -— 2,341 20CCCl 2, 0 155 39.05 2-27-79 -- 2,441 ZODACl 2,.32 1,310 38.70 1-30—79 35.5 2,343 21BBC1 2,300 850 +17.12 1-30—79 38.5 2,317 21DAA1 2,262 345 +17.52 2—12-79 24.0 2,280 28DBA1 2,346 420 8.83 1-29—79 49.0 2,337 35ABC1 2,262 1,035 9.65a 2-27-79 44.5 2,252 lS-3w- lBBCl 2,290 1,800 +8.82 2—26-79 39.0 2,299 9ACBl 2,600 600 104 R 9-25-66 37.0 2,496 DL 9ADA1 2,540 550 157.50 1—26-79 24.0 R 2,386 DL 9CAC1 2,672 1,310 201.40 2—26—79 —— 2,471 llDACl 2,420 410 79.20 2—26—79 —— - 2,341 lS-2W-18CCA1 2,375 —- 46.00 2—27—79 —— 2,329 18CDD1 2,370 960 40 R 1— 2—79 30.0 R 2,330 29AD81 2,364 —— +0.40 1-23—79 26.0 2,364 34CBC1 2,380 —- +7.82 2—23—79 37.5 2,388 25-3W—ISBBC1 4,163 -- 39.17 2-20—79 -- 4,124 17AAD1 4,160 640 36.67 2—20-79 -— 4,123 DL lBDACl 3,817 -- +2.00 2—20-79 11.0 3,819 18DAC2 3,830 —- 17.25 2-20-79 -- 3,813 ZS-ZW- 2CADl 2,455 850 5.77 2-22-79 36.5 2,449 DL 3CAA1 2,430 900 +6.28 2-22-79 43.0 2,436 JCBBl 2,466 900 27.65 2-22-79 36.0 R 2,438 10CCC1 2,686 -- 136.85 2-22—79 —- 2,549 35ACBl 2,945 1,187 412.69 2-22-79 40.0 2,532 DL 36CDC1 2,930 1,182 359.20 2-21—79 23.5 R 2,571 DL 25-1W— 60DC1 2,525 —— 180.70 2-16—79 -- 2,344 16AAA1 2,920 615 573 R 3—31—70 -- 2,347 DL 23CCC1 3,027 726 557.02 2—23-79 26.5 2,470 DL 3S-4w-2BAD1 4,000 880 20.87 2-20-79 -- 3,979 DL 3S-2w—28AAC1 .3,524 355 250.00 2—16-79 -— 3,274 DL 36DDBl 3,131 -- 19.35 1— 8-79 -- 3,112 3s—1w— 4DCC1 3,100 749 587 R 1- 1-75 -- 2,513 15DCC1 2,735 250 58.20 2—22-79 -- 2,677 18DCC1 3,240 738 365.15 2-23-79 -- 2,875 30CDA1 3,088 265 230.00 2- 1-79 -- 2,858 4S-2W—11AEA1 3,250 419 20.49 2-21-79 -- 3,230 4S-lE—ZSBBD1 2,500 1,600 +ll.72 3— 8-79 32.5 2,512 29CCD1 2,685 3,040 +92 R 11-16—59 70.0 2,777 DL 34BAD1 2,570 2,980 +243 R 7- 8—62 76.5 2,813 DL 4S-2E—19ACBl 2,575 3,080 +164 R -— 73.0 R 2,739 DL 29DBC1 2,440 1,600 +4.02 3- 8—79 27.5 2,444 55—18- 3AAB1 2,585 1,900 +10.02 3- 8—79 32.0 2,595 lOBDDl 2,650 2,960 +122 3- 8—79 64.5 2,772 DL 21CBD1 2,799 660 +6.92 3- 9—79 65.5 2,806 DL 24ADBl 2,719 3,120 +58 3- 9-79 66.0 2,777 DL )A J14 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS . TABLE 5.—Record of wells—Continued Altitude of land Water level surface (feet above Reported Altitude of National depth of Feet potentiometric Geodetic well (feet above (+) surface (feet Vertical below or below Water above National Datum land land Date temperature Geodetic Vertical Well number of 1929) surface) surface measured (°C) Datum of 1929) Remarks 2408A1 2,745 2,480 +15.02 3—12-79 66.0 2,760 DL SS—ZE- lBBCl 2,395 1,800 +112 3- 9—79 50.0 2,507 58001 2,530 2,009 +45.02 3-12—79 42.0 2,575 DL ZSAAAl 2,615 3,025 10.15 3-29-79 34.5 2,605 DL SS-3E—2OBBBl 2,465 —— +4.42 3-12-79 27.0 2,469 258881 2,400 1,320 +3.12 3—12-79 18.0 2,403 268C81 2,485 2,970 +148 3—12-79 84.5 2,633 35ccc1 2,500 2,570 +141 3-12-79 71.5 2,641 55-108-28CA81 2,544 1,100 +1.82 3-27-79 -- 2,546 32ADA1 2,510 -- +40 4—25—67 -- 2,550 65— 3E- SCACl 2,670 3,600 +20.02 3-13-79 61.0 2,690 DL 9Acc1 2,635 1,425 48.50 3-13—79 39.0 2,587 DL 23c001 2,740 1,241 88.90 3—13-79 30.0 2,651 DL 33CACl 2,875 1,176 208.30 3-13-79 30.5 2,667 DL 65- 4E-14ABC1 2,665 1,905 36.54 3—13-79 54.5 2,638 DL 36ccc1 2,675 2,000 +11.92 3—14-79 40.0 2,687 DL 65— 5E-10DDD1 2,520 1,667 +2.5 R 3—15-79 38.5 2,522 DL ZOAABl 2,510 -~ +4.42 3—15-79 41.0 2,514 , 6s— 6E—12CCD1 2,510 990 240 R 6- -79 37.0 2,270 DL '-5 328001 2,570 1,402 +7.02 3-15-79 37.5 2,577 DL 6s- 7E- 2c001 2,485 1,350 +6.32 3—16-79 35.0 2,491 ‘ 6s- 8E—33ABA1 3,080 4,000 317.40 3-16-79 59.5 R 2,763 DL es— 9E- 20881 2,620 1,235 75.97 7-16—62 -— 2,544 ‘ 68-10E-12CDD1 3,080 990 535 R -- -- 2,545 ‘ 300ca1 2,990 437 251.45 3—26-79 -- 2,739 . 75— 3E- 4ADC1 2,935 804 249.40 3-14—79 33.5 2,686 DL IZBADl 3,000 -- 319.56 3—14-79 36.0 2,680 75- 4E- 1c0c1 2,670 —- +6.82 3-31-79 42.0 2,677 44~ 108081 2,755 1,145 76.92 3—14-79 37.5 2,678 DL 278cc1 2,770 1,390 95.00 3-14-79 27.0 2,675 DL ( . 75- SE- 7ABBl 2,604 1,625 +79 3—15—79 39.5 2,683 DL 13CB81 2,771 1,954 113.22 3—15-79 31.5 2,658 DL 28ACD1 2,810 1,003 94 R 8-12-53 33.5 2,716 DL “ 75- 6E— SBADZ 2,580 960 +46.42 3—15-79 51.0 2,626 A 16c0c1 2,595 513 +72 3-15-79 43.0 2,667 DL . 22AAD1 2,640 1,410 +28.92 3—15-79 46.5 2,669 0L 34DADl 2,725 300 +14.92 3-15—79 32.5 2,740 DL Y a BS-llE-ZlBAl 3,160 305 40.80 3-26—79 -— 3,119 DL 31CABl 3,168 290 168.84 4— 4-79 -— 2,878 DL 85—12E-24CCC1 3,480 —— 240.28 3— 6-79 -- 3,240 85-13E-Z3CCD1 3,390 -— 69.28 2-22—79 -- 3,321 ‘ 4 35ACD1 3,479 700 182.43 3- 8-79 —- 3,297 DL 95- 1W- 4ADDI 5,800 474 317.13 5-10-72 -— 5,483 9S— SE- 40801 3,633 2,502 850 R 7—13-62 53.0 R 2,783 DL 95-12E—JSBCD1 3,705 797 343.14 2—21—79 -- 3,362 DL 95-13E-20CCDl 3,805 920 448.88 3—27-79 -- 3,356 DL 220001 3,700 575 417.43 3— 6-79 -— 3,283 DL ZSADDI 3,680 400 175.80 3- 8—79 -- 3,504 DL 3100c1 3,819 840 458.89 3- 6-79 -- 3,360 DL 95-15E-31CCB1 3,812 1,058 83 R 3—27-79 26.5 R 3,729 95-17E-Z9ACDl 3,150 730 +490 R 12—10-70 42.0 3,640 DL 3ZDDA1 3,637 1,295 +88 R 3—28-79 40.0 3,725 lOS-IZE- 2CBA3 3,733 —- 353.06 4- 5-79 —- 3,377 IZCDCI 3,742 500 366.48 2-21—79 -— 3,376 IDS-l3E- SCCAl 3,823 575 448.88 3— 6—79 -— 3,374 DL 105-14E- lCAAl 3,840 688 58.22 2-21-79 -- 3,782 105-163- 8CDA1 3,782 953 24.04 3—27-79 27.0 R 3,758 DL 105-17E—I4CCDI 3,788 1,200 48.38 3-28—79 29.0 R 3,740 115- 7E-ZSDABl 4,345 1,400 842 7- -63 34.5 3,503 DL llS—lls— 7CCAl 4,577 550 Dry —— —- —— llS-13E-llBDDI 3,979 420 205 R 10—12—60 -— 3,774 llS-IGE- GDBAl 4,149 975 266.80 3-28-79 20.0 R 3,882 DL ZlDAI 4,296 866 529 R ll-l9—62 25.5 R 3,767 DL 125- 83- 6ADA1 4,450 1,400 881 R 9- 1—63 47.0 3,569 0L 125-133-253331 4,589 1,200 790 R -- 23.0 3,799 IZS-ISE-10CBBI 4,482 1,017 605 R 2-20-79 —- 3,877 DL 22391 4,513 638 513.17 11-30-60 -- 4,000 lZS-IGE- 9BD1 4,411 1,185 599.25 7-21—58 -- 3,812 125-185- 1A551 4,225 850 56.94 4- 5—67 30.5 R 4,168 135-113-12CDC1 5,661 1,000 485 R -- -- 5,176 8 L ‘A THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J15 TABLE 5.—Record of wells—Continued Altitude of land Water level surface (feet above Reported Altitude of National depth of Feet potentiometric Geodetic well (feet above (+) surface (feet Vertical below or below Water above National Datum land land Date temperature Geodetic Vertical Well number of 1929) surface) surface measured (°C) Datum of 1929) Remarks 263331 5,259 1,050 600 R -- 20.0 4,659 13S-12E-3lBDA1 5,366 600 340 R —— 21.5 5,026 36ABD1 5,245 620 575 R -- —- 4,670 13S-14E-35CDDl 5,120 518 232.67 8-18-59 —- 4,887 13S-16E-12DAA1 4,742 700 122.94 3-28-79 36.5 R 4,619 DL 14S- 98- ZBAA1 5,250 953 623 R 7-16-56 26.5 R 4,627 DL 14S-12E-12ACC1 5,516 850 540 R -- —— 4,976 158CC1 5,518 810 395 R -— -— 5,123 14S—15E-285AD2 4,979 455 111.65 3-20-79 19.5 R 4,867 155- GE-IOBBAI 5,104 850 727 R —- -— 4,377 DL 155- 9E-34BBC1 5,595 850 705 R -- —— 4,890 155-112-3OBAA1 5,378 550 500 R -- -— 4,878 155-14E- 2ACD1 5,273 815 145.27 4- 4—79 -— 5,128 DL the Snake River. The artesian head is generally below land surface to the south, where drillers report that when water is encountered, the water level in the wells rises several feet to several tens of feet. A generalized potentiometric surface for the deep hydrothermal system, based on measured or reported water levels in 148 wells, is shown in figure 2. Water levels used to define this surface are from wells com- pleted in volcanic rocks and from the deepest wells in the area. The measured or reported water levels are given in table 5. An attempt was made to convert water levels to a cold-water equivalent head; however, insufficient data to do this for many wells resulted in a loss of data points. Therefore, the potentiometric surface shown in figure 2 is not temperature corrected. The general direction of regional ground-water move- ment can be inferred from the potentiometric-surface map (fig. 2). Movement is down the hydraulic gradient and roughly perpendicular to the potentiometric con- tours, from areas of recharge to areas of discharge. TEMPERATURE DISTRIBUTION AND HEAT FLOW The widespread occurrence of thermal ground water in southwestern Idaho and north-central Nevada suggests a broad heat source. Heat-flow distribution on the Snake River Plain is interpreted by Brott, Blackwell, and Mitchell (1978) to be related to combined effects of crustal thermal refraction and a large crustal heat source emplaced under the western part of the plain 10—15 mil- lion years ago. Any discussion of the geothermal system, therefore, must consider deep circulation of meteoric water in an area of above-normal heat flow. Heat-flow values reported by Brott, Blackwell, and Mitchell (1978) ranged from 3.3 HFU (heat-flow unit; 1 HFU equals 1 microcalorie per centimeter squared per second) in granitic rocks near the margin of the western Snake River Plain to 1.5 HFU near the Snake River. Sass and others (1971) reported heat flow in north-central Nevada to be 3.3 HFU. McIntyre (1979) calculated a heat flow of 1.7 HFU, using average thermal- conductivity values of Brott, Blackwell, and Mitchell (1978) and a bottom-hole temperature of 150°C in an 11,125-ft test hole. Subsequently, a series of bottom-hole temperature measurements were made at intervals of time and extrapolated to infinity in the manner described by Parasnis (1971) to obtain a bottom-hole temperature of 196°C and a corresponding thermal gradient of 1.7°C per 100 ft. This thermal gradient is less than the 2.0°C per 100 ft estimated by Young and Whitehead (1975) and is greater than the 1.4°C per 100 ft estimated by McIntyre (1979), due to use in the com- putation of a greater depth (11,125 ft) and higher bottom-hole temperature (196°C), respectively. Using a thermal gradient of 1.7°C per 100 ft and the thermal-conductivity value used by McIntyre, a heat flow of 2.2 HFU was obtained, which is within the 1.5—3.3 HFU range reported by Brott, Blackwell, and Mitchell ( 1978) for the Snake River Plain. If a heat flow of 2.2 HFU over the entire 1,520 mi2 of the system is assumed (see following section on “Conceptual Model of the System”, a net heat flux into the system of 8.7X107 cal/s is indicated. Heat is dis- charged naturally from the system conductively at the surface, convectively in thermal-spring discharge, and advectively in thermal ground water moving northward out of the area. Conductive and convective discharge can J16 GEOHYDROLOGY 0F GEOTHERMAL SYSTEMS be estimated and the advective discharge determined from the relation: Q advective= Q input_ Q conductive_ Qconvective' Heat discharged conductively at the surface was estimated using temperature data in Young and Whitehead (1975) from wells less than about 2,000 ft deep completed in sedimentary deposits. Average ther- mal gradient in six wells was 09°C per 100 ft. Thermal conductivity measured in 15 core samples of sand and clay was 3.6><10‘3 [cal/(cm s 0 °C)] (Brott and others, 1978, p. 28). Using these values, a conductive heat flow of 1.1 HFU is obtained, or about 4.3)(107 cal/s for the 1,520 mi2 system. Convective heat flow discharged naturally in thermal springs was 1.6><107 cal/s on the basis of temperature and discharge measurements from springs included in Stearns, Stearns, and Waring (1937). Using the above equation, the amount of heat discharged advectively in thermal ground water leaving the system is (8.7X107)—(4.3><107)—(1.6X107), or 2.8X107 cal/s. On the basis of the conceptual model proposed, heat is discharged advectively from the system in water moving across the northern boundary, which is nearly coincident with the course of the Snake River. If the temperature of the water is known, the volume of discharge, Q, neces- sary to transport 2.8><107 cal/s from the system can be calculated from the relation, Net heat flux Heat content of the water - Using the heat content of the thermal water, the volume of discharge across the boundary necessary to transport the residual heat from the system is estimated for various water temperatures as shown in table 6. TABLE 6.——Estimated thermal-water discharge, on the basis of enthalpy Discharge, Q Temperature of thermal water (°C) (ft3/s) (acre-ft/yr) 90 12.8 9,300 80 14.5 10,500 70 16.8 12,200 60 20.1 14,600 50 25.0 18,100 40 33.2 24,000 Thermal-water temperatures along the northern boun- dary range from 35° to 81°C and average about 50°C. If an average temperature of about 50°C is assumed for the ' thermal water all along the northern boundary, from table 6, about 25 ft3/s, or nearly 18,000 acre-ft/yr, is re- quired to transport convectively, 2.8)(107 cal/s of heat from the system. Using an estimate of recharge (thermal-water outflow across the northern boundary, plus hot-spring discharge) and the volume of the geothermal reservoir, the residence time of water in the system can be calculated. From the equation (Pearson and Truesdell, 1978) for calculating tritium content (T) at any point in a well- mixed reservoir for some transit time (t), dT Q Q dt T° V T V W ’ where To, Q, V, 'yt are tritium concentration in recharge water, input (= output). reservoir pore volume and decay constant (In 2/half- life), respectively, the ratio V/Q is the turnover time for the system. Bolin and Rodhe (1972) have shown that this turnover time is identical to the average transit time or average age of particles leaving the reservoir, which they prefer to term “residence time.” Applying the relation t=V/ Q to the southwestern Idaho north-central Nevada geothermal system, and us- ing the values for reservoir volume and recharge of 576 mi3 and 39.8 ft3/s, respectively (25.0 ft3/s ground-water outflow to the north, plus 14.8 ft3/s hot-spring discharge), residence time is about 3,400 years for a reservoir having an effective porosity of 0.05; residence time is about 6,800 years for a reservoir having an effective porosity of 0.10. Brott, Blackwell, and Mitchell (1978) obtained porosities of 0.05 from laboratory analyses of core sam- ples from silicic volcanic rock. This is probably a minimum for the reservoir, considering the secondary porosity that likely occurs due to faulting. Calculations for the geothermal system apply to an average reservoir thickness of 2,000 ft. If the geothermal reservoir volume is less than 576 mi3, or if the recharge is greater than 39.8 ft3/s, calculated residence time is reduced; if the reservoir volume is greater or if the dis- charge is less, calculated residence time is greater. In spring 1982, water samples were obtained from several of the hottest wells by the authors and W. H. Low (U.S. Geological Survey) for analysis of carbon isotopes 120, 13C, and 14C. On the basis of analysis of stable-isotope data discussed earlier, water from these wells has not mixed with shallower, local ground water and probably is representative of hot water from the deep system. Although 13C to 120 ratios obtained from carbon-isotope analysis of water from upland cold THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA springs are not presently available, preliminary esti- mates on the basis of carbon-14 determinations indicate the geothermal waters in the area have an apparent age in the range of 18,000—25,000 years (oral commun., W. H. Low, 1982). Reservoir thermal energy in the Bruneau—Grand View system was estimated by Brook and others (1979) using the relation: q, =pC(a)(d)(t-t,ef) , where q,=reservoir thermal energy, pc =volumetric specific heat of rock plus water, a =reservoir area, d =reservoir thickness, t=reservoir temperature, and tref =reference temperature. From this equation, and using the conceptual model described in the following section, the reservoir thermal energy for the geothermal system in southwestern Idaho and north-central Nevada is about 130x1018 cal. This differs from the value of 107x10la cal obtained by Brook and others (1979), due to the larger reservoir vol- ume assumed. CONCEPTUAL MODEL OF THE SYSTEM The conceptual model is based on available geologic, geophysical, and hydrologic data. It is, of necessity, generalized and probably an oversimplification of what is undoubtedly a complex system. This model is not uni- que but can be considered a first approximation, es- tablished within existing constraints. As additional data are generated, more specific models will be developed. An interpretation by Mabey (1976) of seismic, magnetic, and gravity data indicates about 3,280 ft of in- terbedded basalt and sedimentary rocks of Pliocene to Holocene age overlie about 1,800 ft of silicic volcanic rocks of Miocene and Pliocene age. Older basalts of Ter- tiary age and rocks of the upper and lower crust underlie the silicic volcanic rocks. McIntyre (1979) examined well cuttings from the Anschutz Federal No. 1 drill hole located in NEIASWIA sec. 13, T. 5 S., R. 1 E., and described a 2,000-ft section of rhyolite between 2,720 and 4,750 ft. Drillers’ logs for wells in the area indicate that hot water is encountered in rhyolite at a depth of about 2,900 ft. On the basis of this evidence, the rhyolite is con- sidered to be the principal reservoir rock in the northern part of the study area. For simplicity, a single aquifer about 2,000 ft thick, at a depth of 3,000 to 5,000 ft, is proposed to be in the silicic volcanic rocks of the Idavada Volcanics. Considering the J17 following three items, that the southern boundary of the system is the edge of the Snake River Plain—approx- imated using the aeromagnetic map of Zietz, Gilbert, and Kirby (1978)—that the northern boundary is coinci- dent with the course of the Snake River, and that the east-west boundaries are as shown on plate 1, then, the surface area of the hot-water reservoir is about 1,520 miz, and the volume of rock is 576 mi3. There is no evidence to indicate the existence of a shallow, local heat source. Available data indicate only slightly above-normal heat flow over most of the Snake River Plain. In the western part of the plain, however, heat-flow values range from about 1.5 HFU to 3.3 HFU along the southern margin (Brott and others, 1978). This model differs from previously described geothermal convection models where meteoric water is circulated to depths near 10,000 ft, is heated, and then, due to density difference between the hot and cold water, rises along faults to ultimately fill the available pore space in some reservoir rock. The proposed model suggests that for this part of the Snake River Plain having a heat-flow value of about 1.7 HFU, water temperatures equivalent to those estimated using geothermometers can be attained at depths near the base of the geothermal reservoir at about 5,000 ft in rocks of the Idavada Volcanics, and deeper circulation of water below that depth is not required. Considerable transfer of heat convectively within the reservoir is implied, however, by the occurrence of 80°C water in two wells at a depth of about 2,900 ft. In a solely conductive regime, and on the basis of data from the deep drill hole, water temperatures of 80°C would occur at a depth of about 5,000 ft. For water temperatures of 80°C to occur at a depth of only 2, 900 ft, a heat flow near 3.2 HFU would be required; such a high heat flow would result in a bottom- hole temperature of about 270°C in the deep drill hole. Neither of these is consistent with a conceptual model of the geothermal system developed based on available data. On the basis of thermal considerations, recharge to the system is about 40 ft3/s. Using the stable-isotope data, it is not possible to suggest a principal recharge area for the convective system. However, the generalized potentiometric contours (pl. 2) indicate that the recharge area lies somewhere to the south and may be considerably removed from the area of discharge. Water in the thermal system is meteoric in origin and most likely fell as rain or snow in the distant past when the regional climate was much colder than at present. Water is heated by circulation to depths near 5,000 ft. Lack of significant amounts of tritium in the water indi- cates minimum residence time in the system is more than 30 years. On the basis of estimates of reservoir vol- ume and amount of recharge, the residence time in the system is probably at least 3,400,—6,800 years, and in view of recent preliminary carbon-14 analyses, perhaps as long as 25,000 years. J18 The dilute nature of the hot water (less than about 400 mg/L dissolved solids) in the system could indicate one of three possibilities: (1) rapid water movement and short residence time, hence, little water-rock reaction; (2) reservoir temperature sufficiently low so that water- rock reaction occurs very slowly; or (3) a very old sys- tem, in which most of the minerals in the rock have been leached, owing to long contact with circulating hot water. None of the above reasons are sufficient by themselves to adequately explain the dilute nature of the thermal water. Interpretation of the tritium and stable-isotope data and of the subsurface temperatures calculated using chemical geothermometers precludes using either possibility 1 or 2 above as the basis for a conceptual model of the geothermal system. Concentra- tions of fluoride, lithium, and boron in the thermal water indicate that relatively soluble minerals still per- sist in the reservoir rock and are still being dissolved. If possibility 3 were true, the more soluble minerals al- ready would have been dissolved and concentrations of fluoride, lithium, and boron in the thermal water would be much less. Preliminary interpretation of carbon-14 analyses of the thermal water has extended the estimates of resi- dence time in the system. Consideration of a longer resi- dence time will alter the conceptual model proposed. With increased time constraints, in order to address the thermal and hydraulic considerations, the new model must include (1) an increase in the size or effective porosity of the reservoir; (2) a reduction in the outflow, hence, total recharge to the system; (3) an increase in the temperature of the water that discharges from the system; or (4) some combination of the above. In any case, the proposed generalized conceptual model is one of an old system with long residence time where circulation has occurred for thousands or tens of thousands of years. The volume of reservoir rock involved is sufficiently large so that soluble min- erals continue to be dissolved, and the chemistry of the water reflects more the type of rock through which the water has passed than the residence time or the reservoir temperature. SUMMARY Rocks of igneous, sedimentary, and metamorphic origin in southwestern Idaho and north-central Nevada range in age from Precambrian to Holocene. Rocks related to the occurrence of thermal ground water in- clude (1) sedimentary rocks of the Idaho Group of Quaternary and Tertiary age; (2) basalt of Quaternary and Tertiary age, which consists chiefly of the Banbury Basalt of Tertiary age; and (3) silicic volcanic rocks of Tertiary age, which include the Idavada Volcanics. GEOHYDROLOGY OF GEO’I‘HERMAL SYSTEMS The silicic volcanic rocks and the Banbury Basalt are thought to be the chief reservoir rocks of the thermal ground-water system(s) in the study area. Low con- centration of magnesium in the thermal water seems to favor silicic volcanic rocks as the principal reservoir rock. These rocks have been fractured and displaced by a system of northwest-trending faults, particularly in the area adjacent to the Snake River. Thermal ground water generally occurs under artesian conditions; water temperatures range from 20°C to more than 80°C. The artesian heads are variable and range from a few feet above the water-bearing zone to several hundred feet above land surface. Thermal ground water in the study area occurs, for the most part, in discon- tinuous, or at least, compartmented zones. Ground- water movement in the western part of the area is generally northeastward. Ground water in the eastern part of the area tends to move northward and northwestward. Nonthermal ground waters are generally a calcium bicarbonate type and have dissolved-solids concentra- tions generally less than 120 mg/L. The hot waters, which are thought to be representative of the deep con- vective system, are generally a sodium carbonate or bicarbonate type and have dissolved-solids concentra- tions generally less than about 400 mg/L. Estimated reservoir temperatures using the Na-K-Ca and H38i04‘ corrected silica geothermometers range from about 80° to 96°C in southwestern Idaho and from about 51° to 106°C in north-central Nevada. The sulfate water isotope geothermometer indicates maximum reservoir temperatures in southwestern Idaho are near 100°C. Tritium concentrations in the hot water in southwestern Idaho are near zero, which indicates a minimum resi- dence time of 30 years. Tritium concentrations of cold springs in the southern part of the study area indicate residence times as short as 2 years or as long as 20 years. Stable isotopes are depleted in the hot waters relative to present-day meteoric waters and were proba- bly recharged several thousand years ago during one of the Holocene glacial advances when ambient temper- tures were about 3.5°C colder than at present. On the basis of estimated recharge rate and the pore volume of the geothermal reservoir, minimum residence time in the system may be 3,400—6,800 years; on the basis of preliminary estimates of carbon-14 determinations, resi- dence time could be as long as 25,000 years. Widespread occurrence of thermal ground water in southwestern Idaho and north-central Nevada suggests a broad heat source. Reported heat-flow values range from 1.5 to 3.3 HFU. Extrapolation of data from a deep test hole in the northern part of the study area indicates a conductive heat flow of 2.2 HFU. A heat flow of 2.2 HFU would be associated with a temperature gradient of 1.7°C per 100 ft. Reservoir temperatures of 80°—96°C THERMAL GROUND WATER, SOUTHWESTERN IDAHO, NORTH-CENTRAL NEVADA J 19 could be attained by circulation of water to depths between about 4,000 and 5,000 ft. Additional data from the test hole indicate a 2,000-ft section of rhyolite between 2,720 and 4,750 ft, which is considered to be the principal reservoir rock (Idavada Volcanics) in the northern part of the study area. Using an average conductive heat flow of 2.2 HFU for the study area, heat flux into the system is estimated to be 8.7X107 cal/s. Considering heat losses of 4.3)(107 and 2.8X107 cal/s to conduction and thermal-spring dis- charge, respectively, the net amount _of heat that is transported convectively through the system is about 2.8)(107 cal/s. Assuming the heat is discharged advec- tively from the system in water moving across the northern boundary, about 25.0 ft3/s, or 18,000 acre-ft/yr, of 50°C water would be required to transport 2.8><107 cal/s of heat from the system. The reservoir thermal energy for the geothermal system is about 130x1018 cal. The conceptual model proposed for southwestern Idaho is one of an old system, where water has circulated for thousands, even tens of thousands of years. The prin- cipal reservoir rock is in the Idavada Volcanics at depths between about 3,000 and 5,000 ft. Water in the system is meteoric in origin and fell as rain or snow in the distant past. Water is heated within the reservoir by circulation to depths near 5,000 ft. Residence times near 3,000 years are indicated on the basis of climatic evidence, as well as physical reservoir parameters; however, residence times of as long as 25,000 years are not precluded. SELECTED REFERENCES Benseman, R. F., 1959, Estimating the total heat output of natural thermal regions: Journal of Geophysical Research, v. 64, no. 8, p. 1057-1062. Bolin, Bert, and Rodhe, Henning, 1972, A note on the concepts of age distribution and transit time in natural reservoirs: Tellus 25 (1973), v. 1, p. 58—62. Bond, J. G., compiler, 1978, Geologic map of Idaho: Moscow, Idaho, Idaho Bureau of Mines and Geology, 1 sheet, 1:500,000. Brook, C. A., and others, 1979, Hydrothermal convection systems with reservoir temperatures 290°C, in Assessment of geothermal resources of the United States: U.S. Geological Survey Circular 790, 163 p. Brott, C. A., Blackwell, D. D., and Mitchell, J. C., 1978, Tectonic implications of the heat flow of the western Snake River Plain, Idaho: Geological Society of American Bulletin, v. 89, p. 1697—1707. Chapman, S. L., and Ralston, D. R., 1970, Ground-water resources of the Blue Gulch area in eastern Owyhee and western Twin Falls Counties, Idaho: Idaho Department of Water Administration, Water Information Bulletin no. 20, 36 p. Craig, Harmon, 1963, The isotopic geochemistry of water and carbon in geothermal areas, E. Tongiori, ed., in Spolette Conference on Nuclear Geology and Geothermal Areas, Spolette, 1963: Rome, Consiglio Nazionale delle Recerche, 17 p. Crandell, D. R., and Miller, R. D., 1964, Post-hypsithermal glacier ad- vances at Mount Rainier, Washington: U.S. Geological Survey Professional Paper 501—D, p. 110—114. Crosthwaite, E. G., 1977, Stable-isotope analyses: Unpublished data on file in Idaho District ofiice, U.S. Geological Survey. Dansgaard, W., 1964, Stable isotopes in precipitation: Tellus, v. 16, p. 436—468. Denton, G. H., and Karlen, W., 1973, Holocene climatic varia- tions—their pattern and possible cause: Quaternary Research, v. 3, p. 155—205. Ekren, E. B., McIntyre, D. H., and Bennett, E. H., 1978, Preliminary geologic map of the west half of Owyhee County, Idaho: U.S. Geological Survey Open-File Report 78—341, 14 p. Ellis, A. J., and Mahon, W. A. J., 1977, Chemistry and geothermal systems: New York, Academic Press, 392 p. Fournier, R. 0., 1979, Geochemical and hydrologic considerations and the use of enthalpy—chloride diagrams in the prediction of un- derground conditions in hot-spring systems: Journal of Volcanology and Geothermal Research, v. 5, pl. 16. Fournier, R. 0., and Potter, R. W., II, 1979, Magnesium correction to the Na-K-Ca chemical geothermometer: Geochimica et Cosmochimica Acta, v. 43, p. 1543—1550. Fournier, R. 0., and Rowe, J. J., 1966, Estimation of underground temperatures from the silica content of water from hot springs and wet steam wells: American Journal of Science, v. 264, p. 685—695. Foumier, R. 0., Sorey, M. L., Mariner, R. L., and Truesdell, A. H., 1979, Chemical and isotopic prediction of aquifer temperatures in the geothermal system at Long Valley, California: Journal of Volcanology and Geothermal Research, v. 5, p. 17—34. Foumier, R. 0., and Truesdell, A. H., 1973, An empirical Na-K-Ca geothermometer for natural waters: Geochimica et Cosmochimica Acta, v. 36., p. 1255—1275. Gat, J. R., 1971, Comments on the stable isotope method in regional groundwater investigations: Water Resources Research, v. 7, no. 4, p. 980—993. Hobba, W. A., Jr., Fisher, D. W., Pearson, F. J., Jr., and Chemerys, J. C., 1979, Hydrology and geochemistry of thermal springs of the Appalachians: U.S. Geological Survey Professional Paper 1044—E, 36 p. Hoover, D. B., and Tippens, C. L., 1975, A reconnaissance audio- magnetotelluric survey, Bruneau—Grand View area, Idaho, in H. W. Young and R. L. Whitehead, Geothermal investigations in Idaho, Part 2, An evaluation of thermal water in the Bruneau—Grand View area, southwest Idaho: Idaho Department of Water Resources, Water Information Bulletin no. 30, 126 p. Kharaka, Y. K., and Barnes, Ivan, 1973, SOLMNEQ: Solution- mineral equilibrium computations: U.S. Geological Survey Com- puter Contributions, PB 215—899, 82 p. Mabey, D. R., 1976, Interpretation of a gravity profile across the western Snake River Plain, Idaho: Geology, v. 4, no. 1, p. 53—55. McIntyre, D. H., 1979, Preliminary description of Anschutz Federal No. 1 drill hole, Owyhee County, Idaho: U.S. Geological Survey Open-File Report 79—651, 15 p. McKenzie, W. F., and Truesdell, A. H., 1977, Geothermal reservoir temperatures estimated from the oxygen isotope compositions of dissolved sulfate and water from hot springs and shallow drill holesz'Geothermics, v. 5, p. 51—61. Nir, A., 1964, On the interpretation of tritium “age” measurements of groundwater: Journal of Geophysical Research, v. 69, no. 12, p. 2589—2595. O’Neil, J. R., and Kharaka, Y. K., 1976, Hydrogen and oxygen isotope exchange reactions between clay minerals and water: Geochimica et Cosmochimica Acta,v. 40, p. 241—246. Parasnis, D. S., 1971, Temperature extrapolation to infinite time in geothermal measurements: Geophysical Prospecting, v. 19, no. 4, p. 612—614. . Pearson, F. J., and Truesdell, A. H., 1978, Tritium in the waters of Yel- lowstone National Park, in Robert E. Zartman, ed., Short Papers of the Fourth International Conference, Geochronology, Cosmochronology, Isotope Geology 1978, Colorado, August 20—25, 1978: U.S. Geological Survey Open-File Report 78—701, p. 327-329. J20 GEOHYDROLOGY OF GEOTHERMAL SYSTEMS Renner, J. L., White, D. E., and Williams, D. L., 1975, Hydrothermal convection systems, in D. E. White and D. L. Williams, eds., As- sessment of geothermal resources of the United States—1975: U.S. Geological Survey Circular 726, 155 p. Richmond, G. M., 1962, Quaternary stratigraphy of the La Sal Moun- tains, Utah: U.S. Geological Survey Professional Paper 324, 135 p. Rightmire, C. T., Young, H. W., and Whitehead, R. L., 1976, Geother- mal investigations in Idaho, Part 4, Isotopic and geochemical analyses of water from the Bruneau—Grand View and Weiser areas, southwest Idaho: Idaho Department of Water Resources, Water Information Bulletin no. 30, 28 p. Ross, S. H., 1971, Geothermal potential of Idaho: Moscow, Idaho, Idaho Bureau of Mines and Geology Pamphlet 150, 72 p. Sass, J. H., Lachenbruch, A. H., Munroe, R. J., Greene, G. W., and Moses, T. H., Jr., 1971, Heat flow in the Western United States: Journal of Geophysical Research, v. 76, no. 26, p. 6376—6413. Steams, N. D., Stearns, H. T., and Waring, G. A., 1937, Thermal springs in the United States: US: Geological Survey Water- Supply Paper 679—B, 200 p. Steward, J. H., and Carlson, U. E., compilers, 1978, Geologic map of Nevada: U.S. Geological Survey, 1 sheet, 1:500,000. Thomas, C. A., Broom, H. C., and Cummins, J. E., 1963, Magnitude and frequency of floods in the United States, Part 13. Snake River basin: U.S. Geological Survey Water-Supply Paper 1688, 250 p. Toth, J., 1962, A theoretical analysis of ground-water flow in small drainage basins: Journal of Geophysical Research, v. 68, no. 16, p. 4795—4812. US. GOVERNMENT PRINTING OFFICE: 576-034 Print Order 3 Truesdell, A. H., and Hulston, J. R., 1980, Isotopic evidence on en- vironments of geothermal systems, in P. Fritz and J. Ch. Fontes, eds., Handbook of environmental isotope geochemistry: New York, Elsevier Scientific Publishing Co., v. 1. 546 p. White, D. E., 1970, Geochemistry applied to the discovery, evaluation, and exploitation of geothermal energy resources, in United Na- tions Symposium on the Development and Utilization of Geother- mal Energy, Pisa, 1970, Proceedings: Geothermics, v. 1, pt. 2, Special Issue 2. White, D. E., and Williams, D. L., eds., 1975, Assessment of geother- mal resources of the United States—1975: U.S. Geological Survey Circular 726, 155 p. Young, H. W., Lewis, R. E., and Backsen, R. L., 1979, Ground-water withdrawals and associated convective heat flux in the Bruneau—Grand View area, southwest Idaho: U.S. Geological Survey Water-Resources Investigations 79—62/Open-File Report, 17 p. Young, H. W., and Mitchell, J. C., 1973, Geothermal investigations in Idaho, Part 1, Geochemistry and geologic setting of selected ther- mal waters: Idaho Department of Water Resources, Water Infor- mation Bulletin no. 30, 43 p. ' Young, H. W., and Whitehead, R. L., 1975, Geothermal investigations in Idaho, Part 2, An evaluation of thermal water in the Bruneau-Grand View area, southwest Idaho: Idaho Department of Water Resources, Water Information Bulletin no. 30, 126 p. Zietz, Isidore, Gilbert, F. P., and Kirby, J. R., Jr., 1978, Aeromagnetic map of Idaho: U.S. Geological Survey Geophysical Investigations Map, GP—920, 1 sheet, 1:1,000,000. UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY ll7°00’ T. 2 N. 430301...“ —- 43°30’ N. .2 Malhuer Co T. 5 S. I. 43°OO'W Jordan El Valley f” OREGON R. 3 W. Givens Hot Springs OREGON CORRELATION OF MAP UNITS I NEVADA UTAH INDEX MAP OF STUDY AREA Boundary of Bruneau—Grand View area Castle Creek Known Geothermal Resources Area 1l6°00’ R.4E. J. Strike Dam Bruneau Known Geothermal Resources Area Holocene and Pleistocene Pleistocene, Pliocene and Miocene Pliocene and Miocene Miocene and Cretaceous QUATERNARY TERTIARY CRETACEOUS > PRE-CRETACEOUS King Hill I 3’” 1 lilllv’l (PER 104 DESCRIPTION OF MAP UNITS QUATERNARY SEDIMENTARY ROCKS—Chiefly unconsolidated de- posits of clay, silt, sand, and gravel, primarily of fluvial, lacustrine, and glacial origins QUATERNARY AND TERTIARY SEDIMENTARY ROCKS—Poorly to well-sorted fluvial and lacustrine deposits ranging from clay to coarse gravel with intercalated basaltic lava flows and ash. Include Idaho Group and Payette Formation in Idaho and tuffaceous sedimentary rocks in northern Nevada QUATERNARY AND TERTIARY BASALT—Chiefly plateau and canyon filling olivine basalt flows in places, intercalated with elastic sedimentary rocks. Includes basalts of the Bruneau Formation and Banbury Basalt in Idaho and Banbury Basalt in northern Nevada TERTIARY SILICIC VOLCANIC ROCKS—Rhyolitic, latitic, and andesitic k a. n E. w . RIVER R. :2 E, Bliss o rocks, massive and dense, with jointing ranging from platy to columnar; include ldavada Volcanics in Idaho and northern Nevada mag; > TERTIARY AND CRETACEOUS INTRUSIVE ROCKS—Intrusive granitic rocks of comparable age and composition to the Idaho batholith PRE-CRETACEOUS SEDIMENTARY AND METAMORPHIC ROCKS—Undifferentiated elastic and nonclastic sedimentary and metamorphic rocks CONTACT ——I-— KNOWN FAULT—Bar and ball on downthrown side ---------- - INFERRED OR CONCEALED FAULT DRAINAGE BOUNDARY __._._ APPROXIMATE BOUNDARY OF CONCEPTUAL MODEL ll5°00' + 43°00' R, 13 E, .‘1 7" ,3 Malhuer Co 5 l; 42°00' m fT T. 47 N. f T. 44 N, _' o -, U E : o D , E n :3 . I ;. .‘l 4:. (A? Z ll7°00’ Base from US. Geological Survey l:500,000 scale map T. 42 N. ' 4l°30' W R. 47 E. D Hagerman T. 7 S. 114030’ T. 42 N. 41°30’ R. 53 E, ll6°00' EDGE T0 17°30’ GENERALIZED GEOLOGY OF SOUTHWESTERN IDAHO FOR THE NORTHWESTERN EDGE, 30 l l 50 R. 56 E. 114°30’ ' T. 44 N. -- -—-—-41°30‘ R. 64 E. Idaho geology modified from Bond, 1978. 5'0 MILES Nevada geology modified from Steward 1982 MAGNETIC DECLINATION FOR THIS SHEET VARIES FROM 16°OD’ FOR THE SOUTHEASTERN MEAN ANNUAL CHANGE IS 0°06’ WESTERLY AND NORTH-CENTRAL NEVADA Cassia Co l and Carlson, 1978. Structure in southwestern 80 KILOMETERS - Idaho from Ekren, McIntyre, and Bennett, 1978 GE.75 Pb v. EOW'J UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 1044~J GEOLOGICAL SURVEY PLATE 2 ll7"llU’ EXPLANATION _3500— Potentiometric—surface contour, spring 1979, dashed where approximately located. Contour interval is var- iable. National Geodetic Vertical Datum of 1929 0 Well 34BAD' . Well and number for which water-quality analysis is given (table 1) 33CBC|$ (? Spring and number for which water-quality analysis is given (table 1) O C.) Si . 1E»: Drainage boundary Lg. IDAHO z o T, E S. 8 l I: o T T‘ 2 3' NEVADA UTAH ~ 2 84l Swan Fafls \ o Dam \\ Murphy D . '—\ INDEX MAP OF STUDY AREA . ,. . M ..L...35AQBI...O..,._.. E O E ”ML. we I E \O “II—'0‘— T. 3 S T. 3 s. \e \ \ l \ DES/6:“ R E E. E M E I \ 00 E E l “PO R. 2 E. E s w” E Oak// % . : TE T. 4 s. -_ ‘: 5%.. u) E —-"TQE2V* ~ T. 4 s. . E U V300 /./ ">Catherine C’ ’9 T“- VVVVV w W l i .7 :i‘ D 955% / 1’ TI 5 8. Silver 34BADI/ I O m ’ 430 EN .8ADAI my ,/ g L Q ng Hill El: 09 E E 00 * - - (,3: o x , 1 30/ 4300 E] ' 0 6,) Grand Vze . . ‘ Jordan 0 \ \ Valley / 5“} e 8 0\ E—— \\ C J 3 k R 6E H tt TIES. o d3 m‘ er . I, , amme ' 'f U 00 263cm Reservoir K 51mm; R. 8 E. U W...“ ._ 9- H E- C 1 RIVER E. E2 E, Bliss U : ,,_, , _ o . ‘ MN E E T»: Z 2600\ \Q 0 \ / 5’ O\ O E 6 S \ 2 \ R. 13 E. ' ' _\ \Egfimcsms s ‘ ‘ E l f ,f' E l :50 \s g \D Bruneau / / I ’\ T. 6 S, \ 3’ $1 \3; , I l E W ~— ./ E .«y/ OE 0 SE 7ABBI \ E T. 7 S. l l E ,I/ .E E {$0 0 l— ‘_ 3 DHagerman K i / V/ " . \\ E T. 7 S, .. M .1 .. .._.. W . . m / E flat I / 9.2700- .- Elmore Co E / ’ 1 ”gym,” ‘ E South \ \ E T E / i I 1"" «WW9 C0 ‘- \_—»./ E 185. ’ Miuntain I} E‘ I ‘ E E/ E EE/ l ; ’ \\ \ ’1’ w \MEM/ E. . E I E E - l l / m 0 O 04 b E 1 5 M E ?ZOCCAIS ’ ; K l /1800 __’::Eg~ \ 5'19 / l E i ‘: I 5 E0 E f E l ’ ‘z r" \: / z , E .- : o / A. g 9 E 1 . E \ \// \T s \ \ \Ag 2 . SW .l 13CBCIS ? ’ \ v \ T. 8 S. l gofi‘flfork Owyhee ‘Z‘E, . . E \ / / / _.—- -— . O\‘ . MEWS » g - ~- : I . . i . E .. . l 3 E . = / \‘x U 0 / E /6" ‘3700 .E_ «9 Wm / E E 5 fl . E/ 734,00 0A 0 c4860 '_ ‘1 flex E margins w l E / .‘ 100 / j ‘l‘ _ “ / E E o T. 10 S. E \ E / / ’13: , / i E E \ E /f E, / - \ Castleford /3800-.. \%G 36/”‘T E E ‘ Eff 8 (/ E E3 1/ O \ ‘_ _ \T EEEEE l ‘ 26ABEs \o " E A ’ ‘ <2 8 /’ ; A E \9’ . . E “EH: x° X ' 0 T. EE S. / .-‘ 5, LL .53” ‘“\ +— Juniper ( / C3) E _. \\\ / ’ ’ mjvfipynmin- 5 I ’ T / / ’9 K... E , x 00 E E“ f“) T O /’ A0 a’ ’ T. E2 SI K 00/ E /” / EM \. ,. I ,/EEO/ \ \ \\ .12CDDE W4 7E,// ,r/AA // TESS. \ * /”\ ; \ l E /,/ 0/ E TESS. / \Au .3 Marvs~cjzegfi~) EE ' ‘ A50 i3 l T. I l. E //\/k\ 5/” \ \ Til/’7 . ‘C’chf—fii a // \ 9 4300 , / m. T M s 0% , E . /« E! E“ E _ I, .3 E ET. H s. ’5 i E [I / ’ £3“ ’ ’ 8 3 8'8 3 E : E 0 . o \ o l f . l . ~ U. Q ~~~~~ ——~-—--~. ~~~~~~ ~- . , o % ‘ s: 0 =3 T: / \ “RE ’ E— rl’ ’“ 0 'l ” “S ‘ 2 O I ; A! , 3! f E m E \Q E / ‘E \ 3/ l 3? Salmon FLT/E's g r"! :3 0 / $313 ( EKM’T' 0% f E Q Creek Resemozr £511 T 15 S a i {’3 L63: E \ $0 E L. M...” _ E w my“ . E g I EE E .m E E E l ) T E \ j LWApproxima‘ce boundar of l N E T' ‘6 3' I (3% ' E 5ng E E 0&6) E lDuckValleylndian Restart/ration {£24336 T , F ll C ( E / V NDA‘iQM . \f ___E Owyhee E}; E E, E __. I a... E DU '7— VALLEY _._.l. E EW 1 EmoflyE-Eee CO ‘ 7; _ EM W1.” ff 0 :1- 19.5% E “42“00' EVADA K/ ‘ EH“) CO E s E INDIAN f RESEM/ATTION E .. Elke Co ‘ \ * Ell“? C0 \NEVADA E E E E l l \E E E f E T. 47 N. \ E l E l 57 17CBC| E , l E Owyhee E g: . i \. l E E E \ l \ ’ ”m E \ a R, 67 E, E \ E z E \\ E E (E; [-— w 13ACClES} 23BBDlS T. 46 N. 1 E E E K " E \\ E E "i. ...... .__....... ”N Mountain City E / \ j ’ E ”" "“ E“”‘”'“' E” 7 ., 1 K..- T / E 714er / A E E E : \\ \k E /"’/ l E . E ‘E E\ ?25DAAIS E \\ E // I , E .5 r 7 t E "’ ' ‘ T“: / ‘ E 5 \\ E \ 2 . T 44 N / ~ E E‘\_\ “\\E\ :E \\ \ . WWWWWMW EU \J \\ E\ 3' \\ E T. M N, MM “ E O E E \ “Kw WiidHorse . «(Mum 30 L: E E b E E “*\\ \ Reservoir \ ‘ ‘ \ \T“\\Nw/ 3: ’ 3 \w/ R“ \ \ E. 43 N. R. 46 E2. \ E \ E1143 N. u\ \~\ \ ...._E ‘ R~ 55 E' \\ . \ . ’\ \ N- ,,,,,,,,,, c . ,, -- E E E /‘ ' ‘ I} R. 81 E. E l E. 42 N. , E 42 N. x. f M 4310' I 2.; E k Sf W.“ t . H. 57 E. L _ . 5 ~~~~~~ «a E n . E \\ E E E E T, 41 N E *’ T. 41 N. T.st N E E’ “X \ f E E E i . .. .‘ .. i K, E E E . R Eek”;’j’i.f_im T‘\ E ' \ ‘E ‘ R. 53 E. W H . * “x i' Base from us Geological Survey ‘5“ 80 F E D 10 20 30 40 ‘ 50 MILES 1:500,000 scale map E. 43 N. l ” T 40 N’ TLL 1 ' l J [E l l l I E E J E IE I l E E J I E 0 10 20 30 40 50 60 70 80 KILOMETEHS A M W A Em” “A 1982 MAGNETEC DECLINATIDN FOR THIS SHEET VARIES FROM 15°on' FOR THE SOUTHEASTERN .' j E EDGE T0 17°30' FOR THE NORTHWESTERN EDGE. MEAN ANNUAL CHANGE IS 0°06' WESTEHLY l , E SE; N /, E Er. 39 N, / GENERALIZED POTENTIOMETRIC SURFACE OF THERMAL WATER, T. 38 M. AND HYDROLOGIC-DATA SITE LOCATIONS IN SOUTHWESTERN ~ 38 E .. E IDAHO AND NORTH-CENTRAL NEVADA