PUBLIC HEALTH ASSESSMENT
PALMER, HAMPDEN COUNTY, MASSACHUSETTS
To identify possible facilities that could contribute to the air, surface water, and soil contamination near the PSC site, the MDPH searched the Toxic Release Inventory (TRI) database for 1987 and 1988. TRI is developed by the EPA from the chemical information provided by certain industries. TRI included two industries within a one-mile radius of the site. All reported releases were to the air and consisted of the following chemicals: hydrochloric acid, sulfuric acid, sodium hydroxide, and nitric acid.
The following sections summarize environmental data and identify compounds of concern in the study area. The environmental data were taken primarily from the remedial investigation completed in 1992 (HMM 1992). Compounds of concern are those compounds whose concentration in at least one environmental media exceeded the public health comparison values established by the ATSDR. The compounds of concern listed in this section will be further assessed in the following section. The presence of a chemical on the compound of concern list does not mean that a human health threat exists. Some contaminants may be eliminated as a concern in the subsequent sections for other reasons (e.g., incomplete exposure pathway). Contamination levels may be reduced due to recent cleanup activities at the site.
1. Ambient Air
The nature and extent of contaminant releases to the ambient air from sources located at the PSC site were investigated in a two-phase ambient air sampling program conducted in conjunction with the Remedial Investigation of the site during the months of July and August 1988 (HMM 1992).
The Phase I air sampling program only screened for total volatile organic compounds (VOCs). Based on the results of the Phase I air sampling program, Phase II air sampling was designed to identify and quantify specific VOCs. Samples were taken over a three-day period from August 16-18, 1988. Samples were collected prior to test excavation (pre-excavation) and during soil disturbances (during excavation). Pre-excavation sampling locations were as follows: 1 upwind (off-property) station and 1 downwind of where the intrusive activities (onsite) would take place. During excavation sample stations consisted of: 2 upwind (off-property) stations and 4 downwind (onsite) of site excavation area stations, 2 downwind sampling stations approximately 100 feet northwest of the test excavation pits, and 2 downwind stations 5 to 40 feet northeast of test excavations pits. All samples were collected from a height of approximately four and one-half feet above ground surface. Samples were collected over a two-hour sampling period during test pit excavation and all samples were analyzed for VOCs (Table 1).
The only compound detected at maximum concentrations exceeding the ATSDR comparison values in pre-excavation samples was benzene. However, the maximum concentration of benzene (1.2 µg/m3) was lower than typical ambient air levels of this ubiquitous chemical in the United States (about 15-110 µg/m3 [ATSDR 1993a]). While trichloroethene (TCE) and tetrachloroethene (PCE) exceeded ATSDR comparison values in the sampling done during excavation activities, these concentrations do not reflect undisturbed site conditions but rather potential temporary conditions during future excavations. Neither PCE nor TCE were detected in pre-excavation sampling.
In addition, methylene chloride was detected at elevated concentrations (15 µg/m3). However, the compound, which is a common laboratory contaminant, was also detected in laboratory blanks (up to 14 µg/m3) that were analyzed in conjunction with the samples collected during this sampling period. Detection of methylene chloride in these blank samples indicates that methylene chloride in the air samples was likely due to laboratory contamination and not to the site.
2. Lagoon and Catch Basin Surface Water
The onsite lagoon was originally used as an earthen bermed area designed to contain the release of materials from the adjacent vertical storage tanks. Sediments within the lagoon are a result of the disposal of sludge from the storage tanks. Repetitive sludge disposal produced the accumulation of a layer of sludge over the base of the earthen bermed area. This layer prevented the infiltration of rain water which resulted in ponding of water and the eventual development of the onsite lagoon. The lagoon contained up to 210,000 gallons of surface water. The catch basin is another ponded area located in the northern portion of the site.
Lagoon surface water and catch basin water samples were collected in September 1988 and November 1990 as part of the RI. A total of five samples were analyzed for a variety of organic and inorganic compounds (results are shown in Tables 4-23 through 4-25 in the RI). The RI reported that volatile organic compounds were generally less than about 15 µg/L, while semivolatile organic compounds were less than about 50 µg/L. The RI also reported some elevations in lead (maximum concentration 800 µg/L) and arsenic (maximum concentration about 9 µg/L). ATSDR does not have comparison values for chemicals in surface waters that are not used for drinking water purposes. The likelihood of these surface waters being a source of chemical exposure to humans will be discussed in the exposure pathways analysis section later in this document.
3. Lagoon and Catch Basin Sediment
Analyses of the sediments within the lagoon and catch basin were conducted as part of the RI through three sampling rounds: Phase I (0-3 in) in September 1988, Phase II (sample depth not specified) in November 1990, and additional EPA sampling in November 1990 (data summarized in Tables (4-23 through 4-25 and 4-27 in the RI). ATSDR does not have comparison values for sediment. While soil comparison values can be used to screen for possibly elevated levels of chemicals in sediment, the soil comparison values assume daily incidental ingestion of chemicals in soil by young children (< 6 years old), a situation highly unlikely to have occurred or to be occurring for these sediments. The RI noted that elevated concentrations of some polyaromatic hydrocarbon (PAH) compounds, polychlorinated biphenyls (PCBs), and some VOCs (e.g., toluene) were found in these sediments. The likelihood of these sediments serving as a source of exposure to young children or others will be discussed in the exposure pathways analysis section.
4. Subsurface Soil
As part of the 1988 remedial investigation, eleven grab soil samples were collected from ten test pits and analyzed for organic and inorganic compounds. In April 1990, subsurface soil samples were taken from five onsite soil borings. Samples collected at two-foot intervals from five on-site soil borings advanced to depths of 12 to 16 feet were screened in the field for total VOCs, pesticides and PCBs, nine of these samples were submitted for confirmatory laboratory analyses for pesticides and PCBs (data summarized in Tables 4-10 through 4-14 and 4-18 in RI).
The RI noted that the soil borings on the property indicated elevated concentrations of some VOCs (e.g., trichloroethene), PAHs, PCBs, and lead. The likelihood of exposures to chemicals in subsurface soil are discussed in the exposure pathways analysis section.
5. Surface Soil (0-6 inches)
ATSDR defines surface soil as 0-3 in depth. No samples of this depth were available. However, four soil (0-6 inches) samples were collected onsite in November 1990 and analyzed for PCDDs and PCDFs. These samples were collected in the vicinity of the soil borings containing elevated PCB concentrations. All samples were below ATSDR's comparison value for this group of compounds.
Twenty composite soil (0-6 inches) samples were collected in July 1991 for lead analysis. Lead concentrations exceeded an EPA soil screening guidance level of 400 ppm over the majority of the property, with a maximum concentration over 39,000 ppm. Lead is therefore a compound of concern for surface soil onsite. The highest concentrations were found in samples taken near the former vertical storage tanks. The lead is believed to be related to waste oil and sludge that were contained in the storage tanks onsite.
Five rounds of groundwater monitoring in two onsite wells (one shallow and one deep overburden well) were conducted between October 1988 and October 1990. The likelihood of exposure to these chemicals in groundwater will be discussed in the exposure pathways analysis section. The RI reported that the primary contaminants detected in groundwater were volatile organic (e.g., tetrachloroethene) and semivolatile organic compounds (e.g., methylphenols) (see Tables 4-44 through 4-46 in the RI). The highest levels of these compounds were detected in the shallow groundwater in the wetlands area immediately downgradient from the property boundary.
1. Quaboag River Surface Water
Phase I surface water sampling of the Quaboag River was conducted in September 1988. Four surface water samples were collected as follows: 1 sample upstream of the site, 2 samples from an area adjacent to the site, and 1 sample downstream of the site. The four samples were analyzed for organic and inorganic compounds. No VOCs, SVOCs, pesticides, or PCBs were detected in the surface water samples. Nine inorganic compounds (i.e., barium, calcium, copper, iron, magnesium, manganese, potassium, sodium, zinc) were detected. As noted above, no comparison values exist for surface waters not used as drinking water. However, when comparing the concentrations of these inorganic compounds in the Quaboag River to drinking water comparison values (a very health-protective assumption), no compound exceeded these ATSDR values.
2. Quaboag River and Wetland Sediment
Phase I sediment sampling of the Quaboag River and of the wetland was conducted in September 1988. Samples in the Quaboag were from the same locations as the Phase I surface water samples and were analyzed for organic and inorganic compounds. Phase II sediment sampling was conducted in November 1990. The five Phase II samples were analyzed for the same parameters as the Phase I samples. One additional sediment sample was collected in November 1990 and analyzed for PCDDs/PCDFs, pesticides, and PCBs. Sediment comparison values were not available. However, when the chemical concentrations in sediment were compared to soil comparison values (a very health-protective assumption), all Quaboag River sediment concentrations were either below the soil comparison values or comparable to typical background levels reported for the chemicals in soil or sediment.
Phase I wetland sediment sampling conducted in September 1988 included the collection of grab samples from 35 locations. Each sample was collected from a 0-3-inch depth. Sample locations correlated with surface water drainage ways and areas of sediment transport of deposition. The wetland sediment samples were screened for total VOC concentrations. Based on the results of field screening, 17 samples were submitted for analysis of VOCs, semi-VOCs, pesticides, PCBs, and metals.
Fifteen sediment samples from the Phase II sampling round were collected in November 1990 and analyzed for semi-VOCs, pesticides, PCBs, and metals. Three additional wetland sediment samples were collected and analyzed for dioxins, PCDF, pesticides, and PCBs. Three additional samples were collected from the spill area and analyzed for VOCs. The spill area is believed to be a site where significant releases of oil and waste material occurred.
As discussed above, no ATSDR comparison values are available for sediment. While soil comparison values can be used to compare with levels of chemicals in sediment, the soil comparison values assume daily incidental ingestion of chemicals in soil by young children (< 6 years old). Thus, such a comparison is extremely health-protective. Chemicals that exceeded soil comparison values were benzo(a)pyrene, benzo(a)anthracene, beryllium, lead, and PCBs. The maximum concentration of beryllium was 0.8 ppm, which is only slightly higher than the soil comparison value (0.2 ppm) and the typical background concentrations in Massachusetts soil (0.4 ppm), and well below typical soil concentrations in the eastern U.S. (from less than 1 to 7 ppm; HMM 1992) (see Table 2). Thus, beryllium was not selected as a compound of concern. Although benz(a)anthracene was greater than its comparison value, it was well within typical background concentrations in soil or sediment (0.005-59 mg/kg) (ATSDR 1993b). Hence, it was not selected as a compound of concern.
The remaining three compounds, benzo(a)pyrene, lead, and PCBs, were selected as compounds of concern from wetland sediment. The PAH compounds were found at elevated concentrations in the spill area, which is believed to be a site where significant releases of oil and waste material occurred. Total PAH concentrations decreased from 122.5 mg/kg in the spill area to 10 mg/kg in the drainage ways and areas of lower elevation. PCBs also were detected in concentrations greater than their comparison value within and adjacent to the spill area. PCB concentrations decrease to less than 1 mg/kg as one moves along suspected drainage pathways leading from the spill area. Lead concentrations exceeding the EPA soil screening level were detected in and adjacent to the spill area. Lead concentrations decreased from 3,270 mg/kg in the immediate vicinity of the spill area to less than 100 mg/kg along the drainage pathway that extended toward the Quaboag River. The highest lead concentration (7,970 mg/kg) was found in a sample taken adjacent to a mound of fill deposits of assorted debris, located just northeast of the spill area. The distribution of lead, PAHs, and PCBs appears to follow similar trends and are likely related to historical waste oil spills and contaminated surface water runoff from on-site soils. Elevated concentrations appear to be confined to the property and the spill area section of the adjacent wetland.
3. Subsurface Soil
Subsurface soil samples were collected from one boring located to the west of the property which was installed in April 1990 and from six monitoring wells installed to the west, south and north of the property. Continuous split spoon samples were collected at intervals of two feet through the upper twenty feet of overburden and at five-foot intervals thereafter. Most of the samples were taken from below the water table (less than one foot to eight feet below ground surface). Samples were screened for VOCs and based on the results of the screen, samples were analyzed for organic and inorganic compounds (data summarized in Tables 4-19 and 4-20 in the RI). The RI reported that PAHs and other semivolatile organic compounds and lead were found at elevated concentrations in subsurface soil. The likelihood of exposures to chemicals in subsurface soil are discussed in the exposure pathways analysis section.
4. Surface Soil (0-6 inches)
Off-property surface soil (0-6 inches) sampling consisted of five composite samples collected on March 18, 1991 from an athletic field north of the site and four samples collected on July 1990 from the wetland to the south of the site. Samples from the athletic field were analyzed for organic and inorganic compounds. No compound was detected at levels greater than the ATSDR soil comparison values.
Four samples (0-6 in) collected from the spill area section of the wetland, directly adjacent to the property boundaries, were analyzed for lead. Lead concentrations ranged between 190 mg/kg and 50,100 mg/kg. Concentrations in three of four samples exceeded the 400 mg/kg EPA soil screening level for lead in soils.
Monitoring wells were installed in two phases during the Remedial Investigation. Fifteen groundwater monitoring wells were installed in six locations in August and September 1988 during the Phase I investigation. Two or three wells at one location allowed the screening of contaminants in the over burden or bedrock material. The Phase II groundwater monitoring well installation program was conducted during March and April of 1990. The location of the Phase II wells was based on the results of the Phase I monitoring program.
As discussed for onsite groundwater monitoring, five rounds of groundwater sampling were conducted during which off-property groundwater samples were collected. A total of 85 samples was collected to the west, east, south, and north of the site with the majority of the samples collected from the wetland to the west of the site. The samples were analyzed for organic and inorganic compounds.
The highest VOC concentrations were found in the wetland area immediately downgradient of the site. VOCs appear to be limited to the shallow overburden samples because no VOCs had been detected in bedrock, nor were they generally found in deep overburden monitoring wells. The shallow VOC plume appears to be centered in the wetland along the southwestern edge of the site. The plume extends to the northeast toward Water Street and to the northwest toward the river. Based on a comparison of the distribution of VOC contaminants over the five sampling rounds, the plume appears to be getting smaller and receding back toward the site.
6. Fish in the Quaboag River
Fish sampling in the Quaboag River to support an ecological risk assessment (e.g., risks to predators who consume an entire fish) for the remedial investigation was conducted in November 1990. Two species of edible bottom-feeding fish (tessellated darter and white sucker) were sampled from three reaches of the river. Composite whole-body samples were analyzed for semi-VOCs, pesticides, PCBs, and metals. No PCBs or pesticides were present in the fish samples. Because the samples were whole-body samples and not edible-tissue samples, detected concentrations do not represent what actual exposure could be to individuals eating the fish. However, the data is adequate to rule out significant human health risk as a result of consumption of these species of fish from the Quaboag River in the vicinity of the site.
Each of the major investigations summarized in the previous sections regarding the environmental contaminants has its own quality assurance and quality control (QA/QC) procedures. Based on the available information, the sampling, analytical techniques, and data collected were generally adequate for the purpose of this public health assessment. In terms of potential exposure, a data gap for the site pertains to lack of knowledge about possible exposure patterns or chemical concentrations in edible tissue of fish in the Quaboag River. This issue will be further discussed in the pathways analysis section below.
The presence of physical and other hazards was investigated during the July 18, 1991, site visit. Based on that visit, we believed that the site represented a physical hazard to any individual present onsite. These physical hazards included heavy equipment, scrap metal, 55-gallon drums, the waste lagoon, and miscellaneous debris (e.g., tires, wood, scrap metal). Some of these items (e.g., debris) could have attracted the attention of older children who may have accessed the site.
While the site was accessible to unauthorized individuals prior to September 1991 via several holes in the fence around the site, it is not known to what extent individuals accessed the site. During the July 1991 site visit, some human activity was evident onsite (e.g., evidence of unauthorized dumping). However, during September 1991, the EPA constructed a new fence around the already existing fence. A second site visit, conducted by the MDPH in December of 1991, indicated that individuals were still accessing the site. Ice skate tracks were evident across the frozen waste lagoon and junked car parts were present inside the fence. All the physical hazards present in July 1991 were still present on-site at the time of the second site visit.
EPA officials report that in January 1992 site access was further restricted by chaining and locking the access gates and that no further evidence of trespassing within the fenced area has been noted. EPA also reported that the 55-gallon drums that were observed on the site were relocated behind the secured fence. The access gates are now reportedly well maintained and warning signs are prominently placed on the gates and around the fence line. The fence perimeter is monitored by the EPA and the Performing Settling Defendants to check for evidence of fence damage and unauthorized entry.
To determine whether individuals are possibly exposed to contaminants, the factors influencing human exposure were evaluated. Exposure pathways are identified as completed or potential. Completed pathways indicate that exposure to a contaminant has occurred in the past, is currently occurring, or will occur in the future. Potential pathways, however, indicate that exposure to a contaminant may have occurred in the past or may occur in the future.
Property and Wetland Soil (0-6 inches)
Past exposures were possible to contaminants in property or wetland surface soil (0-6 inches) in that portion of the wetland in and near the spill area. Exposures to property soil were most likely for workers at the site. While there is some evidence of non-employee use of the property in the past (e.g., for authorized dumping, skating on property ponds), the evidence does not suggest that these uses would have been frequent or would have presented significant opportunities for exposure to chemicals on the property. Past exposures to the spill area section of the wetland could have occurred to nearby residents. However, this area is densely vegetated, and thus it is unlikely that frequent contact occurred with these soils. Access to the property and contaminated areas of the wetlands is currently restricted, therefore, current and future contact exposure of local residents to contaminants in onsite soils is not expected.
Possible exposure routes to surface soil include soil ingestion and dermal contact with the soil. Likely receptors to wetland soils would be adult or older children residents. For property soils, the likely exposure routes for workers would be soil ingestion and dermal contact with the soil.
Wetland Sediment Pathway
Past exposures were possible from the contamination of the sediment in and near the spill area section of the wetland. Past exposures to the spill area of the wetland could have occurred to nearby residents. However, this area is densely vegetated, and thus it is unlikely that frequent contact occurred with these soils. As noted above, access to the spill area of the wetlands is mitigated by the presence of a fence around the area. Exposure routes could include incidental ingestion of or dermal contact with the sediment.
Past or current exposures to chemicals in the groundwater did not occur and is not occuring because the groundwater in the vicinity is not a source of municipal drinking water and no known private wells are in use at downgradient locations in the vicinity of the site. Future exposure to contaminants in groundwater (e.g., volatile organic compounds) could occur if private wells were installed in the site vicinity, prior to the achievement of groundwater cleanup remedies. However, institutional controls restricting groundwater use in the immediate area of the site will be implemented until cleanup levels have been achieved and maintained. These restrictions will minimize the opportunity for future exposure.
It has been reported that the Quaboag River is used for fishing. However, it is not known whether and to what extent individuals consume fish from which part of the river. While the tesselated darter and white sucker are both edible fish species, neither species is heavily used by anglers in Massachusetts. Data are not available for fish species that are more likely to be consumed by individuals who fish from the Quaboag River.
Subsurface Soil Pathway
Future exposure to subsurface soil contaminants could occur during intrusive activities for onsite remedial workers if the proper health and safety procedures are not followed. In addition, should future construction occur at the site and the current levels of contamination remain in place, future exposures during site construction would be possible.
Lagoon and Catch Basin Surface Water and Sediment
The property lagoon and catch basin are contaminated; however, the likelihood of significant direct contact with these surface waters or sediment is highly unlikely given their undesireable qualities (e.g., oil sheen on surface of lagoon water). While skating on the lagoon has been reported, such activity would not result in exposure to contaminants in the water.
Several pathways were eliminated based on the fact that no compounds exceeded ATSDR comparison values or other health-protective values. These pathways included the Quaboag River surface water and sediment and athletic field surface soil.
In this section we will discuss the health effects in persons exposed to specific contaminants, evaluate state and local health databases, and address specific community health concerns. To evaluate health effects, ATSDR has developed minimal risk levels (MRLs) for contaminants commonly found at hazardous waste sites. The MRL is an estimate of daily human exposure to a contaminant below which non-cancer, adverse health effects are unlikely to occur. The MRLs are developed for each route of exposure, such as ingestion and inhalation, and for length of exposure, such as acute (less than 14 days), intermediate (15 to 364 days), and chronic (greater than 365 days). When MRLs are not available, reference doses (RfD) provided by the EPA are evaluated.
Property and Wetland Surface Soil--Ingestion and Dermal Contact
The only compound of concern in property or wetland surface soil (0-6 in) was lead. Surface soil lead concentrations on the property were very high. Of 16 surface soil samples from the property, all but one had a concentration of at least 2,390 ppm, with an average concentration of 12,800 ppm and a maximum concentration of 39,200 ppm. Four surface soil samples from an adjacent area (called the spill area) in the wetland showed concentrations of 190, 768, 1,760, and 50,100 ppm.
ATSDR does not have an MRL for lead. Young children (i.e., less than 6 years old) and the unborn fetus are most sensitive to possible effects from lead exposures. Levels of concern for lead are based on blood lead levels. Substantial human data exist in the literature that show dose-response relationships between health effects (e.g., learning disabilities) and blood lead levels in young children. The U.S. Centers for Disease Control (1991) stated that "epidemiological studies have identified harmful effects of lead in children at blood lead levels at least as low as 10 µg/dL" (page 2). Considerable uncertainty exists, however, on the relationship between exposures to lead in soil and resulting blood lead levels. While ATSDR (1992) has noted that in general, blood lead levels may rise about 3-7 µg/dL for every 1000 ppm increase in soil or dust concentrations, this relationship depends on many factors including, the age of the individual (e.g., young children tend to ingest more soil than adults and tend to absorb lead from soil into the bloodstream more readily than adults), the bioavailability of the lead in the soil (e.g., some soil lead may not be as easily absorbed into the body as other soil lead), the frequency of actual contact with and hence possible incidental ingestion of the soil containing the lead (e.g., daily contact versus occasional contact, grassy area versus bare dirt), and the nutritional state of the individual (e.g., poorly nourished individuals may absorb lead more readily than well-nourished individuals).
Should young children be exposed (or have been exposed in the past) regularly to lead concentrations found on the property, they would be at risk of elevated blood lead levels and hence adverse health effects, such as decreased IQ scores. However, as noted in the exposure pathways section, it is unlikely that young children less than 6 years old regularly accessed either the property or the spill area section of the wetlands and hence, it is unlikely that these young children would have had much opportunity for exposure to these lead in soil concentrations. Because the unborn fetus is also sensitive to lead exposures, it is possible that female employees at the site in the past may have been exposed to lead in soil. If female employees regularly came into contact with soils at the site, these individuals would have been at risk of having elevated blood lead levels. Effects on the unborn fetus may include premature births or decreased mental ability in the infant.
Other adult employees may have had opportunities for exposure to these lead concentrations. Health effects from lead exposure to adults include neurological effects (e.g., impairment on neurobeahvioral tests), gastrointestinal effects (e.g., colic), and effects on the blood-forming system (e.g., changing the activity of certain enyzmes).
The area of elevated lead concentrations in wetland appears confined to what is known as the spill area. If individuals had frequented this spill area in the past, then they could have had the opportunity for exposure to these concentrations. With sufficient opportunities for exposure (e.g., daily contact with the soil), the health effects noted above may have been possible.
Wetland Sediment--Ingestion and Dermal Contact
Compounds of concern in wetland sediment included benzo(a)pyrene, lead, and PCBs. As with potential exposure to property and wetland soil, past exposures to wetland sediments in the spill area adjacent to the property would likely have been intermittent, if at all, due to the dense vegetation around the contaminated area. Thus, it is not likely that there was frequent opportunity for exposure (e.g., via incidental ingestion) to these chemicals in sediment. Regulatory guidance levels for contaminants in soil or sediment are generally based on daily contact (e.g., via incidental ingestion of the soil or sediment), but if such contact does not occur as often as assumed when developing regulatory guidance, the potential for health effects is likewise reduced. For example, in 1984 the MDPH conducted a survey in Norwood, Massachusetts, regarding possible contact with a site contaminated with PCBs. The ATSDR comparison value for PCBs in soil is 1 ppm, a value that assumes daily contact with and high ingestion rates of soil. PCB concentrations in soil at Norwood were as high as 220,000 ppm. Individuals with the greatest potential for exposure had accessed the site for camping, running, playing, and biking. Results of blood analysis for PCBs (a good indicator of whether exposure had occurred) on individuals with the greatest use of the site did not indicate elevated blood PCB levels (MDPH 1984). Hence, the intermittent use of areas with elevated PCB concentrations in soil did not result in exposure as reflected in the blood PCB concentrations.
The same concerns expressed with lead in soil (property or wetland) pertain to lead in wetland sediment. The area of the wetland with elevated lead concentrations was relatively confined (i.e., the spill area), and it is unlikely that individuals accessed this area on a regular basis. However, if individuals did access the area regularly and had opportunities to ingest soil containing high levels of lead, it is possible that their blood lead levels increased, thereby increasing the potential for health effects, such as neurological effects.
Exposure to benzo(a)pyrene (maximum concentration of 9.1 ppm) is unlikely to result in any noncancer effects, even assuming daily contact by a young child. Benzo(a)pyrene is considered carcinogenic based on results in experimental animal studies (cancers of the stomach, esophagus, larynx). Of the 33 sediment samples (including duplicate samples; data taken from Tabel 4-28 of RI) analyzed for PAHs and other compounds, only seven samples had benzo(a)pyrene concentrations greater than typical background concentrations (shown in ATSDR 1993b to be as high as 1.2 ppm). These seven samples were in or adjacent to the spill area, with one sample located about 150 ft southwest of the spill area. As discussed in the exposure pathways section, it is likely that potential contact with this area of the site would have been intermittent. Intermittent exposure to these soils is not likely to have significantly increased cancer risks to potential receptors (e.g., adults, older children).
1. Health Studies
One health evaluation conducted prior to this investigation exists for the PSC Resources area. The study was conducted by the MDPH to review cancer incidence and mortality for selected sites in Monson (MDPH 1988). The MDPH calculated cancer incidence and mortality for selected childhood cancers and bladder cancer. Lung cancer mortality rates were also calculated. Place of residence at the time of diagnosis or death was examined for each case to determine any possible geographic concentration. Slight elevations in mortality and incidence were observed for several cancers. However, an environmental association could not be determined due to the small number of cancer cases/deaths in Monson (MDPH 1988).
2. Current Descriptive Health Study
The Community Assessment Unit (CAU) of the MDPH conducted an investigation of cancer incidence in the towns of Palmer and Monson for the period 1982-1992. Standardized Incidence Ratios (SIRs) were calculated for cancer of the liver, bladder, kidney, stomach, and lungs, as well as for leukemia. Cancers were selected for investigation based upon documented community concerns or as a result of a potential relationship to the types of contaminants present at the site.
For the purpose of this assessment, particular emphasis was placed on census tract 8101 in Palmer and the areas in Monson located within a one-mile radius of the PSC Resources site.
Cancer incidence data were obtained from the Massachusetts Cancer Registry (MCR) of the MDPH, Bureau of Health Statistics, Research and Evaluation. The MCR has been monitoring cancer incidence in the Commonwealth since 1982.
In order to determine whether an elevated rate of a specific type of cancer exists in Palmer and Monson and in each of the three Palmer census tracts (CTs 8101, 8102, 8103), data were adjusted by age and sex and analyzed to compare the actual (or observed) number of cancer cases to the number that would have been expected based on the statewide cancer incidence experience. Reliable population data are necessary to calculate incidence rates. The population figures for the state of Massachusetts, for Palmer and Monson, and for each Palmer census tract were generated by calculating the percentage of increase or decrease in total population for these towns and each census tract based on 1980 federal census data and 1990 MARS (Modified Age, Race, and Sex) file, an adjusted listing of the 1990 federal census data (U.S. Bureau of Census 1990). For the purpose of consistency and the gap between census years, an assumption is made that the change in population occurs at a constant rate throughout the ten-year census period. From these calculations, the midyear estimates are obtained and used to calculate state, town, and census tract populations.
A standardized incidence ratio (SIR) is an estimate of the occurrence of disease in a population in relation to what might be expected if the population had the same cancer experience as some larger population designated as "normal" or average. This normal population is usually selected to be the state as a whole. Specifically, an SIR is the ratio of the observed number of cancer cases to the expected number of cases, which is then multiplied by one hundred (100). An SIR equal to 100 indicates that the number of cancer cases occurring in the population being evaluated is equal to the number of cases expected in a "normal" population. An SIR greater than 100 indicates that more cancer cases occurred than expected; an SIR less than 100 means that fewer cases occurred than expected. Accordingly, an SIR of 150 is interpreted as an excess of 50 percent of cases over the expected number; an SIR of 90 indicates 10 percent fewer cases than expected.
Caution should be exercised, however, when interpreting an SIR. The interpretation depends on both the size and the stability of the SIR. Two SIRs can be the same size, but not have the same stability. An SIR of 150 based on two expected cases and three observed cases indicates a 50 percent excess of cancer. The excess is attributed to only one extra case, and that case may have occurred by chance alone. Conversely, an SIR of 150 based on 200 expected cases and 300 observed cases shows the same 50 percent excess in cancer, but because the SIR is based on a greater number of cases, the estimate is more stable. It is unlikely that 100 excess cases would occur by chance alone.
To determine if the observed number of cases is significantly different from the expected number, or if the difference may be due to chance, a 95 percent confidence interval (CI) is calculated (Rothman and Boice 1982). A 95 percent CI is the range of estimated SIR values so that the range has a 95 percent probability of including the true SIR for the population. If the CI range does not include the value 100, then the study population is significantly different from the normal population. If the CI range excludes 100, and the observed SIR is greater than 100, there is a significant excess of the number of cancers. Similarly, if an observed SIR is less than 100, and the CI range excludes 100, the number of cancer cases is significantly lower than expected. If the CI range includes 100, then the true SIR may be 100, and it cannot be concluded with sufficient confidence that the observed SIR reflects a true cancer excess (or deficit). Statistical significance is not assessed when fewer than five cancer cases are observed.
The width of the CI also reflects the stability of the SIR estimate. For example, a narrow CI (e.g., 103-115) allows a fair level of certainty that the observed SIR is close to the true SIR value for the population. A wide interval (e.g., 82-450) leaves considerable doubt about the true SIR, which could be much lower or higher than the observed SIR. This would indicate an unstable statistic.
Place of residence at the time of diagnosis was evaluated for each case (by primary site) to assess any possible geographic concentraiton of cases. Mapping the cases allowed the determination of the number of cases within each census tract; hence census tract-specific cancer rates could be calculated. Residence at the time of diagnosis is self-reported to the diagnostic facility, which subsequently reports to the MCR. When bladder, kidney, and lung cancer cases were mapped, it was determined that some lung and kidney cancer cases actually lived in towns contiguous to Palmer and Monson. This information was confirmed by town street lists and local phone book listings for the appropriate time periods.
Because population data, which are required to calculate SIRs, are not readily accessible in finer units than census tracts, rates were only calculated for the towns as a whole and for each census tract. MDPH staff examined the geographic distribution of each cancer for each census tract to qualitatively determine if there appeared to be some concentration on a finer scale (e.g., a street, an area of several blocks). This latter examination would address citizen concerns about possible cconcentration in areas of concern.
MDPH also evaluated the available occupational information, which uses the 1980 census occupational classification system, and the reported smoking status of individuals diagnosed with specific cancer types investigated for this report (U.S. Bureau of Census 1980). This was accomplished by using existing cancer registry data.
The public comment release draft of this health assessment included cancer incidence data for the years 1982-1988. These 1982-1988 data showed that cancer incidence in Palmer and Monson occurred essentially as would be expected for most cancer types. In Palmer, statistically significant elevations were observed among males in bladder cancer incidence and among males in lung cancer incidence. In census tract 8108 in Palmer, where the PSC Resources site is located, bladder cancer in males and males and females combined were statistically significantly elevated. In Monson, there was a statitically significant elevation in lung cancer incidence among males and females combined. The increase was attributed to a statistically significant increase in male lung cancer incidence.
Since the public comment release draft of this health assessment, the MCR has released data that include the period 1982-1992.
The MCR publication and this report display the most recent cancer incidence data from 1982-1992 in two separate time periods: 1982-1986 and 1987-1992. These trends are useful to determine if significant elevations or decreases in cancer incidence during these periods, which may not be apparent in an inclusive analysis. Tables 3-7 present the results of the cancer incidence analysis.
These most recent data show that statistically significant elevations in the cancer types reviewed for this report are as follows:
|-||In Palmer, census tract 8101, during 1982-1986, bladder cancer in males in was statistically significantly elevated (12 cases observed/ 4.2 cases expected; SIR=289; 95%CI=[149-505]). The elevation in males explained the statistically significant elevation in males and females combined in this period in census tract 8101 (14 obs/ 5.9 exp; SIR=237; 95% CI=[129-398]).|
|-||In Palmer, census tract 8101, during 1982-1986, lung cancer in males was statistically significantly elevated (21 obs/12.2 exp; SIR=172; 95% CI =[106-263]).|
|-||In Palmer, census tract 8101, during 1987-1992, kidney cancer in males and females combined was statistically significantly elevated (9 obs/ 4.0 exp; SIR=225; 95% CI=[103-429]).|
|-||In Monson, no statistically significant elevations were observed for either time period.|
A description of these analyses follows.
Table 3 summarizes bladder cancer incidence in Palmer. During the 1982-1986 period, overall bladder cancer incidence in males and females combined for Palmer as a whole was slightly elevated (18 obs/12.8 exp; SIR=141; 95% CI=[84-223]). This elevation was found among males (16 obs/9.3 exp; SIR=172; 95% CI=[98-279]). Female cases occurred at a less than expected rate for Palmer (2 obs/3.5 exp; SIR=57). Bladder cancer in males and females combined in census tract 8101 was statistically significantly elevated. In census tract 8101 male bladder cancer incidence was statistically significantly elevated (12 cases observed/ 4.2 cases expected; SIR=289; 95%CI=[149-505]). Female cases in census tract 8101 and bladder cancer in census tracts 8102 and 8103 occurred at expected rates.
During the 1987-1992 period, bladder cancer incidence was slightly elevated in males and females combined for Palmer as a whole (19 obs/14.3 exp; SIR=133; 95% CI=[80-208]). This elevation was attributed once again to an elevation in males cases (17 obs/10.4 exp; SIR=163; 95% CI=[95-261]). Female bladder cancer occurred at the expected rate (2 obs/ 1.7 exp). In census tract 8101 for males and females combined, bladder cancer was slightly elevated (8 obs/6.6 exp; SIR=121; 95% CI=[52-238]). This elevation was due to about one extra case. Bladder cancer incidence among males also appeared to be slightly elevated (8 obs/4.7 exp; SIR=171), however this elevation was attributed to about three extra cases and was not statistically significant.
Table 4 summarizes kidney cancer incidence in Palmer. During the 1982-1986 period for males and females combined for Palmer as a whole kidney cancer occurred at less than expected rates (2 obs/5.2 exp; SIR=38). Male incidence (1 obs/3.1 exp; SIR=32) and female incidence (1 obs/2.1 exp; SIR=47) were at less than expected rates for the 1982-1986 period. Analysis of individual census tracts in this time period indicates kidney cancer incidence at less than expected rates.
During the 1987-1992 period for Palmer overall in males and females combined, an elevation in kidney cancer was observed (15 obs/8.6 exp; SIR=174; 95% CI=[98-288]). This elevation was attributed to about four extra male cases (9 obs/5.3 exp; SIR=171; 95% CI=[78-325]) and about three extra female cases (6 obs/3.4 exp; SIR=177; 95% CI=[64-384]). These elevations were not statistically significant. In census tract 8101 among males and females combined, a significant elevation was observed (9 obs/4.0 exp; SIR=226; 95% CI=[103-429]). This elevation was attributed to about two extra male cases in census tract 8101 (6 obs/2.3 exp; SIR=213) and two extra female cases (4 obs/1.6 exp; SIR=246). The elevations among males and females were not significant. Kidney cancer in census tracts 8102 appeared slightly elevated, however these elevations were based on small numbers and were not significant. Census tract 8103 analysis indicated less than expected rates for kidney cancer (1 obs/2.6 exp; SIR=39).
Table 5 summarizes lung cancer incidence in Palmer. During the 1982-1986 period in Palmer overall for males and females combined lung cancer incidence occurred at a slightly higher rate than expected (47 obs/41.2 exp; SIR=114; 95% CI=[84-152]). This elevation was attributed to male cases (34 obs/27.0 exp; SIR=126; 95% CI=[87-176]). Female cases occurred as would be expected (13 obs/14.3 exp; SIR=91). Analysis of individual census tracts shows that this elevation is attributed to males in census tract 8101 (21 obs/12.2 exp; SIR=172; 95% CI=[106-263]) which was statistically significant. Lung cancer in females in census tract 8101 and males and females in census tracts 8102 and 8103 occurred at expected rates.
During the 1987-1992 period for Palmer as a whole for males and females combined lung cancer incidence occurred as would be expected (52 obs/53.5 exp; SIR=99; 95% CI=[74-129]). Lung cancer for males (31 obs/32.2 exp; SIR=96; 95% CI=[65-136]) and females (22 obs/21.6 exp; SIR=102; 95% CI=[64-154]) occurred at expected rates. A slight elevation was observed in census tract 8101 for males and females combined (27 obs/24.8 exp; SIR=109; 95% CI=[72-159]). This elevation was attributed to about one extra male case (16 obs/14.4 exp; SIR=111) and one extra female case (11 obs/10.3 exp; SIR=107). Lung cancer in census tracts 8102 and 8103 occurred as would be expected.
Leukemia, Liver and Stomach Cancer
Tables 6 and 7 summarize cancer incidence data for the following cancers: leukemia, liver and stomach for Palmer and Monson. For each of these cancer types, no statistically significant elevations in 1982-1986 or 1987-1992 for Palmer or Monson were seen.
Leukemia incidence for Palmer was slightly elevated for males and females combined in the 1982-1986 period (10 obs/5.6 exp; SIR=180; 95% CI=[86-331]). This elevation was not statistically significant and was attributed to four extra male cases (7 obs/3.1 exp; SIR=229; 95% CI=[92-472]). Leukemia incidence in Palmer during the 1987-1992 period for males and females combined occurred as would be expected (7 obs/6.3 exp; SIR=112; 95% CI=[45-231]). In Monson for males and females combined during 1982-1986 leukemia appeared slightly elevated (4 obs/3.0 exp; SIR=131). This elevation was due to one extra male case (3 obs/1.7 exp; SIR=174) and was not statistically significant. Leukemia incidence in Monson during 1987-1992 also appeared elevated (6 obs/3.4 exp; SIR=175; 95% CI=[64-381]). This elevation was due to two extra males cases (4 obs/2.0 exp; SIR=198) and was not statistically significant.
Liver cancer in Palmer occurred as would be expected during both the 1982-1986 period (1 obs/1.4 exp) and the 1987-1992 period (2 obs/2.0 exp). No cases of liver cancer were reported for Monson.
Stomach cancer incidence for males and females combined in Palmer during 1982-1986 appeared slightly elevated (12 obs/7.0 exp; SIR=171; 95% CI=[88-299]). This elevation was not statistically significant and was attributed to two extra male cases (7 obs/4.2 exp) and two extra female cases (5 obs/2.8 exp). During the 1987-1992 period a slight excess for males and females combined for Palmer was noted (10 obs/7.4 exp; SIR=136; 95% CI=[65-250]). Again, this elevation was not significant. About one extra male case (6 obs/4.4 exp) and one extra female case (4.0 obs/2.9 exp) was observed for this period. In Monson, stomach cancer occurred slightly less than would be expected during both periods.
Appendix A contains brief discussions on risk factors for each of the cancers discussed in this section.
c. Geographic Distribution
MDPH's earlier analysis of the geographic distribution of 1982-1988 cancer cases in Palmer, including census tract 8101 where the site is located, and in the area of Monson within a one-mile radius of the PSC Resources site showed no unusual concentration of cases in any area of Palmer or in relation to the site. The most recent cancer incidence data revealed that for the 1982-1986 time period, bladder and lung cancer incidence among males in census tract 8101 is elevated. The 1987-1992 time period analysis revealed that kidney cancer in census tract 8101 for males and females combined is elevated. Thus, MDPH evaluated the geographic distribution of these three cancer types for the entire 1982-1992 period.
Particular attention was shown to census tract 8101 in Palmer where PSC Resources is located. PSC Resources is located at 10 Water Street in Palmer on the east side of the Quaboag River, which divides Palmer from Monson in the south. According to the U.S. Bureau of the Census, Palmer includes three localities, Three Rivers, Thorndike, and Bondsville. The Bondsville locality runs across the border to include part of Belchertown in Hampshire County. The town of Monson has one locality, Monson Center. The majority of streets in Palmer and Monson are clumped in each of these localities which act like town centers. Each locality in Palmer has a separate zip code which further defines these small areas. However, the locality boundaries are not defined by census tract borders, which intersect the localities. This report analyzed census tracts because population data was available for these units.
Analysis of the geographic distribution of 1982-1992 cancer cases (bladder, kidney, lung) in Palmer and Monson showed no unusual concentration of cases. It appeared that cases were concentrated in the south of Palmer, where PSC Resources is located. However, this area also contains the largest group of streets in Palmer and is densely populated. Thus, one would expect that the majority of cases would reside in this area.
d. Smoking Status
Cigarette smoking is known to be the principle causative factor in the development of lung cancer, and has been strongly associated with the development of bladder cancer and kidney cancer. Since significant elevations were noted in lung, bladder and kidney cancers in Palmer in addition to an elevation in lung cancer incidence in Monson, the smoking status of individual cases was evaluated.
In Palmer, 57% (n=21) of bladder cancer cases were current or former smokers at the time of diagnosis. Further evaluation for the significant elevation in male cases in census tract 8101 for the 1982-1986 period showed that 5/12 cases had a positive smoking history, and 3/12 cases never smoked. Smoking information was not available for the other 4 cases. Kidney cancer analysis in Palmer indicated that 55% (n=12) were current or former smokers at the time of diagnosis. In census tract 8101 for the 1987-1992 period 55% (n=11) reported a positive smoking status. Thus, it is likely that smoking played a role in some of the cases noted in these cancer elevations.
Analysis of smoking status for lung cancer cases for Palmer revealed that 88% (n=99) were current or former smokers at the time of diagnosis. Nine percent of the cases (n=10) reported never having smoked, and for 3% (n=4) the smoking status was unknown. Census tract 8101 for the 1982-1986 time period for males was analyzed further because of the significant elevation noted. Ninety percent (n=19) of males reported a current or former smoking status at the time of diagnosis. Additionally, these cases were diagnosed at the average age of 67 possibly indicating many years of exposure to tobacco smoke. Lung cancer was elevated in Monson during both 1982-1986 and 1987-1992, but these elevations were not significant. Analysis of smoking status shows that 89% (n=54) of these Monson cases reported being current or former smokers at the time of diagnosis. It is likely that smoking played a role in the development of these cases in both Palmer and Monson.
A variety of occupational exposures are either suspected or known to be associated with the development of specific types of cancer. Occupation as reported to the MCR at the time of diagnosis was reviewed for cases of bladder, lung and kidney cancer cases in Palmer, and for lung cancers in Monson. The majority of cases reported occupation as "retired" or "at home". However, one kidney cancer case, one bladder cancer case, and five lung cancer cases reported occupations that may have resulted in exposures to chemicals related to the development of these diseases. The occupational data are incomplete and do not include specific job-duty information that could further define exposure potential. Thus, it could not be determined what role occupation may have played in the development of these cancers.
Based upon the environmental and personal risk factor information reviewed, it cannot be determined if the environment played a role in the elevations in lung, kidney and bladder cancer incidence. Due to the elevations in census tract 8101, a discussion of risk factors potentially associated with these cancers is provided below.
Cigarette smoking is the principal risk factor for bladder cancer (ACS 1992, Morrison 1984, Schottenfeld and Fraumeni 1982, Matanoski and Elliot 1981). In Palmer, 57% (n=21) of the cases reported they were current or former smokers. In census tract 8101 for the 82-86 period 42% (n=5) were current or former smokers at the time of diagnosis. Thirty-three percent (n=4) of cases for this period did not report smoking history. More than half of the bladder cancer cases reported a positive smoking history which may have contributed to the development of these cases.
Occupations associated with a high risk of developing bladder cancer include workers in the dyestuffs and rubber industries (due to exposure to aromatic amine compounds) (Page and Asire 1985, Schottenfeld and Fraumeni 1982, Matanoski and Elliot 1981), and in the leather, metal, and textile industries (ACS 1990, Page and Asire 1985, Schottenfeld and Fraumeni 1982). There is some evidence that suggests that a combination of occupational exposures and cigarette smoking may increase the risk of developing bladder cancer (Matanoski and Elliot 1981, Vineis et al. 1981). However, review of the available occupational data for bladder cancer cases in Palmer (33 percent of the cases did not report any occupational information) indicated that only one male case reported an occupation that may have involved exposure to chemicals associated with the development of bladder cancer; this case also reported being a former smoker at the time of diagnosis. Thus, available data do not demonstrate that occupational risk factors may have played an important role in bladder cancer cases.
Epidemiological studies have also suggested other risk factors for bladder cancer. These include residence in urban areas, chronic bladder inflammation, and metabolites of certain foods (ACS 1990, Page and Asire 1985).
When the bladder cancer cases were mapped, there was no evident concentration of cases. It may appear that some cases are concentrated in the area of Palmer where PSC Resources in located, however this area is more densely populated that other parts of Palmer and more cases would be expected in areas of population density.
Cigarette smoking is the most important known risk factor for kidney cancer. Analysis of smoking status in Palmer indicated that 55% (n=12) were current or former smokers at the time of diagnosis. In census tract 8101 for the 1987-1992 period 55% (n=11) reported a positive smoking status. A positive smoking history likely played a role in the development of kidney cancer for these cases.
Kidney cancer is more common among males than females and usually occurs between the ages of 50 and 60. Epidemiologic studies have shown that long-term use of pain relievers such as phenacetin and acetaminophen may increase the risk for cancer or the renal pelvis (ACS 1990). An association has been established between the incidence of von Hippel-Lindau disease in families and the development of renal cell carcinoma, which suggests a genetic factor (ACS 1990, Schottenfeld and Fraumeni 1982). Renal cell carcinoma has also been associated with obesity among women but not among men; some researchers suspect that this may be related to the excess estrogen levels commonly observed in female obesity (LaVecchia et al. 1990, Page and Asire 1985). A diet high in animal fats has been implicated in several epidemiologic studies as a potential risk factor for renal cell carcinoma, however, an association has not been fully established (Schottenfeld and Fraumeni 1982).
Occupational exposures to asbestos as well as exposures in the petroleum industry are also suspected risk factors in the development of renal cell carcinoma (Page and Asire 1985, Smith et al. 1989). Review of the occupational data for Palmer (33 percent of the cases did not report any occupational information) indicated that only one male case reported an occupation that may have involved exposure to chemicals associated with the development of kidney cancer. This case reported a negative smoking history. Thus, available data do not demonstrate that occupational risk factors may have played an important role in kidney cancer cases.
Lung cancer represents approximately 15 percent of all cancers nationwide. Lung cancer incidence increases sharply with age in the 45-74 year age group. Male lung cancer incidence peaks in the 75-84 age group, and female incidence peaks in the 65-74 age group. The incidence is usually 4 to six times higher in males than in females (Schottenfeld and Fraumeni 1982).
Smoking is the most important risk factor for the development of lung cancer. The risk of developing lung cancer depends on the intensity of one's smoking habits (i.e., duration of habit, amount smoked, and tar yield of cigarette). Fortunately, the risk of lung cancer declines after smoking cessation. However, investigators estimate that it can take more than ten years of not smoking for long-term, heavy smokers to reduce their risk to a level similar to that of someone who has never smoked (USDHEW 1989). Lung cancer trends reveal that lung cancer has become a disease increasingly associated with populations of lower socioeconomic status, since these individuals may smoke more (Schottenfeld and Fraumeni 1982). There is evidence that passive smoking (exposure to secondhand smoke) may increase an individual's risk of developing lung cancer. The degree of this risk has not yet been fully established (Janerich et al. 1990, Page and Asire 1985, USEPA 1992). However, recent studies have concluded that environmental tobacco smoke, containing the same carcinogens, is associated with lung cancer in adults and respiratory ailments in children.
The lung cancer analysis of smoking status for cases in Palmer revealed that 88% were current or former smokers at the time of diagnosis. Analysis of census tract 8101, where a significant elevation was observed, revealed that 90% of males reported a current or former smoking status at the time of diagnosis. In Monson, were elevations were noted, 89% of the cases reported being current or former smokers at the time of diagnosis. It is likely that a positive smoking history played a role in the development of these cases in both Palmer and Monson.
Epidemiologic studies indicate that several occupations are associated with an increased risk of developing lung cancer, particularly those involving exposure to asbestos (i.e., mill workers, miners, textile, maintenance, and shipyard workers, roofers, and insulation workers) (Page and Asire 1985). Other known occupational exposures include radon and radon daughters (uranium mining), polycyclic hydrocarbons (coke ovens, roofing, smelting), chromium (plating, spray painting), nickel (smelting, roasting, electrolysis), inorganic arsenic (cooper smelting, production and use of pesticides and herbicides), chloromethyl ethers (chemical industry) and ionizing radiation (mining, X-ray, gamma ray) (Sandler et al. 1992, Page and Asire 1985). Vinyl chloride and beryllium are suspected occupational agents, although epidemiologic evidence is preliminary (Page and Asire 1985). A combination of these occupational exposures and regular cigarette smoking dramatically increases the risk of developing lung cancer (ACS 1990, Page and Asire 1985, Schottenfeld and Fraumeni 1982).
In Palmer, five lung cancer cases reported occupations which may have led to chemical exposures associated with the development of lung cancer. Fifteen percent of the cases in Palmer reported occupation as "retired" or "at home". Job capacity and length of time in that capacity could not be determined. Thus, it could not be concluded what role occupation may have played with these cases.
Urban air pollutants have long been suspected in the etiology of lung cancer, however, no association has been substantiated. It is extremely difficult, at best, to measure low-level effects of any carcinogen in a general population (Page and Asire 1985, Schottenfeld and Fraumeni 1982).
3. Blood Lead Data
Massachusetts law requires annual blood lead testing, at a minimum, for all children between the ages of 1 and 4 years. The MDPH Childhood Lead Poisoning Prevention Program (CLPPP) receives the data on the child's blood lead levels and has data, beginning with October 1, 1986, for each town in the Commonwealth. BEHA staff reviewed these data for Palmer and Monson to determine whether they had test results for a one-year-old child, who resided near the site, reported to BEHA staff as having "elevated lead levels" since the age of 2 months. CLPPP has a record of one test result from the child, taken in 1992 when the child was a little over a year old. The result of blood lead testing revealed that lead was not detected in the blood sample (i.e., blood lead was less than the level of detection for the analytical method [5 µg/dL]). No other test results are available for the child. Thus, BEHA staff were unable to confirm that this child had elevated levels of lead in blood.
We have addressed each of the community concerns about health as follows:
- Will the groundwater around the site result in subsequent contamination of area groundwater wells used for drinking water purposes?
The municipal drinking water supplies for Palmer and Monson are located upgradient from the site. Therefore, they are not currently affected by the contaminants in the aquifer around the site. It is not expected that site related contaminants will ever reach the municipal drinking water supplies. Nor is it expected that the municipal drinking water supplies will ever be drawn from the aquifer underneath or around the site.
Many residents in both town use private groundwater wells as their drinking water source. A groundwater contamination plume extending from the site to the Quaboag River has been identified. The area between the site and river is zoned as agricultural/low density residential and allows for future residences and private drinking water supply wells to be installed in the area of the groundwater contamination plume. However, based on the water distribution information available from the local water departments, there are not any wells in either town that are located in the area which is currently or will likely in the future be impacted by the site related contaminants. In addition, the 1992 Record of Decision provides for the establishment of institutional controls to restrict onsite groundwater use and land development until the site clean-up levels have been achieved and maintained. Cleanup activities include measures to mitigate continued releases to the groundwater and address off-site migration of contamination in groundwater. Long term groundwater monitoring is planned to ensure that cleanup activities are effectively reducing groundwater contamination.
- What are the health risks associated with the site?
The greatest opportunities for exposures to site contaminants may have been to soil or sediment on the property or adjacent wetlands. Unless individuals, particularly young children (e.g., less than 6 years old) regularly accessed contaminated areas of the property or adjacent wetlands in the past, health effects were not likely to result from contact with these areas, despite elevated concentrations of some chemicals, especially lead. In the future, should groundwater wells be installed into areas with contaminated water, residents could become exposed to contaminants in these wells. Appropriate controls, site-cleanup efforts and monitoring should limit the risk of health effects via these pathways.
- Is the athletic field to the north of the site affected by the contaminants? If so, are there any health risks associated with the use of the field?
The surface soil samples taken from the athletic field did not reveal concentrations of health concern for any compound. Thus, we would not expect contact with athletic field soils to pose increased risk of health effects.
- Does the site pose a health risk with respect to household lead exposure?
One case of a young child (age 1 year) with elevated lead levels was reported to MDPH. However, MDPH was unable to confirm that the child had elevated lead levels, instead determining that the only blood lead test result available for the child, taken when the child was a little over one year old, showed no detectable lead in the blood. Elevated lead concentrations have been detected in sediments of surface water bodies on and near the site. It has also been detected in on-site soils. Although no data on soil lead concentrations are available for offsite residential yards, the data do show rapidly decreasing concentrations of lead in soil and sediment as samples were taken further away from the source areas. Hence, the data suggest that elevated soil lead concentrations are likely restricted to the property and areas immediately adjacent to the property (e.g., the spill area).