- In-placa Cleaning - of - lVIilk Pasteurizing - Equipment E gggil‘ 9325s; N5 Kwsjakanxzzcsa m gaplenzém l9 g TEXAS AGRICULTURAL EXPERIMENT STATIO R. D. LEWIS. DIRECTOR, COLLEGE STATION, TExAs DIGEST The cleaning in-place 0f stainless steel milk lines, using acid and alkaline rinses, and sanitiz- ing them with chlorine, is a satisfactory proced- ure. In the study reported in this bulletin the critical part of the vat system used for pasteuri- zation was the preheater in which the tempera- ture gradient was high enough to cause a milk stone deposit. In-place cleaning and sanitizing of a plate type regenerator-pasteurizer-cooler is satisfac- tory, if following the water-acid-water-alkali rinse sequence, the plates in the unit are inspected and freed of milk residues by light brushing or hos- ing with cold water. When acid rinses are used each time the unit is cleaned, there are no milk stone deposits on plates when the flow rate of solutions is at least 28 gallons per minute, although short sections of 2-inch line are in the same circuit with ll/g-illCh line. A rinse of 200 parts per million (p.p.m.) chlo- rine is adequate for a system, previously cleaned in-place, after an interval of 40 hours. Acid and alkaline rinses at 165° F. are sat- isfactory in a system in whichmilk has been heat- ed to 163 - 164° F. E . k r In-place Cleaning of Milk Pasteurizing Equipment A. V. MOORE and F. E. POTTER* UNTIL THE PAST FEW YEARS, milk processing uipment has been traditionally washed and san- Etlized by methods that required dismantling (1, , , 3)- Cleaning in-place saves time and labor. It i lso prevents possible damage that may result if F-ssijpe lines are dropped. i; l“ leaning milk equipment is a process with l clear-cut objectives, other than that of ining high quality. These are the emulsi- of fat, softening of protein, neutralizing -* cid and preventing the formation and de- ‘ 1 n of a film. The latter objective is neces- " ecause of the minerals in the milk, the wash 1 A and the cleaning compounds used. All these » fits of deposited minerals are of greater con- n heated equipment surfaces. Such films it green referred to for many years as “milk . 1Q‘ ’ 4 p Following the cleaning process, there ~ ‘i be a sanitizing treatment that will insure ’ ‘imum of bacteriological contamination in ceeding milk processed. Sanitizing, as used “complete” sterilizing. -place cleaning of dairy processing equip- -was first recognized by the U. S. Public T I Service in its 1953 Milk Ordinance and y: ¢(4). It outlines certain recommendations r ‘iitoperly cleaning in-place. While the prin- .c~ié;qof satisfactorily cleaning and sanitizing » i911 lipe lines and plate type heat exchangers re. same, the method of treating pipes must u: iiglittlified when a plate type heat exchanger or ‘sixty-time press” is included in the circuit. ‘ibis study was conducted in two phases. The l irst phase covered a 10-week period during which he milk was vat pasteurized each day except Sun- l ay. The second phase covered an 8-month period uring which a high-temperature, short-time unit as used each Monday, Wednesday and Friday. Source of Milk The cleaning and sanitizing procedures in this btudy were carefully planned, since the milk pro- teased was the regular College Creamery supply. This study preceded the publication of the ‘I953 Ordinance and Code, and since the Texas ‘fRespectively, professor and formerly instructor, Depart- , ment of Dairy Husbandry. bulletin, implies “practical,” but not neces- a l -~0=74n Q3 State Health Department follows the ordinance, per se, for all standards, it was necessary to ob- tainpermission to clean in-place while continuing to label milk “Grade A.” The State Health De- partment granted this permission so that a rea- sonable supply of milk could be handled through the equipment being studied. The raw milk supply was from the College herd and averaged about 450 gallons daily. It was produced under excellent sanitary conditions through the use of mechanical milkers, a closed- pipe system including a filter, and a water-cool- ed, enclosed surface cooler. The milk was piped directly from the surface cooler to an insulated stainless steel-lined truck tank and delivered to the Creamery once daily. Upon arrival at the Creamery, the milk was usually below 50° F. and contained fewer than 15,000 bacteria per milliliter. The tank was emptied into a refrigerated storage vat. Cleaning Procedure for the Pipe System After the day’s run of milk had been process- ed, the cloth bag was removed from the filter, the homogenizer by-passed and all valves removed and scrubbed by hand. With the valves replaced, rinse solutions prepared in the receiving vat were circulated through the box-tube pre-heater, the filter and the entire line up to the holding vats. This was a total of about 180 feet of ll/g-iIlCh stainless steel line. Rinses were conducted through a 50-foot length of 1-inch Tygon tubing which connected the end of the sanitary line and the re- CONTENTS Page Digest .......................................................................... .. 2 Introduction ............................................. .. ............... .. 3 Source of Milk... ........................................................ .. 3 Cleaning Procedure for the Pipe System ............ .. 3 Results with the Pipe System .............................. __ 4 Beginning and End-of-run Bacteria Counts ...... .. 5 Saturday to Monday Bacteria Counts .................. .. 5 Cleaning and Sanitizing the High-temper- ature, Short-time System In-place __________________ _. 5 General Observations on Cleaning the H.T.S.T. System .................................................... __ 6 Bacteria Counts of Bottled Milk __________________________ __ 6 Acknowledgment ______________________________________________________ __ 6 References ....................................................... _; ________ __ 7 C0 ceiving vat. The receiving vat milk pump pro- vided a flow velocity of 25 gallons per minute throughout the system. Before the experiment, paper gaskets were used throughout the milk-processing equipment. During the study, all but three of these were re- placed with neoprene gaskets. The three paper gaskets were examined daily as the experiment progressed and were replaced every 2 to 3 days as they showed evidence of becoming soft and dis- integrated. The neoprene gaskets were examined every few days, and at the end of 4 weeks they were still resilient and in satisfactory condition. The following sequence of rinse solutions was used: 1. Water at 110-120° F., flushed through, but not recirculated; this rinse was continued un- til it no longer appeared milky. 2. Thirty gallons of acid solution were made up in the receiving vat, according to the manu- facturer’s directions, and recirculated at 125° F. for 4O minutes. This solution was then discard- ed. 3. Thirty gallons of water at 110° F. recir- culated for 5 minutes, then discarded. 4. Thirty gallons of alkaline solution, made up according to the manufacturer’s directions, re- circulated for 40 minutes at 125° F., then dis- carded. 5. Cold water flushed through, but not re- circulated, for 5 minutes. The system then stood idle until the next morning, an interval of about 17 hours. There was an additional idle period of 24 hours each Sunday. 6. Before processing began, a new filter cloth was installed, the homogenizer connected into the circuit and 30 gallons of 200 p.p.m. hypo- chlorite solution pumped through the system. This was not recirculated, but went on through the holding vats, pump, plate cooler, surge tank and bottle filler. The hypochlorite was finally discarded as some of it was “milked” out of each filler valve. Results With the Pipe System Milk was processed daily, except Sunday, throughout the 10-week study. Each day one bot- tle of milk was collected at the filler from the be- ginning of the run and one from the end of the run. Standard procedures were used to test these samples for total bacteria count and for coliform organisms. The results of analyses for bacteria in the milk and swab tests of lines, gaskets and elbows the first 3 days of the experiment, indi- cated that further use of swabs would be of no benefit, since they were all sterile. After 15 days of using the washing procedure outlined, the ex- cellent condition of the lines and the consistently 4 low bacteria counts of the milk indicated that th< time of recirculating the acid and alkali solution: could be reduced. The washing procedure was modified for 4 consecutive days; the 40-minute recirculating per iods for the acid and alkali solutions were reducee to 20-minute periods. This resulted in a mill stone accumulation that grew progressively wors< in the preheater during the 4 days; in addition pinpoint colonies, indicating heat-tolerant organ isms, appeared in the end-of-the run milk samples The preheater tubes were thoroughly brushed ant cleaned with acid and all gaskets were replaced The 40-minute period for acid rinsing was re sumed, but the alkali rinsing period was held t1 20 minutes For the duration of this part of the study — 39 operating days — the procedure wa: as originally outlined, except that the alkali rins ing time was 20 minutes. All pipe continued t: be bright and clean looking during this treatment After 4 weeks of continuous use, a set of neoprene gaskets was removed and examined bacteriologi cally. None was contaminated and all were re silient. No estimate of their probable maximun life was determined. Tables 1, 2 and 3 summarize the bacteriologi cal quality of the homogenized-pasteurized-bot tled milk when the lines from the receiving va" to the pasteuring vats were cleaned in-place b1 three slightly different procedures. There wa: one instance of a count exceeding the Grade A standard of 30,000. This occurred at the begin ning of one run. At the end of the run on th< same day, the count was 2,650. Since all line: and other milk surfaces beyond the pasteurizin; vats were conventionally washed by dismantling this one high count may have come as a result o. a bottle being contaminated. Table 1. Standard plate counts and coliform counts o vat-pasteurized, homogenized milk when lines were cleaned in-place—acid and alkali rinses each circulated 40 minutesl Total count Coliform count Degree Beginning of run End of run Beginning of run End of run None in 1 m None in 1 m None in 1 ml. None in 1 ml. Maximum 520 1020 Minimum 25 30 % ‘Fifteen consecutive operating days, except Sunday. Table 2. Standard plate counts and coliform counts o vat-pasteurized, homogenized milk when lines were cleaned in-place—acid and alkali rinses each circulated 20 minutes1 Total count >_ Coliform count Day Beginning of run End of run Beginning of run End of run 1 200 12002 None in 1 ml. None in 1 m 2 130 5002 None in 1 ml. None in 1 m 3 140 3002 None in 1 ml. None in 1 m 4 80 51002 None in 1 ml. None in 1 m ‘Four consecutive operating days. QUncountabIe pinpoints in 1-10 and 1-100 dilutions. Table 3. Standard plate counts and coliform counts of vat-pasteurized, homogenized milk when lines were cleaned in-place—acid rinse circulated 40 min- utes; alkali rinse circulated 20 minutes1 Total count ___ Coliform count Beginning of run End of run Beginning of run End of run Count i Number of days Less than 1 0 0 32 28 1 0 0 4 7 2 to 9 0 0 2 3 10 to 99 7 5 1 1 100 to 499 25 26 0 0 500 to 999 2 3 0 0 1,000 to 4,999 2 4 0 0 5,000 to 10,000 2 1 0 0 Over 10,000 l ($2.000) 0 ' 0 0 lThirty-nine consecutive operating days, except Sundays Beginning and End-of-run Bacteria Counts The only period during which there was a significant increase in the bacteria count in milk at the end of the run over that at the beginning, was when both the acid and alkali rinses were used for 20 minutes. This was caused by heat- tolerant bacteria that produced pinpoint colonies. When both the acid and alkali rinses were used for 40 minutes, and in the third period of the study when acid was used for 40 minutes and al- kali for 20 minutes, there were about the same number of counts higher as there were lower at the end of the run. Since most of the counts were low, those that were higher at the end of the run may be considered unimportant. Saturday to Monday Bacteria Counts The system was idle from each Saturday noon until Monday morning. N0 change in the man- ner of cleaning or sanitizing was made becauseof this extra day between runs. Any buildup of bac- terial growth that might have occurred was ade- quately eliminated by the chlorine rinse that im- mediately preceded processing. Any buildup, how- ever, was most unlikely in view of the excellent condition of the lines and gaskets. The Saturday to Monday counts for 7 weeks were: End of run, Saturday Beginning of run, Monday 550 100 150 30 280 200 . 110 65001 120 170 300 150 610 150 Cleaning and Sanitizing the High-temperature, Short-time System In-place After the raw milk was standardized in the holding vat, it was routed through an in-line fil- ter, 3-section regenerator-pasteurizer-cooler with a 16-second holding tube, homogenizer, surge tank and filler. Homogenization took place before pas- ‘Contaminated bottle suspected. teurization. This circuit included about 100 feet of ll/g-inch stainless steel pipe. The acid and alkaline rinses were applied to the entire system up to the surge tank, excluding the homogenizer and the flow diversion valve. The chlorine rinse was applied to the entire system. At the surge tank, a 50-foot length of 1-inch Ty- gon tubing was used to complete the circuit back to the solution vat. The use of Tygon tubing was approved for this research but it does not meet the requirements of the U. S. Public Health Service even though such tubing is not used to convey milk. The 1953 code states: “Return, re- circulating lines are installed, and are of the same or equivalent material and construction as the milk line.” (p. 89, par. 5c). Previous experience (8) indicated that neo- prene gaskets are satisfactory and may be left in place indefinitely. These were used in parts of the system; in places that were inspected daily, paper gaskets were used, but for one day only. During this second phase of the study, the acid and alkaline detergents manufactured by 5 suppliers were used. There areno significant dif- ferences in the recommendations for using these materials. The procedure followed was a com- posite of the recommendations made by the de- tergent suppliers and those of a dairy industry group (6)1. After the flow of milk was stopped, both heating and cooling media to the plate unit were shut off. Sanitary valves were removed, hand brushed and replaced. Milk remaining in the press was drained. The inline filter element was remov- ed and the filter casing reconnected. Pressure on the plate gaskets was increased slightly to avoid leakage while the solutions were pumped through. The Tygon tubing was connected at the surge tank. A temperature-time recorder bulb was placed in the make-up vat. 1. The flow diversion valve was disconnected and the homogenizer by-passed. 2. Water at 100° F. was flushed through the system and wasted on the receiving room floor until it ran clear. The system was drained free of rinse water. 3. An organic acid solution was prepared in the make-up vat to maintain a level of 18 inches above the outlet after the system was full. The milk pump on the vat also served as a solution- lThe first of several “Tentative Suggested Procedures for Cleaned in-place Pipe Lines Used in Plants Handling Milk and Milk Products” was prepared by a Dairy In- dustry Supply Association task committee in June 1951. (5) The committee included representatives of the Inter- national Association of Milk and Food Sanitarians, the U. S. Public Health Service, and a Dairy Industry Com- mittee. A later report of the group that includes the lay- out and engineering phases of in-place cleaning, as well as the suggested cleaning procedure, was published in the March-April 1953 issue of the Journal of Milk and Food Technology, Vol. 16, No. 2, pp. 77-78. 5 recirculating pump. The acid solution was ad- justed to 165° F. and circulated for 30 minutes, after which it was discarded. 4. Water from the hose at 120° F. was then used to flush out the acid solution. This requir- ed about 10 minutes. 5. Alkaline detergent solution at 165° F. was circulated for 30 minutes. 6. Cold water was again flushed through until the alkali was removed. This required about 10 minutes. At this point, the press was opened for inspection. The plates were brushed lightly and cold water was used to rinse them. 7. The press was again tightened and new paper gaskets were installed. 8. After complete draining of the final cold water rinse, the system stood without further treatment until the next milk was to be process- ed. In most cases, this was a 40-hour interval, on a 3-day-a-Week processing schedule. 9. Immediately before processing began, a 200 p.p.m. chlorine solution was pumped through entire system, including the surge tank and 1 er. General Observations on Cleaning the H.T.S.T. System As would be expected, the critical parts of the system were the heating and regenerating sections of the press. On opening the press, it was found in many instances that there were flakes of semi-jellied milk solids on the plates in these sections, particularly on those in the heat- ing section. These flakes were not, however, stuck to the plates. They were easily removed by light brushing or by hosing with water, but they did have to be so removed. This required only a few minutes. Milk stone never accumulated on the plates, even though the speed of the solutions was not sufficient to flush out all of these flakes. Milk stone did form in the thermometer well, a short 3-inch section of pipe immediately in front of the flow diversion valve. This section was hand-scrub- bed each time the press was opened. Owing to excessive pressure on the plate gas- kets, it was impractical to maintain the recom- mended flow rate of 5 feet per second. The 2- inch, 16-second holding tube would require a dis- charge of 43 gallons per minute at a 5-foot-per- second velocity. Therefore, it was necessary to lower this volume 35 percent, or to 28 gallons per minute, a velocity 3.25 feet per second. It has been reported (7) that, under proper conditions, velocities as low as 1.5 feet per second are satis- factory. The physical condition of the rinsed surfaces and the bacteriological condition of the milk justify this reduced velocity. The daily use of the acid rinse may have been responsible for these good results. While the D.I.S.A. task committee recom- mends the acid rinse on an “if necessary” basis, 6 experience during the previous study here (8) in- dicated that it was a daily necessity, even though the water supply at the Creamery is soft (about 4 p.p.m. of calcium carbonate). There was no at- tempt to eliminate the acid rinse in this study, nor to use it intermittently. The object was to avoid hand scrubbing. There was no apparent damage to any part of the system as a result of the above cleaning procedure. The strengths of acid and alkaline rinses were maintained according to the recom- mendations of the manufacturers. At no time was it necessary to resort to hand cleaning, ex- cept as noted above. The acid and alkali detergents furnished by the five manufacturers were apparently equal in efficiency of cleaning. Each was used for 3 or more weeks. Bacteria Counts of Bottled Milk Routine standard plate counts for total or- ganisms and for coliforms were made on bottled milk samples from the beginning and end of each day’s processing. During most of the study per- iod, the plant operated 3 days a week. Samples were usually plated 48 hours after bottling in or- der that counts would be comparable with those reported by local inspection agencies. Table 4. Standard plate and coliform countslof homogen- ized-pasteurized bottled milk, processed in a plate type pasteurizer, cleaned in-placel Total count Coliform count Beginning of run End of run Beginning of run End of run Number of days Number Less than 1 0 0 64 60 Le-s than 10 0 0 76 79 11-20 0 0 5 4 21-30 0 0 3 1 100-499 43 48 0 0 500-999 32 31 0 0 1,000-3,000 7 5 0 0 Over 3,000 22 0 0 0 1 84 operating days. 2 3,800, 3,760. All of the total counts were within the limit (30,000) for Grade A pasteurized milk. The coli- form count exceeded 10 in 13 instances, but it is doubtful that there is any relation between the incidence of coliforms’ and the method used to clean and sanitize the equipment. There was no practical difference between total counts or coli- forms found at the beginning and at the end of runs. Acknowledgment This study and the one (8) reported in Pro- gress Report 1472, “In-place Cleaning of Steel Sanitary Milk Processing Lines,” were supported by a grant from the Dairy Products Institute of Texas, Austin, Texas. (1) (2) (3) (4) References Moore, D. R., Tracy, P. H. and Ordal, A. J. Perman- ent Pipe Lines for Dairy Plants. J. Dairy Sci., 34: 804-814. 1951. i Baker, M. P., and Fortney, C. G. The Use of Stain- less Steel for Permanent Pipe Lines. Paper pre- sented at Milk Industry Foundation Convention, De- troit, Mich., Oct. 1951. Private Communication, Lamb, N. W., The Borden Co., Houston, Texas. Milk Ordinance and Code. Public Health Service Publication No. 229. U. S. Gov’t Printing Office, Washington, D. C. 1953. (5) (6) (7) (3) Tentative Standards for Cleaned In-place Milk Lines. D.I.S.A. Task Committee. Nov. 1951. A Suggested Method for the Installation and Clean- ing of Cleaned In-place Sanitary Milk Pipe Lines for Use in Milk and Milk Products Plants. I.A.M.F.S., U.S.P.H.S. and D.I.C. Journal of Milk and Food Technology. Vol. 16, No. 2, pp. 77-78. March-April 1953. Clean In-place Pipelines. H. P. Hodes. Dealer. Vol. 42, No. 5, p. 44. Feb. 1953. Moore, A.V. In-place Cleaning of Steel Sanitary Milk Processing Lines. Texas Agr. Expt. Sta. Prog- ress Report 1472. June 20, 1952. The Milk