ess. J *-s>0 s USNS Potomac Oil Spill Melville Bay, Greenland August 5, 1977 A Joint Report on Scientific Studies and Impact Assessment by the NOAA-USCG Spilled Oil Research Team and the Greenland Fisheries Investigations, Ministry for Greenland. August 1979 U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administration United States of America MINISTRY FOR GREENLAND Greenland Fisheries Investigations Denmark USNS Potomac Oil Spill Melville Bay, Greenland 5 August 1977 A Joint Report on the Scientific Studies and Impact Assessment by the NOAA-USCG Spilled Oil Research Team and the Greenland Fisheries Investigations, Ministry for Greenland Peter L. Grose and James S. Mattson Center for Environmental Assessment Services Environmental Data and Information Service National Oceanic and Atmospheric Administration U.S. Department of Commerce Washington, D.C. 20235 USA Hanne Petersen Greenland Fisheries Investigations The Ministry for Greenland Jaegersborg Alle 1 B DK-2920 Charlottenlund, Denmark Major Contributors Paul Boehm and David Feist Energy Resources Co. 185 Alewife Brook Parkway Cambridge, MA 02138 Vibeke Jensen, Nis Hansen, and K. Krongaard Kristensen Water Quality Institute Agern Alle 1 1 DK-2970 Hoersholm, Denmark Martin Ahnoff and Goran Eklund Department of Analytical Chemistry University of Gothenburg GU-CTH S-41296 Goteborg, Sweden I August 1979 Washington, D.C. NOTICE The National Oceanic and Atmospheric Administration does not approve, recommend, or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to the National Oceanic and Atmospheric Administration or to this publication furnished by the National Oceanic and. Atmospheric Administration in any advertising or sales promotion which would indicate or imply that the National Oceanic and Atmospheric Administration approves, recommends, or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirectly the advertised product to be used or purchased because of this National Oceanic and Atmospheric Administration publication. 11 PREFACE The spill of Bunker-C fuel from the USNS POTOMAC on August 5, 1977, in Melville Bay, Greenland, is among the few oil spills studied in Arctic waters. This report presents the studies and findings of the NQAA-USCG Spilled Oil Research and Gron lands Fisker iundersogelser (Greenland Fisheries Investigations, Ministry for Greenland) teams which responded to study the spill. This spill is noteworthy for three reasons: 1) the spill occurred in the pristine waters of the Arctic which have very low background levels of petroleum hydrocarbons and a fragile ecology, 2) the fuel oil, (a blend containing 55 percent pitch with a specific gravity of 1.054) remained on the surface until it weathered sufficiently to sink, and 3) a fairly comprehensive sampling program was undertaken in spite of the remoteness and logistic problems, which served as the basis for comprehensive studies on the fate and impact of the spilled oil. ili ACKNOWLEDGMENTS This report represents the efforts of many people and organizations without whose support and dedication it could not have been generated. The following scientists participated in the field investigations: P. L. Grose, Physical Oceanographer , NOAA; J. S. Mattson. Marine Chemist, NOAA; E. I. Chan, Marine Ecologist, NORA; G. Mueller, Arctic Biologist, University of Alaska; H. Petersen, Environmental Biologist, Greenland Fisheries Investigations (GF) ; S. A. Horsted, Fishery Biologist, GF; J. Chr istensen, Marine Ecologist, Marin ID. The field program was supported by the U.S. Coast Guard Cutter WESTWIND, the Military Seal if t ship MIRFAK, and the Greenland research vessel ADOLF JENSEN. While valuable support was received from all of the personnel on these vessels- specific acknowledgment of assistance is due to: Crndr. D. Super, Captain: Marine Science Technician (MST) Chief L. Pierce; B. Grahm, MST2; and G. Davis, MST3; and Lt. Crndr. J. Carruthers, Executive Officer, all from the WESTWIND as well as Captain J. Petersen and the entire crew of the ADOLF JENSEN. Many useful observations and oil samples were received from the U.S. Coast Guard Atlantic Strike Team and the U.S. Navy Superintendent of Salvage representatives involved in the cleanup effort. The biological samples were sorted and analyzed by R. Maurer and J. Kane of the Narragansett Laboratory of the NOAA National Marine Fisheries Service and 0. Norden Andersen of Marin ID, Denmark. Chemical analyses were performed by Energy Resources Co. for the NOAA samples and by the Water Quality Institute and University of Gothenburg (Sweden) for the Danish samples. Microbiological studies were performed by the Water Quality Institute. Funding for the studies was supplied by NOAA's Environmental Data Service, U.S. Navy Superintendent of Salvage, and the Ministry for Greenland. Publication costs were underwritten by the NOAA Hazardous Materials Response Program. IV TABLE OF CONTENTS PREFACE . 1 i i ACKNOWLEDGMENTS iv 1.0 INTRODUCTION 1-1 2.0 SCIENTIFIC RESPONSE TO THE OIL SPILL 2-1 2 . 1 NOAA Response 2-1 2.2 Danish Response 2-3 2.3 On-scene Coordination . . . . 2-4 3.0 FIELD PROGRAM 3-1 4.0 MOVEMENT OF THE OIL 4-1 4.1 Circulation And General Climatology .Of Baffin Bay 4-1 4.2 Currents And Horizontal Advection~6f The Oil 4-2 4.2. 1 Geostrophic Currents 4-2 4.2.2 Direct Measurements 4-4 4.2.3 Wind Stress Currents 4-5 4.2.4 Summary of Currents 4-6 4.2.5 Horizontal Advection of OIL 4-6 4.3 Physical Observations 4-7 A. A Vertical Movement . 4-8 4.5 Summary Of Oil Movement 4-9 5.0 WEATHERING OF SURFACE OIL 5-1 5.1 Ana lut ica I Procedures . 5-4 5.2 Results 5-7 5.2.1 The Reference Samp le 5-7 5.2.2 Gas Chromatography of the Surface Oil Samples 5-8 5.2.3 Spectrof luorometry, NOAA Samples 5-17 5.2.4 Spectrof luorometry and Mass Fragmentography, Danish Samples .. 5-18 5.2.5 Aspha I tenes 5-26 5.3 Summary For Weathering Of Surface Oil 5-22 6.0 MICROBIOLOGICAL STUDIES 6-1 6.1 Counting And Identification Of Microorganisms 6-1 6.1.1 Analytical Procedures 6-1 6.1.1.1 Total Count on Agar Plates 6-1 6.1.1.2 Yeast and Fungi on Agar Plates 6-2 6.1.1.3 Oil Degrading Bacteria on Agar Plates 6-2 6.1.1.4 Oil Degrading Yeast and Fungi on Agar Plates 6-2 6.1.1.5 Oil Degrading Microorganisms by Most Probable Numbers 6-3 6.1.2 Identification of Isolated Oil Degrading Bacteria 6-3 6.1.3 Pesul ts 6-3 6.2 Degradation Rates By Natural Cultures Of Microorganisms 6-5 6.2. 1 Procedures 6-5 6.2.2 Results 6-7 6.3 Degradation Of Oil By Isolated Monocultures 6-11 6.3.1 Procedure 6-11 6.3.2 Results 6-13 6.4 Summary Of Biodegradat ion Studies 6-13 7.0 ACCOMMODATION OF OIL INTO THE WATER COLUMN 7-1 7 . 1 Samp I ing Procedures 7-1 7.1.1 The Danish Samples T- 1 7.1.2 The NOAA Samp I es 7-2 7.2 Analytical Methods 7-3 7.2.1 Danish Samples 7-3 7.2.1.1 Ut^-Spectrof luorometry 7-3 7.2.1.2 Gas Chromatography/Mass Spectrometry 7-4 7.2.2 NOAA Samples r-6 7.3 Results 7-7 7.3.1 The Danish Analyses 7-7 TABLE OF CONTENTS (CONTINUED) ?. 3. 1.1 Quantitative Ul^-Spectrof luorometry r~? 7.3.1.2 UV Comparison of the Danish and NOAA Sampling Methods 7-13 7.3.1.3 Mass Spectrometr ic (MS) Analysis 7-14 7.3 . 1.4 Compar ison of MS and UV Resui ts 7-16 7.3.2 The NOAA analyses 7-17 7.3.2. 1 Gas Chromatography „ 7-28 7.3.2.2 Ul^-F luorescence 7-25 7.4 Summary Of Accommodation Into The Water Column 7-27 e.G BIOLOGICAL STUDIES ON PLANKTON AND FISH 8-1 8. 1 Samp I ing Procedure 8-1 6.1.1 Dan ish Samp I ing 8-1 8.1.2 U.S. Samp ling ' r-A 8.2 Species Compos ition 8-4 8.2.1 The Dan i sh Samp I es 8-4 8.2.2 The U.S. Samples 8-6 8.2.3 White Particles on the Sea Surface 8-16 6.3 Oil Contamination Of Zooplankton And Fish 6-10 8.3.1 Chemical Examinations 8-16 8.3.1.1 Analutical Procedure 8-10 8.3.1.2 Results 8-11 8.3.1.3 Discussion 8-13 6.3.2 Physical Examinations 8-14 8.3.2. 1 Procedure 6-14 8.3.2.2 Resu Its 8-14 8.3.2.3 Discussion 8-16 8 . 4 Summary 8-16 9.6 MARINE MAMMALS AND SEABIRDS 9-1 9. 1 Observations Of Marine Mammals 9-1 9.2 Oil Contamination Of Sealskins 9-1 9.2. 1 Analytical Procedure . 9-2 9.2.2 Results 9-2 9.3 Seabird Observations 9-7 16.6 IMPACT ASSESSMENT 1 Pi- 1 10. 1 Fate Of The Sp i I led Oil 10-1 16.2 Impact Of The Spilled Oil On Biota 10-3 10.3 Conclusions 16-4 11.0 SIGNIFICANT SCIENTIFIC FINDINGS 11-1 11.1 Observations On The Behavior And Fate Of The Spilled Oil 11-1 11.1.1 Sinking of the Oil , 11-1 11.1.2 Weathering Rates 11-2 11.2 Biological Findings 11-3 11.2.1 B iodegradat ion .• 11-3 11.2.2 Zooplankton 11-4 11.2.3 Birds and Mamma Is 11-4 11.3 Conclusions 11-5 12.0 REFERENCES 12-1 13.0 APPENDIX 13-1 13. 1 Marine Agar 13-1 13.2 Agar Substrate 13-2 13.3 Bunch Substrate (used for MPH method) 13-3 13.4 ColweU Substrate (used for MPN method) 13-4 vi LIST OF FIGURES 1-1 USNS POTOMAC oil spill site and surrounding area 1-2 3-1 Locations of WESTWIND sampling stations 3-4 3-2 Locations of ADOLF JENSEN sampling stations 3-5 3-3 Detailed locations of NOAA and Danish stations 3-6 4- 1 Dynamic topography of the sea surface 4-3 4-2 Vertical density distribution at three stations 4-4 4-3 Wind runs for 10-12 and 15-19 August 4-5 5-1 Distillation curve for the cutter stock 5-3 5-2 Analytical scheme for the NOAA surface oil samples 5-6 5-3 Gas chromatogram of fl (saturate) fraction of POTOMAC fuel 5-8 5-4 Gas chromatogram of f2 (aromatic) fraction of POTOMAC fuel 5-8 5-5 Relative peak abundances of n-alkanes from NOAA analyses 5-9 5-6 Relative peak abundances of n-alkanes from Danish analyses 5-9 5-7 A time series of gas chromatograms of surface oil samples 5-11 5-8 GC/MS results for f2 (aromatic) fraction of POTOMAC fuel. 5-12 5-9 GC/MS results for f2 fraction of 18 August surface sample 5-13 5-10 GC/MS results for f2 fraction of 19 August subsurface sample 5-13 5-11 Ratio of nC l?/pr istane for spilled oil samples 5-14 5-12 Synchronous fluorescence scan of POTOMAC fuel 5-15 5-13 Relation between peak height and fuel concentration 5-16 5-14 Excitation and emmission spectra of a surface oil sample 5-19 5-15 Comparison of concentrations of selected aromatics 5-21 5-16 Change in asphaltene content with time 5-21 6-1 Gas chromatograms from first experiment 6-8 6-2 Gas chromatograms from second experiment (control) 6-9 6-3 Gas chromatograms from third experiment (added nutrients) 6-18 7-1 Analytical scheme used for NOAA water samples 7-8 7-2 Oil concentration as a function of depth at four stations 7-12 7-3 Fluorescence spectra of a sample containing spilled oil 7-13 7-4 Fluorescence spectra of a sample containing cooling water 7-14 7-5 Fluorescence spectra of uncontaminated water 7-15 vil LIST OF FIGURES (CONTINUED) 7-6 Fluorescence spectra of NOftA sample. 7-17 7-7 Example of a mass fragmentogram 7-19 7-8 Relative aromatic concentrations vs. molecular weight. 7-20 7-9 Trimodal gas chroma togr am of extract 'd, fl fraction 7-25 7-18 Trimodal gas chromatogram of extract 2, f2 fraction. 7-25 7-11 Cap contamination gas chromatogram of extract 10-2b. fl fraction. . 7-26 7-12 Cap contamination gas chromatogram of extract 2B-2b, f2 fraction. . 7-26 7-13 Gas chromatogram of vial cap, f0 (unfract ionated) . 7-27 7-14 Low level gas chromatogram of extract 30-lb, f0 (unfract ionated) . . 7-27 8-1 Comparison of species compositions from neuston and bongo ........ 8-9 8-2 Gas chromatogram of Boreomys is from station 5465 8-11 8-3 Gas chromatogram of Themlsto from station 5465. ................... 8-12 8-4 Gas chromatogram of Copepod from station 5470. 8-12 8-5 Gas chromatogram of a Copepod 50 nmi from spill (station 5477). ... 8-13 9-1 Capture locations of analyzed seals. 9-3 9-2 Gas chromatogram of seal 1 (heavily oiled) 9-4 9-3 Gas chromatogram of seal 2 (heavily oiled). 9-4 9-4 Gas chromatogram of seal 3 (spots on neck) . 9-5 9-5 Gas chromatogram of seal 4 (far from spill site). 9-5 9-6 Gas chromatogram of seal 10 (spots on necK) . 9-6 9-7 Gas chromatogram of seal 13 (far from spill site). 9-6 vili LIST OF TABLES 3-1 Scientific stations occupied during the POTOMAC oil spill 3-2 3-1 Scientific stations (continued) 3-3 5-1 Surface oil samples col lected 5-2 5-2 Concentrations of naphthalenes and phenanthrenes 5-12 5-3 Concentrations of 2-, 3-, 4- , and 5 ring aromatic compounds 5-16 5-4 Characterizations by spectrof luorometry of surface samples 5-18 5-5 Concentrations of ten aromatic isomeric groups 5-26 6-1 Enumeration results from microbiological examinations 6-4 6-2 Biochemical test results for isolated strains 6-6 6-3 Results from GC analyses of experiments with natural cultures 6-7 6-4 Growth of natural cultures , 6-11 6-5 Growth of isolated monocultures in time and temperature 6-12 7-1 Trace impurity concentrations from solvent in Danish samples 7-5 7-2 List of samples for which results are not reported 7-9 7-3 Oil concentrations in Danish water samples. 7-10 7-4 Amounts of interfering substances within solvents 7-11 7-5 Extraction efficiency of Danish procedures 7-15 7-6 Results of intercompar ison sampling 7-16 ?-? Samples analyzed by mass fragmentography 7-18 7-8 Concentrations of selected aromatic compounds in water samples. ... 7-18 7-9 Comparison of GC/MS and UV analyses 7-2 1 7-10 Seawater extract analyses- NOAA 7-22 7-10 Seawater extract analyses, NORA (continued) 7-23 7- 1 1 Group ings of NOAA seawater samp les 7-24 8-1 Summary of Danish biological stations 8-2 8-2 Summary of NOAA biological stations. 8-3 8-3 Zooplankton enumeration from Danish stations 8-5 8-4 Fish larvae collected at Danish stations 8-6 8-5 Zooplankton enumeration from NOAA stations 8-7 8-6 Occcurrence of oil contamination on dominant plankton groups 8-15 9-1 Date and locations of seal captures 9-2 ix 1.0 INTRODUCTION On the morning of August 5, 1977, the U.S. Navy ship POTOMAC, bound for ThuLe Air Force Base with a cargo of arctic-blend diesel fuel, was being escorted in intermittent dense fog by the U.S. Coast Guard cutter WESTWIND through the scattered sea ice of Melville Bay in the northeastern part of Baffin Bay off western Greenland. The WESTWIND moved into a large ice floe at 9 to 10 kn, followed by the POTOMAC. When both vessels were in the ice floe, a crewman on the POTOMAC sighted oil on the water at 0430 local time. The WESTWIND was immediately informed. The POTOMAC'S position at the time of the sighting was 74°52'N, 61°13'W (Figure 1-1). It was discovered that the No 2 Bunker (deep port) tank, containing about 107,000 us gallons of Bunker-C fuel, had been "holed" by an iceberg or "growler". Almost all of the fuel in the holed tank was eventually spilled. Calm seas with waves of to 2 ft and light winds of to 7 kn kept the oil from dispersing for several days after the spill. While the WESTWIND escorted the POTOMAC to Thule with a water bottom in Uer holed tank, representatives of the U.S. Coast Guard (USCG) National Strike Force and the U.S. Navy Military Seal if t Command flew to Thule aboard a USCG C-130, arriving on August B. Immediately upon arrival, they left for the scene of the spill onboard the WESTWIND. From August 10 to 12, the C-130 flew over the entire area, carrying its airborne oil surveillance system for mapping the extent of the oil slick. Helicopters from the WESTWIND also participated in the search. The C-130 located the major concentration of the" oil on August 12, and directed the Coast Guard helicopters and the icebreaker WESTWIND to the scene. 1-1 1-2 90° 80° 70° 60° 50° 40° 30° 20 75° USNS POTOMAC oil spill site and surrounding area. The spill site is marked by + in Melville Bay. 2.0 SCIENTIFIC RESPONSE TO THE OIL SPILL As is standard for oil spills, the scientific response to the POTOMAC oil spill was an ad hoc effort by concerned scientists. The U.S. and Danish responses were not fully preplanned nor were they coordinated before arrival on scene. This section summarizes the events and planning that culminated with their arrival on scene and the objectives for their responses. Summaries of the impact of the spill and the significant scientific findings are presented in Chapters 19 and 11. 2. 1 NOAA Response On August 11 and 12, the Office of the Chief of Naval Operations called upon the Office of Environmental Monitoring, National Oceanic and Atmospheric Administration (NOAA), regarding possible advice or assistance with the POTOMAC oil spill. After consultation within NOAA, the Navy was informed that the NOAA-USCG Spilled Oil Research Team (SOR) could lend assistance. On August 14, a meeting was held at the U.S. Navy Military Seal if t Command in Washington D.C. attended by representatives of the Navy Superintendent of Salvage, Chief of Naval Operations, and the NOAA SOR team. At this meeting a description of the water circulation in the vicinity of the spill was presented along with the most recent "pollution report" from on scene. This report stated that there had been a "fairly rapid dispersion and breakup considering [the: nature of C the oil]... the rainbow streamers Chad] disappeared leaving chocolate mousse streamers ... [weather] calm ... sea conditions have not accelerated this breakup ... C the oil] will not float forever; it will breaKup and dissipate through evaporation and Cchemical] breakdown with subsequent dispersion into [the] water column," and "oil within streamers on 12 Aug showed 50,003 gals of recoverable oil ... a 50 percent recovery would be optimistic." They cautioned against overconf idence, continuing "a [weather] change with [increased] wind could disperse the oil so widely that recovery could be reduced to 1-5,000 gals." The message finished with, "CcJonversely, the oil could be pushed ashore contaminating 20 miles of ice and coastline with 10-25,000 gals ...," and since "impact would be in a 2-1 2-4 2.3 On-scene Coordination On August 17, the NOAA and Danish scientists met onboard the ADOLF JENSEN to exchange plans and discuss scientific coordination. At this meeting it was agreed that the chemistry sampling was all to toe done from the ADOLF JENSEN because of better logistic support . Consequently, one of the NOAA team joined the ADOLF JENSEN with the required sampling and analytical equipment. It ojas also agreed that the NOAA team remaining on the WESTWIND would focus its efforts on the movement of the oil and on obtaining neuston and bongo tows for biological samples at or near the surface. The Danish team on the ADOLF JENSEN would concentrate on the subsurface biological sampling in addition to the chemistry. Daily contact was maintained between the two teams to ensure coordination and to discuss current findings. 3,0 FIELD PROGRAM The field program in response to the USNS POTOMAC oil spilt was concentrated from August 13 to 21 for the vast majority of the sampling. Forty-six stations were occupied from the U.S. Coast Guard cutter WESTWIND and the Greenland research vessel ADOLF JENSEN. The locations of these stations are presented in TaPle 3-1 and Figures 3-1 to 3-4. Also noted in Table 3-1 are the sampling programs under taKen at each station. These sampling stations were the basis for the physical, chemical, and biological studies conducted by the NOAA and Danish scientists. In support of the physical studies, 11 temperature profiles were obtained to depths of 459 m using expendable bathythermographs (XBTs), eight stations collected hydrographic temperature and salinity data using Hansen bottles and reversing thermometers, and, at two stations, expendable current measuring probes were deployed. The physical programs were under taKen to determine the movement of the oil and are contained in Chapter 4. Thirteen surface oil samples were collected Tor studies on the weathering of the surface oil. Atiquots of these samples were analysed as part of the NOAA and Danish programs and are discussed in Chapter 5. To ascertain the impact of microbiological degradation of the spilled oil, cultures were grown on various substrates from samples collected at five stations by Danish scientists. The results of these studies are contained in Chapter 6. Water samples were collected at IS stations to determine how much oil was accommodated into the water column. More than 58 samples at depths down to 30 m were obtained. The NOAA team collected their samples using a NisKin sterile bag sampler while the Danish team collected their samples with a 1.0 l glass bottle lowered in an stainless steel frame. These bottles were fitted with a teflon-lined cap which was pierced by a spike when a messenger was dropped, allowing the bottle to be filled. The bottles were not resealed before being brought hack through the surface. Except for two reference stations acquired on the return to Thule by 3-1 3-2 rH •H a CO O O H O WD 3 •H 3 0) •H P- 3 U o o co c o •H 4J CO i-i co o •H c 0) -H CO I CO H o u •K -a CD 12 ■u rJ O as a 3 to o E CO U Sh 00 O 53 !h Pu x cO txO 0) 4-1 •H E2 Ml 3 o rJ CD X) 4J •H 53 CO ►J CO 4-1 CO a) •H 2 d c d CD CD * K^ K*S K*N P*N * CN o rH o H o H rH o rH o rH H ^o m m oo cn cn o o co h rH rH rH O rH ^O \0 \C O \D m on cm o cm m co o o <-\ H o H m H rH rH H H O H \D vD \0 \D ^D ^D ^D o m m m st H co i— i i— i cn ^O ^O vD *X) ^O CN O o CO m o H m CM CN rH CM H CO H st .-H o o CN o n o m o st o m O m m o m o m o o m o m o LO O rH CN CO H CM ^D \D VO \0 I I St st - r^. m O o CO O O O CN m o o rH rH o rH CO st CO r-^ H o >4D CM CO rH H CM r-l CM r-{ ■H o O rH rH CM GO 00 00 00 00 00 00 a) r-H 3 3 3 3 3 3 3 4J r~ <3 cu H •H g cu r» 4J r^ CO CT> Q rH 3 CU CO c CU EC <^i 3 •rl e co co co co 3 3 J-i 3 o J-l co co 3 3D cu cu 4-> cu ai cu co S3 S3 en 2 S3 S3 pq tJ N N tsl N N o S3 c o co •H 5-i CO CX f3 o a u cu 4-1 3 S3 § o M bO ^ o CO Xi QJ 4-1 CO cO 3-3 3 ^ o u CO o I 1 o 3 o 4J CO 3 a' S3 O 4J CO 3 CU S3 o 4J CJ CJ co co m co co co co co co H ^CM CN N O O S3 O S3 s s o X! CO N CM CN CO CM CN O-d-o-d-asr^ooin rHrHr-IOr-lr-ICOOr-l r- 1 i — I i— Irlr !HOHH o H CM CM O O m iH 00 m co o H o o CN O o o uo CTi o r-« vO o 00 LO m m en m CfiO^DCMXCOCMinvO HCMCMCMHHCMHH ininuiminioinmin vf m r- I vo r-^ oo -> 3 <3 oor^oininoNOoo . £1 •u cO pq cu H XI cO -3 .. c CO cu a. cO X 4J O 3 CU CO 3 cO S3 v —' 4J 3 cO CU 4J CO CU -o CJ I M bO O U (U X •K J-l 3 cj co •rl CO J3 CU & CO U 4J bO 3 CU >-i J-i T3 J-4 S^ 3 S3 CJ 1 I S3 cj CU s cO CO .3 J-i CO cu ■H T3 3 P bO •H O O H O CU -H a x co o 3 3 CU J-l CJ CO cu •3 < CO O 4-1 S3 cu I 3 S3 cu ex CO OJ & •H O £XH S cO 3 co O 4J M A! CU 3 4-1 cO cfl tH 3 Ph CO Jg I I I I o a 3 ni 3-4 011 CAPE ty MELVILLE * O SAVIGSIVIK CAPE YORK O 10 CAPE WALKER 9 O 1 17 AUG . 2 - • A— V ( fc 75' cS^-\ • 5469 5470 .'r.i. 5468 5467 17 AUG 5462 12 AUG L& 15 AUG (5471 IV i5466 ''5473 \P" 5464 \i5474 ••5463 m * % #5472 5461 5465 '< 5460 SPILL SITE '5475 • 5476 _L__ (5477 1 62° 61° 60° 59° 58° 57° Figure 3-2. Locations of stations occupied by the ADOLF JENSEN, 76° 75° 3-6 62' 40' 20 30' " T " 17 AUG 61° 40' 5462 2 17 AUG 20' 10' 75 °N 12 AUG £N WW-6 X-34I WW-4 ••X-36 r WW-3 •5466 ^WW-7»\ • \l WW-2« X • WW-8 ww-iV 5473 >5474 X-32, 33 15 AUG 5461 W2 5460 + SPILL SITE J W1 A 5465 • 62° Figure 3-3. 40' 20' 61 W 40' 20' — 20' 5475 30' 10' 75°N 20' Detailed locations of NOAA and Danish stations in the area of highest oil concentration. See Table 3-1. 3-7 net performed oblique tows from 200 m depth for about 30 minutes at 1.5 kn speed- while the Hensen net was hauled vertically from 50 m at 0.33 m/s. The results of the species composition and contamination studies are contained in Chapter 8. Limited observations on marine mammals (seals) and seabirds were made during the study period. Oiled skins of seals, shot later in the year by local Eskimo hunters- were made availible to the Danish team for chemical analyses. These studies are presented in Chapter 9. As with any ad hoc scientific response, not all the desired information was collected during the field program. In retrospect, there should have been more direct measurements of the surface currents using the current probes, and it would have been desirable to have conducted the zooplankton sampling to account properly for diurnal variability. Probably the most important data not collected were numerous samples of the subsurface oil for weathering studies. However, given the short notice before departure, it is probably significant that the sampling was as complete as it was. 4.8 MO^MENT OF THE OIL To assist the On-scene-Coordinator and to assess ecological effects. It was necessary to determine and predict the movement or the spilled oil. This was done by both direct measurements and modeling. The horizontal advection of spilled oil is influenced by the general circulation or permanent currents upon which the local effects of wind stress and waves on the oil are superimposed. These local effects tend to move the oil independently of the surrounding watpr. The approach used was to determine separate contlbutions for the three time scales of motion — days for the permanent currents, hours for wind stress, and minutes for the wave interaction — and then draw conclusions based on their relative magnitudes. Extensive use was made of a report by Muench (1971) and supporting data from U.S. Coast Guard icebreaker cruises in 1968, 1969, and 1978 (Muench et al., 1971; Moynlhan and Muencn, 1971; and Muench, 1972) . 4.1 Circulation find General Climatology Of Baffin Bay Baffin Bay is a semienclosed body of water with limited access to the Arctic Ocean to the north and to the Atlantic Ocean to the south through Davis Strait. The waters in the bay consist of four layers: a surface layer extending to a depth of about 5 to 18 m generated by ice melt, a cold subsurface *layer (< 8 C) from 28 to 188 m deep generated from the mixing of Arctic water and Atlantic water, a deeper layer of warm water (> 8 C) from 150 to 400 m deep originating from the Atlantic, and a bottom layer of cold water (< C) below 500 m. The circulation of Melville Bay is dominated by the West Greenland Current, which flows to the north along the west coast of Greenland. This current appears to be driven primarily by runoff and ice melt as the climatological winds are not very strong. The current primarily is contained in the upper layer of the bay (down to 20 m) . 4-1 4-2 4.2 Currents find Horizontal Rejection Of The Oil 4.2.1 Geostrophic Currents Historical, data (Muench, 1972) indicate that the predominant surface currents in Melville Bay are roughly parallel to the coast. In the area of the spill, the current is to the northwest, changing to west off Cape YorK. The geostrophic currents follow the contours of dynamic topography shown in Figure 4-1. Since these currents are based on isolated hydrographic stations acquired before 1371, the geostrophic currents were estimated during the spill from three hydrographic stations occupied on August 18 and 15 by the WESTWIND. fit these stations, temperatures and salinities were measured with Nansen bottles and reversing thermometers at seven depths from the surface to 158 m. The locations of the stations are shown in Figure 3-4. The primary purpose of the stations was to define the density structure at the top and bottom of the subsurface Arctic water as an aid in predicting the location of the oil in the water column in the event the oil sanK. Thus the acquired data did not adequately define the major density structure at the surface required for geostrophic calculations. Additional levels were interpolated for, based on the temperature-salinity relationship derived from the Nansen casts and the temperature structure derived from XBTs at each station. The density structure was found to be nearly constant below 4Q m (Figure 4-2), in accord with historical data, and the surface currents were computed relative to a depth of 40 m. The calculations indicated a westerly flow C260°T) at 3 em's, which is only a crude estimate because of t\r\e interpolation used to supplement the Nansen data and the lacK of synoptic ity. The error bounds for the speed were estimated to be -3 to +5 cm/s and +/- 30 degrees for direction. Since the origin of the lighter surface water is ice melt, the derived current is consistent with the greater concentration of ice near the shore. Several hydrographic stations were occupied by the Danish researchers from the ADOLF JENSEN, but simultaneous XBT data were not availible to allow interpolation of the required additional levels to define the density structure. Also, all but one of the ADOLF JENSEN stations were more than 30 nmi from the area of the spill or were separated in time by more than 36 hours from the WESTWIND stations. The one comparable station (Station 5466) was consistent with the three WESTWIND stations. 4-3 60° 50° • 1928 STATIONS X 1940 STATIONS Figure 4-1 Dynamic topography of the surface with respect to 1,500 decibars (from Barnes, 1941). Contour interval is 0.05 dynamic meters (from Muench, 1972). 4-4 Figure 4-2. Vertical density distribution at three hydrographic stations. Nine additional XBT stations in the area indicated that the top of the subsurface Arctic water (6 C isotherm) was between IB and 38 m deep. However, as the sensity at temperatures near C are dominated by salinity effects, these XBT data did not yield any further information for estimating the geostrophic current. 4.2.2 Direct Measurements On August 20, an attempt was made to measure surface currents directly with three Richardson current probes deployed from a helicopter. Only one was partially successful; the other two failed to release any subsurface floats. From the single float released from the one partially successful probe, the surface current was computed to be approximately 8 cm/s in a southwesterly direction <220°T) . This current would correspond to the currents at 1 m depth and is relative to the net transport of the entire water column. Based on measurements of the elongation of the dye patches caused primarily by Stokes drift and only slightly by windage, the uppermost surface waters (5 cm) were calculated to be moving at 2 cm/s relative to the water at a depth of 1 m and at a right angle to it (125°T). 4-5 4.2.3 Wind Stress Currents A second method for assessing the magnitude of the surface currents is to model the currents from wind observations. The wind model indicates that the wind driven surface currents were directly downwind with a speed of about 3 percent of the wind speed. Wind observations were made on the WE3TWIND at 6-hour intervals while the ship was in the vicinity of the spill. Tbese wind data and visual estimates of wave heights are contained in Mattson and Grose (1973) . Figure 4-3 shows the wind runs for the two periods when the WESTWIND was o so ioo .15 AUG. 1200 GMT 12 AUG. 0000 GMT %. V \ 408NM 128* TRUE s 18 AUG. 0000 GMT *\ S 18 AUG. 0000 GMT 19 AUG. 1800 GMT 11 AUG. 0000 GMT I 10 AUG. 0000 GMT Figure 4-3. Wind runs for 10-12 and 15-19 August. The nominal position is yS^lS'N, 61* W, at the scene of the spill. From these runs, one can model that the oil would have moved only 12 nmi in a nortwesterly direction (340°T) from August 10 to 12 and about 12 nmi in a southwesterly direction (120°T) from August 15 to 13. Oil locations for August 15 and 17 (Figure 3-4) indicate a net drift of 4 nrni to the northwest (310°T), and (for this time period) the model predicts a drift of 2 nmi to the south (160°T) at an average speed of 2 cm/s. From this difference between the slicK locations and the wind model results, one can 4-6 infer a general circulation of 6 cnvs in a northwesterly direction (320°T). 4.2.4 Summary of Currents Estimates of the three time scales that contribute to the local surface current are as follows: 1-General circulation. a-Historical values of less than 29 cnvs to the west. c-Geostroptc currents of 3 cnvs to the west (2C0°T) measured realtime to 40 m depth. c-Inferred currents from the wind model of 6 cm/s to the northwest (320°T) 2-Surface currents at 1 m. a-Current of 8 cm/s to the southwest (217°T) measured with Richardson probe. fa-Average currents of 2 cm/s from wind stress in various directions. 3-Surface currents at 5 cm (measured Stokes drift of 2 cm/s). 4.2.5 Horizontal Advection of Oil Large errors, on the order of a factor of 2, are associated with all the measurements described above. However, there is no doubt that the surface currents are of small magnitude. The two shorter time scales probably contribute less than 5 cm/s and tend to average to because of the random directions. The long-scale motions appear to be westerly at about 5 cm/s, which indicate that the oil was not transported far from the site of the spill, i.e., less than 40 nmi over a 2-week period. The elongated slick pattern was probably caused by spillage over a period of time (1-hr duration at 8 kn = 8 rimi length). Once generated, this pattern was then advected by local currents in random directions, depending primarily on winds with a net set to the northwest from the permanent currents. Oil observed in neuston tows northwest of the main spill site during the return to Thule on August 20 probably stemmed from continued leakage by the USNS POTOMAC as she made her way to the same port via the same route, because the observed currents were not strong enough to have carried the prlnciplal spill that far. Advection of the oil by the surface currents did, however, allow a large surface area to be exposed to the oil at one time or another. The exposed area 2 weeks after the spill was estimated at 500 sq ml based on oil locations 4-7 and current measurements. 4.3 Physical Observations Shortly after the spill, the oil was reported to he in the form of small pancakes 10 to 25 cm in diameter and 9.5 cm thick, which were organized in windrows about 3 m wide presumably from wind-induced Langrnuir circulation. Sheen, a visible but thin slicK, was seen emanating from these pancaKes, and major concentrations were easy to spot from the air. By August 13, 14 days after the spill, the oil was no longer noticeable from the air. Vessels at the site still reported some pancakes 1 to 13 cm in diameter on the surface, in streamers or rows several hundred meters long and about 78 m wide. Less than 1 percent of the surface in these windrows, however, was covered with oil. One windrow observed on August 30 was estimated to contain less than 350 gal in an area of 4,003 sq m. Eighty percent of the pancakes no longer were surrounded by sheen, and more than 55 percent of the area of the individual pancakes was submerged. Many pieces the size and shape of cornf.lakes were observed at the water surface and in the water column. On the previous day, a large number of subsurface flakes were also reported. By this time the surface oil had become spongy in texture, but it was not undergoing water-in-oil emulsif icat ion or "mousse" formation. Even after several days of weathering, less than 5 percent water was found in the surface oil after it had been heated to 80 C. The spilled oil did not reach shore during the spill response. Icebergs in the vicinity of the spill were examined by the U.S. Coast Guard to see if oil was adhering to them. It was noted that the oil stayed away from the icebergs probably because they were actively melting. Over the next 3 months, isolated reports of oil sightings and samples were received from biologists on expeditions CE. Born and T. Kristensen, personal, communication to H. Petersen, 1973) and from local hunters. One oil sample (October 1) was confirmed as coming from the POTOMAC and some of the seal skins (Chapter 9) may have been contaminated by POTOMAC oil; whether this contamination occurred locally where captured or during their passage through Melville Bay is unknown. Oil clumps or tarbails were also reported to have been taken in seal nets north of Thule at Moriussaq (Figure 9-D during the Autumn of 1977 and during April or May of 1978. A Green lander hunting polar 4-8 bear in Melville Bay (southeast of Savigsivik) reported oil above and below the sea ice. This hunter noted that the oil was coming up through cracks in the ice. Nith the exceptions noted.- none of these reports were confirmed as coming from the POTOMAC spill by chemical analysis because the Green landers did not collect samples. However, for lack of another source, they may very likely have come from this spill. While the evidence (next section) strongly indicates that most of the oil sank, it is apparent that at least some of the oil remained on the sea surface and may have been advected to distant places over the next few months. 4.4 Vertical Movement The initial specific gravity (s.g.) of the spilled oil was 0.976 which was a blend of 55 percent pitch with a s.g. of 1.054 and 45 percent cutter stock with a s.g. of 0.883, As the cutter stock lost its light fractions through evaporation, the s.g. of the remaining oil increased. A computation based on the percent n-alkanes remaining (54 percent) in a surface sample acquired on August 18 and the density of the original cutter stock (0.883) indicated that, after 13 days of weathering, the s.g. had increased to 1.802. On this date small flakes of oil were observed beneath the sea surface. The water column in the area can be considered a two layer fluid system as seen in Figure 4-2 with the top layer being 10 to 20 m thick. The s.g. of the two layers were 1.023 and 1.027 respectively. The s.g. of the bulk weathered oil (1.002) indicates that it should still ha'-je been floating and indeed it was sampled from the surface. It is hypothesized that the skin of the oil pancakes was even more depleted in light fractions and, through an exfoliation process not fully understood, separated from the pancakes in flakes and sank because the s.g. was greater than 1.023. Calculations similiar to those \ mentioned above indicate that when 73 percent of the cutter stock had evaporated, the residue blend would have a s.g. of 1.023 and would start to sink. It cannot be assumed that the s.g. stopped increasing (because weathering stopped) as soon as the oil sank below the surface; chemical analysis of the one subsurface (bongo) sample (Chapter 5.2.2) indicated that weathering in fact increased through dissolution of the sparingly soluble aromatic fractions. It is believed that a gre?t amount of the weathered oil 4-9 (67 percent of the 107,000 gallons spilled) eventually sanK to the bottom In the area where the spill was observed. Below 20 m, sinking whould have been accelerated by the greater compressibility of the oil compared with the water and by the nearly uniform density of the lower layer. Historical geostrophic current data indicate that the velocities are low (2 percent of the surface currents) in this deeper layer which actually consists of three water masses as noted earlier. Thus once the oil had sunk below 23 m, it would not tend to move horizontally during the several weeks that it would take to reacn the bottom about 1000 m below. Pit a mean fall velocity of l cm/mi n, it would take less than 50 days for the oil to reach the bottom in most areas of Melville Bay. 4.5 Summary Of Oil Movement The spilled oil was transported on the sea surface by variable winds and a slow circulation to the west or northwest. The variable winds probably caused little or no net movement while the slow circulation transported the oil a maximum distance of 48 nmi. The surface oil impacted approximately 500 sq mi of Melville Bay, and a small plume was left in the wake cf the POTOMAC from continued light leakage while she proceeded to Thule. After sufficient evaporation of the cutter stock (about 33 percent of the total volume), it is expected that a great amount of the residue sank in the form of small (1 cm) flakes (small flakes were observed sinking) to the bottom of Melville Bay. The water depth in the area of the spill was about 1,000 m and the detrital residue was scattered over a large area (~500 sq mi). It is expected that this petroleum detritus will remain Indefinitely in the bottom sediments. 5.0 WEATHERING OF SURFACE OIL Both the Danish and the NOAA teams analyzed a time-series of surface oil ("tar") samples (Table 5-1), with the objectives of determining the degradation rate and the ultimate fate of the spilled bunKer fuel. Bunker fuel is usually a very heavy material and in this case the fuel was comprised of a blend of 55 percent pitch Cs.g. = 1.054) and 45 percent "cutter stock" (s.g. - 0.883). The pitch originates from the residuum of the refinery distillation process and would not be usable as ship fuel in its unblended form. Cutter stock is added to the pitch in order to reduce the viscosity of the blend sufficiently to allow it to be pumped, after heating, from the ship's fuel tanks to the burning orifices. Cutter stock can originate anywhere in the refining process and would normally be a distillate material, exhibiting complete volatilization over a discrete temperature range. In the case of the POTOMAC'S spilled fuel, the cutter stock volatilzed completely over a temperature range of 154 C to 388 C, as shown in Figure 5-1. Because of its distillate nature, the cutter stock component of the Bunker fuel will evaporate, partially or entirely, on exposure to natural weathering processes at the sea surface. In measuring this process for the POTOMAC spill, both teams employed a combination of gas chromatography (GO and gas chromatography/mass spectrometry (GC/MS) in their efforts to address the following questions, among others: 1-How did the composition and density of the surface oil vary with time? 2-Did the low temperature and calm sea state produce slower evaporation than one would expect in temperate zones and rougher seas? 3-Ulhat was the "cutoff point" for the evaporative process in terms of relative vapor pressures or boiling points, e.g., did n-alkanes with boiling points of up to 250 C exhibit measurable evaporative losses? 300 C? 350 C? 4-Did dissolution of the low molecular weight aromatics occur at a measurable rate? Could it be distinguished from evaporation if losses 5-1 5-2 r-- r*- o> H en 3 00 3 < '-n •rl a CO o o H o Ph en 3 rH a S CO en •H O o> y 03 >4H l-l 3 I lo crj H en •w 3 0) o u 3 n 3 4-1 •H ^ 60 C o 1) "3 3 4-> •H 2 CO O w 3 AS rH en •H 01 O M 3 *3 CJ QJ • 3 H H 3 i— l • (X •H > P- H rH T3 en *« at CU 0) 3 C 3 w ■U 43 c 3 !-( 3 rJ u O 3 o O 01 V-i 01 ai 3 M en to 4J 4H w 4J S-' < M u 3 3 0) 0) a) QJ e QJ 6 e 1 r", a ex 3 U 3 3 S-l •H 3 •H ■H 3 g g T3 O "3 -3 ^ •H H H 6 s a 3 MH B 13 CJ 4= o u 3 0) H •u ^3 Di GO CO lo •H m lo T3 3 j| CXI & rH rH OJ O * ^ •* * rH X >% r-, en CJ CO CO o X X 42 OJ •H 3 0) X. w 4*: rH JX ^ 3 3 3 en 3 3 e 6 3 0) CJ O O o o -^ ^ 3 rH 3 3 u u CO cO 3 •H 3 3 E* Pn H H pM o PL, P-, 3 4= H 3 •H 3 ■u 3 Pi ^ >, 3 3 cq eq en QJ QJ 3 rH rH ^ rH rH 3 •H •H CN ^ > > * UH r-H 3 rH QJ £ e 33 s o CJ H 3 c CN •H •H S-i 3 QJ O 43 0J 3 s 4-J ■M rH rH § en •H a a •H 3 s o o ■*: 3 3 0) en 25 3 3 P. 3- e 6 S-i r-i rH o o 3 3 3 S-i S-i 3 CJ 4H HH 3 O o 3 rH <-{ T3 T3 3 3 3 >, >, •U u •iH X> JO CJ o 3 3 3 3 3 rH rH 3 3 3 <-\ rH ^ ^ ^ O O 3 3 3 CJ cj H H H O LO CN CO CN o en o O rH o rH o rH o <-{ o vO LO CN ^O >>0 vO vO o rH CN CO X> O ^ ^D ^D CN CN CN CN rH CN o CN O rH c CN >J3 o LO ro CN O Q rH r~- O CN CO o o LU cc - s 30 - 20 10 - i i i i i i i i i i i i i i i i 150 Figure 5-1, 200 250 300 TEMPERATURE CELSIUS 350 400 Distillation curve for the cutter stock of the Bunker-C fuel spilled from the USNS POTOMAC. were observed? 5-Did microbial degradation taKe place to any noticeable extent? Could microbial degradation have occurred at all in this relatively "pristine" environment? 6-Did the oil sinK? Uhere did it go? The USNS POTOMAC oil spill studies did provide answers to most of these questions. In some instances the answers were surprising, and in all instances the answers were welcome and provided field confirmation of one or another model of oil spill weathering processes. This chapter describes both the Danish and the NOAR approaches to these questions. 5-4 5.1 Analytical Procedures The NOAA team entered into a contract with Energy Resources Company, Inc. (EECO) of Cambridge, Mass., for the analyses of Its POTOMAC spill samples. The principal investigator on the ERCO contract was P. Boehm, and the coauthor of EPCO' s report to NOAA was D. Feist. Greenland Fisheries Investigations had arranged with the Danish Water Quality Institute and the University of Gothenburg, Sweden, for the analyses of their oil samples. ERCO used similar procedures in analyzing both the seawater extracts (Chapter 7) and the surface oil samples. The latter were analyzed by silica gel^alumlna column chromatography, glass capillary gas chromatography, mass spectrometry, spectrof luorometry, and determination of asphaltene content. The silica gel/alumina column chromatography procedure was the same for both the surface oil and the seawater extract samples, except that 1.0 g of copper metal powder was layered on top of the columns used for the surface oil samples to remove elemental sulfur from the oil. The silica gel (7.5 g) /alumina (2.5 g) column was then eluted with 18 ml of hexane and 25 ml of benzene to yield saturated (fl) and unsaturated (f2) fractions of each oil and water sample. The fractions were evaporated by rotary evaporation and an aliquot was subsequently weighed on a Cahn electrobalance. The column chromatography fractions were then characterized by glass capillary gas chromatography using a Hewlett Packard Model 5840A gas chromatograph equipped with a flame ionization detector and interfaced to a PDP-10 computer. Samples were analyzed on a 15 m SE-30 column, with temperature programed from 50 to 260 C at 3 C/min. Retention indexes (RI) of Individual components of each fraction were calculated by comparing observed retention times with the retention times of n-alkanes in a standard mixture. The Danish Water Quality Institute (Hansen, et al., 1976) employed both SCOT (Support Coated Open Tubular) and packed columns on a Hewlett Packard Model 5830A and 5840A gas chromatographs, respectively. The SCOT column was 56 m long, packed with 01/- 101, and was temperature programed from 85 C to 275 C at 4 C/min. None of the surface oil samples analyzed by Danish researchers were fractionated prior to injection into the gas chromatograph. 5-5 Selected spilled oil aromatic fractions (fP) were analyzed by ERCO by GC/MS using a Hewlett Packard 5700A chromatograph interfaced to a 5980A mass spectrometer with an electron ionization source and a 5934ft data system. The gas chromatograph was equipped with a 15 m SE-30 glass capillary column and was temperature programmed from 80 C to 250 C at 4 C/min. Mass chromatograms were reconstructed from mass fragments characteristic of particular compounds, and peaks were electronically in:egrated to yield absolute concentrations. Unfract ionated surface slick and selected f2 column chromatography fractions also were characterized by ERCO using spectrof luorometry with a Farrand MK- 1 spectrof luorometer equipped with corrected emission and excitation modules. Two types of spectra were obtained. Emission spectra from 280 to 480 nm, with excitation of the sample at 254 nm, were used for "matching" with the Bunker fuel carried by the POTOMAC (Jademac, 1977). Synchronous scans, from 250 to 500 nm emission wavelength and 225 to 475 nm excitation wavelength, were used to examine compositional changes in the fluorescent compounds of the oil (Wakeham, 1977). Sample concentrations used for the fluorescence spectra ranged from 3.1 mg/ml for synchronous scans to 20.0 mg/rnl for emission scans. The analytical procedures used by ERCO for the surface oil samples are shown in Figure 5-2. Asphaltenes were determined by ERCO using ASTM Standard Method D893-69, except that hexane was substituted for pentane. One gram subsamples of oil were repetitively dissolved in 10 ml of hexane and centrifuged at a relative centrifugal force of 600 to 700 g's. Approximately 10 washes were required to completely remove the hexane-soluble components of the oil. The residual "asphaltene" was dried at 110 C and its final weight was expressed as a percentage of the initial weight of the oil. The Gothenburg University (Ahnoff and Eklund, 1979) analyzed oil samples and surface water samples suspected of containing oil particles or films. Fluorescence spectra were recorded using an Aminco-Bowman SPF Spectrof luorometer . For the qualitative comparison of different samples, intensities at three wavelength combinations were measured, 230/340 nm, 270/360 .nm, and 310/400 nm, respectively. The Bunker-C fuel and one surface oil sample were further investigated by mass fragmentography, employing glass capillary gas chromatography combined with a computerized mass spectrometer 5-6 SURFACE OIL SLICK SAMPLE WEIGH ASUBSAMPLE; DETERMINE ASPHALTENES BY ASTM METHOD DISSOLVE IN DICHLOROMETHANE; WEIGH ALIQUOT ON CAHN BALANCE ASPHALTENES DISSOLVE IN CYCLOHEXANE; WEIGH ALIQUOT ON CAHN BALANCE TOTAL OIL EXTRACT (HEXANE) TOTAL OIL EXTRACT (CYCLOHEXANE) FRACTION ON A SILICA GEL/ ALUMINA COLUMN MAKEUP A SOLUTION OF KNOWN CONCEN- TRATION; FINGER- PRINT TOTAL EXTRACT BY EMISSION AND SYNCHRONOUS SPEC- TROFLUORESCENCE FRACTION 1 (f,) (SATURATES) FRACTION 2 (f 2 ) (AROMATICS) ROTARY EVAPORATE WITH N 2 WEIGH ALIQUOT ON CAHN BALANCE; FINGERPRINT BY SE-30 GLASS CAPILLARY GAS CHROMATOGRAPHY ROTARY EVAPORATE WITH N 2 WEIGH ALIQUOT ON CAHN BALANCE; FINGERPRINT BY SE-30 GLASS CAPILLARY GAS CHROMATOGRAPHY Figure 5-2. Analytical scheme for the surface oil samples used by ERCO for the NOAA samples. 5-7 (Parian Mat 112 - Spectrosystem 100 MS). Mass fragmentography was used to determine the concentrations of selected aromatic hydrocarbons. Due td the high sensitivity of the mass spectrometer when used in the nonscanning mode, the same technique could he used to analyze subsurface water samples. Technical details are contained in Chapter 7.2.1.2, as well as an example of a mass f ragmen togram (Figure 7-7). 5.2 Results 5.2.1 The Reference Sample A sample of the BunKer-C fuel carried by the USNS POTOMAC was supplied by EXXON from the originating refinery. Both groups used the EXXON sample as a standard. From this sample two fractions were prepared by ERCO; an fl containing the saturated hydrocarbons (alKanes) and an f2 containing naphthenoaromatic and aromatic hydrocarbons. These were analyzed by glass capillary gas chromatography. The chromatograms, shown in Figures 5-3 and 5-4, indicate that the fl fraction is dominated by a series of normal and branched alKanes from n-Cll to n-C30, with a maximum abundance at n-C16 and by a significant proportion of unresolved components with simiiiar boiling range and maximum detector response. The observed n-alKane composition is quite simiiiar to an analysis of BunKer-C oil published by ClarK and Brown (1977). The f2 fraction is dominated by an unresolved complex mixture (UCM) with only small relative proportions of resolved components. The major resolved components as determined by GC/MS are methyl- and dimethyl-naphthalenes, phenanthrene, and methyl and dimethyl phenanthranes. A sample of the BunKer fuel from the damaged tanK was collected by the Danish la i son officer at Thule shortly after the POTOMAC arrived in port. Details of exactly how and where the sample was obtained within the tanK are not clear. This "Thule sample" was analyzed by the Water Quality Institute and found NOT to compare with either the EXXON sample or the samples collected from the sea surface of Melville Bay. The difference in composition between the "Thule" and other oil samples may be due to weathering of the oil which remained in the tanK during the transit from the spill site, which is unliKely, or, more probably, the oil sampled from the tanK was a residue from a previous fuel carried in the tanK which had coated the tanK sides. 5-8 OJ) h o el Figure 5-3.. Gas chromatogram of fl (saturate) fraction of the spilled fuel. Figure 5-4. Gas chromatogram of f2 (aromatic) fraction of the spilled fuel. 5.2.2 Gas Chromatography of the Surface Oi l Samples The glass capillary gas chroma tograms of the spilled oil samples are quite similiar, except for some variability in the relative abundance of the more vol i tie and soluble components. The abundance of the n-alKanes in the fl fraction for each spilled oil sample was normalized to that of n-C20, allowing one to compare relative changes in the chemical composition of the spilled oil 5-< in O Q Z a HI > Ui BUNKER C FUEL 10 AUGUST 12 AUGUST o— o 14 AUGUST X -* 18 AUGUST 12 14 16 18 20 22 n-ALKANE CARBON NUMBER 24 Figure 5-5. Relative peak abundance of n-alkanes for spilled oil samples analyzed by ERCO. 2.0 - Cx Ill o 1.8 „ C 20 z < 1 6 o 1.4 •4" / < 1.2 / *: "*H* y < LU 1.0 n LU 0.8 > \- 0.6 < _i LU 0.4 — -x LL 0.2 .$>" ' ' • 11 12 13 + BUNKER C FUEL O 10. AUGUST A 14. AUGUST • 20. AUGUST ■X 1. OCTOBER J. J, JL X 17 18 20 n - ALKANE CARBON NUMBER J. 21 22 23 24 25 26 Figure 5-6. Relative peak abundance of n-alkanes for spilled oil samples analyzed by Water Quality Institute. 5-10 samples over time under the assumption that the concentration of n~C20 was constant. Since the primary' short term weathering mechanisms, evaporation and dissolution, preferentially deplete lower molecular weight compounds, normalization to a relatively high molecular weight compound is reasonable, ft graphic representation of the n-alKane relative peaK abundances for selected samples is given in Figures 5-5 (ERCO) and 5-6 (Water Quality Institute). The scatter of the points for n-alKanes less than n-ci8 is much greater than for those components above n-C18 and shows a trend of depletion of lower molecular weight compounds as function of time. The observed depletion of the lower molecular weight n-alkanes with time is the result of selective evaporation and dissolution of the lower weight alKanes with vapor pressures greater (boiling points less) than that for n-C17. The cold surface temperature (3 to 4 C), light winds (l to 3 m/s), and the relatively thicK (ca. 6 mm) form of the pancakes should tend to slow evaporation rates relative to warm water spills, thus suggesting losses of only the lightest compounds. This was not the case, however, as Figures 5-5 and 5-6 indicate. In fact, n-alkanes with boiling points, of up to 300 C (n-C17; boiling point = 303 C) showed measurable losses after 2 weeks of weathering (Figure 5-7) . The small amounts of resolved components in the gas chromatograms of the f2 fractions of the spilled oil samples made EHCO's quantification of individual components difficult by GC/T1S, and mass chromatograms were reconstructed to quantify individual compounds. Absolute concentrations were calculated by normalization to an internal standard. Five aromatic isomeric groups of compounds, methyl naphthalenes (M/E 142), dimethyl naphthalenes (M/E 192), phenanthrenes CfVE 178), methyl phenanthrenes (M/E 192), and dimethyl phenanthrenes (M/E 206) were quantified in the Bunker — C fuel and in the spilled oil samples collected on August 17 and 19. The results are shown in Figures 5-8, 5-9, and 5-10 and in Table 5-2. The August 18 surface sample (Figure 5-8) is depleted in methyl naphthalenes and dimethyl naphthalenes, as one would expect from the known relatively high vapor pressure and solubility of these compounds. Absolute concentrations of phenanthrenes and methyl phenanthrenes in the oil slick samples appeared to increase slightly with time. The evaporation and dissolution losses of the lower boiling compounds probably accounts for the apparent increase in the methyl phenanthrenes. 5-11 s- 0) 4-1 3 4-1 -H ■!-' CO d H >. ■U •H H cfl O* CD 4-> Cfl 13 (A -o OJ N H Cfl 9 CO 01 H I CO •H O CD CJ cd i-H S-l S to I IH CO n 00 CO 4-J cfl a o M U OJ cfl 60 CU 4J o m rH CU CO I CU u d 60 •H 5-12 Table 5-2. Concentrations of naphthalenes and phenanthrenes in selected oil samples analyzed by ERCO. M/E Compound group Concentrations (ug/mg oil) Reference August 18 August 19 fuel slick sample bongo sample (01-32) (01-23) (01-30) 142 Methyl naphthalenes 156 Dimethyl naphthalenes 178 Phenanthrenes 192 Methyl phenanthrenes 206 Dimethyl phenanthrenes 1.7 0.5 0.0 6.1 3.2 0.9 1.7 1.7 1.7 3.3 3.8 3.8 3.5 5.6 4.9 * Concentrations reflect the sum of all isomers of the compound group, \lf^l#^^ ■ ■ e " Va' ' ! 'ts ' ' ' '?'«' ' ' ' ^' ' ' 30 1 ! ' SB ' ' g ' ' '^ ' ' " n ^ rrrT 5 5 Figure 5-8. GC/MS total ionization current for the Bunker-C fuel, f2 (aromatic) fraction. Labels refer to alkyl-napthalenes (N) and phenanthrenes (P) . 5-1" TI i^lWi* *** " k" ■ sg ! ; I ,L I ■ | ffi ' ' IE ! ' SB ! ! g ' ' SB ! ' SB ' ' BE ' ! g " ' ■ Figure 5-9. GC/MS total ionization current for the August 18 surface sample, f? (aromatic) fraction. Figure 5-10. GC/MS total ionization current for the August 19 subsurface sample, f2 (aromatic) fraction. The bongo net sample of oil (Figure 5-10), collected on August 18 during a biological tow, shows simil iar but more extensive loss of the methyl and dimethyl phenanthrenes. The n-alkane ratios from the fl fraction gas 5-14 chroma togram for this sample are similiar to those of the August IS surface sample. Thus, the more extensile loss of the suPstituted naphthalenes in this sample is not explained Py evaporation alone. Boehm and Feist suggest that, once Proken into smaller particles and dispersed into the water column, the spilled oil weathered more rapidly than surface oil, primarily due to dissolution of the aromatic compounds. 2.5 r w X 2 a. uj z ■ WATER QUALITY INSTITUTE 0.5 0.0 t i 10 15 NO. OF DAYS AFTER SPILL Figure 5-11. Ratio of n-Ci7/pristane for the surface oil samples. (*) sample from October 1 — 56 days after the spill, Because microbial degradation of oil results in a depletion of n-alKanes relative to branched alkanes (ZoPelL. 1969), the ratios of n-C17/pristane and n-C13/phytane can be used to monitor biochemical degradation processes (Elumer and Sass, 1972). The n-Cir/pr istane ratio for the spilled oil samples shows statistically significant fluctuations but does not show a consistent trend with time. The Danish SCOT column data show the same consistancy over almost 5-15 2 months. Both sets of data are shown in Figure 5-11. If biochemical degradation of the spilled oil uias extensive, the n-C17/pr istane ratio would show a significant decrease with time. The absence of microbial degradation of the oil over several weeKs after the spill is consistent with slow initial microbial activity because of the low water temperature and also because of the small surface area/volume ratio of the oil pancakes. Both of these factors inhibit large-scale microbial degradation. It is difficult to say how long the October 1 sample was actually exposed to a potentially degrading environment, as we do not Know the precise circumstances of its collection. Suffice it to say, though, that no microbial degradation tooK place for the first 2 weeks after the sp i I L , and it is probable that none took place for the initial 4 to 8 weeks. Further discussion of microbial biodegradation is contained in Chapter 6. 250 400 450 500 CO IXI 250 300 350 400 WAVE LENGTH (NM) 450 500 Figure 5-12. Synchronus fluorescence scan of POTOMAC fuel by ERCO, 5-16 Table 5-3. Concentrations of two-, three-, four-, and five-ring aromatic com- pounds in selected spilled oil samples Sample Concentration (yg/ml) 307-nm (2-ring) Relative peak height * 365-nm 405-nm (3- and 4- ring) ( >5-ring) Bunker C fuel 10 August 12 August 15 August 18 August 20 August 2 8 3.9 3.3 3.2 3.5 3.4 102 98 102 107 105 114 100 100 100 100 100 100 83 83 82 78 77 76 * The analysis was done by measuring peak height maxima on a synchronous fluorescence scan from 250- to 500-nm emission with excitation at a wave- length 25 nm shorter. 60 50 40 30 20 10 - 0.0 • 307nm (2 RING) x 365nm (3,4 RING) a 405nm (£5 RING) 1.0 2.0 3.0 CONCENTRATION (pg/ml) 4.0 Figure 5-13, Relationship between peak height and Bunker-C fuel concentration for synchronous fluorescence scans. 5-1? 5.2.3 Spectrof luorometry, NOflft Samples Unfractionated spilled oil samples were dissolved in cyclohexane (2 to 28 ug/ml) and characterized by emission and synchronous fluorescence scans by EPCO as described earlier. As the spilled oil weathered, the emission scans showed a slight enhancement of the concentration of aromatics fluorescing at wavelengths shorter than the maximum emission and a diminuation of those fluorescing at longer wavelengths. This trend is consistent with weathering patterns observed by U.S. Coast Guard investigators in oils with an emisson maximum between 240 nm and 400 nm (Eastwood, 197?) . The synchronous scans (Figure 5-12) are considerably more detailed quantitatively thus providing more chemical information than single wavelength emission scans. One can theoretically discriminate between one-, two-, three-, four-, and five-ring aromatic compounds based on discrete fluorescence bands for each compound type (Lloyd, 1971). Synchronous fluorescence bands are 280 to 290 nm for benzenes, 310 to 330 nm for naphthalenes, 340 to 380 nm for three- and four-ring compounds, and >405 nm for five-ring compounds. The relation between peak height and the Bunker-C oil concentration is linear for the peaks corresponding to two-, three-, four-, and five-ring aromatic compounds over a concentration range of to 4.0 ug/ml of oil (Figure 5-13). The relative concentrations of three- and four-ring compounds and five-ring compounds appeared to decrease about 10 and 20 percent faster, respectively, than two-ring compounds over a period of 2 weeks (Table 5-3), a result which appears anomalous. This trend contradicts the observations by GC/MS (Table 5-2) and University of Gothenburg fluorescence analyses (Table 5-4) that two-ring aromatics (naphthalenes) decrease at a greater rate than three-, four-, and five-ring aromatics. In light of the observation that no biological degradation took place during the same time period (Figure 5-11), these synchronous scan fluorescence results cannot be explained away as being due to the formation and dissolution of polar-substituted three-, four-, and five-ring aromatics. Hence the quantitative aspects of synchronous scan fluorescence analyses must be considered suspect. 5-13 5.2.4 Spectrof luorometry and Mass Fragmentography of Danish Samples Oil samples and surface water samples, found to contain oil in amounts indicating the presence of particles or films, were characterized by spectrof luorometry and mass fragmentography. Fluorescence intensities at the three wavelength combinations, 230/340 mn, 270/360 nrn, and 310/400 nm, were compared with the reference oil. Complete agreement would yield relative intensities of 1.0, 1.0, and 1.0 respectively . The results are shown in Table 5-4. Table 5-4. Characterization by spectrof luorometry of surface samples analyzed by the University of Gothenburg. Sample Relative intensities* 230/340 270/360 310/400 Bunker C fuel 1.0 1.0 1.0 10 August 0.92 0.99 1.0 13 August 0.87 0.96 1.0 14 August 0.75 0.99 1.0 18 August 0.83 1.0 0.99 20 August** 0.74 1.0 0.96 * Intensities are given relative to the Bunker C fuel and are normal- ized so that the highest value of the three is set to unity. ** See also Figure 5-14. Pi typical set of spectra is shown in Figure 5-14. While intensities at the longer wavelengths stayed almost constant during the 15 days of weathering, a gradual decrease down to about 0.75 relative intensity is seen at the shortest wavelength combination of 230/240 nm. This is interpreted as a loss of the low molecular weight components and is in accord with the more specific mass fragmentographic analyses carried out on the reference fuel and one surface oil sample collected on August 10. Ten aromatic isomeric groups of compounds were quantitated. The concentrations found in the two analyzed samples are listed in Table 5-5. A comparison of these two samples is shown graphically in Figure 5-15. It is seen that, while the concentrations of dimethyl naphthalenes CM/E 156) and lower compounds have decreased, the concentrations of all other groups lie very near those in the BunKer-C fuel. 5-19 Figure 5-14. Excitation and emission spectra of a surface oil sample collected at Station 5473 on August 20 generated by University of Gothenburg. Intensities at 230/340 ran, 270/360 nm and 310/400 nm were used for qualitative comparison of the different samples. ft good fit between a spill sample and a reference oil can be taken as a strong indication of the origin of the spilled oil as has been shewn by Grahl-Nielsen (1976) in connection with another spill incident. 5-20 Table 5-5. Concentrations of naphthalenes, phenanthrenes , and dibenzothiophenes In two oil samples analyzed by the University of Gothenburg. Concentrations (ug/mg oil) M/E Compound group Reference August 10 fuel slick sample * 0.44 0.085 0.90 0.20 2.90 0.90 1.30 1.25 0.28 0.25 0.51 0.48 0.80 0.80 0.20 0.21 0.55 0.50 0.65 0.60 8.09 5.19 0.81 0.52 128 Naphthalene 142 Methyl naphthalenes 156 Dimethyl naphthalenes 170 Trimethyl naphthalenes 178 Phenanthrene 184 Methyl phenanthrenes 192 Dimethyl phenanthrenes 198 Dibenzothiophene 206 Methyl dibenzothiophenes 212 Dimethyl dibenzothiophenes Total ** * Concentrations are not corrected for a few percent of water present in this sample ** Naphthalene excluded 5.2.5 Asphaltenes flsphaltenes were measured gravimetrical ly as the hexane-insolubie fraction of the oil by ERCO. in contrast to shipboard estimates of asphaltenes made by the SOR Team, the Laboratory data show no consistent changes over time. fl slight decrease in the percentage of asphaltenes from 15 to 13.9 percent after 7 days followed by a gradual increase to 15.5 percent after 16 days were noted (Figure 5-15). No major changes in the asphaltene content are suggested by the data, flsphaltenes would not be expected to separate out by gravity settling since their density is approximately 1.0 and their surface activity tends to Keep them dispersed in the oil diagram, 1977). 5-21 100% 50- >< PHENANTHRENES DIBENZOTIOPHENES S NAPHTALENES 100 — r— 150 —i — 200 -iMW 250 Figure 5-15 Comparison of concentrations of selected aromatic hydrocarbons in the surface oil sample collected August 10 with the POTOMAC Bunker-C fuel. Horizontal axis is molecular weight. Vertical axis is the ratio of the sample to the Bunker-C fuel in percent 20- 1S = 3 u. o I 10 IT 1 5 10 NO. OF DAYS AFTER SPILL Figure 5-16. Change in asphaltene content of surface oil samples as a function of time. 5-22 5.3 Summary Fo* Weathering Of Surface OH Gas chromatography, mass spectrometry, and spectrof luorornetry have deen used to fingerprint and to investigate chemical changes in the spilled oil samples- Both gas chromatography and spectrof luorornetry confirmed that all of the samples consisted of oil carried by the POTOMAC. The primary weathering mechanisms for the first 2 weeks are evaporation and dissolution. N-alKanes with boiling points less than that of n-Clf, and substituted naphthalenes, are depleted by 50 to 190 percent after 15 days of weathering. Dissolution may play a significant role in weathering once the oil is dispersed in the water column; a subsurface sample collected from a bongo net contained significantly smaller amounts of methyl and dimethyl naphthalenes than did a surface slick sample collected at aproximately the same time that contained similar amounts of n-alKanes with boiling points similar to those of the alkyl naphthalenes. The synchronous scan fluorescence data on the NOAA samples failed to corroberate the GC/MS and single wavelength fluorescence data and hence the quantitative aspects of this analysis method must be considered suspect. Microbial degradation of the oil appears to proceed slowly, if at all, since the n-C17/pr istane ratio varied only slightly with time. There seems to have been no significant change in the asphaitene content of the oil samples 15 days after the spill. Rsphaltenes did not appear to have been preferentially settling from the spilled oil. The tar flakes observed to be settling in the water column may have been "sloughing off" of the highly weathered "skin" that is known to form at the air-oil interface of thick lenses of oil (pancakes). 6.0 MICROBIOLOGICAL STUDIES Sterile water samples were collected by the Danish team and sent to the Water Quality Institute where microbiological studies were undertaKen to determine the presence of oil degrading microorganisms at the spill site and their effect on the spilled BunKer-C fuel. Three studies were completed. The first was to determine which microorganisms were present in the area and whether their abundance was increased by the spilled oil. The second study was to determine the rate of degradation of oil by the natural culture found at the site while the third was to study the rate of degradation for each component of the natural culture. 6.1 Counting And Identification Of Microorganisms Surface water samples collected in the spill area on August 14 and 21 were used as the basis of this study. The sampling procedure is described in Chapter 3. While it was realized that the lag time for growth of oil degrading microorganisms would not be completed during the 1-weeK interval, an attempt was made to determine whether or not the natural populations were enhanced by the spilled oil. Five different counting techniques were used to measure the abundance of microorganisms and several tests were performed on each sample to isolate and identify the oil degrading microorganisms. 6.1.1 Analytical Procedures 6.1.1.1 Total Count on Agar Plates Total counts were measured by culturing the water samples on marine agar (Difco Bacto marine agar 2216). The composition of this substrate is given in the Appendix. Sterile physiological sodium chloride solution (9 g NaCl disolved in 1 I of water) was used for the preparations of the dilutions of the seawater samples. 1 ml of the diluted samples was transfered to sterile petri dishes; then liquid marine agar, cooled to 45 Z, was added to each dish and incubated for ? days at 23 C. 6-1 6-2 6.1.1.2 Veast and Fungi on Agar Plates Counts of yeast and fungi were performed using marine agar as the substrate. Bacteria were inhibited by the addition of Streptomycine and Tetracycline (both at egual concentrations of 50 ug/ml) and the pH was adjusted to 5.0. The substrate and incubation were identical to that used for the Total Count described above. 6.1.1.3 Oil Degrading Bacteria on Agar Plates Counts of oil degrading bacteria were measured on Colwell's MPN-medium solidified by the addition of 1.5 percent agar. Col well's medium was modified by reduction of the amount of MgS04, 7H20 to avoid precipitation of phosphate. The composition of the actual substrate used is contained in the Appendix. Shell i/itrea oil 27 was used as the single carbon source for these cultures (0.5 percent on a volumetric basis ). Shell l/itrea oil 27 is free from detergents and has a composition of 70 percent alKanes, 24 percent cycloalKanes, and 6 percent aromatics. A special technique as described by Baruah et al. (1967) was used to insure that the oil was homogeneously incorporated into the substrate. This technique called for dissolving 5 g of the oil in 15 ml of ethyl ether and then carefully mixing with 5 g of silica (Cab-O-Sil) in a morter. The ether was then evaporated in a rotary evaporator and the remaining powder (silica and oil) was easily distributed homogeneously on the substrate. Fungizone (at the rate of 10 ug/ml) was added to suppress yeast and fungi growth. Base agar (without the oil) was produced and used as control. Both oiled and unoiled plates were incubated at 20 C and counted after 14 and 23 days. 6.1.1.4 Oil Degrading Yeast and Fungi on Agar Plates Counts of oil degrading yeast and fungi were made using the same substrate as for bacteria described above except that streptomycine and tetracycline (50 ^ig/ml each) were added to inhibit the bacteria and the pH was adjusted to 5.0. As with the bacteria counts, Shell l/itrea oil 27 was added as the single source (0.5 percent by volume) and incubation was at 20 C. Counts were made after 14 and 23 days. 6-3 6.1.1.5 Oil Degrading Microorganisms by Most Probable Numbers Counts of Oil degrading microorganisms by Most Probable Numbers (MPN method) were made using two mediums. The two mediums are described by Bunch and Harland C 1976) and Mills, Brezil, and Colwell CIST'S) and are contained in the Appendix. The amount of MgS04, H20 in Colwell's medium was reduced as mentioned earlier. Portions of these mediums (5 ml) were distributed into screw cap tubes. After sterilization, 56 ul of oil (Shell P-'ltrea oil 27) or paraffins were added to each tube. The substrates were tested by the following petroleum degrading microorganisms: Nocardia, Pseudomonas , Arthrobacter , Streotomuces , Graph i lum , Pentamyces, and Candida . Both were found to be suitable for MPN investigations. Of the two, the Colwell medium was preferred. In all the MPN investigations, triplicate tubes were placed In a rotary shaker at 198 rpm during incubation at 15 C. The tubes were read weekly for six weeks. 6.1.2 Identification of Isolated Oil Degrading Bacteria Qualitative examinations of the isolated bacteria were carried out for identification using the following tests: Haemolysis, pigmentation, cytochrome oxydase, catalase, tween 28, tween S3, gelatine, O/F test, Simmon's citrate, motility (S.I.M.), indole (S.I.M.), H2S (S.I.M.), l percent tryptone, 1 percent tryptone with 4 percent NaCl, 1 percent tryptone with 7 percent NaCl, lAP. broth, arginine, lysine, urea, nitrate, arabinose, cellobiose, lactose, sucrose, glycerol, starch, xylose, chitin, veal infusion broth with 1 percent NaCl at 5 C, 10 C, 15 C, and 30 C, antibiotlca test (ptevidine), and gram staining, fill the reactions were read daily over a period of l to 7 days. Unless stated above, all the incubation temperatures were at 20 C. 6.1.3 Results The results from the microbiological examinations are presented in Table 6-1 for all five counting methods. In all cases and for all samples the observed counts are small. The concentrations of oil degrading microorganisms estimated by the MPN method are less than 1580 per liter, which corresponds to less than 1 percent of the total count. As indicated by the values from Station 5474, the total count as well as the numbers of oil degrading bacteria do not appear to be a function of depth. 6-4 Table 6-1. Results of enumerations from microbiological examinations of water samples . Number of Numt >er of Numb er of Number of Station Depth Date Total oil oil c >il bacteria of degrading degrading degr ading on base collec- count mi cro- bacteria yeast and agar tion organisms on agar fung i on pla ites (MPN -method) plates agar plates no. m day/mo. per ml per liter per liter per liter per liter 5462 0.0 14/8 50 < 300 < 5 < 5 < 5 5462 0.0 14/8 2 < 300 < 5 < 5 < 5 5466 * 17/8 560 700 85 ** ** 5468 0.1 18/8 5 < 300 30 < 5 55 5474 0.0 21/8 4 < 300 < 5 < 5 < 5 5474 0.2 21/8 2200 < 300 < 5 < 5 5 5474 1 21/8 370 700 10 < 5 10 5474 3 21/8 3700 < 300 5 < 5 5 5474 5 21/8 240 < 300 < 5 10 < 5 5474 10 21/8 10 1500 < 5 5 10 5474 30 21/8 2300 1500 205 15 340 5475 0.0 21/8 14 400 30 ** ** * Water surface with visible layer of oil ** No examination The number of oil degrading bacteria on oiled agar plates was the same order of magnitude as the number of bacteria on the unoiled plates. This result corresponds to other studies conducted on water samples collected from Greenland waters and indicates that the MPN method must be preferred to the agar plate method. Q-l An increase in either the total count or tne number was not found in the samples taken on August 21 relative to the values from the samples taken on August 14. This indicated that the spilled oil was probably not significantly biodegraded. This result confirms the findings from the gas chromatography performed on the surface oil samples (Chapter 5.2.2). The results of the identification tests for all the bacteria isolated are presented in Table 6-2. Tentative identification of the isolated strains are as follows: 1-S train 800 seems to be the genus Pseudomonas . 2-Strains 801i and 3012 seem to be Entero bactericae , but are untypical in the catalase test. 3-Stralns 802, 303, and 804 seem to be the genus corunebacter lum according to tentative investigations performed by Dr. R.R. Colwell. The reactions indicate that these strains all could be indentified as the species Corunebacter lum Pseudodlohther lae. 4-Strain 812 does not seem to fit into any Known taxonomical group. 5-Strain 812a could not be placed in a taxonomical group based on the tests conducted. It should be emphasized that none of the isolated strains could be identified as either fish or human pathogenic bacteria. 6.2 Degradation Rates By Natural Cultures Of Microorganisms Three experiments were performed to determine the rate of degradation of oil by the natural cultures of microorganisms found in the Melville Bay area. The experiments were started on August 27, 1977 using seawater sampled sterilely at Station 5475 collected on August 21 and a surface oil sample collected on August 10. The first experiment was Intended to determine the rate of oil degradation using the samples as collected. The second was a control experiment measuring the change of the oil in the absence of microbiological activity (control). The third experiment was to determine the rate of degradation of the samples with the addition of nutrients. 6.2.1 Procedures For all the experiments, 50 ml of the water sample was mixed with 0.250 g of oil in a conical flask. To prevent biological activity for the second experiment, 0.25 percent formaldehyde, by volume, was added. For the third experiment 1 g/l of K2HP04 and 2 g/l NH4N03 were added as nutrients. Multiple 6-6 Table 6-2. Biochemical examinations of strains isolated from substrates using oil as the single carbon source. All reactions read within 7 days, Tests 800 800. 801, Strain No. 802 803 804 812 812a Haemolysis + - - - - + - - Pigmentation - - - - - - - - Cytochrome ox ydase + + - - - - - (+) Catalase + - - - + + - + Tween 20 + - + + + + - + Tween 80 + - + + + + - + Gelatine + - - - - - - - 0/F test Ox F F gl- gl- 81- gl- gl" Simmon's citrate - - - - + + Motility (SIM) - - - - - - - - Indol (SIM) - - - - - - - - HoS (SIM) - - - - - - - - 1% tryptone + + + + + + - + 1% tryptone + 4% NACL (+) (+) + + + + + + 1% tryptone + 7% NACL (+) (+) (+) (+) (+) (+) + (+) V.P. - - - M M M M - Arginine - - - - - - - - Lysine - - - - - - - - Urea + + + + + + - - NO 3 M + + + + + M + Arabinose M - - M M M M M Cellobiose M - - M M M M M Lactose M (+) + M M M M M Sucrose M - - M M M M M Glycerol - _ - - - + - - Starch M - - M M M M M Xylose - - - - - - - - Chitin - - - - - - - - Veal infusion + 1% NACL at broth 5°C (+) „-_- "" (+) (+) + "■" (+) Veal infusion + 1% NACL at broth 10 °C ™~ (+) (+) (+) (+) (+) (+) — Veal infusion + 1% NACL at broth 15°C + (+) (+) (+) (+) (+) + (+) Veal infusion + 1% NACL at broth 30°C + + + + + + + + Antibiotica test (Pteredine) — « + + + + + (+) Gram rod g- rod g" rod g+ rod gr+ rod gr+ rod g" rod gr+ rod M = test not carried out gl = glucose negative Ox = oxidative degradation (+) = weak reaction F = fermentative 6-? flasks were prepared for each experiment: these were maintained at 15 C and carefully shaken (108 rpm) for the 6-week duration of the experiment, fit weekly intervals chemical and microbiological examinations were performed on one of the flasks for each experiment. For the chemical examinations, the total mixture of oil and water from one flask was extracted three times with CC14. The combined extract was diluted to 25 ml with CC14. A few microliters were then injected into a gas chromatographic column. The gas chromotography was performed using a SCOT column as described in Chapter 7.2.1.2. The microbiological examination were performed by the MPN method as described in Chapter 6.1.1. 6.2.2 Results The results from the gas chromatographic analyses on the SCOT column are presented in Table 6-3. The relative amounts of n-Cl?/pr Istane and n-ClB/phytane were determined and are shown in these tables. As previously mentioned, the n-alkanes such as C17 and C18 are degraded faster than pristane and phytane and thus their ratios can be used as indicators of b iodegradat ion. Some of the chroma tograms are included as Figures 6-1 to 6-3. Table 6-3. Relative amounts of n-C, 7 /pristane and n-C. R /phytane determined by GC analyses of natural cultures. Date Incubation Experiment I Experiment II Experiment III 1977 (weeks) C 17 /Pri C 18 /Phy C 17 /Pri C 18 /Phy C 17 /Pri C 18 /Phy Sept 14 2\ 1.45 1.78 1.27 1.77 0.62 0.85 Sept 21 3h 1.28 2.12 Sept 28 kh 1.33 1.85 1.39 1.80 0.41 0.49 Oct 5 5h 1.39 1.72 Oct 12 6h 1.45 1.74 Oct 26 8% 1.29 1.67 1.36 1.76 0.14 0.29 (1) = Water and oil (2) = Water, oil, and formaldehyde (control) (3) = Water, oil, and nutrients 6-8 2 weeks 8 weeks Figure 6-1. Gas chromatograms of oil degraded by natural cultures. From Table 6-3 it is seen that the n-C!7/pr istane and n-ClS/phytane ratios do not indicate degradation of the oil for the first two experiments where oil and water and then oil, water, and forma idehyde were cultured, in contrast- the third experiment with oil, water, and nutrients show a remarkable degradation of the oil after 2.5 weeks. The oil degradation continued during the incubation period. The results of the microbiological examination are presented in Table 6-4. These results also indicate a much faster propagation of the oil degrading microorganisms in the experiments with the added nutrients. The 6-9 2 weeks 8 weeks Figure 6-2. Gas chromatograms from the second oil degradation experiment (control) . first experiment indicates that the propagation of oil degrading microorganisms without added nutrients is a very slow process. The propagation rates in the Melville Bay area where the spill occurred would have been even slower as the ambient temperature there was about 4 C rather than the 15 C used in the experiments. 6-10 ■3 r Mil , A- fe M / ? 2 weeks #' ""W 4 weeks h Mui y ^ A „*HlW 8 weeks Figure 6-3. Gas chromatograms from the third oil degradation experiment which had nutrients added. Table 6-4. Growth of natural seawater cultures as a function of time. 6-11 Incubation time (weeks) Experiment I A B (1) Experiment II A B (2) Experiment III A B (3) 15,800 720,000 4.1xl0 6 19.xl0 6 22.xl0 6 300 900 110,000 460,000 6 120,000 1.2xl0 6 6.3xl0 6 <300 3,900 24,000 46,000 25.xl0 9.3x10 8.8x10 210,000 14,800 4.5xl0 ( 17.xl0 ( 21.xl0 f 24.xl0 f 26.xl0 ( 460,000 2.1xl0 6 2.4x10 24.xl0 1.1x10' 6 A = Total count per ml B = Oil degrading microorganisms per liter (MPN method) (1) = Water and oil (2) = Water, oil, and formaldehyde (control) (3) = Water, oil, and nutrients 6.3 Degradation Of Oil By Isolated Monocultures 6.3.1 Procedure All the oil degrading strains isolated from the natural cultures (Table 6-2) were tested for the ability to degrade the following mixtures of hydrocarbons: 1-Mixture of cyclopentane, cyclohexane, and cycloheptane (cycloalkanes) 2-Mixture of benzene, naphthalene, and phenanthrene (aromatics) 3-Paraffins (alkanes) 4-Shell l^itrea oil 27 (alkanes, cycloalkanes, and aromatics) These investigations were carried out in screwcap tubes. The Colwell medium with reduced MgS04, 7H20 content was used as the medium with a 1 percent hydrocarbon concentration (50 pi hydrocarbon with 5 ml of medium in each tube). The tubes were placed in a rotary shaker (100 rpm) during the incubations at 5, 10, and 15 C. The reactions were followed weekly for 6 weeks. 6-12 Table 6-5. Growth of isolated oil degrading strains in various petroleum hydrocarbon mixtures as functions of time and temperature. Times are in weeks . Bacterial strain 5°C 3 10°C 2 3 15°C 3 800 801 801^ 802 803 804 812 812a ALKANES X X X X X X X X* X X X* X X - X X 800 801 x 8012 802 803 804 812 812a CYCLOALKANES + •f 800 801! 801 2 802 803 804 812 812a AROMATICS 800 801-l 801 2 802 803 804 812 812a SHELL VITREA OIL 27 (ALKANES + CYCLOALKANES + AROMATICS) - - X X - X X X X X - - + X* X X X* X X X X X - - X - = no growth + = doubtful growth X = visual growth X* = visual growth after 1 week 6-13 6.3.2 Results Results of the studies on monocultures are presented in Table 6-5 for each hydrocarbon mixture. Strains 802, 803, and 804 appear to be very active oil degrading strains, especially in the biodegradation of paraffins and Shell l/itrea oil 27. After only 1 week at 15 C, these strains grow in mediums with oil as single carbon source. Only two strains, 382 and 812, degraded oil (paraffins) at 5 C after 6 weeks. In most cases the lag phase at low temperatures seems to be more than 6 weeks for the monocultures and depends on the species and composition of oil. However, in the natural environment the lag phase could be quite different from that determined from the monocultures in the laboratory, because the toxic constituents normally produced could be removed from the oil/water interface at the sea. 6.4 Summary Of Siodegradation Studies The number of microorganisms found in Melville Bay water was small, and oil degrading microorganisms constituted less than l percent of the total amount. Eight different microbial strains were isolated in the water samples using oil as the only carbon source. No increase in total number of the oil degrading microorganisms was found in samples collected 16 days after spill as compared with samples collected 3 days after the spill. fill the microbiological examinations show that in situ biodegradation would have been very slow. The results from the degradation studies with seawater and oil show that, unless nutrients are added, oil degradation is a very slow process. The gas chromatographic analyses did not indicate oil degradation during an 8 week period, and the microbiological analyses showed a slow propagation of oil degrading microorganisms during the same period. This seems to indicate a lag period of more than 3 weeks and, consequently, a slow degradation process. 6-14 The addition of nutrients to the oil and water increased the oil degradation strongly. Both the gas chromatography and the microbiological analyses show that, after l to 2 weeks, a high oil degrading activity was already present. In this case, the lag period was found to he less than 2 weeks. The results from monoculture experiments showed that only two strains, 302 and 312, would degrade paraffins at 5 C after 6 weeks incubation. 7.0 ACCOMMODATION OF OIL INTO THE WATER COLUMN Subsurface water samples were collected after the POTOMAC spill to determine the amount and composition of oil accommodated into the water column. These water samples were collected by both the Danish and NOAA teams, the former using a brown bottle in a stainless steel frame that was opened via a messenger puncturing a Teflon seal, and the latter using a General Oceanics Inc. "Sterile Bag" sampler. Analyses of the samples were carried out using UV- fluorescence spectrometry (UV), gas chromatography (GO, and gas chroma tography/mass spectrometry (GC/MS), by Ahnoff and Eklund (1978) for the Danish samples and by ERCO (Boehm and Feist, 1978) for the NOAA samples. 7.1 Sampling Procedures 7.1.1 The Danish Samples The Danish water samples were acquired with a noncommercial sampler. This device was actually constructed and supplied by M. Ahnoff and G. Eklund of the Department of Analytical Chemistry, University of Gothenburg, Sweden (Ahnoff et al., 1974). The sampler consisted of a 1-liter, wide-neck brown bottle held in a stainless steel frame. The bottle was sealed with a Teflon sheet placed under the screwcap. After lowering the sampler to the desired depth, a messenger was used to operate a mechanism which punched a hole in the Teflon sheet through a hole in the screwcap. The sampler was retrieved without reclosing the opening. Prior to the cruise, all sampling bottles were washed with tap water, distilled acetone, and purified n-hexane until the hexane showed no traces of contamination using uv-spectrof luorometry . During the cruise in Melville Bay, bottles were rinsed with purified n-hexane between samplings, but there was no opportunity to control the bottles. Throughout the sampling program, duplicate samples were taken on separate lowerings for separate analysis. The samples were extracted in the sampling bottle by adding 10 ml of n-hexane and a magnetic stirring bar, stirring for 45 minutes, and then transferring 2 to 4 ml of the organic phase to test tubes sealed with Teflon lined screwcaps. All the test tubes had been quality controlled by 7-1 7-2 spectrof lurometry prior to use. Sample extracts acquired from Station 5471 and subsequent stations were stored in glass vials supplied by NOAA. These samples were used for analyses by UV-spectrof luorometry only. 7.1.2 The NOAA Samples The NOAA team used a commercial sampler made by General Ocean ics. Inc. of Miami Florida. This sampler, called a "Sterile Bag Sampler", consists of a pair of hinged metal plates that, when triggered by a messenger, open a sterile polyethylene bag. The bags are used only once and, after being opened at depth, are resealed after filling. When filled and resealed, the bags contain between 0.8 and 2 liters of water, averaging about 1.5 liters. Each water sample was taken as a "double replicate", in that each depth was sampled twice (on separate lower ings) and two 500 ml aliquots were taken from each bag. Each aliquot was extracted in a separatory funnel with 10 ml of UV-spectroqual ity hexane. No filtration of the samples was attempted before extraction. The extracts were stored in prerinsed glass vials sealed with aluminum foil and screwcaps. Considerable difficulty was experienced because of the closures on the NOAA samples. For those samples which were not sealed tightly, the extract suffered from either spillage or evaporation causing the loss of some samples. Other vials which had been seated too tightly suffered from contamination when the aluminum foil tore and allowed the solvent to extract the waxes in the normal cap liner. Because of the difficulties experienced, the NOAA team recommends that no substitutes for Teflon cap liners be used for any extract containers in the future. Subsequent laboratory experiments (Boehm and Feist, 1978a) indicated that the polyethylene bags adsorb oil from the water sample and leach plastisizers into the sample. Indications are that the bags are inappropriate for water samples with oil concentrations less than 100 ppb or when the sample is held in the bag for much more than 15 minutes. 7-3 7.2 Analytical Methods The water samples collected in Melville Bay were analyzed by UV-spectrof luorometry (III/') , gas chromatography (GO, and gas chromatography mass spectrometry (GC/ffS) . 7.2.1 Danish Samples According to Ahnoff and EKlund (1979)- the reasons for choosing the UV technique for determination of petroleum hydrocarbons are that 1) the sensitivity is sufficient for measuring very low concentrations, down to "natural" background levels and 2) the technique is fast so that all samples can be analyzed within a few days after their arrival at the laboratory. GC/MS, using a glass capillary column, was employed for the following reasons: 1) the high selectivity of the technique permits an unambiguous determination of single petroleum hydrocarbons; 2) the instrumental sesitivity is sufficient to measure concentrations down to "natural" background levels; and 3) compounds to be measured can be selected. By measuring aromatic hydrocarbons, high selectivity can be obtained relative to biogenic hydrocarbons found naturally in water. The procedures had been designed to be as simple as possible to obtain low contamination of the samples. For example, the number of transfers of the sample between different vials was Kept at a minimum, to minimize exposure to surfaces and atmospheres that could cause contamination. Glassware and solvents were checked before use by UV-specrof luorometric measurements on blanks. 7.2.1.1 uV-Spectrof luorometry Oil fluoresces when exposed to UV-light. Fluorescence is the emission of light by previously excited electrons. Fluorescence in petroleum is dominated by aromatic molecules and is extremely complex because of the oil's complexity. A complete characterization of an oil by spectrof luorometry would include registration of fluorescence intensities at a large number of exci tat ion/ emission-wave length combinations and calculation of a three dimensional map of corrected fluorescence intensities. Such a procedure yields more information than is necessary when the object is to screen levels 7-4 of oil in a targe number of samples. It is time-consuming, requires special equipment, and cannot be employed on samples containing oil in trace amounts. On the other hand, a procedure consisting of measuring the fluorescence intensity at a single excitation/emission comhination reduces the information level significantly and makes quantitative evaluation difficult. A procedure which employs more than one excitation/emission comhination is rapid and gives appreciably more information tnan a single point measurement. Measurements at 230/340, 270/360, and 310/400 nanometers Cnm) were employed by the Swedish chemists. The qualitative information ohtalned from this procedure can be used to evaluate each measurement for quantitative determinations. The comhination of measurements snould indicate whether the fluorescence characteristics of the sample appear reasonable or if contamination has occurred. Intensities at the different wavelength combinations were compared with corresponding intensities of a standard solution made from the reference oil. A blank correction was made for the contamination found in the solvent used for extraction. 7.2.1.2 Gas Chroma tography/Mass Spectrometry In this technique, glass-capillary gas chromatography is used to separate volatile petroleum hydrocarhons which are then selectively detected with a mass spectrometer. The technique is simlliar to that described by Grahl-Nielson (1976) although he employed a quadrupole instrument. Masses to Pe monitored were those of naphthalenes, phenanthrenes, and dibenzothiophenes. These aromatic hydrocarbons have low enough molecular weights to be soluble in seawater but sufficiently high molecular weights so as not to readily evaporate. These compounds are known to be biologically active, as they are easily absorbed by living organisms where they exhibit toxic and other detrimental effects. Two milliliter aliquots of hexane extracts of water samples were concentrated to about 50 pi under a stream of purified nitrogen. Five nanograms of l-bromonaphthalene were added to each extract to serve as an internal standard. The background levels of the selected aromatic hydrocarhons in the n-hexane used for extraction were determined from a 10 ml sample concentrated to 50 ul CTable 7-1). 7-5 Table 7-1. — -Background concentrations in Danish water samples due to trace impurities in the solvent used for extraction. Compound nanogram per liter Naphthalene 1.3 Methyl naphthalenes 1.0 Dimethyl naphthalenes 0.30 Trimethyl naphthalenes 0.15 Phenanthrene 0.21 Methyl phenanthrenes 0.24 Dimethyl phenanthrenes 0.20 Dibenzotiophene 0.05 Methyl dibenzotiophenes 0.06 Dimethyl dibenzotiophenes <0.03 Total (naphthalene excluded) 2.3 A combination of a Carlo Erba Fractouap 2101 gas chromatograph ana a Parian MAT 112 mass spectrometer along with a Spectrosystem 100 MS was used. The gas chromatograph was equipped with a spl itless injector. The extracts were analyzed on a 40-m by 0.33-mm (inside-diameter) Ol-'- 101 glass capillary column. The following conditions were chosen: injector block at 250 C, the oven temperature programmed from 100 C to 240 C at 4.5 degrees/minute after an initial isothermal period of 5 min at 25 C, and the carrier gas (He) at a flow rate of 2 ml-'rnin at amPient temperature. Two microliters of the concentrated extracts were injected without stream splitting with a spl itless period of 60 s. The amount of the selected aromatic hydrocarPons was quant itated relative to a standard mixture containing 0.1 ppm naphthalene, phenanthrene, dibenzothiophene, and 1-bromonaphthalene. For the quantitation of the alkulated naphthalenes, phenanthrenes, and dibenzothiophenes, the total ion current per mole was assumed to Pe the same for these compounds as for their nonalkylated homologues. From mass spectral data on pure substances, it was known how large a fraction of the total ion current was made up by the ?-6 measured fragment ion. Thus the amount of each selected aromatic hydrocarbon was calculated. The sum total of selected aromatics was also calculated. An equivalent amount of reference oil was estimated by multiplying these values by a factor obtained from measurements on reference oil samples. This factor expresses the ratio between total weight and weight of the selected aromatics in the reference oil. It must be pointed out that, while the values for the selected aromatics are true values, the equivalent amount of reference oil is a theoretical value that may produce deviations between the total amount of petroleum hydrocarbons reported and the amount actually present in the water samples. The precision in mass spectrometr ic determination of the selected aromatic hydrocarbons was determined by injecting the sample from each station five times and calculating relative standard deviations of the measured amounts. The precision is highly dependent on the magnitude of the ion current from the measured ionic species. Since the ion current from the more branched aromatics is distributed between several peaks, the magnitude of ion current is lower and hence the precision is poorer. The relative standard deviation is 1 to 5 percent for naphthalene, methy inaphthalene, and phenanthrene; 8 to 12 percent for dimethy Inaphthalene, tr imethy Inaphthalene, methy Iphenanthrene, dimethy Iphenanthrene, and dibenzothiophene ; and 20 percent for dimethyldibenzothiophene. The detectabi I i ty of aromatics in seawater by mass spectrometry is limited by the background levels in the extraction solvent and by contamination from glassware. The detection limit is low, about 0.1 picogram injected or 5 picograms per liter of seawater for a single compound. The overall contamination of samples during sampling and processing could not be precisely determined. The lowest values of the sum of the selected aromatics was about 20 ng/l. They are probably close to the detection limit set by the procedure used. 7.2.2 NOAA Samples ERCO (Boehm and Fiest, 1978) perfomed the analyses on the NOAA water samples using procedures similiar to those for the surface oil samples. The details of these procedures are contained in Chapter 5.1 and will not be repeated here. In summary, the hexane extracts of seawater were dried over sodium sulfate, weighed, and selectively characterized by silica gel/alumina ?-? column chromatography, glass capillary gas chromatography, and UV spectrof luorescence. A blocK diagram of the procedure is presented as Figure 7-1 . 7.3 Results 7.3.1 The Danish Analyses 7.3.1.1 Quantitative UV-Spectrof luorometry The spilled oil contained high amounts of heavy aromatic hydrocarbons giving rise to fluorescence at long wavelengths. Its fluorescence characteristics differed significantly from those of lighter oils such as diesel and lubricating oils, and also from the pattern found in apparently unpolluted water. Therefore, fluorescence patterns having characteristics similiar to the spilled oil could be found in many samples, even though the concentrations were quite lout. The fluorescence patterns of the contaminated samples, including those samples with the highest concentrations, deviated significantly from that of the spilled oil. Dissolution and weathering processes produce a different composition of the petroleum hydrocarbons in the water column within a few days of the spill. Since fluorescence at long wavelengths (310/460 nm) was considered typical for the spilled oil, it was used as an indication that the petroleum hydrocarbons originated with the spilled oil. Samples which contained petroleum hydrocarbons in amounts above the baseline level, but did not show the characteristics typical of the reference oil, were considered not to contain oil spilled from the POTOMAC. In at least one sample, the fluorescence spectrum was similiar to the spectrum of the cooling water from the ADOLF JENSEN. Of the 76 subsurface samples that were taken, three were lost and five were considered as contaminated. These eight samples that were not analyzed are listed in Table 7-2. For the corresponding sampling points, results from quantitative spectrof luorometr ic (UV) measurements are based on single samples. For the other 30 sampling points, duplicate samples were analyzed. 7-8 WEIGH A 20 ml ALIQUOT ON CAHN BALANCE TOTAL EXTRACT ABLES Gig) HEXANE EXTRACT OF 500 ml SEAWATER r DRY OVER SODIUM SULFATE; EVAPORATE TO 0.2 ml UNDER NITROGEN I CONCENTRATED HEXANE EXTRACT FRACTION ON A SILICA GEL/ ALUMINA COLUMN FINGERPRINT BY SE-30 GLASS CAPILLARY GAS CHROMATOGRAPHY FRACTION 1 (f-^ (SATURATES) FRACTION 2 (f 2 ) (AROMATICS) ROTARY EVAPORATE WITH N 2 5 WEIGH ALIQUOT ON CAHN BALANCE; FINGERPRINT BY SE-30 GLASS CAPILLARY GAS CHROMATOGRAPHY ROTARY EVAPOPATE WITH N 2 ; WEIGH ALIQUOT ON CAHN BALANCE; FINGERPRINT BY SE-30 GLASS CAPILLAR* GAS CHROMATOGRAPHY Figure 7-1 Analytical scheme for hexane extracts of seawater used by ERCO for NOAA samples. 7-9 Station Depth 5460 10 m 5461 5 m 5466 1 m 5468 5 m 5470 1 m 5470 20 m 5471 20 m 5477 5 m Table 7-2. List of samples for which results are not reported. Code Cause 1 a 8 a 16 a 29 a 39 b 45 c 52 a 71 b__ a ; Strong indication of contamination of sample. Comparatively high levels of oil with fluorescence characteristics strongly deviating from those of the spilled oil. b ; Screwcap on sample vial was not sufficiently tightened. c ; Not delivered As a rough test for similiarity to the spilled oil, a value of at least 0.5 for the following relative fluorescence intensity Cr) was required: r - c3 10/400 C27Q/360 . Here, C310/410 and C270/368 are fluorescence intensities at the indicated wavelengths expressed in reference oil equivalents of ug/l . The results of these tests are shown in TaPle 7-3. Quantification was made with the reference oil from EXXON' s Aruba refinery as the standard. All values are mean values from the duplicate samples, except for the sampling points listed in Table 7-2. The results, corrected for interference from the solvent, are shown in Table 7-4 expressed as equivalent concentraions of the reference oil in ug/l . Some profiles of concentration of petroleum hydrocarbons as a function of depth are contained as Figure 7-2. Fluorescence spectra of different types of water samples are shown in Figure 7-3 (contaminated by Bunker-C fuel), Figure 7-4 (contaminated by cooling water from the ADOLF JENSEN), and Figure 7-5 (apparently uncontaminated water). 7-10 Table 7-3. Petroleum hydrocarbons in subsurface water samples, quantitated as the amount of reference oil that gives rise to the same fluorescence intensity at chosen wavelength combination Station Depth microgram per liter Spectral (m) measured at ( nm) similarity to reference oil 230/340 270/360 310/400 5460 1 0.91 0.53 0.27 + 10 0.22 0.14 0.049 5461 1 4.9 3.5 2.5 + 5 0.74 0.57 0.41 + 10 0.53 0.40 0.29 + 20 0.56 0.32 0.13 5462 1 1.0 0.50 0.30 + 5466 1 1.5 0.90 0.48 + 5 0.54 0.36 0.35 + 10 0.48 0.32 0.19 + 20 0.32 0.21 0.082 5467 1 0.41 0.22 0.081 5468 1 0.48 0.23 0.11 + 5 0,59 0.40 0.20 + 5469 1 0.28 0.16 0.038 5 0.36 0.26 0.13 + 10 0.28 0.20 0.11 20 0.28 0.18 0.065 + 5470 1 0.31 0.14 0.043 5 0.20 0.12 0.032 10 0.24 0.11 0.032 20 0.29 0.12 0.028 5471 1 0.24 0.15 0.041 5 0.74 0.34 0.11 10 1.0 0.49 0.25 + 20 0.23 0.12 0.026 5473 1 0.47 0.27 0.13 + 5 0.91 0.64 0.43 + 10 0.56 0.24 0.10 20 0.53 0.25 0.11 30 0.37 0.25 0.12 5474 1 0.87 0.32 0.11 5476 1 1.7 0.77 0.19 5477 1 1.3 0.73 0.28 5 0.47 0.27 0.072 10 0.63 0.28 0.083 20 0.63 0.19 0.079 7-11 0.27 0.052 0.031 0.30 0.056 0.033 0.75 0.19 0.047 Table 7-4. Amounts of interfering substances in solvents used for extrac- tion, quantitated as micrograms of oil per liter of water sam- ple. Bottle Used at stations micrograms per liter 230/340 270/360 310/400 (SmJ I 5460 - 5468 II 5469 - 5473 III a 5474 - 5477 a American hexane The relative difference between duplicate samples, calculated as (cl - c2) / 0.5(cl + c2), ranged between 2 and 133 percent. The median value was 40 percent and the arithmetic mean difference was 55 percent. The concentrations found at different sampling points ranged from 0.03-8.84 jjg/l at Station 5478 to 2.5 ug/l at Station 5461 at 1 m depth. Thus significant differences could be seen between different stations and between different depths, although the precision was relatively poor. The deviation between duplicates is due partly to the fact that recently polluted water is not homogenous; thus, the duplicate samples do not necessarily contain equal amounts of oil. Judging from the fluorescence characteristics of the subsurface water samples, none of them contained oil which ojas identical in composition to the POTOMAC oil. Therefore, there will be a systematic error when the spilled oil is used as a reference for quantitative evaluation. This is not unique for the fluorescence technique but is a general problem when oil is to be analyzed at the 1 ppb level. Each technique which can be employed at this concentration level suffers from the drawback that it does not have equal sensitivity for all components of the oil. The UV-f luorescence technique is sensitive to aromatic hydrocarbons in oil and has the property of generally being more sensitive to larger molecules. Therefore, if the composition of the oil in the sample is shifted towards the lighter part of the reference oil, the total concentration can be underestimated. This can be partly overcome by choosing excitation and emission wavelengths that are typical of the lower aroma tics, e.g., the naphthalenes and phenanthrenes . 7-12 STATION 5470 20/8 v microgram/ 1 iter I , ' f _ / V">>1 61 ' E 10 STATION 5471 20/8 microgram/ 1 iter 0.5 ■ ■ 1.0 1 \ 5- \ E 10 T3 / / / / / 1 1 20 / I 51 ft 20 0.5 j i i i i_ STATION 5461 13/8 microgram/liter 1.0 *'. Vsr w 59°W 5470 O 5471 5466 C 5461 O SPILL 1 76° N 75° N STATION 5466 17/8 microgram/liter 1 " 5- 10- L_J ■ ■ 0.5 ■ ■ 1.0 E / / / ft / / / 20 1 1 2.0 3.0 J— " I 10- i — f—\ Figure 7-2. Concentrations of petroleum hydrocarbons at four stations as a function of depth. Concentrations by UV-Spectrofluorometry, Values from duplicate samples are indicated. Spectral wave- lengths: excitation 310 nm, emission 400 nm. 7-13 £t*ISsfrn Vto e*Ubk,ti»n VOjO t XU> t XfC^o &®®*m» Figure 7-3. Fluorescence spectra of an extract of subsurface water collected on August 13 at Station 5461, 1 m depth. Left: excitation spectrum from fixed emission wavelength. Right: emission spectra from different excitation wavelengths (see also Table 7-10.). A simple recovery test was made to check the efficiency of the Danish extraction procedure. Tap water was added to a sampling bottle, adjusted to a salinity of 38 g/l, and IB ug of the reference oil dissolved in 18 ul of dichloromethane was added. The water was stirred for 38 min, extracted with hexane, and analyzed using the normal procedure for the water samples. The recovery was close to 188 percent as is seen in Table 7-5. However, fihnoff and Eklund's experience (1979) from real seawater samples that have been extracted in two consecutive steps indicated that' extraction is not 188 percent but somewhere between 88 and 188 percent and that the extraction efficiency is affected by the particulate load of the water. 7.3.1.2 UV Comparison of the Danish and NOfifi Sampling Methods fit Station 5471, water samples were taken using both the Danish and NOfifi procedures. Two of the NOfifi samples which had been extracted by the NOfifi team were analyzed by Ahnoff and Eklund. fi comparison of these two samples with the Danish samples taken at the same depths at this station is presented in Table 7-6. Spectra of these samples are shown in Figure 7-6. The fact that the emission spectra look the same, independent of the excitation wavelength 7-14 Figure 7-4. Fluorescence spectra obtained from a surface water sample, station 5474, that contained waste cooling water from the ADOLF JENSEN. (for explanation see Figure 7-3.) is atypical of oil and suggests the presence of only one or a few similiar fluorescing compounds, possibly naphthalenes. 7.3.1.3 Mass Spectrometr ic CMS) Analysis Table ?-? lists the surface oil and water column samples that were analyzed using the MS method in Sweden. The concentrations of different aromatic hydrocarbons found in these water samples as determined by MS analysis are given in Table 7-8. According to UV analyses, three of these contained less than 1 ug/l total petroleum hydrocarbon concentration as noted in this latter table. These samples contained aromatic hydrocarbons in amounts barely above the practical detection limit of the MS method. This limit was not set by instrument sensitivity, but rather by the amount of contamination introduced into the samples during the analytical procedure. Contamination interfered more strongly with the MS than with the UV analyses. r-15 smnbM s* lurfcxt. ru9 fc*uWH%K %%A t *# t Xp W Figure 7-5. Fluorescence spectra obtained from a surface water sample, Station 5469, collected well to the north of the spill area. The vertical scale is more expanded than in Figures 7-3 and 7-4, Table 7-5. Recovery of Melville reference oil from synthetic seawater using hexane as extractant. Excitation/ emission wavelength (nm) 230/340 270//360 310/400 Recovery (%) 96.1 98.7 96.9 7-16 Table 7-6. Comparison between samples taken by NOAA and Gothenburg Uni- versity sampling methods. Station Depth Method microgram per liter 230/340 270/360 310/400 (nm) 5471/1 0-0. 5m NOAA 6.2 2.1 0.20 0m GU 0.59 0.31 0.20 5471/2 0-0. 5m NOAA 10 2.8 0.40 0m GU 0.24 0.15 0.042 The sample at 1 m depth from Station 546 1, which contained a few ug/l of petroleum hydrocarbons, showed 50 ng/l of selected aromatic hydrocarbons (Table 7-8). Mass fragmentograms are shown in Figure 7-7. In Figure 7-8, the concentrations of different aromatic hydrocarbons are compared with the concentrations found in the surface oil sample collected on August 10. It can be seen in this figure that, for each type of aromatic hydrocarbon, the relative concentrations decrease with increasing molecular weight. This is in accord with the lower solubility of the higher weight compounds. For the naphthalenes, evaporation of the lightest naphthalenes from the oil can also contribute to the differences seen in Figure 7-8. 7.3.1.4 Comparison of MS and UV Results The selected aromatic hydrocarbons make up approximately 1 percent of the total weight of the original FOTOMAC fuel which was spilled (Table 5-5). Assuming that the same relation between the selected aroma tics and the total amounts of petroleum hydrocarbons exists in the water samples (obviously this is not true), values of total petroleum hydrocarbon concentrations can be calculated. Table 7-9 presents such concentrations for the sample at 1 m from Station 5461. Obviously, these values must be maxima since the selected aromatic hydrocarbons belong to the most water soluble components of the oil and thus would be expected to be present in high relative concentrations. The UV analysis shows lower relative concentrations. Compared with the MS technique, the UV method measures a broader spectrum of aromatic hydrocarbons. Higher wavelengths are more selective for the heavy aromatic components, and they 17 H ■'- Oh %70 t 4 XiO t X'H t VO 0&#r pn'^ i ft yr it it ad* Figure 7-6, *}>* StDQi%$t; Fluorescence spectra obtained from a water sample collected in a polyethylene bag by NOAA at Station 5471 just below the surface. also yield lower concentrations. lvalues from measurements at 310/498 nm can be regarded as minimum concentrations. Consequently, it can toe concluded that there is good agreement between the MS and the UV determinations on at least this sample. 7.3.2 The NOAh analyses The NOPA hexane extracts were analyzed by ERCO, and a complete report of their findings is contained in Eoehn and Feist (1373) from which the following material was extracted. 7-13 Table 7-7. Samples analyzed by mass f ragmentography . 1 "Thule oil" acquired from tank of the POTOMAC. 2 Reference oil Bunker-C retain acquired from EXXON refinery at Aruba where POTOMAC last loaded with fuel 3 Surface oil sample collected on August 10, 1977. 4 Station 5460 1 m code=4 5 Station 5461 1 m code=5 6 Station 5461 5 m code=7 7 Station 5469 5 m code=33 Table 7-8. Concentrations of naphthalenes, phenanthrenes, and dibenzothiophenes in subsurface water samples. Compound 5460 (lm) 5461 (lm) 5461 (5m) 6569 (5m) Naphthalene Methyl naphthalenes Dimethyl naphthalenes Trimethyl naphthalenes Phenanthrene Methyl phenanthrenes Dimethyl phenanthrenes Dibenzothiophene Methyl dibenzothiophenes Dimethyl dibenzothiophenes nanogram per liter 4.3 5.7 5.5 6.6 3.2 7.2 1.8 3.4 0.9 12 1.5 3.3 <2.0 8.4 1.8 4.2 5.8 5.5 1.1 1.5 2.3 5.3 0.9 0.6 0.7 4.4 <0.1 <0.1 0.6 1.9 0.2 0.3 0.2 2.7 0.4 0.7 <0.1 2.5 0.5 0.9 Total (naphthalene excluded) 14 50 10 15 Spectrofluorimetric analysis 230nm/340nm 930 4900 740 500 270nm/360nm 560 3200 570 350 310nm/400nm 240 1900 410 180 7-19 K — m/e 206 J — CD E -i « A r E : — j . 4 i § est i • - "• E a r*. — c ; O - © E t I-* x— ^ J r ( i ' LO ^_ ' —to ' l_ i 1 j- ■t-o Q B •xl 03 Vj o 4J c 4-1 0) cd a 60 "d CO cu M 4J IH O CU CO H Cfl a 55 X 08-67 5469-<(,b 108 08-77 5470-<|>A 100 X 08-76 5470-a 12 08-78 10-b 153 X 08-74 ll-4»a 51 08-73 ll-b 98 X 08-90 11-10A 108 X 08-91 11-10B 185 X 08-82 Fisher 765498 118 72 60.9 X X 08-80 MCB Blank 254 149 108.6 X X L = Lost X = GC ed samples. Three of the four procedural blanks were simlliarly contaminated, and the fourth contained small amounts of hydrocarbons (<5 yig) confirming the contaminat ion . The 14 remaining samples were not contaminated by the vial caps but contained only minor amounts (<10 jjg/580ml) of a few resolved components in the unfract ionated sample (Figure 7-14). The small number of resolved components- usually one or two, argues against gross contamination by the POTOMAC oil and suggests a biogenic origin of these components. In no case was major contamination of the water column with POTOMAC oil observed. Selective dissolution should result in a series of substituted naphthalenes, and gross incorporation of oil into the water column should result in an fl 7-24 Sample I.D. Table 7-11 Groupings of seawater samples NOAA Date Station Depth Group (August) No. (m) 17 5466 1 T 18 5467 T 18 5467 T 18 5468 1 C 18 Blank - C 18 5468 LL 18 5468 1 LL 18 5468 1 LL 18 5468 1 LL 19 5469 LL 19 5469 LL 19 5469 LL 19 5469 C 19 5470 C 19 5470 1 C 20 5471 0.2 C 20 5471 0.2 C 20 5471 0.2 C 20 5473 1 C 20 5473 1 C 20 5473 1 LL 20 5473 5 LL 20 5473 10 LL 20 5473 10 C 20 5473 20 LL 20 5473 20 LL 20 5473 30 LL 20 5473 30 LL 21 WW-10 10 C 21 WW-11 C 21 WW-11 10 C 21 WW-11 10 C 1 2 3 4a 5 7b 10 11 12 13 14 15 0a 0a H-la H-lb H-2a H-2b 1-la 1-lb l-2a 5- la 10-lb 10-2 b 20-2b 20-4b 30-lb 30-2b 10-0b ll-0b ll-10a 11-1 0b T = Trimodal unresolved envelope. C = Cap liner contamination. LL = Low level concentrations (< 10ug/5QQ ml). pattern similar to that for the whole oil (Figure 5-3). Neither of these patterns was observed in any of the NOAA water sample extracts analyzed by GC techn iques . 7-25 Figure 7-9. Gas chromatogram of the fl (saturate) fraction from extract 2 showing the trimodal group. Analysis by ERCO. Figure 7-10. Gas chromatogram of the f2 (aromatic) fraction from extract 2 (trimodal group)- Analysis by ERCO. 7.3.2.2 UV-Fiuorescence Three f2 fractions of water extracts, incuding members of the trimodal alkane group, the cap blank group, and a shipboard blank, were analyzed by synchronous-scan spectrof luorometry and compared with an f2 fraction of the reference oil. No conclusions could be drawn from the results because of the similarity of peak shape and concentration of the two water sample extracts and the blank. 7-26 j — LJLJLJ . j Figure 7-11. Gas chromatogram of the fl (saturate) fraction from extract 10-2b showing cap liner contamination. Analysis by ERGO. L--' — "J*. Figure 7-12. Gas chromatogram of the f2 (aromatic) fraction from extract 10- 2b showing cap liner contamination. Analysis by ERCO. 7-2.7 Figure 7-13. Gas chromatogram of a sample vial cap liner (unfractionated) Figure 7-14. Gas chromatogram of the unfractionated extract 30-lb showing the low level group. Analysis by ERGO. 7.4 Summary Of Accommodation Into The Water Column UV-spectrof luorometry, gas chromatography, and mass spectrometry were used to analyze the Danish and NOAA hexane extracts of water samples to investigate chemical changes of the spilled oil incorporated into the water column. Use of these techniques was hindered by contamination of some of the samples, both during collection and during storage. Some of the extracts showed gas chromatography patterns atypical of the oil spilled by the POTOMAC, 7-28 Most of the remaining samples contained small amounts of resolved components (<20 ug/'ml) . uV-spectrof luorometry was used to screen the horizontal and vertical distribution of oil in Melville Bay. Petroleum hydrocarbons were found in the Danish samples at concentrations from 0.03 ug/l up to above 2 )jg/l, using spectral wavelengths typical for quantification of the spilled oil. Using different wavelengths for quantitation, the maximum concentration found was between 2.5 and 4.9 ug/l. The spilled oil could be traced by UV techiques down to about 0.1 ug/l. Depth profiles, taken 8 to 12 days after the spill occurred, showed maximum concentrations near the surface (1 m depth) and a rapid decrease down to 10 to 20 m depth. Profiles taken 13 to 15 days after the spill showed maximum concentrations at 5 to 10 m depth corresponding to the bottom of the surface layer (Chapter 4. 1) . The petroleum hydrocarbons in the water column contained higher relative amounts of the low-molecular-weight aromatic compounds than the reference oil and surface oil samples. Mass spectrometr ic analysis was used on a small number of the Danish water sample extracts as a comparison with their uV-spectrof luorometr ic analysis. Good agreement was found between the two techniques. While large-scale dispersion of the oil into the water column might have occurred during the 8 days before the first water samples were collected, no gross accommodation of the spilled oil into the water column was found in any of the water sample extracts analyzed. 8.0 BIOLOGICPL STUDIES ON PLANKTON AND FISH Samples of zooplankton and fish were collected in the vicinity of the spill as well as in reference areas to examine the impact of the oil (Tables S-I and 3-2) . The Danish samples were forwarded to the Water Quality Institute for hydrocarbon analysis, to Marin ID for analysis of the composition of species and identif ication, and to Greenland Fisheries Investigations (GF) for examination of contaminated plankton. The 13 U.S. biological samples were forwarded to the Fiahkton Ecology Laboratory, National Marine Fisheries Service, Narragansett, P.. I., for analysis of species composition, abundance, and contamination of planKton. 3.1 Sampling Procedure 3.1.1 Danish Sampling Plankton samples were collected with a Stramin net (2 m diameter, mesh 500 threads/m hauled from 200 m to the surface at 1.5 kn over a period of about 33 min) and Hensen net (72 cm diameter. No 3 silk hauled vertically at 3.3 m/s from 50 m to the surface.) (Table 3-1.) One plankton sample was collected with the Stramin net at the spill site (Station 5460) and two samples were collected in areas with oil pancakes on the surface," at Station 5461, the Stramin net broke the surface within the oiled area and at Station 5464, the Stramin net broke the surface outside of the oiled area after hauling inside the oiled area. Three samples were collected on a line north of the center of the oiled area (Stations 5469, 5470, and 5471). One sample was collected in an area with subsurface oil flakes (Station 5473) and one in a reference area (Station 5477). Samples were also collected with the Hensen net at Stations 5460, 5463, 5471, and 5477 (Figure 3-2) . The volume of the plankton samples from the Stramin net were measured immediately (Table 3-1) and the plankton were examined for oil contamination. 3-1 8-2 CD I H > a o a cj rH ex a •H -o H cj C •H CO c o •H .u 03 4J CO H CO cj •H M o o •H X) CO •H C cO O td CO I oo cct H CO •H CO S>s 4-1 rH C CO 0) C to crj cu u Ph CU rH rH •H XI O -H CO rH >H CO CU o e H 3 c cu E 3 rH C > a cu P cu H as cu r^ CO CT\ P rH c o •H 4-1 •H CO O Ph a o •H 4J CO •u CO rJ CO cu u co co cu cu r^ >N o oo MO m o o o CO CU C3 CJ C CO ex o o c CO cu M CO O C Oi MO cm oo o o 0> uo m m CN CN CN CN CN CN I I I o o o UO rH co CN O CN CO CN rH rH O <-\ CO o-i rH r-\ 00 00 X) CN o uO O CN CO co cu r*i CO CN UO o o o o m o O rH O CO CN UO CN CM m co 00 r-^ O CN o CN 00 00 UO CM CM I O o CN O CM OO M0 O uo CO rH O0 -X CO <-\ uo * M0 CO 00 o o o uo in in i I I o o o o CN 00 00 •X CO o CO o u0 00 rH t^- cn cr o c CO 4-1 rH o CO CU U c •H CO 4-1 u CU CJ 3 eg a !h CU rd 8-3 (11 /~\ a CO 3 rH 1 O o - r-i CN rH O 00 r^- d 4-1 rH CN CN ro CN - vD a-* = "3 a ai \-x rl (U 4-1 <-i •rl Pn O > o T-i r-» CN a\ CN o \o vj- LO o 00 rH fl ^■^ o vO o 1^ r~- 00 o-. o o 00 CN vD v£> t^ 4-1 H o o o O o O <-\ T-\ o rH T~i r-i rH CO 4-1 CO CU r^. o o o O o o o o r-i T~i O a\ — ' — ^ ^\ -^ ~-^ ■- — * — , -■^ *^^. *^ cO n r-^ 00 00 00 00 00 00 00 00 00 00 00 00 CO O •H 60 _ - O p: LO LO rH » • • m. O o LO CN o -* >* C^ r- CN o CO LO LO •H s CN lH rH T-i r-i o r-i T~i LO LO LO T-i T-i 42 o o O O o o o o z o z •H r-i T-i rH T-i T-i r-\ r-\ r-i T-i LO r-~ T-i T-i • 4J vO vO v£> vO vO VO vO VD vO vO \o vO vO CO •H • CO 53 P O LO m LO LO LO LO LO MH O vO !-». 00 CTN O CN 00 00 vO r^- CO LO LO rH t-H rH > o o o o o o z o z H m LO LO LO LO LO LO lO LO LO LO LO LO g 3 r^ r^ f«. r-~ r^« r** r~ r^ r^ r^ r^ r^ r-^ 3 O co •H 4-1 r-\ CN CO o r-i rH T-i CN CN CO T-i r-f 4-1 CN 1 CO co LO CO LO 00 CO CO o LO CO CO 0) s • » • • rH M o o o o o 43 CO 4-1 "■^ ^*^ ^^ -■^ CO cu CO o o o o H O cu !3 so 3 O PQ 60 3 O PQ 60 3 o PQ 60 3 O PQ 8-4 Samples of the dominant groups (Copepods, Parathemisto, and Fteropods) mere selected and Kept frozen for hydrocarbon analysis. The remainder of each sample was preserved in 4 percent Formalin for identification, counting, and further examination for oil occurring as external smudges or ingested into the gut. 3.1.2 U.S. Samp I ing Neuston samples were collected at 11 stations with a 0.5 by 1.0 m rectangular frame fitted with a 9.505 mm mesh net. Tows of IB min duration were conducted at speeds of 3.3 Krn-'hr, effectively sampling a surface area of 550 sq m. In addition, at Stations 1 and 2, stepped oblique tows were made with a 61 cm bongo sampler fitted with 0.505 and 0.333 mm mesh nets. Pit Station 1, a 45 min tow was conducted which sampled at 20, 15, and 10 m depth each for 15 min. Pit Station 2, the tow was for l \ir and sampled 20, 15, 10, and 5 m depths each for 15 min. ft L L samples were preserved in 10 percent Formalin. A summary of these tows is contained in Table 8-2. 3.2 Species Composition 8.2.1 The Danish Samples The Stramln net samples were reduced to an aliquot of about 3000 specimens with a sample divider and identified to the lowest possible taxa. The Hensen net samples were totally counted and identified ( Table 8-3). The fish larvae from all the samples were identified (Table 3-4). The copepods were the dominant group in the planKton. Calanus hyperbcreus was the dominant species in all the Stramln net samples accounting for 3? to 82 percent of the specimens. Calanus glacial is was the second most numerous species and occurred at all stations. Pit the reference station, Metridia longa was the dominant species, but only a few or no specimens were present at the other stations. In the Hensen net samples, the small copepod Pseudocalanus minutus was the dominant species, possibly because of the smaller mesh in this net. Because of the Known daily vertical migration of planKton, natural variation present in the samples made comparison between samples collected at different times uncertain. 8-5 Table 8-3. Result s of zooplankton enumerations for Danish stations Station number 5460 5464 5469 5471 5477 5-460 5469 5471 5477 Type of net and depth S 200, 225-0 m, 30 min haul 0,45 m^ Hensen, 50-0 m Sarsia princeps 64 8 70 Leukartiara breviconis 64 Aglantha digitale 5440 2040 4960 3238 128 61 63 54 67 Clione limacina ad. 704 72 1216 70 5 1 " " juv. 8 4 6 5 Limacina helicina ad. 704 24 2144 141 64 3 1 3 " " juv. 9 4 4 20 Hiatella sp.juv. 1 Conohoecia sp. 1 Calanus hyperboreus ? 50880 3920 39360 665 20 26880 120 19 86 35 ii ii y 54720 7600 42240 49984 44720 124 4 109 93 it it jy 9600 3040 20800 5 280 10880 44 9 71 51 ii .. IXI 960 320 1280 - - - - Calanus glacialis $ 7680 1040 1280 5 280 9600 96 3 27 28 n .1 g 80 ii ii y 5120 2840 3200 6688 5 2480 376 21 103 354 II m jy 4 8 4 12 Calanus f inmarchicus ? ii ii s ii ii y 320 3200 60 6 18 50 3 114 320 640 192 14 49 II ,i jy 1 Calanus copepodits IV 856 96 364 819 Calanus nauplii 192 312 196 30 Pseudo calanus minutus $ 280 36 68 75 " " V? 568 390 444 845 ii i. y C5 S3 co O cn O m CM CM in in cn CO CO lO o O O IS •— o t/i O =3 O -J Q. cn •— CM lO tn o cm en o r— f— © r— «3- 1— cn lO CM O o I I I o • I ro i I i I I I I I O r-» en en •a- en r>» CO 3 f> CM I I CM I— r— i r». I I CM i «■ r*. m cn in p— io o «a- r— CM GO f— r— CO •— i i en eo en tn in O I CM I I I en I I «3- I I I I I I I en i i i i u> r>» r»» i i i I I CM I I i i i i i m i i i r«> i i in i i i i i i i r-» i i i CO r- O en cn CO t n f— csj r— to I i ■ i i i i I eft m cm o us m m i i i i i in i i i i in i i i i i i i l/il-Q (5 C. «.i cn *3 cn r> =1 c o c ro 2 — .£ ro £ r0 CJ r rs 'rtf i *"" "D J= c c Ci 4-> o 'C 'u i- »o -O rs o -t-> s- — E "oJ ■^ CJ 4-> Q. cn •a cn u t- cj c Ci. o s_ °i c -ii' CI - o 3 re ° <_) UJ tn r- [_ ro 4-> ro •o o 4-» *o cn 4-> JZ <0 I- 3 CJ Q. E CJ XJ >0 O O 4J CI j= c cn c E t JB C «— U r— a. E CJ T3 O — >> O >-, a. o <_> a: 8-8 Large Calanoid copepods strongly dominated the planKton in the 0.535 mm mesh Bongo samples. Calanus hyperboreus .- one of the largest Known calanoids and one which occurs primarily in Arctic waters, accounted for approximately 7? percent of the total zooplanKton numbers at stations l and 2. Calanus glacial is and Calanus f inmarchicus , smaller, morphologically si miliar; and congeneric species, occurred at both stations but were not numerous enough to be considered dominant. It should be noted that Paratnernisto l ibel lula , a hyper id amphipcd, and smaller copepods were not ranKed high in the B.505 mm mesh bongo samples. C. huperboreus also dominated the smaller mesh (0.333 mm) bongo samples (Table 8-5). The noticeable difference in species composition between the smaller and larger mesh samples was the increase in numbers of smaller species, P_. minutus and Oi thona simi l is , in the smaller mesh samples. P. minutus was ranKed second in numerical, importance for these samples. Meuston samples taKen in the vicinity of the spill (Stations l to 8) were dominated by P. I ibel lu la . In the control area (Stations 3, 10, and 11), P_. I ibel lula was also dominant. The composition of communities in the neuston (surface) and bongo (water column) samples are compared in Figure 3-1. The two dominant species appear to be almost mutually exclusive. The neuston samples are strongly dominated by P_. I ibel lul a (31 percent) while C_. huperboreus (76 percent) dominates the bongo samples. Calanus glacial is , £. f inmarchicus , and Limacina he I icina were of less importance in both biotopes. Sars (icS>0) reported that f\ I ibel lula is a good indicator of the Arctic water and occurs in large numbers at the surface. Populations are Known to be composed primarily of juveniles, as the large adult individuals are seldom encountered. Similarly, in this survey the population was disproportionately juveniles. From 95 to 100 percent of P.. I ibel lula at any one station were juveniles. Hyper id amphipods are Known to be extremely strong swimmers, capable of extensive vertical movement and P^ l Ibel lula has been reported from the surface to 2,500 m. Examination of the diurnal occurrence of F\ I ibul lela in the neuston samples during this survey indicates that this species is from 100 to 1,000 times more abundant at the surface during the Arctic night and twilight periods. 8-9 CALAMUS HYPERBOREUS CALANUSGLACIALIS BONGO NEUSTON CALANUS FINMARCHICU5 LIMACINA HELICINA PARATHEMISTO LIBELLULA 80 60 _ T _ 40 20 1 1 - 20 T T~ 40 t — i — r 60 PERCENT NUMBERS ~1 1 80 Figure 8-1. Comparison of the species composition of communities in bongo (0.505 mm mesh) and neuston (0.505 mm mesh) samples. Stations 1 and 2 were combined for the bongo samples. If a slick was present, as at station 2, then the plankton movement into and out of contaminated waters would provide a pathway for hydrocarbon compounds to enter major plankton and fish communities in the water column. Arnphipods, because of their numerical importance, are considered as a major food resource for fish species and the ringed seals. Whether or not there is a food chain magnification of hydrocarbon compounds as has been shown for other pollutants, e.g., pesticides and heavy metals, has yet to be determined. However, the potential for significant impact exists if these compounds are transferred to the more sensitive deepwater environments. Characteristically, these low productivity regions are inhibited by long-lived slow-growing species with low levels of fecundity; therefore, the carrying capacity for such pollutants in these environments would be expected to be minimal. Fish larvae were virtually absent in the plankton during this survey. The only specimen collected was a radiated shanny, St ichaeus punctatus. 9-10 8.2.3 White Particles on the Sea Surface Eleven days after the spill, small white particles were observed floating on the sea surface. Some of the particles could be identified as the remains of dead zooplanKton (copepods) . These white particles appeared to be globules of white fat or oil. It is suspected that no relation exists between dead planKton and the spilled oil, as the white particles were observed both within oiled areas as well as in areas that the spilled oil did not reach. Furthermore, dead planKton have been reported in this area during 1923 as a natural phenomenon of the copepod Me tr i d 1 a longa CJespersen, 1934). The most probable reason for the association of the dead planKton and the oil is that they were both concentrated by convergent processes such as Langmuir circulation. This explanation does not however preclude an adverse effect of oil on planKton which could conceivably contribute to planKton mortality. 8.3 Oil Contamination Of Zooplankton find Fish 8.3.1 Chemical Examinations Ten samples of zooplanKtcn and fish from Danish hauls were chemically examined for petroleum hydrocarbons by the Water Quality Institute CHansen et al., 1973). 8.3.1.1 Analytical Procedure In general, the procedure of Farrlngton and Tripp (1975) and Farrington and Mederias (1975) was applied for extraction and isolation of the hydrocarbons between n-C12 and n-C36. fin amount of 2 to 20 g (wet weight) of biological material was used for each analysis. After homogenization in a blender, the sample was refluxed for a few hours with 40 g KOH per liter of 90 percent methanol. After cooling, the mixture was filtered with suction if solid materials were present. The residue was washed off the filter with a small volume of pentane. The saponification mixture, if filtration was necessary. was extracted three times with pentane. The extract was evaporated with a rotary evaporator until 1 to 2 ml remained. Column chromatography of the extract was performed by using a column of equal amounts of alumina (A 1203) packed on top of silica (5102). The alumina 8-11 and silica were activated overnight at 250 C and 150 c, respect ively, and both were subsequently deactivated with 5 percent of water. The ratio of column material to nonsaponif iable lipid had to be 100:1 or more for the analysis to continue. The column was eluted with 1.5 column volumes of pentane + benzene (30 percent + 50 percent by volume). The eluate was evaporated to near dryness on a rotary evaporator and then redissolved in a small volume of CC14. A few microliters were injected into the gas chromatographic column. A standard n-alkane mixture of Known concentration was used to measure the detector response per unit weight of alKane. The internal standard used was n-C22. Gas chromatography was performed on a SCOT column as described in Chapter 5.1. 3.3.1.2 Results The results from some of the gas chromatographic CGC) analyses on the SCOT column are shown in Figures 8-2 through 8-5. In all the chroma tograms a few edSa&fcfcJ wJw Figure 8-2. Gas chromatogram of Boreomysis from station 5465. very strong peaks dominated, showing the presence of biogenic hydrocarbon; One of these, pristane, has been found in all the samples. Most of the samples also show the presence of a complex mixture of petroleum hydrocarbons. The amounts of petroleum hydrocarbons are low 3-12 X) * m-+iH-~frHJ '* i f<^_ Figure 8-3. Gas chromatogram of Themis to from station 5465, k '. Wl^-Y¥-^-J^H (/'■ Wf-rt U !/Wj_J*— ^_ Figure 8-4. Gas chromatogram of a Copepod from station 5470. 3-13 V-Wfl Figure 8-5. Gas chromatogram of a Copepod from station 5477. Reference station located 50 nmi southeast of the spill site. compared with the amounts of the biogenic hydrocarbons. Petroleum hydrocarbons were found even in the reference sample from station 5477 (Figure 8-5). Only two samples seemed to be uncontamlnated witn petroleum hydrocarbons. 8.3.1.3 Discussion The copepod sample from Station 5477 was taKen as a reference sample far from areas contaminated by the POTOMAC oil. The small amounts of petroleum hydrocarbons found by GC may indicate inadvertent contamination either during sampling or analysis. However, the analytical method could not distinguish between internal or external oil on the organisms. As a whole, the largest amounts of petroleum hydrocarbons were found in copepocis, which have a rather high lipid content, and the lowest amounts were found in pterapods, which have low lipid content. This indicates that only part of the petroleum hydrocarbons found are related to contamination either during sampling or analysis. 8-14 No correlation was found between the water analysis performed by fluorescence and the zooplankton analyses performed by GC. This could be the result of the fact that the amounts of petroleum hydrocarbons found in tne zooplankton are relatively low and dependent on other factors such as lipid content or the fact that the results indicate the possibility of contamination during either sampling or analysis as mentioned above. Earlier investigations (Johansen et al 1977) in the area off West Greenland which studied invertebrates* fish, and sediments, showed no presence of petroleum hydrocarbons. S.3.2 Physical Examinations 8.3.2.1 Procedure The Danish plankton samples were examined under a dissecting microscope for the presence of oil. The contamination was classified as n external for oil adhering to the cuticle and adhesion to appendages or 2) internal for ingested oil. Specimens suspected to be contaminated in the gut with oil were cleared with lactic acid. 8.3.2.2 Results The occurrence of contamination is shown in Table 8-6. Station 5468: The sample was collected at the spill sits where no oil was visible, but increased hydrocarbon levels were found in the water. The highest percent of oil ingestion was found at this station. The number of plankton with algae in their guts was also high. Station 5461: The sample was collected in an area with oil pancakes on the surface and the plankton net retained both oil and plankton organisms. Many of the organisms had oil adhering to their exoskeletons, but it was impossible to determine whether or not the oil was present before they were caught in the net. It was remarkable that the mandibles were often contaminated. No oil was seen as ingested, and very few organisms had ingested algae. Station 5464: The sample was taken in an area with oil pancakes on the sea surface: however an attempt was made to avoid oil contamination of the plankton net. Even so, contamination of the plankton after capture cannot be excluded. About 1 percent of the copepods were found contaminated with oil 8-15 Table 8-6. Occurrence of oil contamination on dominant plankton groups . Station Plankton Examined Contaminated Type of Contamination No. No. % External Ingested 5460 Copepoda 1739 62 4 5 57 5461 Copepoda 2005 N N 5464 Copepoda 1044 11 1 9 2 Parathemisto 19 2 11 1 1 Chaetognatha 715 2 0.3 2 5473 Copepoda 355 1 0.3 N 1 Parathemisto 520 4 0.8 N 4 _ N Many individuals were observed with external contamination. These were not counted because oil particles were also caught in the net and the contamination may have occurred after collection. and two specimens were found with oil in their guts. Of the 19 Parathemisto in the sample, one was contaminated externally and another internally. Stations 5469. 5470. and 5471: These Stations were acquired well to the north of the oiled area. No oil contamination in the gut or on cuticle was found. Many of the copepods had algae in their guts. Station 5473: Both oil flakes and plankton were contained in the sample so the number of externally contaminated specimens was not determined. No oil at all was found in the guts of pterapods and only two specimens of arrow worms were externally contaminated. The oil particles found in the guts of the copepods were examined with a fluorescence microscope but no fluorescence was seen. Oil was collected together with plankton in all of the NOAA samples. The alimentary tracts of individuals that appeared dark were removed, cleared, and examined to determine whether or not oil had been ingested. No oil was found in any of the alimentary tracts examined. 8-16 8.3.2.3 Discussion The percentage of planktonic organisms affected by oil was small and thus there was probably no major effect on the plankton population. The percentage of individuals with oil in their guts was smaller than that found at the AFGO MERCHANT oil spill (Maurer, 1976). The sample collected at the spill site (Station 5460) was the only one with a high percentage of copepods having oil in their guts. Furthermore, the samples which had oil collected in the net did not show a high incidence of copepods with oil in their guts. The number of Parathernisto I ibel lula was generally low except at station 5473 and the neuston stations; however, the percentage of individuals which had ingested oil was higher. This may be the result of differences in the size of the oil particles availible for consumption. Parker (1969) found that copepods with small particles of a Bunker-C oil in their guts did not show signs of distress. However, larger oil particles might cause a blockage of the guts with fatal effects for the individual specimens. In cases of both internal and external contamination, the growth and reproduction of individuals, if not their very survival, may be affected. Exoskeleton contamination may inhibit sensitive chemo-recept i ve pores used for positioning during reproduction (Fleminger, 1973). Adhesion of oil to appendages could also interfere with the feeding currents and food handling. In addition, those individuals which were contaminated to the extent that they could not make quick escape responses would be more vulnerable to predation. 8.4 Summary Two separate and distinct plankton communities were present in Melville Bay during the post-spill sampling period. The surface layer was strongly dominated by the hyper id amphipod Parathernisto I ibel lula , while the water column plankton were dominated by the copepod Calanus hyperboreu s . White particles, which were determined to be the remains of dead copepods in some cases, were observed floating in long streamers in parts of Melville Bay after the spill. This phenomona appears to be natural and not related to the oil spill. As evidenced by gas chromatography of selected zooplankton, low levels of petroleum hydrocarbons were found in most of the samples Including those 8-17 collected at a reference station which should not have been impacted or exposed by oil from the USNS POTOMAC. This indicates that either the samples at the reference station were inadvertently contaminated by the gear used or were contaminated by the ADOLF JENSEN cooling water. l/isible oil contamination was observed on 11 percent or the Parathemisto at one station. Copepod contamination did not exceed 4 percent at any station. Ingested oil was the dominant contamination only at the spill site (Station 5460); at the other stations, external contamination predominated. Tine effects of this visible contamination, either internal or external, on the survival or reproductive behavior is not known. Since only a small portion of the total zooplanKton population of Melville Bay was believed to be affected by the oil spilled from the USNS POTOMAC, there was probably no major effect on the zooplankton population from this spill. 9.0 MARINE MAMMALS AND SEABIRDS 9.1 Observations Of Marine Mammals Marine mammals were observed from the ADOLF JENSEN by J. Christiansen of Marin ID. In the area north of Upernavik over the period from August 12 to 20, 43 ringed seats ( Pusa hisp ida ) , 4 hooded seals ( Cystophora cr istata ) , 1 bearded seal ( Er iqnathus barbatus ) , and 7 unidentified seals were observed . During two helicopter flights on August 16 covering most of Melville Bay. scientists from the ADOLF JENSEN observed only 2 seals. Of the 43 ringed seals, 32 were observed in association with winter ice and 25 of these 32 were observed before reaching the spill site while passing through a belt of pack ice. The remaining 7 were observed amongst and on rotten ice near Thorns Isle (75°43'N 60°35'W). The other seals were observed in open water, often near icebergs. In an oiled area, 4 seals were observed together near an iceberg, but nothing unusual was observed about their behavior. The sightings of most of the seals near sea ice rather than in open water was quite normal and agreed with investigations made in the area two years earlier. The small number of seals observed in the oiled area can be explained by the lack of ice in the area without recourse to an avoidance behavior of seals for oiled water. No deaths or abnormalities of seals were reported except for some instances of oil contamination on their skins. 9.2 Oil Contamination Of Sealskins While no seals were captured during the spill response, a local hunter did report oiled sealskins about one month after the spill. In total, about 2S oiled skins were reported. The skins of 18 seals, all shot or captured in nets, were delivered to Denmark for analysis (Figure 9-1 and Table 9-1). Sewen of the seals were heavily contaminated with oil on their backs with lesser amounts on their necks (seals 1, 2, 4, 8, 9, 11, and 13). Three seals had spots of oil on their backs or necks (seals 3, 6, and 10). On eight of the seals it was not possible to either see or smell oil (seals 5, 7. 12, 14. 15, 16, 17, and 18). Ten samples from oiled skins and four samples from seals 9-1 9-2 Table 9-1 Date and Location of Seal Catchings. Seal Date Location Position no. 1 9/29/77 Moriussaq 76°48'N 70°05'W 2 9/15/77 Kuvdlorssuaq 74°34' 57°20' 3 11/23/77 Tasiussaq 73°20' 56°05' 4 1/26/78 Godhaven (Parry Skaer 69°10' 53°40' 5-12 Jan-Feb 78 Upernavik area 72°35' 56° 13 April 78 Niaqornarssuk 68°15' 52°50' 14 - 18 Mar-Apr 78 Upernavik area 72°35" 56° with no visible oil were analyzed by the Water Quality Institute to determine whether or not any oil originated from the USNS POTOMAC spill. 9.2.1 Analytical Procedure The oil was mechanically isolated from the hair and the fat. For one sample (seal 1), the isolated oil was dissolved in a small volume of CC14 and a few microliters were injected into a gas chromatographic column. For ail other samples, a cleanup procedure was necessary to isolate the hydrocarbons from interfering components. The cleanup procedure was similar to that described in Chapter 8.3.1.1. 9.2.2 Results Gas chromatograms were obtained on a SCOT column and a few examples are presented in Figures 9-2 to 9-7. It was not possible to detect petroleum hydrocarbons on seals 5, ?, 12, or 14 confirming the visual and olfactory examinations. Petroleum hydrocarbons were detected in the extracts made from seals L 2, 3, 4, S, 8, 9, 10, and 11. The compositions of the hydrocarbons from seats 1 and 2 (Figures 9-2 •and 9-3) were similar to the composition of the hydrocarbons collected on the sea surface at the spilt site. In these cases it seems probable that the contamination of these two seals originated from the POTOMAC spill. 9-3 90° 80° 70° 60° 50° 40" 30° 20° 10° 0' 80' 75' 70' 65' 60' 70' Figure 9-1. Capture 60 ° 50 ° 40 ° locations of the analyzed seals. See Table 9-1 9-4 -! \hjj^l Figure 9-2. Gas chromatogram from seal 1. Heavily oiled on back. Collected north of Thule, Greenland on September 29, 1977 Figure 9-3. Gas chromatogram from seal 2. Heavily oiled on back. Collected east of the spill site on September 15, 1977 The compositions of the hydrocarbons from seals 4 (Figure 9-4), 6, 9, 10 (Figure 9-5), and 13 (Figure 9-6) were different from the surface oil samples. The sharp peaks found in the surface oil samples (Figure 5-4) were lacking in the chroma tograms from the seals. This difference may be due to biodegradation, and it should be considered that these petroleum hydrocarbons might have originated from the POTOMAC spill. The missing peaks are also evident in the chromatogram from seal 3 (Figure 9-7). Furthermore, the composition of the hydrocarbons in samples 9-5 Figure 9-4. Gas chromatogram from seal 3. Spots of oil on neck. Collected southeast of the spill site on November 23, 1977 Figure 9-5. Gas chromatogram from seal 4. Heavily oiled on back. Collected well south of spill site on January 26, 1978, 9-6 I jR SW-f- 7 if U-i Figure 9-6. Gas chromatogram from seal 10. Spots of oil on neck. Collected southeast of the spill site in January 1978, Figure 9-7. Gas chromatogram from seal 13. Heavily oiled on back Collected well south of the spill site in April 1978. 9-7 from seal 3 showed more n-alKanes with carbon numbers above 20 than did the surface oil samples. Some n-alkanes with carbon numbers above 20 were also found in extracts from seals 8 and 11. The presence of these n-alKanes could be due to a contamination source other than the POTOMAC. It was not possible to determine if the other hydrocarbons on these seals came from the POTOMAC oil spill, because the analytical procedure was not definitive enough to distinguish between the seal contamination and the USNS POTOMAC sllcK samples. Histological examinations were conducted on the skin underneath the oiled areas on some backs of the seals. No damage was seen during these examinations. In controlled experiments with heavily oiled seals (Engelhardt, 1973), the only long term effect of oil contamination found was the appearence of cornea lesions. Although the seals were totally coated with oil during Engelhardt's experiment, they were completely clean after 6 days in an oil free environment. The persistent external contamination of seals from the West Coast of Greenland after the POTOMAC oil spill might be attributed to differences in the composition of the contaminating oils. All the oiled seals reported were either shot or caught in nets: however, it is not certain whether or not the oiled seals ware more susceptible to capture. 9.3 Seabird Observations Observations of seabirds were routinely conducted from the ADOLF JENSEN, yery few birds were observed in the oiled areas during the August study period. Apart from a few flocKs of Little AuKs ( P lotus a 1 1 e ) and Kitti wakes (Rissa trldactyla ) , only solitary birds such as Gulls (Larus sp_. ) , Guillemots (Cegphus gryl le ) , and Fulmars ( Fulmar us glacial is ) were observed. Some of the birds were seen floating on the sea surface and observed to taKe off in areas where an oil film covered the surface; however, no smudged birds were observed and the birds in the oiled areas behaved identically to those in unolled areas. Birds were not observed in direct contact with oil pancakes nor were any troubled or dead birds observed. Furthermore, observations from helicopters did not indicate any effect of the oil on the birds. Except near birdcliffs, none of which were located In the vicinity of the spilU very few birds were observed either in the spill area or along the cruise track to the south. The stomach contents of a single Fulmar, shot in the spill vicinity, showed no signs of petroleum hydrocarbons with gas 9-8 chromatographic analysis. P young Gull with black legs and breast was reported on October 2, 197? at SavlgsiviK (76° 02'N 65°00'W) c~92 nml to the northwest of the spill site), fts the vicinity of the spill area is generally deserted, it is understandable that few oiled birds were reported. The spill occurred in a location and season such that no harm to birds was observed. However, during another season or near blrdcllffs, a spill of the same size as that of the POTOMAC might have caused extensive damage, especially when young birds might be leaving their nests. 10.0 IMPACT ASSESSMENT 10.1 Fate Of The Spilled Oil On August 6, 1977, approximately 107,038 U.S. gallons C~380 tons) of Bunker-C fuel from the USNS POTOMAC was spilled into Melville Bay, Greenland after a fuel tank was "holed" Py a small icePerg. This fuel was a PI end of 55 percent pitch (specific gravity 1.054) .and 45 percent cutter stock of No. 2 fuel (specific gravity 0.883) which initially remained on the sea surface because its specific gravity of 0.976 was less than that of the surface seawater, 1.024. The floating oil was advected to the north and west Py the slow general circulation in the area. Highly variaPle light winds tended to disperse the oil over a large area. Based on various measurements, it is estimated that the oil was not advected more than 40 nmi from the spill site, with the initial direction being north then shifting to the west after one week. The traces of surface oil found in neuston tows during the return to Thule, Greenland of the USCG icebreaker WESTWIND are believed to have originated by continued light leakage from the holed fuel tank as the POTOMAC continued to its destination at Thule. The return path of the WESTWIND was identical to that of the POTOMAC so that these samples would have Paen taken where the probability of finding oil from continued leakage was highest. During the two weeks following the spill, the composition of the oil on the sea surface changed through evaporation such that, by August 20, almost all of the. low boiling point fraction of the cutter stock, components up to n-C17 (boiling point 300 C), had disappeared. This evaporative loss amounted to about 33 percent of the total spill (35,000 gallons). It is estimated that a great part of the remaining oil (66 percent of the total) sank 1,000 m to the bottom over a large area (~500 sq mi) of Melville Bay, because of the Increase in specific gravity of the oil remaining after evaporation. Sinking was confirmed by the visual observation of small flakes (l cm diameter) within the water column after August 13. It is hypothesized that the flakes were originally the skin of oil lenses which, after depletion of the more volitile components, ware sloughed off or exfoliated from the pancakes. 10-1 10-2 By August 20, only small amounts of oil were observed on the sea surface in the spill area. The oil was in the form of small pancakes (maximum diameter of 5 cm) which had lost almost all of the sheen which had surrounded them earlier. These pancakes were found in windrows several hundred meters long and a few tens of meters wide with a mean spacing of 1 or 2 meters between pancakes. The total volume of oil in each of these windrows (300 m by 36 m size) was estimated to be about 20 gallons. On August 21 windrows of oil were sighted at 74°55'N, 60°12'W from the ADOLF JENSEN. These were the most southerly observations of the oil. Some, if not all, of this oil may U3 and R.I. Mateles, 1967: "incorporation of Liquid Hydrocarbons into Agar Media," Applied Microbiology, vl5, p 961. Blumer, M.M., and J. Sass, 1972: "Oil Fo Hut ion: Persistence and Degradation of Spilled Fuel Oil," Science, vl76, pp 1126-1122. Blumer, M.M., M. Erhardt, and J.H. Jones, 1973: "The Environmental Fate of Stranded Crude Oil," Deep-Sea Research, v2Q, pp 239-259. Boehrn, P., and D. Feist, 1973: "Final Report for Analysis or Greenland Oil Spill Samples," NOAA Contract 78-4050, January 1973. Energy Resources Company, Cambridge, Mass (unpublished manuscript). Boehm, P., and D. Feist, 1973a: "Analyses of the Water Samples From the TSESIS Oil Spill and Laboratory Experiments on the Use of Niskln Bacteriological Sterile Bag Samples," NOAA Contract 03-A01-8-4178, Energy Resources Company, Cambridge, Mass (unpublished manuscript). Bunch, J.N. and R.C. Harland, 1976: "Biodegradation of Crude Petroleum by the Indigenous Microbial Flora of the Beaufort Sea, " Technical Report 10 (unpublished manuscript). Clark, R.C, Jr., and D.w. Brown, 1977: "Petroleum: Properties and Analyses of Biotic and Abiotic Systems, " in chapter l in Effects of Petroleum on Arctic and Subarctic Environments and Organisms, L^o L 1, Nature and Fate of Petroleum, Academic Press, Inc., New York, 321 pp. Eastwood, D., 1977: Personal communication with P. Boehm, U.S. Coast Guard Research and Development Center, Avery Point, Groton, Connecticut. Engelhardt, F.R., 1378: "Petroleum Hydrocarbons in Arctic Ringed Seals, Pusa hinida . Following Experimental Oil Exposure," Proceedings of Conference on Assessment of Ecological impacts of Oil Spills, Keystone, Colorado, June 14-17, 1978, American Institute of Biological Sciences. Farrington, J.W., and B.W. Tripp, 1975: "A Comparison of Analysis Methods for Hydrocarbons in Surface Sediments," ACS Symposium Series, No 18. Marine Chemistry in the Coastal Environment, pp 267-284. Farrington, J.W. and G.C. Medeiros, 1975: "Evaluation of Some Methods of Analysis for Petroleum Hydrocarbons in Marine Organisms, " Proceedings of the 1975 Conference on Prevention and Control of Oil Pollution. American Petroleum Institute, Washington, D.C. pp 115-121. Fleminger, A., 1973: "Pattern, Number, Variability, and Taxonomic Significance of In tegumental Organs (Sens ilia and Glandular Pores) In the Genus Eucalanus (Copepoda Calanoida) , " Fishery Bulletin, v71, n4, pp. 965-1010. Grahl-Nlelson, 0., 1976: Proceed inqs of 12th Nordic Symposium on Water Pollution, Nordforsk, Helsinki. 12-1 12-2 Hansen, N., V.2. Jensen, and K.K. Kristensen, 1378: "The Oil Spill in Melville Bay, Greenland: Results of the Chemical and Microbiological Studies. Draft Report May 1, 1978. Nater Quality Institute, Horsholrrr, Denmark. 55pp. (unpublished manuscripts. Jademac, R., 1977: Personal communication, U.S. Coast Guard Research and Development Center- Avery Point, Groton, Connecticut. Jespersen, P., 1934: "The Godthaab Expedition 1923," Meddelelser on Gronland, v79, nie. Copenhagen 1934. Johansen, V.V., B. Jensen and A. Buchert, 1977: Hydrocarbons in Marine Organisms and Sediments off West Greenland," Cedited by R.G. Ackman) , Fish. Mar. Serv. Tech. Rep. 729, 33p. Lloyd, J.B.F., 1971: "The Nature and Evidential lvalue of the Luminescence of Automobile Engine Oil and Related Materials — I. Sunchronous Excitation of Fluorescence Emission," Journal of the Forensic Science Society, vil, pp. 33-94. Mattson, J.S., and P.L. Grose, 1973: "Impact Assessment of the USNS POTOMAC oil spill, Melville Bay, Greenland, August 6, 1977," Interim Report, National Oceanic and Atmospheric Administration, Environmental Data and Information Services, Center for Environmental Assessment, Washington D.C. May 1973. (unpublished manuscript). Maurer, R.O., 1976: " A Preliminary Report of Zooplankton in the Vicinity of the Argo Merchant Oil Spill," in "The Argo Merchant Oil Spill: A Preliminary Scientific Report", (Grose, P.L. and J.S. Mattson, editors) NOAA Special Report, National Oceanic and Atmospheric Administration, U.S. Department of Commerce, Washington, D.C, 275 pp. Maurer, R., and J. Kane, 137S: "Zoop lank ton in the Vicinity of the USNS Potomac Oil Spill (Baffin Bay, August 5, 1977)." Laboratory Reference No. 78-07, National Marine Fisheries Service, Northeast Fisheries Center, Woods Hole, Massachusetts, 17 pp. Mi I gram, J., 1977: Personal communication with P. Boehm, Department of Ocean Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts. Mills, A.L., C. Brezil and R.R. Colwell, 1978: "Enumeration of Petroleum Degrading Marine and Estuarine Microorganisms by the Most Probable Number Method," (in press). Moyniham, M.J., and R.D. Muench, 1971 : "Oceanographlc Observations in Kane Basin and Baffin Bay, May and August-October 1969, " U.S. Coast Guard Oceanographic Report No. 44, CG 373-44, Washington, D.C. Muench, R.D., 1971: "The Physical Oceanography of the Northern Baffin Bay Region, " Baffin Bay-North Water Project Report No. l, Arctic Institute, Washington, D.C, 105 pp. Muench, R.D., 1972: "Oceanographic Conditions in the Northern Baffin Bay Region, July-August 1B70, " U.S. Coast Guard Oceanoqraphic Report No. 54, CG 373-54, Washington, D.C Muench, R.D., M.J. Moyniham, E.J. Tennyson, Jr.. W.G. Tidmonsh, and R.D. Theroux, 1971: "Oceanographic Observations in Baffin Bay during July- September 1968," U.S. Coast Guard Oceanographic Report No. 37, CG 373-37, Washington, D.C Parker, C.A., 1969: "The Ultimate Fate of Crude Oil at Sea Interim Report No. 5: Uptake of Oil by Zooplankton, " A.M.L. Report No. B/193 (M) , Admiralty Materials Laboratory, Poole, England, 16 pp. Sars, G.O., 1890: "An Account of the Crustacea of Nonajay: Arnphipoda ," volume I, Universi tetsfor lage t, Bergen and Oslo, Norway. Wakeham, S.G., 1977: "Synchronous Fluorescence Spectroscopy and Its Application to Indigenous and Petroleum-Derived Hydrocarbons in Lacustrine Sediments," Environmental Science and technology, vll, pp. 272-276. 12-3 ZoBell, C.E.. 1969: "Microbial Modif iction of Crude Oil in the Sea," Conference on Prevention and Control of Oil Spills, American Petroleum Institute, New York, pp 317-326. 13.0 APPENDIX 13. 1 Marine Agar Bacto-peptone 5. g Bacto-yeast extract 1. g FeCi3 0.1 g NaCL 19.45 g Na2S04 3.24 g MgCl2 8.8 g CaCL2 1.8 g KCL 0.55 g NaHC03 0.1G g KBr 0.08 g SrCL2 0.034 g H3B03 0.022 g Na2Si03 0.004 g NaF 0.0024 g NH4N03 0.0016 g Na2HP04 3.008 g Bacto-agar 15. g To rehydrate the medium, suspend 55.1 g In 1,000 ml of distilled water and heat to dissolve the medium completely. Sterilize in the autoclave for 15 min at 121 C. Adjust pH to 7.6. 13-1 13-2 13.2 Agar Substrate NACl 24. g MgS04,7H20 0.5 g KCl 0.7 g KH2P04 2.0 g Na2HP04 3.0 g NH4N03 1.0 g Agar 15. g To rehydrate the medium- suspend 46.2 g in 1,000 ml of distilled water and heat to boiling to dissolve the medium completely. Sterilize for 10 min at 115 C. Adjust pH to 7.1. After Coluiell (Mills et al., 1978) 13-3 13.3 Bunch Substrate fused for MPN method) NaCL 5.53 g MgCl2,6H20 2.54 g KC L 0.1 g CaCl2,2H20 0.37 g Tris buffer (Sigma) 7.69 g NH4N03 1.8 g K2HP04 0.1 g Add 1.9 ml chelated solution of metal salts (below). The compounds are dissolved in 1,000 ml distilled water. The pH is adjusted to 7.5. After Bunch and Harland, (1976). Chelated solution of metal salts COC12.6H20 0.004 g CUS04, 5H20 0.004 g FeCl3.6H20 1.0 g ZnS04, 7H20 0.3 g MnS04,H20 0.6 g Na2Mo04, 2H20 0.15 g E.D.T.fl. 6.0 g These compounds are dissolved in 1.000 ml distilled water and the pH adjusted to 7.5. 13-4 13.4 Colwell Substrate (used for MPN method) NaCL 24. g MgS04,7H20 0.5 g KCl 0.7 g KH2P04 2.0 g Na2HP04 3.0 g NH4N03 1.0 g To rehydrate the medium, suspend 31.2 g in 1,000 ml distilled water and heat to boiling until the medium is dissolved completely. Sterilize for 10 min at 115 C. The pH is adjusted to 7.1. After Mills et at. C1978). 17268 *U.S. GOVERNMENT PRINTING OFFICE: 1979 C, .! I / V \ O 2^ «" c \ ^4 J ^TES O* h J NOAA--S/T 79-202 PENN STATE UNIVERSITY LIBRARIES AQQD0?mL|LnL|5