Document jy7wpGydp1KOQGvvEoy2pp7KZ

rtS. (Bull. Environm. Contain. Toxicol. 25, 639-645 (1980)^SoV> TfttrflChlftrnuthylpnA Pnntaminatlon nf Drinking Water \ by Vinyl-Pentad Anhestns-Pemai^ pIPa elffitiiart fi Wakeham1 Alan C Davis', Richard T WittJ. Bruce W Tripp', and Nelson M Frew' 'Department ot Chemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543. 'Department ot Public Works, Falmouth, MA 02540 Tetrachloroethylene (perchloroethylene; 1,1,2,2-tetrachloroethene) is a common solvent used in dry clearing, metal degreas ing, and chemical manufacture. Substantial quantities are produced annually (7 x 108 tons in the U.S. in 1973 (U.S. I.T.C. 1975)). Due to the relatively high volatility (b.p. 121C), solubility in water (150 mg/L), and dispersive use of this chemical, losses to the environment are inevitable. Tetrachloro ethylene has been found as a predominant volatile trace organic constituent in rivers (ZURCHER 6 GIGER 1976, STEIGL1TZ et al. 1976, DOWTY et al. 1975), lakes (GROB 6 GROB 1974, GIGER et al. 1978, SCHWARZENBACH et al. 1979a), and seawater (PEARSON 6 McCONNEL 197S, MURRAY 6 RILEY 1975, SCHWARZENBACH et al. 1979b). Traces of tetrachloroethylene have also been found in finished drinking waters (DOWTY et al. 1975, GIGER 6 MOLNAR-KUBICA 1978, REPORT TO CONGRESS 1975, SYMONDS et al. 1975). Effects on human health of exposure to tetrachloroethylene are poorly documented. However, the NATIONAL CANCER INSTITUTE (1977) has determined that ingestion may cause canceTS in mice, but not in rats. In the absence of a formal drinking water standard for tetrachloro ethylene, the U.S. EPA has suggested a no adverse response level (SNARL) of 40 ppb over a lifetime of exposure, but this level is currently under review. Recently it has been discovered that drinking water transported in vinyl-coated asbestos-cement (A-C) pipes often contains elevated concentrations of tetrachloroethylene. Vinyl-coated A-C pipe has been used in parts of the northeastern U.S. over the past decade in response to concern that waters carried in un coated A-C pipes could contain potentially hazardous asbestos fibers. Tetrachloroethylene is used as solvent during applica tion of the vinyl coating to the asbestos-cement pipe during manufacture, and residual solvent leaches into water carried in these pipes. . We Teport here results of a preliminary study to assess tetra chloroethylene contamination of drinking waters flowing in vinylcoated A-C pipes in the Town of Falmouth, Massachusetts. Data obtained from a brief survey of the Town's water distribution system and in a pipe flushing and recovery experiment are CO o o -cjr- 0007-4861/80/0025-0639 SOI.40 1980 Springer-Verlag New York Inc. presented to describe the problems encountered in trying to re duce this potential health hazard. MATERIALS AM) METHODS Water samples were collected from dead-end (a worst case situa tion since the water is replaced either only slowly or not at all) vinyl-coated A-C pipes of different dimensions, time of in stallation, and degree of usage (Table 1). Samples were obtained via bleeder valves at the downstream ends of the vinyl-coated pipes after flushing the bleeder for 1 min. One-L glass-stoppered bottles were thoroughly rinsed with the water being sampled and were completely filled and closed without leaving a headspace volume. Samples were returned to the laboratory for analysis within 24 h; any storage was at 4C. Tetrachloroethylene (and other volatile organic compounds) was purged from 50 or 100 mL of water using the closed-loop gas stripping method of GROB S ZURCHER (1976). The procedure is similar to the purge-and-trap methodology (BELLAR 6 LICHTENBERG 1974) recommended by the U.S. EPA for routine drinking water analyses, except that our system uses recycled headspace air as the purging gas and a micro-charcoal trap instead of Tenax, Prior to stripping, 1-10 ug of l-chlorohexane (in acetone) was added as an internal standard. Samples were stripped for 30 min at 40C, after which the charcoal trap was extracted with 20 yL of distilled CS2. Stripping for an additional 30 min recovered no more tetrachloroethylene and l-chlorohexane. CS2 extracts were analyzed by glass capillary gas chromatography using a flame ionization detector. Injections were made with the splitter open and the column at ambient temperature. A 40 m x 0.3 mm i.d. column coated with Pluronics 121 (GROB 5 GROB 1977) was used with hydrogen carrier gas. Concentrations of tetra chloroethylene were determined by measurement of peak heights compared to l-chlorohexane and applying the appropriate detector response factor. Tetrachloroethylene was confirmed by GC/MS. Analytical precision determined by analysis of replicate samples and subsamples at levels of 400 and 10,000 ppb was hotter than f 10%. RESULTS AND DISCUSSION Tetrachloroethylene was the most abundant volatile organic com pound detected in the water samples examined. Concentrations in waters from vinyl-coated A-C pipes at various locations in Falmouth are given in Table 1. Levels of this compound range from 140 to 18,000 ppb (yg/L) in unflushed vinyl A-C pipes. For comparison, tetrachloroethylene concentrations in other parts of the Falmouth distribution system were less than 2 ppb. Initial tetrachloroethylene levels appear to be lower in older pipes. Location 6 was a major distribution line leading from a ground 640 S T0044142 h flow ) *l! 'i'lOOlS TABLE 1. Tetrachloroethylene Concentrations (ppb) in Water Samples from Various Vinyl A-C Pipes. it 4-t d C -O O h C V> w tn V Eu (J H "e - * lo 0) t> Ifl XW J <]> 4-i C 03 0> tJQ O A 3VUE* CU Qa H C 0- ft. w S' o XX r4 0 X oo rsi f- C o *3 DO DO DO DO V G CO CX C DO CX X r C C .H t- >V) 5 .H 3U 3^ 3V 3 OX O> OX o ^ VI ^O 1-4 VI <-1 o 4-4 -- <J - 3 U c OV O^ OV 3 U, U. U. CC LL. IL, U. BC Tetrachloroethylene. ppb 1 (1977) 240/20 (9 services) 560 190 265 2 (1972) 340/1S (8 services) 170 5 na J (1976) 670/20 2,180 19 na (no services) 4 (1972) 300/20 (4 services) 140 2 na S (1975) 6 (1975) *"* 520/20 (10 services) 250/30 (no services) 2,860 1,570 s na 4 na 7 (1977) 500/20 (no services) 150 na na 8 (1979)+ 400/20 1,510 56 na (no services) 9 (1979) 364/20 18,000 na na (no services) 10 (1923)* 15 cm cast iron 106 na na 58 na na na na na na na na na na 200 na na na na na na na na * Number of active services (houses) on each 1 me;; average usage estimated at 200 gal/day/ service. ** i a x io6 flow over 5 yr period followed by no flow for 9 months. +Location of flushing/recovery study (Table 2). ^Distribution Line feeding to location 1. *n.a. = Not analyzed. 641 water well and had a flow of about 1,1 x 106 m3 (3 x 10e gal) over a 5-year period following installation. Nine months prior to this sampling, the well was closed because of ground water contamination by detergents. Despite the extensive usage, tetrachloroethylene levels had built up to 1570 ppb in the nine months after the well was closed. Leaching of tetrachlorocthylene con tinued even after additional flushing (Table 1). In general, flushing of a vinyl A-C pipe in lengths varying from 240-670 m for 24 h through a bleeder valve at rates of about 510 m3/h (2S-45 gal/min) leads to a reduction of the level of tetrachloroethylene (Table 1). Data from location 1, however, show that flushing, even for 96 h, does not always lower the level of contamination. The reason for this anomalous behavior became apparent after a closer examination of the pipe network leading to location 1, a schematic of which is shown in Figure 1A'. After four samples from location 1 repeatedly showed high amounts A S 99 I i 364 m/20 cm j ^ vinyl A-C X 1730 m/30 cm S vinyl A-C y B. 240 m/20 cm vinyl A-C 10 j < -e--------1-------------- - Z x - 2 = 3550 m cost iron y - Z - 2790 m cost iron ST0044 144 FIGURE 1. Schematic diagram (not to scale) showing pipe networks leading to sampling locations 1, 9, and 10 (A) and location 8 (B). ------ Vinyl-coated A-C; ------- cast iron. 642 of tetrachloroethylene, we analyzed a sample of water (location 10) flowing through 6S-year-old cast iron pipe leading to the vinyl A-C section and obtained a concentration of 106 ppb. An investigation of pipes upstream of location 1 showed two new sections (sections X-Y and S-T in Figure 1A; 1979 installation) jof vinyl A-C pipe that are apparently the source of these janomalous levels. Analysis of water at location 9 (Figure 1A) revealed a tetrachloroethylene concentration of 18,000 ppb (high 'enough to be detectable by odor). Even though water flowing jtowards locations 1 and 10 probably does not flow through either of these new installations of vinyl A-C pipe, a sufficient quan tity of contaminated water from these pipes apparently mixes into the mainstream flow. Clearly then, no amount of flushing at 'location 1 could ever reduce the contamination as long as tetra chloroethylene was mixing in from the contaminated connecting i>ipes located approximately 3 km upstream in the network. Since Falmouth contains some 250 sections (v 60 km) of vinyl|coated A-C pipe which would require flushing to reduce potential jhuman health hazard, it was quickly realized that insufficient water supplies were available for 24-h flushings. Therefore, we undertook a simple flushing and recovery experiment to de termine the length of time (or volume of water) required to re duce levels to an acceptable level (v 40 ppb) and the .ate at which concentrations return to elevated levels. Results for the flushing and recovery series at location 8 are given in Table 2. I i TABLE 2. Tetrachloroethylene in a Flushing and Recovery Series; Location 8 . Tetrachloroethylene, ppb Unflushed ^ 1 h flushing' 2 h flushing 6 h flushing 16 h flushing 24 h flushing 24 h recovery 48 h recovery 1,510 1,620 130 67 66 S6 3,500 2,700 * 400 m Vinyl-coated A-C pipe; 20 cm diameter. ^Flushing at v 9.8 m3/h (v 40 gal/min); 0.77 pipe volumes/h). ST004 4145 After 1 h flushing at about 9.8 m3/h (about 4S gal/min) no significant concentration change was observed. This is consis tent with the fact that the volume of the pipe (400 m x 0.2 m diameter = 12.6 ra3) was such that a minimum of 1.3 h would be needed to completely replace the initial, contaminated water in 643 the pipe. Thus by hour 2 (1.5 pipe volumes), the concentration of tetrachloroethylene had been significantly reduced. Flushing for more than 6 h did not greatly change the level, suggesting ad ditional contamination from some source upstream. After a 24 h recovery (no water flow) period, the contamination level jumped to 3,500 ppb, twice as high as the initial concen tration. We believe that the loop (Figure IB), which was not sampled, must contain contaminated water from the vinyl A-C leadin line. During the flushing, one end of the loop was isolated (valve in Figure IB closed), but the entire loop was not flushed. Only the 400 m length of vinyl A-C pipe was flushed. During the subsequent 24 h recovery, contaminated water from the loop could have mixed with "cleaner" water in the flushed section, although we cannot account for. the higher levels after recovery compared to before flushing. These results demonstrate the complexity of flushing processes. It would be necessary to flush all vinyl pipes upstream of a given vinyl section, not simply a single vinyl section, in order to eliminate the contaminant. This case study shows that vinyl-coated asbestos-cement pipe is apparently a source of widespread contamination of drinking water by tetrachloroethylene. In some cases, flushing of a vinylcoated pipe leads to a reduction of the tetrachloroethylene level in that pipe, but after flushing is terminated the level of con tamination will gradually increase. However, flushing is not a simple process merely involving replacement of water in a vinyl pipe section since mixing with contaminated water from other sections of the distribution system can lead to contamination of water in non-vinyl-coated pipes. Furthermore, the measurements at location 6 suggest that even after five years of heavy use (1.1 x 106 m3 of water), tetrachloroethylene may continue to leach out of the vinyl lining. Elevated concentrations of tetr chloroethylene were found in pipe sections installed for over eight years. Acknowledgements. We thank C. Lee and J. Whelan for comments o the manuscript. Woods Hole Oceanographic Institution Contribut Number 4650. 644 ST00H |( , 6 rt C REFERENCES BELLAR, T. A., and J. J. LICHTENBERG: J.A.W.W.A. 66, 739 (1974). DOWTY, B., D. CARLISLE, J. L. LASETER, and J. STORER: Science 187, 7S (1975). GIGER, W., and E. MOLNAR-KUBICA: Bull. Environ. Contam. Toxicol. 12, 47S (1978). GIGER, W., E. MOLNAR-KUBICA, and S. G. WAKEHAM: in "Aquatic Pollutants", Hutiinger, 0., ed. Oxford: Pergamon 1978, pp. 101-123. GROB, K., and G. GROB: J. Chromatogr. 9, 303 (1974). GROB, K., JR., and K. GROB: J. Chromatogr. 140, 257 (1977). GROB, K., and F. ZURCHER: J. Chromatogr. 117, 285 (1976). NATIONAL CANCER INSTITUTE: Bioassay of Tetrachloroethylene for Possible Carcinogenicity. Carcinogenesis Testing Division, NCI, Bethesda, Maryland (1977). MURRAY, A. J., and J. P. RILEY: Nature 24, 37 (1973). PEARSON, C. R., and G. McCONNELL: Proc. R. Soc. Lond. B. 189, 305 (1975). REPORT TO CONGRESS: Preliminary Assessment of Suspected Carcino gens in Drinking Water, Environmental Protection Agency, Washington (1975). SCHWARZENBACH, R. P., E. MOLNAR-KUBICA, W. GIGER, and S. G. WAKEHAM: Environ. Sci. Technol. 1_3, 1367 (1979a). SCHWARZENBACH, R. P., R. H. BROMUND, P. M. GSCKWEND, and 0. C. ZAFIRIOU: Org. Geochera. , 93 (1979b). STIEGLITZ, L., W. ROTH, W. kUh.N, and W. LEGER: Vom Wasser 47, 37 (1976). SYMONDS, J. M., T. A. BELLAR, J. K. CARSWELL, J. DEMASCO, K. L. KROOP, G. G. ROBECK, 0. R. SEEGER, C. J. SLOCUM, B. L. SMITH, and A. A. STEVENS: J.A.W.W.A. 63, 634 (1975). U.S. INTERNATIONAL TRADE COMMISSION: Synthetic Organic Chemicals, fi U.S. Production and Sales (1975). ZuRCHER, F., and W. GIGER: Von Wasser 7, 37 (1976). LA O -C. 645