DUXBURY WATER DISTRIBUTION
DUXBURY, PLYMOUTH COUNTY, MASSACHUSETTS
Measured PCE concentrations in the Duxbury water distribution system
In order to evaluate the potential for adverse health effects among Duxbury residents who live on streets with VLAC pipe, MDPH first compiled all the available testing results for PCE in the Duxbury water distribution system to determine the opportunities for exposure that may have existed in the past. Records of PCE test results were received from the Duxbury Water Department, the Duxbury Board of Health, MDEP, and the Duxbury Ad-Hoc Committee on Water Quality (see Appendix B for a list of sources). All the results were compiled in a central database and cross-checked for consistency. Discrepancies between the different sources were resolved by consulting the laboratory datasheet (if available), or through discussions with MDEP or the Duxbury Water Department.
According to the records supplied to MDPH, four hundred and thirty three (433) water samples from the Duxbury water distribution system were tested for PCE between 1980 and June 1, 1997 (Table 1). Nearly 80% of these tests were performed in 1997. The average of detected concentrations of PCE in all the samples was 6.26 ppb (PCE was detected in 260 of the 433 samples2). The observed range of concentrations was from below detection to 180 ppb.
The PCE concentrations in VLAC pipes should decrease over time as more of the PCE is depleted from the lining and should be highest during seasons of low water demand (e.g., during the winter) (DEQE, 1980; Larsen et al, 1983). For the Duxbury water distribution system, it appears that the earlier test results (taken between 1980 and 1996) are slightly higher than the 1997 test results (Table 1). However, it is unclear whether this difference is due to depletion of PCE in the vinyl lining or due to the implementation of remedial measures (e.g., installation of bleeders). There is not much difference between the test results taken during the summer and those taken during the winter (Table 2). These observations do not change when samples from areas with known active bleeders are excluded. Therefore, the test results do not suggest that the period of exposure to PCE for Duxbury residents can be constrained by the age of the VLAC pipe or by the seasons.
The Town of Duxbury installed bleeders and looped water mains on several streets with VLAC pipe to increase the water flow rate and, hence, decrease the PCE concentrations in the pipes. The testing results show that these remedial actions have been effective at mitigating PCE exposure (Table 3). The average of detected PCE concentrations (if samples in which PCE was not detected are excluded) for dead end streets with VLAC pipe but not bleeders is 7.91 ppb, for dead end streets with VLAC and bleeders is 2.48 ppb, and for streets with VLAC pipe with looped water mains is 1.83 ppb. Therefore, people living on streets with bleeders or with looped water mains should have had reduced opportunities for exposure to PCE from the time since the remedial action was implemented. The test results and configuration (e.g., dead end, looped, etc.) for each street are summarized in Table 4.
Despite the number of test results for PCE from the Duxbury water distribution system, it is unclear whether these data are representative of historical conditions in the Duxbury water distribution system. First, there were several long periods (1969-1979, 1987-1988, 1991-1994) when no locations were tested for PCE. Second, some reports suggest that the pipes were flushed before being sampled on some occasions which would cause the PCE concentrations to be underestimated (Leach and Blatterman, 1997). Third, the majority of the samples (nearly 80%) were collected in 1997, and 102 of the 337 samples from 1997 were collected after remedial measures (e.g., bleeders) were implemented. Fourth, the dates and locations for some early PCE tests and the dates when the original bleeders were installed were not documented. Because of this, it was necessary for MDPH to make some assumptions and estimations during the compilation of the testing results in order to address questions about opportunities for exposure to PCE in the past. Therefore, the PCE data compiled by MDPH cannot be extrapolated to predict with surety what the PCE concentrations were during the entire period of potential exposure (1969 to 1997). However, the available data are useful for documenting current (i.e., 1997) conditions in the Duxbury water distribution system.
Measured PCE concentrations in other water distribution systems in Massachusetts
Sampling by MDEP in 23 communities in 1980 found that the range of PCE concentrations in dead end or low-flow VLAC pipes was between several hundred and several thousand parts per billion (DEQE, 1980). This provides a rough estimate of the maximum concentrations that were possible in Duxbury in the late 1970s and early 1980s. Because the vinyl lining was applied by hand during the manufacture of VLAC pipe (i.e., the process was not automated), the thickness of the liner and, hence, the potential for PCE to leach into the water can vary considerably between different sections of pipe. The potential for PCE leaching is also dependent on the amount of time that the vinyl lining on the section of pipe was allowed to cure (Larsen et al., 1983). Therefore, sampling data from different towns can only provide limited information about the conditions that might have existed in Duxbury.
Estimated range of PCE concentrations in the Duxbury water distribution system
The period over which some Duxbury residents may have had opportunities for exposure to PCE from VLAC pipes is approximately 30 years (1969 to 1997). The available testing results from the Duxbury water distribution system are likely to be representative of the PCE concentrations in the system in 1997. However, the available data are not sufficient to accurately predict what the PCE concentrations would have been during the 1970s and 1980s. The results of MDEP sampling in Massachusetts towns in 1980 gives a rough estimate of the highest concentrations that could have been found in Duxbury during the earlier parts of the exposure period, but concentrations of this magnitude are not likely to have persisted for a long period of time because of water use patterns. Therefore, while the PCE concentration in VLAC pipes in Duxbury is known to have ranged from below detection to 180 ppb, the concentrations may have ranged considerably higher under certain, worst-case conditions. The highest reported PCE concentration in VLAC pipe in Massachusetts was 3,500 ppb (Larsen et al., 1983).
Toxicity of PCE
Developmental neurotoxicity is the most sensitive non-cancer health endpoint that may occur following low level exposure to PCE. The lowest dose at which adverse health effects have been noted following oral exposure to PCE is from a study in which doses of 5,000 ug/kg/day of PCE administered to young mice led to hyperactivity. By applying a safety factor of 100 to the results of this study to account for animal-to-person and person-to-person variability, ATSDR derived a minimum risk level (MRL) for acute (less than or equal to 14 days of exposure) oral exposure to be 50 ug/kg/day. The MRL is the dose below which adverse non-cancer health effects are not expected. ATSDR has not derived a MRL for longer durations of exposure. However, adverse health effects have not been observed in animal studies at doses below 5,000 ug/kg/day over longer periods of time than the study used to derive the MRL (ATSDR, 1995).
The average dose of PCE that residents on streets with VLAC pipe might have received between 1969 and 1997 can be estimated using standard assumptions about water consumption (2 liters per day for adults, 1 liter per day for children) and body weight (60 kg for female adults, 16 kg for children) and the estimated PCE concentrations. Volatile organic compounds, such as PCE, in drinking water can also evaporate into the indoor air of homes and be inhaled by residents (e.g., during showering and cooking). This additional dose of PCE from the inhalation of indoor air is typically assumed to be equal to the dose received from ingestion of PCE in drinking water. Accounting for both ingestion and inhalation exposures, the highest known dose of PCE for these residents in the past (i.e., exposure to 180 ppb of PCE) would have been 12 ug/kg/d for adults and 23 ug/kg/d for children. Therefore, the available data indicate that the past opportunities for exposure to PCE from VLAC pipe in the Duxbury water distribution system were below the level for which adverse health effects would not be expected (50 ug/kg/d). If it is assumed that PCE concentrations up to 3,500 ppb could have been possible, the potential dose may have been as high as 230 ug/kg/d for adults and 440 ug/kg/d for children for short periods of time. However, the highest possible dose would have always been at least ten times lower than the level at which health effects for the most sensitive endpoint (i.e., developmental neurotoxicity) have been observed in animals (5,000 ug/kg/d)
Furthermore, concerns about development neurotoxicity apply specifically to those who are pregnant or to young children. Concerns for adults in general would be non-cancerous effects on the liver or nervous system. However, even under worst case conditions (i.e., exposure to 3500 ppb of PCE), the estimated exposure for adults (230 ug/kg/d) would be over four hundred times lower than the lowest level that has been shown to cause liver or other problems in experimental animals (100,000 ug/kg/d). Evidence from exposures to humans is consistent with the results from animal studies (ATSDR, 1995).
Therefore, non-cancer adverse health effects are not expected for adult residents of streets with VLAC pipe. It is also unlikely that young or unborn children could have been adversely affected during development as a result of opportunities for exposure to PCE. However, the possibility of adverse developmental effects for some children cannot be ruled out completely because the PCE concentrations in the Duxbury water distribution system over the last 30 years have not been well documented.
Carcinogenicity of PCE
The results of some animal studies have associated exposure to PCE with the development of cancer (ATSDR, 1995). Assuming that PCE does present carcinogenic risks to humans, the lifetime cancer risks for residents of streets with VLAC pipe can be estimated based on the results of animal studies. Currently, the cancer risk for residents of streets with VLAC pipe should not be elevated because the average of detected concentrations (2.22 ppb) is less than the drinking water standard established by MDEP (5 ppb) (Table 1). In the past, the PCE concentration is known to have been as high as 180 ppb for some period of time at some locations. If this concentration had been constant for 30 years throughout all areas with VLAC pipe, individuals in areas with VLAC pipe might have a slightly elevated risk of cancer. Theoretically, it is possible that the PCE concentrations could have been as high as 3,500 ppb under certain, worst-case conditions in the past. If these conditions persisted for 30 years continuously, exposed individuals could be at a higher risk for developing cancer. While unlikely because of the water flow patterns caused by normal water usage (i.e., active flow because of withdrawls), this circumstance cannot be eliminated from consideration because many streets with VLAC pipe were not tested for PCE until 1997.
Several epidemiologic studies have been done to investigate the relationship between oral exposures to PCE and cancer incidence in humans, some of which have been reviewed by the Ad-Hoc Committee on Water Quality in relation to the VLAC pipe issue. One health study on exposure to PCE from VLAC pipes and cancer has been done for the towns on Upper Cape Cod. This case-control study evaluated the relationship between cases of bladder cancer, kidney cancer, and leukemia and estimated exposure to PCE from VLAC pipe. For two types of cancer (i.e., bladder cancer, leukemia), a positive correlation3 between estimated PCE exposures and cancer incidence was found if certain conditions (e.g., the latency period) were assumed (Aschengrau et al., 1993).
Another health study evaluated the cancer incidence rates for all cancers, liver cancer, non-Hodgkins lymphoma, multiple myeloma, and leukemia in two towns in Finland where PCE had been detected in the water distribution system at concentrations as high as 180 ppb. No significant differences between the cancer incidence rates for the two municipalities and the cancer incidence rates for the entire country were found. This study was limited by uncertainty in how long the residents of the towns had opportunities for exposure to PCE (Vartianinen et al., 1993; ATSDR, 1995).
Finally, a Massachusetts Department of Public Health study of residents of Woburn, Massachusetts, found that the relative risk of developing childhood leukemia was greater for those children whose mothers were likely to have consumed water from two contaminated municipal wells during pregnancy. At one time, these wells were tested and found to contain PCE as well as trichloroethylene, chloroform, other organic compounds, and some metals. It is not known whether PCE or the other contaminants were present in these municipal wells during their entire period of operation or to what extent temporal shifts in the concentrations of contaminant components may have occurred. Therefore, the increased risk of childhood leukemia could not be attributed to exposures to a particular chemical or a specific chemical mix but rather suggested that more studies of low level chemical and metal mixtures should be conducted (MDPH, 1997).
In summary, a limited number of health studies have been done pertaining to oral PCE exposures, including one evaluating the relationship between cancer incidence and PCE leached from VLAC pipe. In one of these studies, a positive correlation between estimated PCE exposures and the incidence of certain types of cancer was found. However, as is often the case with epidemiologic studies, definitive conclusions about the relationship between cancer risk in humans and PCE exposures could not be drawn because of limited data.