Document JrQav1a42J5Ze0bBb7nZzDBLa

UNI - BELL PVC PIPE ASSOCIATION 2655 Villa Creek Dr., Suite 150, Dallas, Texas 75234 214/243-3902 CTL015558 VINYL CHLORIDE; The Control of Residual Vinyl Chloride Monomer in PVC Water Pipe A technical report prepared by the UNI-BELL PVC PIPE ASSOCIATION CTL015525 CTL015557 ABSTRACT Vinyl Chloride Monomer (VCM) is the basic building block in the manufacture ofpolyvinyl chloride - commonly called PVC or vinyl. The end product, polyvinyl chloride, produced from VCM in a complex organic chemical re action, is not a carcinogen. On the contrary, it is used extensively in the medical and health-care industries in many highly sensitive applications, e.g., blood bags, plasma tubing, dialysis units, etc. Polyvinyl chloride cannot breakdown or degrade in service into VCM. Since the inception of research and investigation of the carcinogenic properties of VCM, the PVC industry has executed a comprehensive and extensive program that has reduced VCM residuals to the extremely low levels which can be considered clearly safe. As an additional safeguard, beyond those provided in the controlled manufacture of PVC resin, the PVC pipe industry implemented in 1976 an extensive self-regulation program executed by the National Sanitation Foundation (NSF). Through independent third-party inspection, testing, and listing, it is required that residua! VCM in PVC potable water pipe be limited to a level which insures no detectable migration of VCM into potable water using analysis techno logy sensitive to the extremely low level of two parts per billion. (Two parts per billion can be described, as two seconds of time in 31.7 years.) A huge factor of safety has been self-imposed by the PVC pipe industry limiting human exposure to VCM in drinking water. With an understanding of the effort ex pended and cost incurred by the industry to insure that PVC pipe cannot create a health hazard, and with an under standing of the safeguards provided by NSF Standard 14, the concerned citizen can accept without fear the use of PVC potable water pipe. CTL015526 UNI-BELL PVC PIPE ASSOCIATION 2655 Villa Creek Drive, Suite 150 Dallas, Texas 75234 U. S. MEMBERS Can-Tex Industries, Inc. Capco Pipe Company, Inc. Carlon CertainTeed Corporation Clow Corporation Extrusion Technologies, Inc. Gifford-Hill & Company, Inc. Johns-Manviile Sales Corporation Robintech Incorporated Simpson Extruded Plastics Company CANADIAN MEMBERS Flex-Lox Industries Manville Canada Inc. Rehau Industries Inc. Scepter Manufacturing Company, Ltd. The statements contained in this technical report are those of the Uni-Bell PVC Pipe Association and are not warranties, nor are they intended to be warranties. Inquiries for information on specific products, their attributes and recommended uses and the manufacturer's warranty should be directed to member companies. Copyright 1981,1982 by the Uni-Bell PVC Pipe Association All rights reserved Printed in the U.S.A. UNI-TR-4-82 CTL015527 n CTL015556 27 CTL015555 TABLE OF CONTENTS ABSTRACT................................................................................................................. i INTRODUCTION...................................................................................................... 1 VINYL CHLORIDE MONOMER............................................................................. 1 Polyvinyl Chloride is not Carcinogenic.........................................................2 Residual Vinyl Chloride Monomer................................................................ 2 EVOLUTION IN PVC RESIN MANUFACTURING TECHNOLOGY................ 3 OSH A Regulations on Occupational Exposure............................................ 3 Impact of OSH A VCM Regulations..............................................................3 VCM MIGRATION INTO POTABLE WATER...................................................... 5 Vinyl Chloride Migration Basic Research...................................................... 6 Theoretical Model for VCM Migration..........................................................11 Field Research to Evaluate Water Systems using PVC Pipe...................... 15 Assessment of VCM Migration into Potable Water..................................... 16 RVCM CONTROLS FOR PUBLIC HEALTH....................................................... 18 NSF RVCM Monitoring Program................................................................. 19 Industry Control of RVCM in PVC Pipe.................................................... 20 CONCLUSION......................................................................................................... 22 BIBLIOGRAPHY.................................................................................................... 23 CTL015528 CTL015529 IV CTL015554 25 14. Maltoni, C., "Carcinogenicity Bio-Assays of Vinyl Chloride (VC): Current Results", Instituto di Oncologia, Bologna, Italy (Aug. 1974) 15. Mantell, G. J., J. T. Barr, R. K. S. Chan, "Stripping VCM from PVC Resin", Chemical Engineering Progress, (Sept. 1975) 16. McClelland, N. I., "Summary of RVCM in Pipe & Fittings by NSF Analysis", Report to NSF Industry Advisory Committee, National Sanitation Founda tion, Ann Arbor, Ml (Feb. 1978) 17. McClelland, N. I. "Summary Report on RVCM Monitoring Program", Re port to Uni-Bell Plastic Pipe Association (July 1980) 18. Prust, R. S., "Johns-Manville Research Report 404-140" Johns-Manville Corporation, Denver, CO (Mar. 1975) 19. Prust, R. S., "Johns-Manville Research Report 455-T-502", Johns-Manville Corporation, Denver, CO (Oct. 1975) 20. Prust, R. S., "Johns-Manville Research Report 455-T-511", Johns-Manville Corporation, Denver, CO (Mar. 1976) 21. Prust, R. S., "Johns-Manville Research Report 259-1", Johns-Manville Corporation, Denver, CO (April 1976) 22. Prust, R. S., "Detection, Analysis, and Migration Characteristics of RVCM in PVC Compounds and Pipe Products," Johns-Manville Corporation, Den ver, CO (1976) 23. Sachs, F. A., J. D. Banzer, "Vinyl Chloride -- Release from PVC," Technical Papers, Regional Technical Conference, New York City, Meeting (Oct. 1974) 24. "Standard No. 14 for Thermoplastic Materials, Pipe, Fittings, Valves, Traps, and Joining Materials." National Sanitation Foundation, Ann Arbor, Ml (1976)25 25. Uni-Bell Handbook of PVC Pipe: Design and Construction, Uni-Bell Plastic Pipe Association, Dallas, TX (1980) CTL015553 24 VINYL CHLORIDE; The Control of R sidual Vinyl Chlorid Monomer in PVC Water Pipe INTRODUCTION Carginogens, or substances which can induce cancer, deserve serious concern. On January 22, 1974, a major manufacturer of poly vinyl chloride (PVC) resin, the basic raw material for many PVC products including pipe, reported that three plant workers had died from hemangiosarcoma of the liver, a rare form of liver cancer. Over a period of time, careful investigation revealed that these and other workers (less than 40) who had been exposed to high concentrations of vinyl chloride gas over a period of years while working in poly merization plants had contracted the disease. The fact that the inci dence of the disease in a hugh worker population (over one million porkers worldwide employed over a period of years) was extremely Bv offers little comfort. ^ Even though the tragic effects of exposure to vinyl chloride monomer (VCM) were generally limited to polymerization workers, PVC pipe manufacturers immediately became seriously involved in this matter. Although a PVC pipe manufacturer does not produce or process VCM in his pipe plant, he must be concerned about residual vinyl chloride monomer (RVCM) which may remain unreacted in PVC resin received at the pipe plant. RVCM in PVC resin can diffuse or migrate out of the resin. The pipe manufacturer must exercise proper controls to limit exposure to VCM. He must: 1. Limit pipe-plant worker occupational exposure to VCM which could migrate into the air. This is mandated in U. S. Occupational Safety and Health Administration (OSHA) regulations.11 2. Eliminate the potential for significant migra tion of RVCM from PVC water pipe into potable water. This technical report will review and summarize the changes in technology and the stringent controls imposed to insure that PVC pipe cannot be a source of a health hazard, to insure that the con cerned citizen can accept without fear the use of PVC potable water pipe. VINYL CHLORIDE MONOMER Vinyl chloride monomer (VCM) is the basic building block in 1 manufacture of polyvinyl chloride -- commonly called PVC or vinyl.25 VCM is produced by chemically reacting ethylene and chlorine to create a monomer, a small carbon-based molecule. (See i CTL015530 Figure 1) This monomer is a volatile gas which can be polymerized to produce PVC. Polymerization is a reaction which bonds the vinyl chloride monomers into a much larger molecule or polymer known as polyvinyl chloride (PVC).25 Polyvinyl Chloride is not Carcinogenic. The end product, PVC, produced through the polymerization of VCM is not a carcinogen. On the contrary, it is used extensively in the medical and health-care industries in many highly sensitive applications, e.g., blood bags, plasma tubing, food wrap, etc. PVC is, of course, used to manu facture millions of miles of potable water pipe every year. Chemists understand that common table salt is derived through the chemical reaction of two extremely toxic substances - sodium and chlorine. Only through use of a great concentration of energy in a complex chemical reaction can table salt be reduced to its jAk: components. The average citizen is comfortable in his know^^ that the reaction cannot occur in common use. In the same sense, PVC cannot break down or degrade in service into VCM. Therefore, PVC poses no threat as a carcinogen. Residual Vinyl Chloride Monomer. Why then must PVC pipe manufacturers be concerned about VCM? In the process of poly merizing VCM in the manufacture of PVC resin, trace amounts of unreacted monomer may remain entrapped within the PVC resin particles.3 This residual vinyl chloride monomer (RVCM) immediate ly begins diffusion or migration from the PVC.4,5,22 If excessive RVCM exists in PVC resin received at a pipe plant, migration of the monomer into the plant working atmosphere could represent an occupational hazard for workers.11 If PVC pipe were manufactured with high levels of RVCM, migration of the monomer in significant levels into water conveyed by the pipe could be projected. FIGURE 1. VINYL CHLORIDE MONOMER (VCM) CTL015531 2 BIBLIOGRAPHY 1. Banzer, J. D., Senior Research Chemist, Diamond Shamrock Corp., Research Report, Dec. 8, 1975 2. Bellar, T. A., J. J. Lichtenberg, "Determination of Vinyl Chloride at g/l Level in Water by Gas Chromatography, Environmental Science and Tech nology. National Environmental Research Center, U. S. EPA, Cincinnati, OH (Sept. 1976) 3. Berens, A. R., "The Solubility of Vinyl Chloride in Poly (Vinyl Chloride)", Preprints, Polymer Division, American Chemical Society, Atlantic City Meeting (1974) 4. Berens, A. R., "Diffusion of Vinyl Chloride in Poly (Vinyl Chloride)", Pre prints, Polymer Division, American Chemical Society, Atlantic City Meeting (1974) 5. Berens, A. R., G. A. Daniels, "Prediction of Vinyl Chloride Monomer Migra tion from Rigid PVC Pipe." Polymer Engineering and Science, (Aug. 1976) 6. Collins, R. J., Director of Standards Development, National Sanitation Foundation, Letter to NSF Task Committee on VCM, June 29, 1976 7. Daniels, G. A., G. C. Gaeke, "The Diffusivity of VCM in 8237 Bottle Com pound," A Preliminary Report, Ethyl Corporation, Baton Rouge, LA (Dec. 1975) 8. Daniels, G. A., D. E. Proctor, "VCM Extraction from PVC Bottles" Modern Packaging (Apr. 1975) 9. Daniels, G. A., W. N. Parker, "Water Extraction from PVC Pipe", Ethyl Corporation, Baton Rouge, LA (Sept. 1975) 10. Dressman, R. C., E. F. McFarren, "Determination of Vinyl Chloride Migra tion from Polyvinyl Chloride Pipe into Water", Journal AWWA, American Waterworks Association, Denver, CO (Jan. 1978) 11. "Exposure to Vinyl Chloride", Occupational Safety and Health Administra tion, U. S. Dept, of Labor, Federal Register Vol. 39, No. 194, Part II, Wash ington, D. C. (Oct. 4 1974) 12. Heckman, Jerome H., "Regulatory Status of Polyvinyl Chloride, 1980". Presentation to the Third International Symposium on Polyvinyl Chloride, Cleveland, OH (Aug. 1980) 13. Mach, W. A., "VCM Reduction and Control", Chemical Engineering Pro gress (Sept. 1975) CTL015552 23 CONCLUSION Since January 1974, when vinyl chloride was identified as a hazardous material, the PVC industry, working closely with the U. S. Government, has successfully implemented sweeping changes in manufacturing technology. Today, the hazards of occupational exposure to vinyl chloride in PVC manufacturing and processing plants have been successfully controlled and limited to the extremely low levels considered safe by OSHA. Extensive research has been conducted in the laboratory and in the field to determine if the consuming public has been exposed to unsafe levels of vinyl chloride contaminant in potable water conveyed through PVC water pipe. Although laboratory research, theoretical model projections, and thorough field evaluation have demonstrated little cause for concern, the PVC pipe industry prevailed on National Sanitation Foundation in 1976 to implement a tough RVCM moni toring and control program for PVC water pipe production. The NSF certification program remains in force today and has resulted in an impressive improvement in industry control of RVCM in PVC water pipe. Although a huge factor of safety has been self-imposed by the PVC pipe industry to limit human exposure to VCM in drinking water, such extreme precautions are justified when safeguarding human health. With an understanding of the effort expended and cost incurred by the industry to insure that PVC pipe cannot create a health hazard, the concerned citizen can accept the use of PVC water pipe without fear. CTL015551 22 EVOLUTION IN PVC RESIN MANUFACTURING TECHNOLOGY The initial response to the discovery that VCM could be a car cinogen captured great attention from the manufacturers of PVC products, the plant workers and the media. Prior to the discovery of its apparent carcinogenicity, VCM had been used in many human exposure applications such as propellant in aerosol cans and surgical anesthesia with no recorded or known problems. The effort prior to that time to limit RVCM in PVC during the manufacturing process was influenced by economic pressures. Obviously, the PVC resin manufactuerers were motivated to reclaim RVCM as an effort to improve efficiency in the use of raw materials. Their efforts were limited to that extent deemed economically justifiable. The dis covery of apparent VCM carcinogenicity forced immediate changes in technology to safeguard human health.12,14 OSHA Regulations on Occupational Exposure. A major effort s launched by the United States Occupational Safety and Health Ministration (OSHA) to investigate and determine the nature and fent of the occupational hazard created by VCM in the manufac turing facilities which process either the monomer or PVC containing residual monomer. OSHA developed stringent regulations for control of occupational exposure to VCM in October 1974.11 The agency set a maximum permissible exposure limit for VCM at 1 ppm (part per million) averaged over any 8 hour period and a ceiling of 5 ppm averaged over any period not exceeding 15 minutes.11 OSHA ex empted manufacturers from the requirement to implement a number of tough monitoring and control procedures if the levels of airborne VCM were continually held below an "action level" of 0.5 ppm. Immediately, processors of PVC resin, such as pipe manufacturers, wished to comply with the "action level" and demanded PVC resin from their suppliers which contained low levels of RVCM thereby facilitating efforts to limit occupational exposure to VCM to the 0.5 ppm level assumed safe, by OSHA. It should be noted that OSHA's "action level" can be considered safe with a huge factor of safety. Countless PVC resin plant workers through the years worldwide, had been exposed to very high concen trations of VCM in their work environments, concentrations which far exceeded 1000 ppm for extended periods of time. In spite of such serious exposure to a large worker population, the tragic effects of VCM carcinogenicity were apparently limited to less than 40 cases. However, hazard to human health can justify extreme measures. The victims gained little solace from knowledge that their numbers were so few. Impact of OSHA VCM Regulations. The impact of OSHA's ^ropational exposure regulations of the PVC pipe industry has been great as demonstrated by the following data. In June 1974 new PVC 3 CTL015532 resin samples were gathered by Johns-Manville Corporation on a ran dom basis from ten PVC pipe plants. In the ten resin samples, the average RVCM was 822 ppm.21 (See Table 1) PVC resin samples were again gathered and analyzed in December 1974. As shown in Table 1, the average RVCM concentration had fallen to 349 ppm.21 TABLE 1 RVCM CONCENTRATION IN PVC RESIN 22 PVC RESIN SUPPLIER RVCM RESULTS* JUNE 1974 RVCM RESULTS* DECEMBER 1974 A 259 331 1224 873 B 779 135 197 23.2 C 189 44 D 576 1830 2060 E 670 23.5 F 1330 18.6 935, G 40.4 AVG RVCM - 822 AVG RVCM - 349 REPORTED IN PPM ON A WT/WT BASIS In June-August, 1975 investigation revealed that RVCM levels in PVC resin supplied to PVC pipe manufacturers had been further re duced. (See Table 2). Eighty-four (84) PVC resin samples taken from materials supplied by nine PVC resin manufacturers were analyzed for RVCM.21 The average RVCM concentration had fallen to 30.2 ppm.21 Note that only two of the nine manufacturers were lagging behind in improving their RVCM control technology. If those two manufacturers' samples had not been included in the investigation, the average RVCM concentration had fallen to 10.5 ppm in PVC pipe resin.21 By 1977, most PVC pipe manufacturers had established pur chasing specifications for PVC pipe resin which required RVCM levels certified at or below 10 ppm. To meet these requiremen^tfie resin manufacturers dedicated substantial investment and effcflBo the development and implementation of sophisticated RVCM "strip ping" technology in their operations. Today, PVC pipe manufacturers have little difficulty limiting occupational VCM exposure to levels far lower than the 0.5 ppm "action level" established by OSHA. The 4 CTL015533 proper RVCM control programs in production of PVC water pipe experienced some problems passing NSF inspection and testing. Table 11 demonstrates the progress achieved in the NSF RVCM monitoring program incorporated into the Standard 14 PVC pipe certification program. The full impact of the NSF certification program on control of RVCM in PVC water pipe became apparent as early as 1978. Table 12 demonstrates the improvement in RVCM control achieved by PVC pipe manufacturers over the span of one year. It is reasonable to make the following assumptions based on Tables 11 and 12: At least 98% of the PVC pipe manufactured today in North America contains less than 10 ppm RVCM. Most of the PVC pipe manufactured today in North America contains less than 1 ppm RVCM. YEAR 1977 1978 1979 1980' TABLE 11 NSF RVCM MONITORING RESULTS17 NO. SAMPLES TESTED PASS' (%) NON-DETECTABLE (ND)1 (%) 412 423 664 248 93.7 97.4 97.0 98.4 38.1 52.7 46.1 41.1 'Based on MCL^JIO ppm in PVC pipe or fittings (reference: NSF Standard 14); all failed samples are retested and enforcement procedures followed if required. Retest data are not included in these figures. 1 For routine analyses, RVCM <C0.5 ppm is reported as ND. 'Samples through June 30, 1980. TABLE 12 IMPROVEMENT IN RVCM CONTROL NSF ANALYSIS OF RVCM LEVEL IN PVC WATER PIPE16 RVCM NUMBER 5 PPMCKVCM^IO PPM 50 PPU<KVCNK5 PPM MONITORING OF TIME PERIOD SAMPLES NO. % NO. % RVtM^SO PPU NO. % Feb. thru Nov 1977 Dec thru Jan 1978 332 19 80 5 5.7 185 55.7 106 31.9 6.3 20 25.0 51 63.8 CTL015550 21 GRAPH 4 VCM CONCENTRATION IN WATER DUE TO DIFFUSION FROM PVC PIPE SYSTEMS Age of Pipe, log scale (years) the PVC pipe industry limiting human exposure to vinyl chloride in drinking water. In August 1976, the NSF implemented the recommendations on RVCM limits for PVC water pipe and fittings. Since that time, manufacturers of PVC water pipe and fittings who place the NSF seal on their products have been subject to rigorous product testing and unannounced inspections by NSF to insure that PVC water pipe and fittings contain less than 10 ppm RVCM.24 To facilitate compliance with the NSF Standard 14, most PVC pipe manufacturers require that their PVC resin be supplied with RVCM levels lower than 10 ppm. By the time the PVC resin has been processed through com pounding and extrusion into PVC pipe, the cumulative, repeated application of heat in both manufacturing processes has driven off in most operations more than half of the VCM. The end effect of the self-imposed RVCM limit has resulted in an even greater factor of safety than originally considered possible. Industry Control of RVCM in PVC Pipe. Of course, a properly managed laboratory certification program must be tough. Manufac turers who did not immediately succeed in the implementation of 20 CTL015549 TABLE 2 RVCM CONCENTRATION IN PVC RESIN19 (DURING JUNE, JULY, AUGUST, 1975) PVC RESIN SUPPLIER NUMBER OF SAMPLES ANALYZED AVERAGE RVCM LEVEL (PPM) R 5 3.3 S 16 59.9 T 15 14.2 U 1 3.5 V 1 5.4 w 21 12.4 X 10 97.6 Y 8 8.7 z 7 5.7 TOTALS 84 30.2 evolution in RVCM control technology in the manufacture of PVC resin not only had impact on occupational safety in PVC pipe plants, but it also had great impact on PVC pipe product safety. VCM MIGRATION INTO POTABLE WATER While VCM processors, PVC resin manufacturers, and the federal government were martialing efforts to solve the problem of occupa tional exposure to VCM, PVC pipe manufacturers not only were seriously concerned over occupational exposure, but they were im mediately concerned over product safety and consumer health. The PVC pipe industry launched a "crash" program to obtain answers to vital questions: How much RVCM is in PVC water pipe? What is a safe level of RVCM in PVC water pipe? How much RVCM diffuses or migrates from PVC pipe into potable water? What controls are necessary to safe guard public health? Although there was no experimental evidence regarding the carcino genicity of VCM when ingested (as opposed to carcinogenicity when inhaled), in the interest of safety, the PVC pipe industry reacted based on the assumption that VCM could be a carcinogen when in gested. This assumption demanded an emergency reaction to deter mine whether or not vinyl chloride could be migrating in significant ntities into potable water conveyed in PVC pipe. An obvious assumption was made that the level of vinyl chloride migration from PVC pipe depended upon the level of RVCM in the wall of the PVC pipe. Therefore, research and investigation was 5 CTL015534 launched to determine the levels of RVCM remaining in PVC after extrusion into pipe.5 This was no simple project. The industry first had to develop a new method of analysis sufficiently sensitive to detect minute traces of RVCM in PVC pipe. The method developed employed detection by flame-ionization gas chromatography with identification by mass spectrometry. Vinyl Chloride Migration Basic Research. In the summer of 1975, the PVC pipe industry launched initial investigation and analy sis. Table 3 presents RVCM data on pipe produced from PVC resin TABLE 3 RVCM CONCENTRATION IN PVC PIPE SAMPLES19 (DURING SUMMER, 1975) PLANT DESIGNATION PIPE SIZE NOMINAL DIAMETER (IN.) RVCM IN PVC RESIN (PPM) RVCM IN PVC PIPE (PPM) m %vcmjf LOST A* 3 14 or 26 4.7 82 or 66 A 6 14 or 26 2.7 90 or 81 A 8 14 or 26 2.7 90 or 81 A 10 14 or 26 2.9 89 or 79 B 2.5 8.2 2.0 76 B 2.5 8.2 3.4 59 B 2.5 8.2 2.0 76 B 1.5 8.2 3.5 57 C 6 20.3 3.4 83 C 6 20.3 3.4 83 C 6 27.7 3.5 82 C 6 27.7 2.8 86 O 3 5.0 1.9 62 D 3 5.0 2.0 60 D 3 5.0 1.7 66 D 3 5.0 1.1 78 E 4 13.1 2.3 82 E 4 13.1 5.2 60 E 4 13.1 2.8 79 E 4 13.1 1.6 88 F 3 N/A 3.2 _ F 4 N/A 2.4 F 8 N/A 1.7 AVERAGE VALUES 2.7 76 Note: In plant A, tvwo PVC resins were used in manufacture of pipe. Pipe could have been produced from either resin. 6 CTL015535 sophisticated landmark research based on the "worst case" assump tion that VCM could be carcinogenic when ingested as a contaminant in water. The basic research in a number of laboratories, the theo retical model, and the field test site research strongly supported the assumption that RVCM in PVC water pipe has not created a signifi cant health hazard.10,21 However, it is entirely reasonable when con sidering any potential hazard to human health to select a very large factor of safety. In keeping with this commonly accepted philosophy, the PVC pipe industry did not consider the matter closed upon completing research which clearly demonstrated little cause for concern. Al though the U. S. Government did not deem it necessary to regulate RVCM in PVC pipe or to monitor or regulate the use of PVC water pipe, the industry believed strongly that additional safeguards were necessary to increase further the factor of safety guarding consumer health. NSF RVCM Monitoring Program. On August 19, 1976, re presentatives of the PVC resin and pipe industries serving on the National Sanitation Foundation (NSF) Industry Advisory Committee recommended to the NSF that Standard 14 be revised to incorporate maximum permissible limits for RVCM in PVC potable water pipe. Nationally recognized product standards for PVC potable water pipe manufactured in North America specify that the pipe must be pro duced to meet the requirements defined in NSF Standard 14. The Industry Advisory Committee, working closely with the Uni-Bell Plastic Pipe Association and the Plastics Pipe Institute, recommended that "the maximum limit shall be 10 ppm Residual Vinyl Chloride Monomer" in PVC potable water pipe and fittings. The maximum limit of 10 ppm RVCM was established based on the theoretical model for RVCM diffusion, the laboratory research, and the field test site research. A limit of ten (10) ppm RVCM was selected based on the conclusion that vinyl chloride migration from the wall of nominal one-inch diameter Schedule 40 (ASTM D 1785) PVC pipe (containing 10 ppm RVCM) into stagnant water retained in the pipe will be undetectable at a 2.0 ppb sensitivity in analysis.22 Therefore, migration of VCM in larger diameter PVC pipes would be even lower and vastly lower yet if the PVC pipe is conveying flowing potable water. Graph 4 predicts the potential migration of vinyl chloride into potable water conveyed by nominal 6-inch diameter pipe containing 10 ppm RVCM. The predictions are based on the theoretical model. The graph demonstrates that even in extreme conditions where water residence time is 23 days in a nominal 6-inch PVC pipe containing 10 ppm RVCM, any vinyl chloride present in the water should be less than one (1) ppb.9 A huge factor of safety has been self-imposed by t9 CTL015548 developed could not be verified by a mass spectrometer, the effort by USEPA was thorough and impressive. The data must be consid ered valuable to the overall investigation of RVCM. The lower detec tion limit reported by USEPA in the analysis was 30 parts per trillion (ppt). The highest concentration detected was found at the George town, Texas test site. That concentration was reported to be 1.4 ppb.10 It is obvious that the levels of vinyl chloride detected in field test water samples were extremely low by comparison with labora tory test data. The difference is readily explained. Laboratory ana lysis was performed using water samples confined in a stagnant con dition for substantial periods of time. The dilution which naturally occurs in a typical water distribution system is considerable. It should be noted that water demand in the Coolidge, Arizona test site was exceptionally low. USEPA data reported 0.3 ppb VCM in the water collected at that site; however, they declared the analytical results spurious in that they also detected the same concentration of VCM in the source water. In additional tests, USEPA was unable to detect VCM in that distribution system even though flow rates were extremely low.10 With the evolution in vinyl chloride control technology used by the PVC resin manufacturers, it is reasonable to assume that vinyl chloride migration from PVC water pipe manufactured today in North America must be substantially lower than the levels reported in the field research. R. C. Dressman and E. F. McFarren reported in their summary of the USEPA research: "The concentration of VC in the newest, longest system tested (Georgetown) was low compared with the level deemed hazardous to animals. Manufacturers claim that pipe made for water use within the past year has a lower free vinyl chloride content than pipe manufactured earlier. Large factors of safety therefore should be in herent in the use of the pipe."10 R. C. Dressman and E. F. McFarren further elaborated on the levels of vinyl chloride they detected in potable water samples: "These levels are well below those that thus far have been associated with adverse effects in experi mental animals, and these levels should decline as lower residual monomer pipe is put into use."10 RVCM CONTROLS FOR PUBLIC HEALTH The PVC industry and the U. S. Government had crossed new frontiers in trace contaminant analytical technology. The industry and the U. S. Government had conducted a substantial quantity of 18 CTL015547 with RVCM levels evaluated as shown. Twenty-three (23) PVC pipe samples, selected at random from the production of six (6) pipe plants, were analyzed as shown in Table 3. The average RVCM con centration in the PVC pipe walls was 2.7 ppm.21 It should be noted that the heat of extrusion resulted in a substantial reduction in RVCM. It was determined that seventy-six percent (76%) average of the RVCM present in the PVC resin was driven off during extrusion of pipe. In general, the diffusion or migration of an entrapped volatile gas from a solid material increases with increased temperature.4,7 This represents a proper description of the process of RVCM migra tion from PVC.22 Although in 1975 the average RVCM levels in PVC resin still exceeded 30 ppm, the average levels of RVCM in PVC pipe, in most instances, were less than 5 ppm. In April 1976, the National Sanitation Foundation (NSF) initia major investigation to determine the level of RVCM occuring VC water pipe manufactured that year. The data obtained in investigation is shown in Table 4. Two hundred and forty-six (246) PVC pipe samples obtained from 68 PVC pipe plants were analyzed for RVCM content. Eighty-nine and four-tenths percent (89.4%) of the PVC pipe samples analyzed contained RVCM at levels lower than 10 ppm. TABLE 4 RVCM CONCENTRATION IN PVC PIPE AND FITTINGS (STUDY CONDUCTED DURING APRIL, 1976) NUMBER OF SAMPLES ANALYZED RVCM <10 PPM NO. % ! 0 PPM < RVCM <20 PPM 20 PPM <RVCM<50 PPM NO. % NO. % RVCM >50 PPM NO. % PVC Pipe- 246 220 89.4 PVC Fittings43 21 48.8 18 7.3 4 9.3 1 2.8 7 16.3 1 0.4 11 25.6 289 SAMPLES PARTICIPATED (ABOUT 15% FITTINGS) 68 PVC PIPING FABRICATOR PLANTS PARTICIPATED Having established a clear understanding of the RVCM levels ally found in PVC water pipe manufactured in 1976, the indus- * efforts then had to be directed toward research to define the level" of RVCM in water pipe. This research required careful investigation of the RVCM diffusion process from a solid PVC 7 CTL015536 matrix. The relationship between the level of RVCM present in a PVC pipe wall and the level of migration of vinyl chloride into water conveyed in the pipe had to be established. A series of laboratory research projects was launched by the industry without delay to define the rate of RVCM diffusion or mi gration from the walls of PVC water pipe. As the protocols for these research projects were developed, the PVC industry worked closely with the U. S. Food and Drug Administration (FDA) and the U. S. Environmental Protection Agency (USEPA). Research on RVCM dif fusion from PVC pipe was conducted by the following organizations: B. F. Goodrich Chemical Company Diamond Shamrock Corporation Ethyl Corporation Johns-Manville Corporation USEPA (Cincinnati Research Laboratory) Tests were conducted to evaluate monomer migration from new"and old PVC pipe. The effects of pipe size, temperature, water pressure and water quality variation were evaluated. Research again required the development of new analytical technology. Substantial effort and investment were dedicated to development of analytical techniques and apparatus capable of detecting the presence of vinyl chloride in water with a sensitivity as low as 2 ppb (parts per billion). To dem onstrate the extremely minute level of this sensitivity to trace materials, it could be helpful to quantify 1 ppb. One part per billion represents approximately one minute taken from the entire passage of time since the Emperor Nero persecuted the early Christians in ancient Rome. Thirteen years ago when mankind had reached the level of technological advancement which made possible a voyage to the moon, scientists were inspired by the "incredible" sensitivity of "new" analytical techniques which could find traces of materials as small as 50 ppm (parts per million). This concentration is 50,000 times greater than 1 ppb (one part per billion) which can now be de tected with conventional instrumentation. It should be noted that the research on RVCM reported by USEPA was conducted with instrumentation which provided levels of sensitivity in the range of 30 parts per trillion.2,10 USEPA used gas chromatography with a microcoulometic titration detector. Un fortunately, at the incredibly low levels of detection employed by USEPA scientists, the contaminants detected could not be conclu sively labeled with accuracy by mass spectrometry.21 Howevei^^ie USEPA data is valuable. The research results revealed that migration of vinyl chloride from a solid PVC matrix does occur and is predictable.4,5,9 Table 5 presents data obtained in research by Johns-Manville Corporation. The vinyl chloride detected in water is reported in ppm for a pipe 8 CTL015537 field sites and tested at the Johns-Manville Research Center. With analytical technology sensitive to 2 ppb, no vinyl chloride was de tected in water conveyed through PVC pipe.19 TABLE 9 IN-FIELD VCM WATER SURVEY20 COMMUNITY AMOUNT OF PVC PIPE (FT) SAMPLING POINT VCM DETECTED IN THE WATER (PPM) COOLIDGE, AZ 6280 BEGINNING OF PVC NON-DETECTABLE* (INSTALLED 1964) END OF PVC NON-DETECTABLE* PIONEER, CA 81900 BEGINNING OF PVC NON-DETECTABLE* (INSTALLED 1966) END OF PVC NON-DETECTABLE* ROSEBURG, OR 18480 (INSTALLED 1966-67) BEGINNING OF PVC NON-DETECTABLE* END OF PVC NON-DETECTABLE* GEORGETOWN, TX 66000 (INSTALLED FEB 1975) BEGINNING OF PVC NON-DETECTABLE* END OF PVC NON-DETECTABLE* SALADO,TX 2800 BEGINNING OF PVC NON-DETECTABLE* (INSTALLED 1969) END OF PVC NON-DETECTABLE* *non-detectable at 0.002 ppm Table 10 presents data obtained in analysis of duplicated water samples taken and tested by USEPA. Although the analytical data TABLE 10 PVC PIPE SYSTEMS STUDIED10 Parameter WATER UTILITY Coolidge, George Pioneer, Roseburg, Ariz. town, Tx. Calif. Ore. Salado, Tx. Year pipe manufactured Pipe Length-Arm Length-m; Size-in. Total wall area-m1 Total wall area-sg ft Water temperature-C Sampling point Source Well head Treatment plant End of pipe Sampling date VC concentration*- tig/I Sampling date VC concentration*- fig/I ~1964 1975 ~1966 1966-67 <1 39 km. 1971) ~1968 1.6 1.1 4.8 15 7.6 3.7 1.0 0.69 3.0 9.5 4.7 2.3 3 2 6&8 4&5 6 4 560 8200 4800 6000 88000 52000 ~20 22-25 22-24 4.94 0.68 3.07 0.42 32 1300 14000 14 0.18 0.48 0.18 0.11 0.30 0.11 862 370 4000 21-23 X X 8/22/75 0.3t 0.3t 9/22/75 ** X X 10/10/75 1.4 10/28/75 1.2 X X 9/16/75 0.06 - X X 10/17/75 0.03 -- X X 10/10/75 ~* * 10/28/75 --** Average of triplicate analyses, corrected for recovery efficiency tProbably spurious None detected at a detection limit of 0.03 pg// C__T_L_0_1_5_5_4_6 17 FIGURE 2 VCM IN Ufi WITH HEAVY IMPURITIES IMPURITIES TIME (Minutes) inants can clutter or mask analysis. Figure 2 shows a typical gas chromatograph response in test of a field sample of water. In fact, several trace contaminants are extremely difficult to discern from vinyl chloride with gas chromatography. Mass spectrometry is re quired to identify the trace contaminant detected by the chromato graph. However, a mass spectrometer cannot analyze with practical reliability at levels of concentration lower than 2 ppb. It should be noted that test samples at each field site were collected from water flowing through the PVC pipe and from source water prior to conveyance through PVC pipe.10 USEPA reported vinyl chloride in the source water at one test site. Although the va lidity of this data has been questioned, it must be emphasized that when searching for 30 parts per trillion in analysis of trace contam inants the unexpected should be expected. Thirty (30) parts per tril lion can be described as one second of time in about 1,056 years. Assessment of VCM Migration into Potable Water. Table 9 pre sents the data obtained in analysis of water samples collected at the 16 CTL015545 SAMPLE 1 2 3 4 5 6 7 8 9 Aio 12 13 14 15 16 17 18 19 TABLE 5 RVCM DIFFUSION FROM PVC PIPE22 AGE OF PIPE (DAYS) DIFFUSION PERIOD (DAYS) RVCM IN PIPE (PPM) VCM DETECTED IN 11,0 FOR S/V = 1.0/IN. (PPM) 250 250 250 230 180 220 180 240 300 300 280 300 240 240 180 270 300 180 180 7 <1 7 <1 7 <1 7 3.8 7 3.9 7 26 7 29.2 7 32 7 52 7 70 7 90 7 101 7 131 7 151 7 179 7 294 7 300 7 307 7 307 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.0015 < 0.001 0.001 0.0022 0.0019 0.0038 0.0038 0.0051 0.0050 0.5063 0.0087 0.0099 0.0095 0.0099 NOTE: 0.001 ppm = 1.0 ppb R. S. Prust Data surface to water volume ratio (s/v) of 1.0 in2/in3. The data indicated that the migration of vinyl chloride from PVC pipe into water is directly proportional to the s/v ratio. A s/v ratio of 1.0 in2/in3 ap proximates the conditions offered in nominal 4 inch PVC pipe. In nominal 12 inch pipe, the s/v ratio can be approximated at 0.33. The data also indicated that the rate of monomer migration was directly proportional to time of exposure. It should be noted that the JohnsManville testing simulated a condition wherein water is retained for seven (7) days in a PVC water main. In Johns-Manville Corporation's initial research, nineteen (19) PVC pipe samples filled with distilled water were analyzed. The data revealed two vital items of informa tion: The levels of vinyl chloride detected in the water were directly proportional to levels of RVCM detected in the PVC pipe wall.22 When RVCM levels in the PVC pipe wall were less than 10 ppm, migration of vinyl chloride 9 CTL01553Q into the water over a period of seven days was less than 1 ppb.20 Table 6 presents the data on vinyl chloride migration generated in research by Diamond Shamrock Corporation. In this research nomi nal 1 inch diameter PVC pipe (s/v - 4.1 in2/in3) was tested. Again, the levels of vinyl chloride detected in stagnant water were extremely low.1,21 It should be noted that selected samples of PVC pipe with extremely high levels (greater than 100 ppm) of RVCM were ana lyzed.1 TABLE 6 RVCM DIFFUSION FROM PVC PIPE1 SAMPLE AGE OF PIPE (DAYS) DIFFUSION PERIOD (DAYS) RVCM IN PIPE (PPM) VCM DETECTED]^ H,0 FOR S/V = 4.1^^ (PPM) 1 86 2 86 3 111 4 72 57 67 77 89 99 10 9 11 56 12 56 13 56 14 99 15 95 16 127 17 77 18 116 19 116 1 2.5 1 28 1 179 1 283 2 <1 2 14.8 2 34.8 2 <1 2 14.8 2 34.8 2 <1 2 14.8 2 34.8 3 2.5 3 28 3 179 3 283 150 2.5 180 2.5 < 0.001 < 0.001 0.0068 0.021 < 0.001 0.0032 0.0070 < 0.001 0.0026 0.0042 < 0.001 0.0018 0.0066 < 0.001 0.0022 0.0130 0.0410 < 0.001 < 0.001 J. D. Banzer Data Table 7 presents the data on vinyl chloride migration generated in research by B. F. Goodrich Chemical Company. This researd^as also conducted using nominal 1 inch diameter PVC pipe con^Bng extremely high levels of RVCM.5 The correlation of data between the three research projects ap peared excellent. The data generated supported the conclusion that the level of trace vinyl chloride migration from PVC pipe walls is 10 CTL015539 Field Research to Evaluate Water Systems using PVC Pipe. When considering public health significance, a sense of security is not con clusively supported by laboratory experiments and theoretical models however sophisticated. The PVC pipe industry, at this point, was pre pared and had to move its research into the field. The vital question remained unanswered. Was the consuming public exposed to signifi cant levels of vinyl chloride contaminant in potable water conveyed in PVC water pipe? The PVC pipe industry and the USEPA worked closely in this phase of research and investigation. The vital question could be answered only through careful investigation of actual potable water distribution systems in service which conveyed water through PVC pipe. Five field sites were selected for analysis by USEPA. In co operation with USEPA, Johns-Manville Corporation also tested repli cate water samples collected at the field sites. The field test sites were selected to represent relative extremes in climatic conditions. The sites were also selected because specific information was available concerning the age, length and size of the PVC pipe used.10 The specific RVCM levels in the PVC pipe were not analyzed. However, it was established that all of the pipe installed in the distribution systems was manufactured before February 1975. It is reasonable to assume that the original RVCM concentration in the walls of the PVC pipe could vary from 10 ppm to 200 ppm. The logistics and complications involved with this phase of the RVCM research were considerably more taxing that those experi enced in the previous research conducted in controlled laboratory conditions. Vinyl chloride, as volatile gas, is elusive and difficult to analyze in water samples. The moment that the water is exposed to open atmosphere, any vinyl chloride present in the water commences diffusion out of the water into the air. Extreme caution was neces sary to collect representative samples of potable water from the field sites. The water samples were collected in sealed serum reaction bottles filled without air space. The samples were then refrigerated during shipment and in the laboratory. Because heat increases the volatility of VCM, the samples were warmed to room temperature only when ready for immediate analysis.10 In controlled laboratory conditions, gas chromatography could provide analysis of VCM in distilled water with a sensitivity of 1 ppb and lower. However, when analyzing field water samples, a large number of trace contaminants are present and render analytical pro cedures vastly more complicated. In consequence, the PVC industry selected 2 parts-per-billion (2 ppb) as a reasonable detection level. USEPA used analytical procedures reported to be sensitive to about 30 parts per trillion (30 ppt);10 however, when analyzing at these extremely low levels of sensitivity a great number of trace contam- 15 CTL015544 E X P E R IM E N TA L RESULTS |pcm | GRAPH 2 THEORETICAL VS. EXPERIMENTAL* RVCM DIFFUSION RESULTS FOR RESULTS REPORTED IN TABLE 6 22 GRAPH 3 THEORETICAL VS. EXPERIMENTAL* RVCM DIFFUSION RESULTS FOR RESULTS REPORTED IN TABLE 722 14 CTL015543 TABLE 7 RVCM DIFFUSION FROM PVC PIPE5 SAMPLE AGE OF PIPE (MONTHS) DIFFUSION PERIOD (DAYS) RVCM IN PIPE (PPM) VCM DETECTED IN H.O FOR S/V = 4.1/IN. (PPM) 1 2 3 4 5 6 7 8 9 ^10 ~6 ~6 ~6 ~6 ~6 ~6 ~6 6 ~12 3 22 7 22 14 22 3 177 7 177 14 177 3 292 7 292 14 292 14 29 0.0006 0.0022 0.0046 0.0173 0.0335 0.056 0.021 0.0414 0.113 0.0105 A. R. Berens, G. A. Daniels Data extremely low compared to RVCM levels in the PVC pipe walls. The data further indicated that vinyl chloride migration levels into water can be mathematically predicted with a reasonable statistical level of confidence.21 Theoretical Model for VCM Migration. At this point in research and investigation by the PVC industry, effort was dedicated to the development of a reliable theoretical model with which vinyl chlo ride migration from PVC pipe (containing RVCM) could be quanti tatively predicted with acceptable accuracy. Research conducted by the organizations listed above plus others provided extensive data which supported the following theore tical assumptions: 1. The diffusion or migration of vinyl chloride from PVC (containing RVCM) obeys Fick's First Law.22 2. The diffusion coefficient is independent of RVCM concentration.22 3. The diffusion of RVCM in the PVC pipe wall is the rate controlling step.22 The concentration of vinyl chloride in the water is negligible compared to the RVCM concentration in the pipe wall.22 Fick's First Law can be stated: The amount of diffusion of a given material across ii CTL015540 a unit plane area in a unit time span is directly pro portional to the concentration gradient across the plane. Based on the above assumptions supported by research data, the following theoretical model of vinyl chloride diffusion from PVC pipe into conveyed water was developed:21 Where: Mh q = Total diffused vinyl chloride into water, g S = Pipe surface area exposed to water, cm2 Mt = Average RVCM concentration in the PVC pipe wall, g/cm3 D = Diffusivity of vinyl chloride in PVC, 5.7 x 10*8cm2/day* t = Diffusion time period, days t = Pipe age before start of diffusion time period, days * Research by G. A. Daniels supports use of this constant. 9 Table 8 presents actual diffusion levels compared with the levels predicted by the theoretical model. Graph 1 provides a plot of these results. Graph 2 and Graph 3 provide comparison of the Diamond Shamrock and B. F. Goodrich data respectively with theoret^^ly predicted values. When considering the extremely minute lev^Hbf vinyl chloride being detected, the correlation between laboratory data and theoretical predictions was excellent. Research and investi gation today continues to support the reasonable reliability of the theoretical model. 12 CTL015541 TABLE 8 EXPERIMENTAL VS. THEORETICAL RVCM DIFFUSION RESULTS22 VCM DETECTED IN WATER (PPM) SAMPLE i 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 EXPERIMENTAL < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.0015 < 0.001 0.0010 0.0022 0.0019 0.0038 0.0033 0.0051 0.0050 0.0063 0.0087 0.0099 0.0095 0.0099 TIIEORiiTICAL < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.0011 0.0013 0.0013 0.0019 0.0025 0.0035 0.0037 0.0055 0.0061 0.0080 0.0096 0.0113 0.0137 0.0137 GRAPH 1 THEORETICAL VS. EXPERIMENTAL E X P E R IM E N TA L RESULTS (ppm ) 13 CTL015542 1989 ADDENDUM/UPDATE TO: "VINYL CHLORIDE; THE CONTROL OF RESIDUAL VINYL CHLORIDE MONOMER IN PVC WATER PIPE" (UNI-TR-4) Since this Uni-Bell technical report was last printed, i.e., 1982, the U.S. Environmental Protection Agency (EPA) has adopted national drinking-water limits for at least eight volatile organic chemicals (VOC's), including vinyl chloride. The final maximum contaminant level (MCL) for vinyl chloride was set at 2.0 parts per billion on June 24, 1987. The level of vinyl chloride imparted to water conveyed through PVC pipes manufactured in North America is extremely low. In fact, the levels are so low that detection is usually not possible. Consequently, vinyl chloride levels, i.e., residual vinyl chloride monomer (RVCM), continues to be monitored in the pipe wall rather than the extractant water. In North America, RVCM levels in PVC pipe materials are monitored by both the PVC resin manufacturer and the National Sanitation Foundation (NSF). The vast majority of PVC pipe manufactured in North America contain less than 1 part per million RVCM. At such low RVCM levels, it is extremely difficult to detect any vinyl chloride in the extractant water, even after long retention times. The equation provided on page 12 may be used to predict the extractant water concentration where actual measurement is not technologically possible. At the RVCM level of 1.0 part per million in the pipe wall, the predicted extractant water concentration will be less than 0.1 part per billion. In summary, PVC pipes manufactured throughout North America provide a huge factory of safety relative to the vinyl chloride limit for drinking water. The initial vinyl chloride extractant levels are typically less than one twentieth (1/20) of EPA's allowable 2.0 parts per billion limit. Furthermore, the vinyl chloride extractant levels continually diminish with time. Thus, PVC potable water pipe can be used with confidence. The PVC pipe industry is committed to public health and safety. UNI-TR-4-89 CTLOI5559