PETITIONED PUBLIC HEALTH ASSESSMENT
BURLINGTON NORTHERN LIVINGSTON COMPLEX
(a/k/a BURLINGTON NORTHERN RAIL YARD)
LIVINGSTON, PARK COUNTY, MONTANA
ENVIRONMENTAL CONTAMINATION AND OTHER HAZARDS
Tables 3-16 in Appendix B list the contaminants detected in each medium on- and off-site. ATSDR evaluates the contaminants in subsequent sections of the public health assessment and determines whether an exposure to them has public health significance. ATSDR selects and discusses the contaminants based upon the following factors:
Contaminants listed in the On-site and Off-site Contamination subsections will not necessarily cause adverse health effects at the levels detected. Instead, the list indicates which contaminants will be evaluated further in the public health assessment.
ATSDR health comparison values (CVs) are concentrations of contaminants which are media specific (e.g. water, air, or soil). The CVs are considered to be safe under default conditions of exposure and are used as screening values in the preliminary identification of site-specific "contaminants of concern." The "contaminants of concern," listed in the tables in Appendix B, are those contaminants that were detected above the screening CVs and contaminants without CVs. However, the CVs actually listed in the tables in Appendix B reflect the most appropriate CV that allows for 'site-specific' conditions of exposure and not necessarily the screening CV that allows only for 'default' conditions of exposure. Both CVs, whether allowing for site-specific or default conditions of exposure, are considered to be protective of public health. The fact that a contaminant is discussed does not mean that site-specific exposure to the substance will result in adverse health effects. Rather, the contaminant will be evaluated in subsequent discussions in the document. See Appendix C for a description of the comparison values used in this public health assessment.
On-site Groundwater
The two principal types of groundwater contaminants at the BNRY are (1) free-floating petroleum hydrocarbons near or on the water table, known as free product, and (2) dissolved volatile organic compounds (VOCs) in the aquifer (1).
The free-product plume and on-site soils are the primary sources of total petroleum hydrocarbons (TPHs) detected above background levels in groundwater beneath the BNRY. Seasonal fluctuations in the water table mobilize TPHs from the residually contaminated alluvium (1).
Long-term operations at the BNRY resulted in releases of VOCs to the underlying Livingston aquifer. Most of those VOCs originated from cleaning and degreasing agents used in the shops. Dissolved VOCs are present in a plume extending northeast across the BNRY (1).
Groundwater samples collected from eight on-site monitoring wells at the BNRY refueling facility on July 29 and 30, 1987, revealed fuel product in one of the wells, (Well L-87-2). Well L-87-2 contained concentrations of TPHs, 1,2-dichloroethene (DCE), trichloroethylene (TCE), and xylene, with DCE and TCE measuring well above ATSDR comparison values. Well L-87-2 also contained the semi-volatile organic compounds (SVOCs) naphthalene, 1-methylnaphthalene, and 2-methylnaphthalene, with naphthalene measuring above ATSDR comparison values. Table 8 summarizes the results of the above sampling.
Groundwater samples collected from on-site monitoring wells in June 1988 and analyzed for VOCs and SVOCs revealed that concentrations of the compounds 2-chlorotoluene and bis(2-ethylhexyl)-phthalate exceeded ATSDR comparison values (10). Concentrations in 21 additional groundwater samples collected from 12 monitoring on-site wells between August 17 and August 19, 1988 and analyzed for VOCs, indicated tetrachloroethene (PCE), 1,2-dibromo-3-chloropropane, and 1,4-dichlorobenzene (DCB) exceeded ATSDR comparison values. See Table 8 for sampling results.
On-site groundwater samples were analyzed between May 1989 and May 1992 as part of an effort to determine the extent of contamination of the Livingston aquifer. Sampling on-site monitoring wells for VOCs, SVOCs, and polychlorinated biphenyls (PCBs)/Pesticides (1,16) revealed that concentrations of DCE exceeded ATSDR comparison values. No SVOCs or PCBs/pesticides were detected in any of these groundwater samples. Table 8 summarizes the sampling results.
Envirocon sampled 14 on-site monitoring wells August 27 through 30, 1990. All samples were analyzed for TPHs and VOCs (7). None of the compounds detected exceeded ATSDR comparison values and were not included in Table 8.
On-site groundwater samples collected from monitoring wells in May 1989 and February 1990 and analyzed for dissolved metals analyses revealed the presence of arsenic, barium, cadmium, lead, and selenium (1). Of the metals detected, only lead exceeded the comparison values provided in Table 8. Results are included in Table 8.
Twelve groundwater samples, collected from 11 monitoring wells between June 1990 and May 1992 were analyzed for nitrates and nitrites (1,16). Six of the monitoring wells were on-site. Five of six samples collected on-site revealed some level of nitrates present; however, none exceeded ATSDR comparison values. The nitrate levels ranged from 0.17 to 3.87 parts per million (ppm).
On-site groundwater samples were collected during the third and fourth quarters of 1994, and the first and second quarters of 1995. The August 1994 sampling event consisted of 18 primary samples; November 1994 had 21 primary samples; February 1995 had 18 primary samples; and the May 1995 event consisted of 49 primary samples. These samples were collected from 15 on-site monitoring wells and analyzed for PCE, TCE, cis-1,2-DCE, and chlorobenzene. Twelve of the 15 on-site wells contained PCE above ATSDR comparison values; however, these measurements were in much lower concentrations than in previous sampling events. Eight of the 15 wells monitored contained TCE concentrations above ATSDR comparison values. Again, these measurements were much lower than in previous years. One well contained concentrations of cis-1,2-DCE above ATSDR comparison values and there were no wells with chlorobenzene concentrations above ATSDR comparison values.
On-site surficial soil
Envirocon conducted a surficial soil investigation at the BNRY facility during April and May 1992 (3). This investigation, part of the site risk assessment, included areas of potential contamination and potential human exposure. Envirocon separated the BNRY into nine investigation areas. Surficial soil was defined as any unconsolidated material extending from the ground surface to a depth of two inches. Investigators carefully removed layers of ballast, wood chips, or thick vegetation that covered some of the designated sampling locations and sampled the exposed soil at the designated 0 to 2 inch depth. They analyzed samples for VOCs, SVOCs, pesticides, PCBs, and metals. Of the SVOCs detected, benzo(a)pyrene exceeded ATSDR comparison values. The metals detected were below the comparison values provided in Table 13. The levels detected are summarized in Table 13.
On-site subsurface soil
During development of the Interim Remedial Measures Work Plan (IRMWP), the BNRY was divided into 18 investigation areas, where 113 test pits were excavated and 107 subsurface soil samples were taken. The test pits varied in depth from 1 to 15 feet. In the Remedial Investigation (RI) report, the term soil refers to any unconsolidated material, artificial fill or native material that lies above the water table of the Livingston aquifer (16). The soil investigations focused on identifying subsurface sources of groundwater contamination within and around the BNRY. VOCs, SVOCs, petroleum hydrocarbons, and metals were the compounds analyzed. Of the VOCs detected, TCE and vinyl chloride exceeded the ATSDR comparison values provided in Table 16. Of the SVOCs detected, benzo(a)pyrene exceeded the comparison values provided in Table 16. PCB-1248 also exceeded comparison values. The metals detected did not exceed the comparison values provided in Table 16. All the results are summarized in Table 16.
Burlington Northern Rail Yard conducted interim remedial actions at all areas where chlorinated VOCs in vadose zone soils were potentially affecting groundwater quality. Upon completion of the interim remedial actions, a total of 107 confirmation subsurface soil samples were taken from 11 locations. Soil samples were analyzed for PCE, TCE, cis-1,2-DCE, chlorotoluene, chlorobenzene, 1,4-dichlorobenzene, and 1,3-dichlorobenzene. PCE was detected in 7 samples in the electric shop area and TCE was detected in one sample in the electric shop area. A final remedial action plan will be implemented in the electric shop area. Results are included in Table 16.
On-site subsurface soil/Mission Wye
A waste-oil reclamation plant operated at BNRY from 1955 to 1978. The waste products of this process were reclaimed oil, which was stored and sold to waste-oil recyclers, and an acid sludge, which was disposed at Mission Wye from 1955 to 1978. The on-site soil investigation included sampling the acid sludge disposal area and the oil reclamation area at BNRY. Soil samples were taken at a depth of between 1.5 to 7 feet from eight test pit areas targeted for sampling. TPH concentrations were near background concentrations at all eight test pit areas. PCE and several other VOCs were detected at one pit area, although concentrations were below ATSDR comparison values. Metals were detected at two test pits, although concentrations were below the ATSDR comparison values provided in Table 16. Samples of the test pit area are included in Table 16.
On-site soil gas
Twenty-four soil gas samples were collected during soil investigation activities along on-site drainlines that run from the BNRY shop complex to the Wastewater Treatment Plant (WWTP). The investigation focused on active drainlines or those that received heaviest use (1). Of the six compounds detected, TCE, PCE, and cis-1,2-DCE had levels above ATSDR comparison values. Table 3, Appendix B, contains results for this soil gas sampling.
On-site ambient air
The site-wide ambient air monitoring program at the BNRY site measured particulate matter with diameters of less than 10 micronsa (PM10), total suspended particulates (TSP), and PAHs. The monitoring system comprises one upwind and one downwind air monitoring station. Both stations collect PM10 samples and the downwind station also collects TSP and PAH samples (1). The system began operating in November 1990 and continued through March 1991. See Figure 5 for ambient air monitoring locations.
PAH analyses conducted on samples collected at the downwind station for a 24 hour period on 6 separate days during November and December 1990 detected a total of twelve compounds. None of the compounds were at concentrations above ATSDR comparison values. Table 5, Appendix B, reports the results of the analyses.
Upwind and downwind PM10 filters were analyzed for 34 metals. Six total samples were collected at both stations on the same days during the first 6 sampling rounds (1). Arsenic and cadmium were the only metals detected that exceeded ATSDR comparison values. Table 7, Appendix B, reports the results of the analyses.
Twenty-three samples for PM10 analyses were collected from November 10, 1990 to March 31, 1991. The upwind and downwind samples showed PM10 means of 18 and 16 micrograms per cubic meter (ug/m3) and peaks of 56 and 34 ug/m3, respectively. Total suspended particulates were also collected but only at the downwind station. The mean value was 34 ug/m3 and the peak was 62 ug/m3.
Work-zone ambient air sampling events occurred between November 1989 and June 1991 (16). Envirocon collected air samples during remedial activities that had the potential to generate emissions including sludge handling, test pit excavations, and monitoring well drilling. Several ambient air samples were collected on six different days from each one of six on-site locations (5). Monitoring results at several of the locations, including one station designated as background, were below detection limits. However, substances were detected at locations where sludge was being pumped or monitoring wells were being installed. This suggests that the identified pumping and well installation activities may have contributed to the release of the detected substances from the soil to the surrounding air. Particulate matter was detected at 0.5 - 2.1 mg/m3. Naphthalene and phenanthrene were detected, with naphthalene measuring below comparison values. Table 5, Appendix B, reports the results of the analyses.
Envirocon collected one on-site ambient air sample during February 1992 as part of an indoor air quality study of residences nearest to the site. The sample was analyzed for PCE, TCE, cis-1,2-DCE, and trans-1,2-DCE (4). Analyses of the sample revealed the presence of PCE; however, its concentration did not exceed any comparison values. TCE, cis-1,2-DCE, and trans-1,2-DCE remained below their corresponding detection limits of 0.22 ug/m3. See Table 5, Appendix B, for the sampling results.
Off-site groundwater
Monitoring Wells
This section involves discussion of off-site monitoring wells sampled from June 1988 until May 1995. Monitoring wells were located off-site to the northeast, east, south and southwest areas of BNRY and three wells were located to the east of the Yellowstone River in an effort to characterize off-site groundwater contamination. Contaminants detected most often include PCE, TCE, DCE, toluene, arsenic, barium, cadmium and lead. PCE, TCE, cis-1,2-DCE, arsenic, and lead were above ATSDR comparison values. However, concentrations of these contaminants have continued a long-term decline over the last six years. Almost all detectable concentrations of VOCs extend from the shop area at BNRY eastward to the Yellowstone River. All metal concentrations are likewise concentrated in an east area toward the river. It is safe to assume this data reflects conclusions that VOC contamination in groundwater has migrated off-site and that there is a need for full evaluation of the public health significance of the contaminants.
Groundwater samples obtained from off-site monitoring wells near BNRY in June 1988 were analyzed for VOCs and SVOCs (10). Analyses revealed PCE in six off-site wells and TCE, cis-1,2-DCE, and trans-1,2-DCE in no more than one well each. The trans-1,2-DCE, TCE, and the highest PCE levels were detected in the same well. TCE and PCE were above ATSDR comparison values. See Table 9 for sampling results.
Between August 17 and August 19, 1988, nine off-site monitoring wells were analyzed for VOCs (9). None of the compounds were detected at concentrations that exceeded comparison values. Table 9 summarizes the results of the sampling.
Off-site monitoring wells were analyzed for dissolved metals in May 1989 and February 1990. The May 1989 results revealed arsenic, barium, cadmium, and lead. Arsenic exceeded ATSDR comparison values. Lead exceeded the Environmental Protection Agency's (EPA) action level. February 1990 results revealed arsenic, cadmium, and lead. Again, arsenic exceeded ATSDR comparison values and lead was above EPA's action level. See Table 9 for sampling results.
Envirocon conducted off-site groundwater sampling at twenty off-site wells from August 27-30, 1990 (7). These wells included 13 monitoring wells, 5 private wells, and 2 municipal wells. This paragraph will report data from monitoring wells only. Data on the private and municipal wells appear in later sections. All samples were analyzed for TPHs and VOCs. Available data analyses indicate that TPH in all samples appeared at less than 1 ppm. VOCs detected were DCE, TCE, and PCE. DCE was detected in 2 of 13 monitoring wells, with concentrations of 26 and 3.2 micrograms per liter (µg/l). Neither exceeded ATSDR comparison values. TCE was detected in 2 of 13 monitoring well samples at levels of 12 and 3.6 µg/l. Both of the sample concentrations exceed the 3.0 µg/l cancer risk evaluation guide (CREG) screening value. PCE was detected in 4 of 13 monitoring well samples at concentrations of 158, 40, 15, and 4.6 µg/l. All four sample concentrations are greater than the 0.7 µg/l CREG screening value.
Groundwater samples collected from six off-site monitoring wells between June 1990 and May 1992 were analyzed for nitrates and nitrites. Nitrates were detected in all six wells with concentrations ranging from 0.36 to 2.72 ppm. None exceeded ATSDR comparison values (1).
Groundwater sampling from off-site monitoring wells was performed during January, February and March 1992 (11). January 1992 sampling of two wells for VOCs detected 1,1,1-trichloroethane (1,1,1-TCA) at 0.5 parts per billion (ppb) in one of the samples. This was the first detection of 1,1,1-TCA in any monitoring well; however, a monitoring well sampled during February 1992 indicated 1,1,1-TCA at 0.72 ppb.
In February 1992, a groundwater sampling program collected samples at 11 off-site monitoring wells (11). Samples showed PCE in 10 wells, TCE in 2 wells, and DCE in 2 wells. Levels were above ATSDR comparison values for all three contaminants. See Table 9 for summary results.
In March 1992, samples were collected from three off-site monitoring wells and analyzed for VOCs. None of the target analytes and no other compounds were detected (11).
A groundwater sample collected in May 1992 from Monitoring Well 92-2 on the east side of the Yellowstone River indicated PCE at 15 ppb and TCE at 1.5 ppb. Evidence does not indicate whether underflow beneath the river or VOC sources east of the river were responsible for the presence of the PCE and TCE.
Twenty two off-site monitoring wells and one municipal well (Well B) were sampled during the third and fourth quarters of 1994 and first and second quarters of 1995. A total of 106 samples were analyzed for PCE, TCE, cis-DCE, and chlorobenzene. Thirteen of the 22 off-site monitoring wells (including one well east of the Yellowstone River) contained PCE above ATSDR comparison values, though at much lower concentrations than in the previous six years. TCE was found above ATSDR comparison values in 4 of 22 wells; however, neither of the wells east of the river indicated TCE contamination. Cis-DCE and chlorobenzene were not found in any of the well samples.
The 1997 Annual Ground Water Sampling Report for the BNRY site stated that 43 samples from 28 monitoring and private wells were collected by Envirocon in May 1997. The report summarized data from bi-annual sampling events between August 1995 to May 1997. The referenced concentrations are all below historical maximums.
Municipal Wells
The municipal water supply system was installed in the 1960s and 1970s. In 1988, six wells served the system, including the B Street, D Street, L Street, Q Street, Clarence, and Werner wells. In April 1988, DHES sampling revealed the presence of PCE in the L Street, Q Street, and Werner wells, at concentrations of 0.45, 0.86, and 0.01 parts per billion (ppb), respectively. The PCE concentrations are below the drinking water standard for PCE (5 ppb). The PCE concentration measured in the Q Street well (0.86 ppb) did exceed ATSDR's cancer risk evaluation guide (CREG) screening value of 0.7 ppb.
As a precaution, however, the City of Livingston took the L and Q street wells out of service. These wells were replaced with two new wells (Clinic and Billman Creek), located more than 3,000 feet south of BNRY and upgradient of the plume (i.e., in the opposite direction from which the contamination is moving).
Envirocon conducted additional sampling of the municipal wells (excluding Clinic and Billman Creek wells, but including the out of service wells) between May 1989 and May 1992 (16). PCE was consistently detected in only the Q Street well, at levels ranging from 0.50 to 0.95 ppb. No volatile organic compounds (VOCs) were detected in the B Street, D Street, L Street, Clarence, or Werner municipal wells.
Dissolved metals analyses of groundwater samples obtained in February 1990 revealed arsenic, cadmium, and lead in municipal well water at concentrations of 0.009, 0.001, and 0.01 ppm, respectively (16). However, the arsenic was detected in the L Street Well, which was not in service. Cadmium detected in the B Street Well was below the ATSDR chronic environmental media evaluation guide (EMEG) of 0.007 ppm. The B Street Well continues to be sampled on a semi-annual basis. Lead did not exceed EPA's action level.
Water samples were collected on September 20, 1989, from five points along the municipal water distribution system and from the one-million gallon storage tank on the town's west end (1). No detectable concentrations of chlorinated ethenes were reported in any samples (1).
Additional municipal well sampling for PCE, TCE, and DCE August 27-30, 1990 did not detect any of these contaminants.
Dissolved petroleum hydrocarbons are present in the Livingston aquifer. Groundwater samples were collected from the B, D, L, and Q street wells and from the Clarence and Werner wells between August 1989 and May 1991. Results of the analyses for TPHs were all below detection limits (less than 0.1 ppm) except for one sample in November 1989 detected at 0.1 ppm (1).
The D Street, Clarence, and Werner Wells are being sampled every three years. Quarterly sampling of the B Street well was performed from August 1994 through May 1995. Samples collected were analyzed for PCE, TCE, cis-DCE, and chlorobenzene. No detectable concentrations of these contaminants were reported in any of these samples. Currently, semi-annual groundwater sampling events of the B street well are part of the ongoing site investigation because of it's proximity to the site (approximately 3 blocks upgradient). Sampling results included in the 1997 Annual Ground Water Sampling Report (Envirocon, August 1997) show that no VOCs were detected in the B Street Well during sampling events in November 1996 and May 1997.
Private Wells
Fourteen private water wells were randomly sampled for VOCs during July and August 1989. None of the wells north of BNRY contained detectable concentrations of VOCs. People are continuing to use these wells for drinking water. VOC contaminants were identified in six of the wells located south of BNRY. Of the contaminants detected, PCE, TCE, and cis-1,2-DCE exceeded ATSDR comparison values. None of the private water wells containing these VOCs are currently used for drinking water. Before the 1989 sampling, four of the six wells furnished water for either drinking or domestic use (1). After the 1989 sampling, well owners were notified of possible contamination and provided with municipal water. See Table 11 for the results.
In February 1990, arsenic was detected at a level that exceeded ATSDR comparison values in a private well located south of BNRY near the Yellowstone river. This well is currently used for drinking water and domestic purposes. Overall, dissolved metals have not shown correlation with other groundwater contaminants and occur at concentrations consistent with naturally occurring metal concentrations in the Yellowstone River (16). Specifically, arsenic is considered to be a naturally occurring metal in the river (1). Additionally, the general groundwater flow is either east or north; this well is south of BNRY near the river. These circumstances suggest that the arsenic detected in this well originates from natural sources and not from BNRY. See Table 11 for the level detected.
Three private wells were sampled August 27-30, 1990 for VOCs and two other private wells were sampled for dissolved metals. At the time of sampling, none of the wells were used for drinking water. PCE, TCE, and DCE were found in two of three wells above ATSDR comparison values. The wells with these contaminants are southeast of BNRY but are in close proximity to BNRY. Metals were all below detectable limits. Table 11 presents the results.
Off-site surficial soil investigation
Envirocon conducted a surficial soil investigation at BNRY during April and May 1992 (12). Five off-site background surface soil samples were collected. For this investigation, surficial soil was defined as any unconsolidated material extending from the ground surface to a depth of two inches. Envirocon removed layers of ballast, wood chips, or thick vegetation that covered some of the designated sampling locations and sampled the exposed soil at the designated 0 to 2 inch depth. Samples were analyzed for VOCs, pesticides and PCBs, metals and PAHs. All of the contaminants detected were below the comparison values provided in Table 14. See Table 14 for the levels detected.
Off-site soil gas
Remediation Technologies, Inc. (ReTec) performed a soil gas survey at and in the vicinity of the BNRY facility during August 1988 in an effort to correlate migration of off-site soil gas with that of on-site gas. ReTec collected 168 soil gas samples. The survey detected PCE, TCE, DCE, TCA, benzene, toluene, xylene, and ethylbenzene in varying concentrations in the collected soil gas samples. The VOC concentrations were largest above the source areas. Contour plotsb of the detected VOCs reveal that PCE and, to a lesser degree, TCE have apparently migrated off-site at BNRY (9). The migration patterns are much less apparent for the other detected compounds. In several of the soil gas plots, there are contours that reveal off-site areas of soil gas contamination appearing to be disconnected from the on-site areas. These off-site soil gas contamination areas may or may not be related to the BNRY site. Based on the August 1988 soil gas sampling results, it appears, in general, that it is possible to generate qualitative approximations of soil gas contamination based on detected groundwater contaminants.
The contaminant levels in off-site soil gas appear to correlate poorly with the levels of indoor air contaminants detected in residences located in the general area of soil gas contamination. For example, PCE is thought to be the soil gas contaminant that has most noticeably migrated from the BNRY site (9); however, air samples obtained at houses located in the general area of soil gas contamination (east and west of the site) did not reveal the presence of PCE. However, it must be noted that some of the detection limits of the 1989 and 1990 indoor air sampling rounds were higher than the detection limits used in the soil gas samples.
Indoor residential air
The Montana Department of Health and Environmental Sciences (MDHES) performed indoor air monitoring for four residences during August and October 1989. MDHES also sampled the indoor air of two of the previous four residences plus an additional four residences in November 1989 (1,15,16). The August 1989 analyses by MDHES revealed the presence of 1,1,1-TCA in each house sampled. In October 1989, 1,1,1-TCA was detected in two houses at levels thousands of times lower than the August 1989 detection levels. There was no detection in the remaining two residences. In November 1989, with detection limits lower than in the August monitoring period but higher than in October, 1,1,1-TCA was not detected in any of the six houses sampled. Detection limits for the November sampling ranged from 200 to 450 ug/m3.
TCE was not detected in the MDHES analyses for August, October, and November 1989. Analyses of TCE were not requested in two residences during the August 1989 sampling rounds. However in October 1989, the two residences were analyzed and TCE remained below detection limits. PCE was not detected in any of the MDHES analyses (1,15). Trans-1,2-DCE was detected in the October 1989 sampling round but was not detected in any other sample. Toluene was detected in two residences during the August 1989 sampling round. The same two residences showed no toluene in the October 1989 sampling. See Table 12, Appendix B, for the levels detected.
Envirocon performed air sampling of two residential basements during November 1990 (1). These analyses revealed the presence of ethylbenzene, xylene, and TCE. Only TCE exceeded ATSDR comparison values. All of the analyses for PCE, benzene, 1,2-DCE, and toluene remained below detection limits (1). Detection limits during the November 1990 indoor air sampling round ranged from 50 to 90 ug/m3. See Table 12, Appendix B, for the levels detected.
Envirocon performed additional sampling of indoor air in February 1992, in the basements and upstairs of 17 residences. The 17 homes were distributed over three study areas. Samples were analyzed for PCE, TCE, cis-1,2-DCE, and trans-1,2-DCE. PCE and TCE were the only contaminants detected, and both exceeded ATSDR comparison values. PCE was detected both upstairs and in the basements in all residences. TCE was found upstairs and in the basements in 13 of the homes. Detections of both contaminants were greater in the upstairs of homes than in the basements. One outdoor ambient air sample was also collected from each study area.
A follow-up investigation was performed in March 1992 at selected residences. Analyses were performed for vinyl chloride as well as for the previous target compounds (4). During these sampling rounds, PCE and TCE were detected. The cis-1,2-DCE isomer was detected in four residences, the trans-1,2-DCE isomer was detected in two residences, and vinyl chloride was detected in two residences. The residences containing the vinyl chloride detection overlie an uneven part of the groundwater plume and are across the street from a petroleum storage facility. Those circumstances suggest that vinyl chloride does not originate from BNRY. Table 12, Appendix B, summarizes results for these sampling rounds.
Envirocon collected 68 indoor air samples, including 6 duplicates, at 36 houses and 1 school during January and February 1993 to evaluate the possible contribution of PCE migrating from the BNRY in groundwater or soil gas (6,16). Nine of the 36 residences sampled had detectable levels of PCE in the air and concentrations were above ATSDR comparison values. Review and determinations made through sampling log books indicated that the PCE originated from sources unrelated to BNRY. Table 12, Appendix B, summarizes results for these sampling rounds.
Off-site ambient air
Envirocon collected four off-site ambient air samples on January 27, 1993 (6). Three of the four sampling stations were located upwind of the site. PCE was detected at two of the upwind stations, with both concentrations exceeding ATSDR comparison values. PCE was also detected at the downwind sampling location, but at a concentration less than ATSDR comparison values. The data suggest that the BNRY facility contributions of PCE to the ambient air during the period sampled may have been smaller than the contributions of sources in the Livingston business district. See Table 6, Appendix B, for results of the sampling.
Envirocon took two off-site ambient (outdoor) air samples during February 1992. The samples were analyzed for PCE, TCE, cis-1,2-DCE, and trans-1,2-DCE (4). Analyses of these off-site ambient samples revealed the presence of PCE, although no comparison values were exceeded. TCE, cis-1,2-DCE, and trans-1,2-DCE remained below their corresponding detection limits, which ranged from 0.17 to 0.18 ug/m3. See Table 6, Appendix B, for the results from the sampling described above.
Yellowstone River water samples
One Yellowstone River water sample was collected on September 20, 1989 from a point 50 feet upstream from the 9th Street Bridge. This sample was analyzed for VOCs, TPHs, and dissolved metals. Only TPH, at 0.1 ppm, was detected in this sample.
On March 23, 1990, during a period of low water flow, six Yellowstone River water samples were collected at five locations upstream and downstream of the BNRY. These samples were analyzed for VOCs, SVOCs, PCBs/pesticides, TPHs, and total metals. No SVOCs or PCBs/pesticides were detected in any of these Yellowstone River water samples. One sample taken immediately downstream of the Livingston treatment works plant revealed PCE, 2-chlorotoluene, and TPH. Arsenic was detected in all of the samples, and cadmium was detected in one sample (1). Of the contaminants detected, none exceeded the comparison values provided in Table 15. Arsenic is thought to be naturally occurring and probably originates from warm springs that enter the river in Yellowstone Park. See Table 15, Appendix B, for the levels detected.
Yellowstone stream sediment and river gravel analyses
Oily waste material was released during January 1990 from the wastewater treatment plant (WWTP) discharge line into the gravel river bed lining within the Yellowstone River. Remedial activities included removal of the oily material from the in-line sumps and removal of a portion of the drainline (1).
Stream sediment and gravel were sampled from the Yellowstone River on March 7 and March 21, 1990 at 0.0 and 0.2 ft. depths (See Figure 4 in Appendix A for River sediment sampling locations). Gravel was sampled at the outfall of the abandoned WWTP discharge line, and sediment samples were collected from the banks of the Yellowstone River, Sacajawea slough outfall. Sample locations included areas downstream from the WWTP's discharge line and one location upstream (1). All of the contaminants detected were below the comparison values provided in Table 10. See Table 10 for the results.
Fish tissue samples from the Yellowstone River
Sixteen fish were collected from the Yellowstone River on October 4, 1988. Liver and muscle tissue samples from these fish were submitted to the Montana State University for VOC and SVOC analyses. Results revealed all of the target compounds remained below detection limits. The detection limits were 15 ppm for most of the VOCs and 1 ppm for the SVOCs of interest (13).
Bird tissue samples
Samples of bird tissue were submitted to the Montana State University for VOC and SVOC analyses. Results revealed that all of the target compounds remained below detection limits. The detection limits were 15 ppm for most of the VOCs and 1 ppm for the SVOCs of interest (14).
C. Quality Assurance and Quality Control
ATSDR staff relied on the information provided in the referenced documents to prepare this public health assessment. The Agency assumes that adequate quality assurance and quality control measures were followed with regard to chain-of-custody, laboratory procedures, and data reporting. The validity of the analyses and conclusions drawn for this public health assessment depends upon the reliability of the referenced information.
The entire BNRY property is unfenced. Most hazards at the site are related to normal train yard activities. Train movements within and through the yard occur on a daily basis. A large pile was observed by ATSDR on an earlier site visit. According to DEQ, the cinder pile still remains at the site and is accessible. Based on sampling of the contents of this pile, asbestos was identified. Access to this material will be addressed in the final site plan.
To determine whether nearby residents are exposed to contaminants migrating from the site, ATSDR evaluates the environmental and human components that lead to human exposure. The pathways analysis consists of five elements: a source of contamination, transport through an environmental medium, a point of exposure, a route of human exposure, and an exposed population.
ATSDR categorizes an exposure pathway as a completed, potential, or eliminated exposure pathway. Completed pathways require that the five elements exist and indicate that exposure to the contaminant has occurred, is occurring, or may occur. Potential pathways are those in which one of the five elements is missing, but could exist. Potential pathways indicate that exposure to a contaminant could have occurred, could be occurring, or could occur in the future. In an eliminated exposure pathway, at least one of the five elements is missing and will never be present. The discussion that follows describes only those pathways that are relevant to the site. It also includes information on eliminated exposure pathways. Table 2 presents the completed and potential pathways, and their elements.
A. Completed Exposure Pathways
Off-site groundwater
Groundwater monitoring data indicate that chlorinated solvents (VOCs), metals, and total petroleum hydrocarbons have been identified in the groundwater. Contaminants of concern include trichloroethylene (TCE) and tetrachloroethene (PCE). This contamination originates from spillage of diesel fuel during refueling operations that began in 1947 at BNRY. The volume of diesel fuel on the water table was estimated at 300,000 gallons in 1989. Currently, the volume of diesel floating on the water table is estimated to be 150,000 gallons. Maintenance and cleansing operations also contributed heavily to the contamination. Chlorinated solvents have been identified from wastewater treatment sumps, oil and water separator ponds, and overflow ponds.
Monitoring data indicate that contamination from BNRY has probably migrated off-site. Contaminants can migrate through the groundwater, which flows either east or north toward the Yellowstone River depending on the water table level and the season. Contaminants may also flow beneath the river, although it is inconclusive as to whether VOC detections east of the river originate from BNRY at this time. This groundwater is a source of drinking water.
A past completed exposure pathway was identified for municipal wells which the residents of Livingston utilized for drinking water and other domestic purposes. The two municipal wells with PCE contamination have since been taken off line and replaced with two municipal wells located south of BNRY. Past exposures occurred among residents of Livingston who drank water obtained from the municipal wells where PCE was detected. Of the municipal wells currently in use, five are sampled every three years while one well (B Street) is monitored semi-annually.
A past completed exposure pathway to VOCs was identified for private wells used by residents for drinking water. Past exposures occurred among residents of Livingston who drank from the private wells in which contamination was confirmed. None of the wells where VOC contamination was detected currently provide drinking water. Owners of private wells south and east of BNRY were notified of potential VOC contamination. Well owners who received such notification now use the wells for irrigation and other domestic purposes but not for drinking water.
A past, current, and future exposure pathway exists for residents who use the private well where arsenic was detected. This well is located south of BNRY near the Yellowstone river and is currently used for drinking water and other domestic purposes. To date, dissolved metals have not shown correlation with other groundwater contaminants. The general groundwater flow in the area is either east or north; this well is south of BNRY near the river. Additionally, the Yellowstone river contains arsenic that is thought to be naturally occurring. These circumstances suggest that the arsenic detected in this well originates from natural sources and not from BNRY.
Exposure routes other than drinking water would include dermal absorption and ingestion of foods that have been irrigated or washed with contaminated groundwater. Exposed populations would include residents of Livingston who get their drinking water from this groundwater source.
Off-site ambient air
The source of contamination to the residents of Livingston could be from two origins, BNRY and the Livingston business district. The sampling data available suggest that the business district of Livingston may be a contributor to the air quality of BNRY. However, a completed past, present and future exposure pathway exists from contamination of the air in the vicinity of BNRY.
Contaminants released to the ambient air are dispersed by the winds. Any contaminants emitting from BNRY and the city of Livingston will mix and disperse throughout the area. Off-site sample data indicated the presence of PCE at a higher concentration upwind of BNRY than what was recorded downwind of BNRY. Data collected from off-site ambient air sampling indicated past exposures to persons who live, work, and play in the areas where sampling occurred. Because BNRY continues to operate as a rail yard with refueling by means of a tanker truck and railcar maintenance operations, present and future completed pathways also exist for residents who live within the vicinity of BNRY.
Exposed populations would include residents within the areas where prevailing winds mix the air from BNRY and surrounding area.
Indoor residential air
A past, present, and future completed exposure pathway was identified for residential homes. Various indoor air contaminants, most noticeably PCE; TCE; 1,1,1-trichloroethane (TCA); vinyl chloride; and toluene, were detected during sampling of residences in the city of Livingston. Persons who resided or worked in these homes experienced past exposures from these contaminants. It is not certain, however, that the contaminants detected in indoor residential air are associated with emissions from BNRY. The possibility exists that some of the contaminants detected may have sources originating inside the residences.
Contamination may occur as a result of chemicals inside the home and/or emissions from BNRY. Air currents and wind dispersion may bring contaminant concentrations into any given area, including residences. Residential buildings themselves would help keep chemical vapors from any chemicals stored in the home in a defined area. Sample data suggest that present and future exposures to indoor contaminants are likely to occur when residents of these homes breathe the indoor air directly. Exposed populations include adults and children who reside in these homes.
On-site ambient air
Various ambient air contaminants, most noticeably PCE, arsenic, and cadmium, were detected during on-site ambient air sampling at the BNRY. Some of the contaminants may have been introduced into the breathing zone by activities which disturbed the soil. Other contaminants may have resulted from outside emissions unaffiliated with BNRY. Regardless of the sources, measurements of the air flow through BNRY indicate that workers and on-site personnel have been exposed to on-site contamination in the past. Contaminants emitting from on-site operations, cleaning operations, and maintenance, could result in present and future exposure to on-site contamination.
Points of exposure would include workers and others who entered the area of the BNRY or visitors who had direct contact with windblown dust or with contaminants that volatilized from disturbances of the soil.
Routes of exposure include inhalation of ambient air in the surrounding areas of BNRY. Exposure may occur through inhalation of contaminants in airborne dust from contaminated soil or in emission of contaminants from the facility itself. Workers and visitors to BNRY are the receptor populations.
B. Potential Exposure Pathways
Off-site groundwater
A future potential pathway exists for persons north of BNRY who currently use private wells for drinking water. Residents who are currently consuming water from private wells north of BNRY could experience exposure if the groundwater plume area expands. Currently there are monitoring wells located just northeast of BNRY. Quarterly monitoring of these wells indicate contaminant levels that do not present a threat to human health. Private wells will be sampled should any monitoring wells in the north area reveal contamination levels that may be of health concern.
Yellowstone River fish
Fishermen use the Yellowstone River extensively. It is therefore very likely that people have consumed fish caught in the Yellowstone River. Sampling data from fish tissue, however, revealed that VOCs and SVOCs were below laboratory detection limits. The detection limits were 15 ppm and 1 ppm, respectively. However, the detection limits for this particular sampling exercise were well above ATSDR comparison values. This would indicate that a past, current, and future exposure pathway may exist from the ingestion of fish. Potential exposures for individuals periodically consuming such fish are not expected to be of health concern.
On-site soil gas
Twenty-four soil gas samples were collected during soil investigation activities along on-site drainlines that run from the BNRY shop complex to the Wastewater Treatment Plan (WWTP). It is reasonable to assume that a given gaseous contaminant released from the soil into the air would be diluted to a lower concentration in the air space above the soil. Samples taken between January 21, 1990, and March 4, 1990, revealed airborne substances at stations where sludge was being pumped or monitoring wells were being installed. The levels detected were orders of magnitude higher than the levels detected during the ambient air sampling program. This finding suggests that the mentioned site activities may have contributed to the release of the detected substances from the soil to the surrounding air. The data thus appear to represent a transient or temporary situation.
The contaminants detected could have resulted in a past exposure pathway for workers to on-site soil gas. However, sampling data indicate that contamination may be temporary. No basement sampling of the buildings on-site was performed. A present or future exposure pathway could result if soil gas contaminants are found to be stationary as a result of on-site activities. Soil vapor extraction wells were included as a part of a MDHES interim remediation plan. The well system has minimized and/or eliminated the soil gas contaminant pathway for workers at BNRY.
Off-site soil gas
Soil gas surveys in 1988 revealed that various VOCs have apparently migrated off-site. PCE and TCE, which slightly exceeded ATSDR comparison values, are most prominent among those VOCs. Contour plots performed during the survey suggest that migration patterns are strongest in areas where groundwater contamination is known to exist. However, the contour plots fail to show that off-site soil gas originated at BNRY. Data from samples also fail to show a connection between off-site soil gas and on-site gas, indicating off-site gas in these particular homes may be of other origins.
Residents and workers of homes in the sampled area could have experienced past exposures from elevated levels of VOCs. Present and future exposures for residents and workers could exist because of such soil-disturbing activities as drilling, building, or excavation of basements. Additional sampling during these soil-disturbing activities is advised.
C. Eliminated Exposure Pathways
On-site groundwater
On-site groundwater was not used for human consumption in the past. The final clean up remedy will insure on-site groundwater is not used for human consumption in the future. Exposure to this medium is not likely, and this pathway is eliminated.
On-site subsurface soil
Subsurface soil pits varied in depth from 1 to 15 feet. Exposure to contaminated soils found at depths of 1 to 15 feet is highly unlikely; therefore, ATSDR concludes this is an incomplete exposure pathway.
Yellowstone River sediment
It is unlikely that recreational users may have ingested sediments in quantities of concern from the Yellowstone River. ATSDR has thus eliminated this exposure pathway.
Bird tissue
Montana State University staff analyzed bird tissue for volatile organic compounds (VOCs) and semivolatile organic compounds (SVOCs). All of the target compounds remained below detection limits. The species of bird analyzed, the circumstances under which bird samples were obtained, and the relationship the bird may have had to the site under evaluation are not known. The bird species is not thought to be part of the human food chain; therefore, ATSDR has classified bird tissue as an eliminated pathway.
Yellowstone River water
Recreational users of the Yellowstone River would be unlikely to ingest water in quantities large enough to be of health concern. Sampling data indicated small amounts of contaminants upstream of BNRY, thus suggesting these contaminants are of other origin or naturally occurring. ATSDR has thus eliminated this exposure pathway.
On-site surficial soil
Sampling data indicate levels of contaminants that are not thought likely to produce illness. It is unlikely that workers would have direct contact with on-site soil in quantities large enough to produce illness. ATSDR has thus eliminated this exposure pathway.
Off-site surficial soil
Low concentrations of VOCs, SVOCs, and metals were detected in off-site soils at
levels that were not thought sufficient to cause illness. ATSDR has thus eliminated
this exposure pathway.
Table 2: COMPLETED EXPOSURE PATHWAYS
| Pathway name |
|
Time | ||||
| Source | Media | Point of exposure | Route of exposure | Exposed population |
||
| Off-site groundwater | BNRY | MUNICIPAL WELL WATER | RESIDENTIAL WATER USE | INGESTION, INHALATION, DERMAL | MUNICIPAL WELL WATER CONSUMERS | PAST |
| Off-site groundwater | BNRY/ NATURALLY OCCURRING (METALS) | PRIVATE WELL WATER | RESIDENTIAL WATER USE | INGESTION, INHALATION, DERMAL | PRIVATE WELL WATER CONSUMERS | PAST CURRENT FUTURE |
| Off-site ambient air | BNRY/ LIVINGSTON BUSINESS DISTRICT | AIR | RESIDENTIAL AREAS | INHALATION | RESIDENTS OF LIVINGSTON | PAST PRESENT FUTURE |
| Indoor air | BNRY/ RESIDENCES | AIR | RESIDENCES (INDOOR) | INHALATION | RESIDENTS OF LIVINGSTON | PAST PRESENT FUTURE |
| On-site ambient air | BNRY | AIR | ON-SITE | INHALATION | WORKERS AND CLIENTS | PAST PRESENT FUTURE |
| Off-site ground- water | BNRY | PRIVATE WELL WATER | RESIDENTIAL USE | INGESTION | RESIDENTS | FUTURE |
| Yellowstone River fish | BNRY | BIOTA | FISH CONSUMPTION | INGESTION | FISH CONSUMERS | PAST PRESENT FUTURE |
| Off-site soil gas | BNRY | AIR | OFF-SITE | INHALATION | RESIDENTS, WORKERS |
PAST PRESENT FUTURE |
| On-site soil gas | BNRY | AIR | ON-SITE | INHALATION | WORKERS | PAST PRESENT FUTURE |
This section contains discussions of health effects that could plausibly result from exposures to site contaminants. While the relative toxicity of a chemical is important, the response of the human body to a chemical exposure is determined by several additional factors, including the concentration (how much), the duration of exposure (how long), and the route of exposure (breathing, eating, drinking, or skin contact). Lifestyle factors (e.g., occupation and personal habits) have a major impact on the likelihood, magnitude, and duration of exposure. Individual characteristics such as age, sex, nutritional status, overall health, and genetic constitution affect how a contaminant is absorbed, distributed, metabolized, and eliminated from the body. A unique combination of all these factors will determine the individual's physiological response to chemical contaminants and any adverse health effects the individual may suffer as a result of the chemical exposure.
ATSDR has determined levels of chemicals that can reasonably (and conservatively) be regarded as harmless, based on the scientific data the agency has collected in its Toxicological Profiles. The resulting comparison values and health guidelines (which include ample safety factors to ensure protection of sensitive populations) are used for screening contaminant concentrations at a site, and to select substances that warrant closer scrutiny by agency health assessors and toxicologists. (See Appendix C for a more complete description of ATSDR's comparison values, health guidelines, and other values ATSDR uses to screen site contaminants.)
ATSDR's comparison values and health guidelines do not represent thresholds of toxicity. They are screening values used to facilitate the initial selection of site-specific chemical substances (known as "contaminants of concern") for further evaluation of potential health effects. After the contaminants of concern at a site have been identified, they must be individually scrutinized in more detail (including consideration of all the various factors mentioned in the first paragraph of this section) to determine whether or not, under site-specific conditions, they represent a realistic threat to human health. Although concentrations at or below ATSDR's comparison values may reasonably be considered safe, it does not automatically follow that any concentration above a comparison value will necessarily produce toxic effects. Comparison values are intentionally designed to be lower, often orders of magnitude lower, than the corresponding no-effect levels determined in laboratory experiments.
Solely for the purpose of screening individual contaminants, ATSDR typically compares the lowest comparison value available (i.e., CREGs or other chronic exposure values) for the most sensitive, potentially exposed individuals (usually children or pica children) to the highest single concentration of a contaminant detected. This high degree of conservatism results in the selection of many contaminants as "chemicals of concern" that will not, upon closer scrutiny, be judged to pose any hazard to human health. However, ATSDR judges it prudent to use a screen that "lets through" many harmless contaminants rather than one that overlooks even a single potential hazard to public health. Even those contaminants of concern that are ultimately labeled in the toxicological evaluation as potential public health hazards are so identified solely on the basis of the maximum concentration detected. The reader should keep in mind the protectiveness of this approach when considering the potential health implications of ATSDR's toxicological evaluations.
In this section, only those contaminants that have been detected at concentrations exceeding the relevant comparison values will be discussed in any detail. Note that all contaminants at BNRY were initially screened by comparing their maximum detected concentrations to the lowest available comparison values, as described above. However, the data tables in the Appendix generally list only the more relevant, site-specific comparison values, where available, in order to provide a more useful perspective.
Since a contaminant must first enter the body before it can produce any effect, adverse or otherwise, on the body, this evaluation will further focus on those contaminants associated with completed pathways of exposure. Soil-gas, for example, is not discussed in this toxicological evaluation because soil-gas measurements are not representative of levels in the breathing zone (i.e., do not reflect a completed exposure pathway) and are therefore not directly relevant to human exposure. The only completed pathways at BNRY were on- and off-site air, indoor air, and off-site groundwater. (See Parts B and C of the Pathways Analysis section for discussions of potential and eliminated pathways.)
PCE, TCE, arsenic, and cadmium were detected in completed pathways at concentrations that exceeded relevant comparison values. Each of these 4 contaminants is discussed separately below. Toluene, lead, cis-1,2-DCE, and 1,1,1-TCA were detected less frequently and/or at generally lower levels when they were detected, but their maximum concentrations did exceed some of ATSDR's comparison values. These four contaminants are discussed together under the last subsection entitled "Other Contaminants".
PCE
PCE (perchloroethylene, tetrachloroethylene, or tetrachloroethene) was detected at elevated levels in off-site groundwater, and indoor air. However, ATSDR considers that current levels of exposure are unlikely to produce any adverse health effects.
PCE is a chlorinated hydrocarbon used primarily as a dry-cleaning solvent, a vapor-degreasing solvent, and a drying agent for metals; it is also used in the manufacture of fluorocarbons (17). Not known to occur naturally, PCE enters the environment from sources such as vaporization losses from dry cleaning and metal degreasing industries, and leachate from vinyl liners in asbestos-cement water pipelines used for water distribution. The general population can be exposed to PCE through inhalation of contaminated ambient air and ingestion of contaminated drinking water, especially from polluted groundwater sources. Most absorbed PCE is slowly eliminated unchanged (half-life = 65 hours) in the breath.
PCE is slightly to moderately toxic in laboratory animals (17). In humans, ingestion of small amounts of PCE is unlikely to cause permanent injury. In fact, PCE was formerly used as a remedy for intestinal worms, until that use was discontinued because of inebriating side effects. The known human health effects of PCE have usually been the result of occupational exposure to high concentrations, primarily by inhalation. The threshold limit value (TLV) for PCE in air is 25 ppm or 170 ug/m3. In excess of 100 parts per million (ppm), PCE is irritating to mucous membranes and the respiratory tract and may produce largely reversible effects in the liver. At 200 to 500 ppm, PCE causes symptoms of central nervous system depression, e.g., dizziness, headache, vertigo, inebriation and unconsciousness. However, in men or women repeatedly exposed to 100 ppm for 7 hours per day, no adverse neurological effects were identified by a battery of behavioral and neurological tests. PCE has been declared a probable/ possible (B2-C) human carcinogen by the Environmental Protection Agency (EPA), based on sufficient evidence in animals and inadequate evidence in humans. However, the relevance of the animal data to humans is now being questioned because the induction of cancers in rodents required such high doses and involved elements of rodent biology not shared by humans. Therefore, ATSDR considers that cancer effects are not the most appropriate basis for an assessment of this chemical's potential impact on public health at Livingston. (See Appendix D for a discussion of the carcinogenicity of PCE.)
Off-Site Groundwater
Because the city regularly monitors the public water supply, municipal well water is not now, nor is it likely to become, a significant source of PCE exposure for the general public. Although the maximum level of PCE in private well water (96 ppb or 0.096 ppm) was significantly higher than the maximum contaminant level (MCL) of 5 ppb, it was still much too low to produce any noncancerous adverse health effects in persons using this water as their sole source of drinking water. The maximum level of PCE detected in off-site groundwater (530 ppb) was measured in Monitor Well 89-4 near the northeast edge of the facility, in May 1990. PCE in this water, which is not used for drinking, only marginally exceeded EPA's drinking water equivalent level (DWEL). The DWEL is a lifetime exposure level specific for drinking water (assuming that all exposure is from that medium) at which noncarcinogenic adverse health effects would not be expected to occur (34). The DWEL for PCE is 500 micrograms per liter (µg/L) and includes an uncertainty factor of 1,000, i.e., the DWEL was set at a level 1,000 times lower than the No Adverse Effect Level (NOAEL) observed in animal studies.
Off-Site Air
The levels of PCE in off-site ambient air were too low to be a health concern. More PCE was detected upwind than downwind of the site, which suggests the existence of sources for the PCE in the air other than BNRY.
The highest concentration detected in indoor air at Livingston (82.1 micrograms per cubic meter (µg/m3) or 12 ppb in a basement) marginally exceeded ATSDR's intermediate environmental media evaluation guide (EMEG) of 9 ppb, which is a level to which one could be exposed for 2 weeks to a year without deleterious effects. However, real exposures (in this case, in a basement) are likely to be intermittent and of much shorter duration. For comparison, 12 ppb is at least 2,000 times lower than the American Conference of Governmental Industrial Hygienists (ACGIH's) threshold limit value of 25 ppm PCE, a level considered safe to work in for 8 hours per day, 40 hours per week. Since the first draft of this assessment was written, the intermediate EMEG for PCE has been replaced with a chronic EMEG of 40 ppb or 270 ug/m3, which is not exceeded by any concentration of PCE measured in off-site air.
Inhalation exposure may also occur during bathing. However, based on a worst-case scenario (assuming 500 µg PCE/L domestic water), ATSDR has estimated that peak transitory levels of volatilized PCE that might occur while bathing or showering in PCE-contaminated water would be too low to produce any acute health effects such as drowsiness. Neither would intermittent exposure to such low levels for relatively short periods of time be expected to have any adverse chronic effects.
Fish
The levels found in fish were all below the detection limits of the assay methods used. That means that we do not really know how much, if any, of these compounds was present. The detection limit that applied to most VOCs was 15 ppm. This detection limit is much higher than EPA Region III's Risk-Based Concentration (RBC) of 0.061 ppm PCE in fish. However, the latter limit is based on 2 assumptions that may not apply at BNRY and the surrounding communities. The first is the assumption that PCE may cause cancer in humans. A number of epidemiological studies of occupationally exposed men and women have not identified an increased risk of cancer attributable to PCE, and the cancers induced in laboratory animals at very high doses of PCE by species-specific mechanisms have little or no relevance for human risk evaluation at environmental levels of exposure many orders of magnitude lower. Secondly, the RBC is based on the assumption of a default fish ingestion rate of 54 grams or about 2 ounces/day. When one considers that local patterns of consumption of fish from the Yellowstone River are likely to be intermittent in nature, and that the concentrations of any solvents in the fish would be greatly reduced during cooking, it is highly unlikely that the undetectable levels of PCE that may exist in fish would represent a potential health hazard to individuals periodically consuming such fish.
Arsenic
Arsenic was present in some air and water samples at levels that exceeded ATSDR's lowest comparison value for this metal. However, ATSDR considers that current levels of exposure are unlikely to produce any adverse health effects.
Arsenic, which may be either trivalent or pentavalent, is a metalloid element that is widely distributed in the earth's crust, with an abundance of about 5 micrograms per gram (µg/g or ppm) (18). The manufacture of arsenical pesticides represents the major source of occupational exposure, mainly through the inhalation of fumes and dusts. Arsenic compounds are also used in some pharmaceutical preparations. The threshold limit value/time-weighted average (TLV-TWA) for arsenic and its soluble compounds in the workplace is 0.01 milligrams per cubic meter (mg/m3). Cigarette smokers incur an additional exposure to arsenic as shown by the fact that they have mean blood arsenic levels approximately 50% higher than nonsmokers. The primary route of nonoccupational exposure is by ingestion of food and water; the daily dietary intake of arsenic in the U.S. ranges from <0.04 to 0.2 mg/day, depending on the amount of seafood in the diet. Organic forms of arsenic, which are less toxic and more rapidly excreted, predominate in fish and seafood. There is some evidence that arsenic acts as an essential nutrient with anticancer value. Nutritional essentiality has been demonstrated in at least three different species (rats, goats, and minipigs), but not yet in humans.
Occupational exposure to airborne arsenic is associated with lung cancer, and chronic exposure to high levels in drinking water has been associated with a unique form of skin cancer. In contrast to most human carcinogens, however, arsenic does not cause cancer in laboratory animals when administered orally. At low levels of exposure, toxic inorganic arsenic compounds are effectively detoxified by methylation and excreted in the urine. Only after the methylation capacity of the liver is exceeded do blood arsenic levels increase and adverse health effects occur. Saturation of this detoxification mechanism may provide an explanation for the observation that both the cancerous and noncancerous effects of arsenic exhibit a threshold somewhere between 250 and 500 µg/day (45,46).
Current cancer-based limits on arsenic in drinking water are based on a large Taiwanese study in which consumption of arsenic-contaminated well water (170 to 800 ppb) was associated with increased skin cancer. However, in the U.S., where levels of arsenic in drinking water are much lower (average 5 µg/L), no excess skin cancer incidence has been observed in people consuming relatively high levels of arsenic in drinking water (18). It now appears that arsenic exposure was underestimated in the Taiwan study, leading to an overestimation of risk. It is also possible that the protein- and methionine-deficient population studied in Taiwan was more sensitive than typical U.S. populations, due to a compromised ability to detoxify (i.e., methylate) ingested arsenic (45,46).
Off-Site Groundwater
Arsenic was below the MCL of 50 ppb in all water samples analyzed, including on-site (7 ppb) and off-site groundwater (5 ppb), and Yellowstone river water. In particular, the maximum level of arsenic in private well water (15 ppb) was well below the MCL. Although the latter concentration does exceed ATSDR's cancer risk evaluation guide (CREG) and ATSDR's chronic EMEGs/RMEGs, ATSDR's comparison values are not predictive of adverse health effects (see the discussion on arsenic and cancer in the preceding section). The daily arsenic intake (30 µg/day) from drinking 2 liters of this water every day is an order of magnitude below the apparent threshold for the adverse effects (carcinogenic or non-carcinogenic) of arsenic (approx. 400 µg/day). The daily arsenic dose for a 70 kilogram per day (kg) adult consuming 2 L/day of water containing 15 ppb arsenic would be 0.4 µg/kg/day which is within the range of currently recommended reference dose (RfD) values, i.e., 0.1 to 0.8 µg/kg/day (34). Thus, ATSDR considers that no adverse health effects, including cancer, are likely to result from drinking water containing arsenic at the levels detected in private wells in Livingston, Montana.
On-site Air
The mean levels of arsenic detected in on-site ambient air, i.e., 0.001 µg/m3 upwind and 0.006 µg/m3 downwind (Table 7), exceed ATSDR's CREG of 0.0002 µg/m3 by factors of 5 and 30, respectively. However, ATSDR's CREGs are based on EPA cancer risk assessments which are in turn based on a risk model (i.e., the Linear Multistage Model) that does not take into account the existence of thresholds for promoters like arsenic. In such cases, linear extrapolation from high occupational exposures will significantly overestimate risk for the general population.
The maximum level of arsenic detected in air at BNRY (0.0146 ug/m3) was 75 times lower than EPA's risk-based concentration (1.1 µg/m3) for noncancer effects, and the corresponding maximal dose of 0.292 ug/day (assuming inhalation of 20 m3 air/day and 100% absorption) would be over 1,000 times lower than a presumed 400 µg/day threshold for cancer effects as well. Thus, based on the available data, ATSDR considers that the levels of arsenic in on-site ambient air at BNRY are too low to produce adverse health effects of any kind.
Cadmium
Cadmium was present in some air samples at levels that exceeded ATSDR's lowest comparison value. However, ATSDR considers that current levels of exposure are unlikely to produce any adverse health effects.
The concentration of cadmium in U.S. topsoil ranges from 0.1 to 1.0 ppm with an average value of 0.26 ppm (19). The major use of cadmium is electroplating metals to inhibit corrosion. Cadmium and its compounds are also used in pigments, plastic heat stabilizers, fungicides, and nickel-cadmium batteries. The principal route of occupational exposure is inhalation. The TLV-TWA for cadmium dusts and salts, as cadmium, in the workplace is 0.01 mg/m3 (TLVs, 1993-1994). Nonoccupational exposure is primarily via ingestion of food and drink and cigarette smoking. The body burden of cadmium can be twice as high in smokers as in nonsmokers. Typical concentrations in ambient air are 0.001 to 0.005 µg/m3 in rural areas and up to 0.05 or 0.06 µg/m3 in urban areas. Absorption of cadmium is 15-30% from the respiratory tract, but only 5 to 8% from the gastrointestinal tract.
Chronic occupational exposure to cadmium can lead to renal dysfunction with proteinuria (19). High, prolonged, cadmium exposures have produced bone disease (e.g., osteomalacia and itai itai disease) in a few industrial workers and in some women who resided in cadmium-contaminated areas of Japan for 50 years. Exposed individuals with deficient intakes of calcium or vitamin D are at increased risk for itai itai disease. Cadmium is classified as a "probable human (B2) carcinogen" based on limited occupational epidemiological studies and sufficient evidence of carcinogenicity in rodents exposed by inhalation. Administered orally, however, cadmium has not caused cancer in laboratory animals. Epidemiological studies suggest that long-term occupational exposure to cadmium may contribute to the risk of lung cancer. However, because of confounding exposures to other metals and cigarette smoke, the observed risk cannot be conclusively attributed to cadmium at this time (19).
On-Site Air
Cadmium was detected in on-site air at levels that exceeded ATSDR's CREG. However, epidemiological studies provide only weak evidence that prolonged inhalation exposure to cadmium is associated with an increased risk of lung cancer in occupational cohorts. Such data do not provide a reasonable basis for predicting cancer effects in nonoccupational populations exposed to much lower environmental levels of cadmium. A comparison value based on established noncancer effects would, therefore, be more appropriate for use under the site-specific conditions that prevail in Livingston. Taking this approach, it is found that mean cadmium levels downwind from BNRY were more than 8 times lower than ATSDR's chronic EMEG of 0.2 µg/m3 and were within the range of concentrations typical of urban areas. The fact that cadmium was not elevated in either on- or off-site soils suggests that long-term deposition of cadmium from air has been insignificant at this site. Therefore, based on the available data, it is concluded that the levels of cadmium in on-site air at BNRY do not represent a public health hazard to exposed residents.
TCE
Trichloroethylene, also called trichloroethene or TCE, was detected at elevated levels in some samples of groundwater and indoor air. However, the generally low concentrations to which residents might actually be exposed are not likely to produce any adverse health effects.
The primary use of TCE is as a solvent for degreasing metals (4). High levels of exposure are expected for workers in degreasing plants due to inhalation of vapors or adsorption through the skin. The TLV is 50 ppm in air or 269 mg/m3 (31). The general public is exposed to low levels of TCE, primarily from ambient air and drinking water. Additional low-level exposure may be incurred by persons living near degreasing plants or spill sites (4).
TCE is an effective central nervous system depressant at high doses (several hundred to 1,000 ppm) and was formerly used as an anesthetic. Chronic exposure to lower doses (27 ppm) may irritate mucous membranes and cause some drowsiness. However, there are no reliable reports of adverse health effects at environmental levels of exposure (i.e., the low parts per billion range).
Various types of cancer have appeared in animals treated with TCE orally and by inhalation (20). However, the positive animal studies are plagued by a number of problems, including high rates of mortality due to high dose toxicity; the use of TCE containing epoxide stabilizers that may themselves cause cancer; high, strain-specific, background rates of specific cancers; and species-specific mechanisms (peroxisome proliferation, alpha-2µ-globulin accumulation, activation of TCE-glutathione conjugates) that do not occur in humans and, hence, have no relevance to human risk assessment. The findings from epidemiological studies of workers exposed by inhalation have been largely negative and provide no clear evidence that TCE causes cancer in humans. Some controversial studies suggest a weak association between TCE in drinking water and cancer (mostly leukemia) in humans. However, because these studies fail to establish a link between actual TCE exposure and the incidence of health effects, their findings have been repeatedly challenged in the scientific literature. Based on this conflicting evidence, International Agency for Research Cancer (IARC) has recently upgraded TCE from Group 3 ("not classifiable") to Group 2A ("probably" carcinogenic to humans). EPA, on the other hand, has downgraded TCE from a B2 ("probable") to a B2-C ("probable/possible") human carcinogen, while the ACGIH designates TCE as an "A5" substance, i.e., an agent "not suspected to be a human carcinogen on the basis of properly conducted epidemiological studies" (31). Therefore, ATSDR considers that cancer effects are not the most appropriate basis for an assessment of this chemical's potential impact on public health at Livingston.
Private Well Water
The maximum level of TCE detected in off-site private well water was 15 ppb. This levels exceeds both the MCL of 5 ppb and ATSDR's CREG of 3 ppb but is 466 times lower than ATSDR's intermediate EMEG (7,000 ppb for children, 20,000 ppb for adults). ATSDR considers that 15 ppb TCE in drinking water is not likely to produce acute or chronic adverse health effects.
Indoor Air
The maximum levels of TCE detected in indoor air (73 and 567 µg/m3 or 14 and 106 ppb) exceeded ATSDR's CREG of 0.6 µg/m3 (0.1 ppb), but were far below ATSDR's intermediate EMEG of 2,000 ppb (10,750 µg/m3). These measurements were taken in a basement crawlspace and 1.5 feet down a basement well, areas which clearly do not represent "living" space or contribute to a person's normal breathing zone. Thus, the intermediate EMEG is more applicable to these measurements than is a CREG, which assumes lifetime exposure. The maximum levels detected in upstairs and basement living spaces were 3.33 µg/m3 (0.6 ppb) and 2.71 µg/m3 (0.5 ppb) respectively. These levels would not be expected to cause adverse health effects of any kind.
Other Contaminants
The maximum levels of toluene in indoor air (8,000 and 45,000 µg/m3 or 2,123 and 11,942 ppb) were above ATSDR's chronic and acute EMEGs of 1,000 and 3,000 ppb, respectively (21). The latter values are 30 times lower than the Lowest Adverse Effect Levels (LOAEL) for minimally adverse human effects on which they are based. Both of the maximum levels were detected in August 1989. However, subsequent measurements in October and November were all below detection limits. ATSDR cannot at this time explain the aberrant high readings in August 1989 and considers that they should not be used in the assessment of any potential long-term health effects. Toluene was not detected at levels of concern in any other environmental samples, either on- or off-site. Based on this data, no adverse health effects attributable to toluene exposure would be expected.
Elevated levels of 1,1,1-TCA were detected in indoor air in August 1989, but not during subsequent sampling events. Nor was there any clear relationship between measurements of 1,1,1-TCA on-site and the isolated occurrence of elevated levels in off-site indoor air. Based on this data, no adverse health effects attributable to 1,1,1-TCA exposure would be expected.
The maximum level of lead detected in off-site groundwater was 20 ppb, which marginally exceeded EPA's Action Level of 15 ppb. However, lead levels were below 15 ppb in both municipal and private wells. Considering the length of time that lead has had to migrate from the site, the levels involved, and the seasonal variation in the direction of groundwater flow, it is not likely that lead on-site will result in levels of public health concern off-site in the future. Thus, although it is desirable to keep lead levels in drinking water below the action level of 15 ppb, ATSDR does not consider that the few elevated readings from off-site monitoring wells reflect a current or future hazard to public health.
Cis-1,2-dichloroethene (DCE), like lead, was detected inconsistently in off-site groundwater, and then usually at levels below comparison values. It was not detected in municipal well water. The maximum level detected off-site (190 ppb) exceeded the MCL and Lifetime Health Advisory (LTHA) of 70 ppb, but not the Child's Longer-Term Health Advisory (CLHA) or child intermediate EMEG of 3,000 ppb. The maximum level detected in a private well was 99 ppb. Considering the seasonal variation in the direction of groundwater flow, the inconsistency with which DCE was detected, and the levels involved, the CLHA and intermediate EMEG, which are based on exposure durations of 1-7 years, would seem to be more appropriate than the LTHA and MCL, which are based on daily exposure over an entire lifetime. Thus, although it is desirable to keep DCE levels in drinking water below the MCL of 70 ppb, ATSDR does not consider that the few elevated readings from off-site wells reflect a current or future hazard to public health.
Finally, it should be noted that the above toxicological evaluations were all based on historical maximum concentrations, and that the concentrations of these contaminants have continued a long-term decline over the last six years, a decline which should continue in light of the on-going remediation activities on-site. Monitoring will continue to verify the decline.
Combined Effects
Since the individual contaminants detected at this site are present at levels that would not be expected to result in adverse health effects, ATSDR considers that the combined effect of all these contaminants is not likely to be of public health concern either. This conclusion is based on 4 studies which suggest that a mixture produces no adverse health effects in dosed animals when the components of that mixture are present at levels below their respective No Observed Adverse Effect Levels (NOAEL), i.e., at concentrations that would have produced no adverse effects in animals treated separately with those component chemicals (47-51). In two of these experiments (48,49), all of the component chemicals affected the same target organ, albeit through different mechanisms. In two others (50,51), the chemicals had different target organs and exhibited different modes of action, as do most chemicals in typical environmental mixtures. Considering that ATSDR Comparison Values are typically 100-1000 times lower than the corresponding NOAELs, it is reasonable to expect that environmental contaminants will not produce any combined effects, even if their individual concentrations exceed their respective EMEGs by a significant fraction of the associated safety factor (which ATSDR refers to as a composite "uncertainty factor").
B. Health Outcome Data Evaluation
The Montana Department of Health and Environmental Sciences, with technical assistance from ATSDR Division of Health Studies, completed a cancer cluster study for Livingston and Park County, Montana (1992). Data from the Montana Cancer Registry indicated apparent increases in digestive tract cancer in Park County over the years 1980-1989. More detailed examination of the data revealed that the increase was specifically in cancer of the pancreas. An increase in pancreatic cancer was observed in Livingston in white males, but not in the remainder of Park County, nor in white females (minorities were present in numbers too low to allow investigation). The mean age of onset is between 70 and 80 years which is the third largest age group in Livingston's total population; pancreatic cancer is more prevalent in areas with older populations. No association with smoking was observed for these cases.
The association between cancer mortality and contaminant exposure from railroad refueling facilities was suggested as a possibility. An investigation by the Montana Department of Health and Environmental Services was completed utilizing death certificate data for populated areas where refueling facilities are located in Montana (1995). No statistical relationship between residential proximity to the refueling facilities and cancer mortality could be identified.
Overall mortality in Park Country was examined using CDC computerized databases. Overall mortality was less than in the State of Montana taken as a whole, on either a crude or an age adjusted basis, over the period 1979-1988.
C. Community Health Concerns Evaluation
On-site and off-site groundwater contained elevated levels of arsenic, TCE, and PCE. People did not consume on-site groundwater. Monitoring data for private wells revealed elevated levels of PCE and TCE. Owners of private wells with elevated levels of contaminants have been notified. Monitoring wells are in place to allow for detection of any contamination present. The groundwater plume is thought to move in a northeastern direction toward the Yellowstone River. Low levels of TCE and PCE that were detected on the east side of the Yellowstone River may be due to underflow beneath the river or may be caused by non-BNRY VOC sources east of the river.
ATSDR and the State of Montana have completed a health study of cancer incidence in the Livingston, Park County, Montana, area. This study revealed an increased incidence of pancreatic cancers in Livingston that was not associated with residential proximity to a railroad facility. Incidence of other cancers appears to be average. Gallstones are often associated with diet. Miscarriages occur considerably more often than is usually acknowledged. Nationally, ten to twenty per hundred pregnancies end in miscarriage. The contamination detected in on-site and off-site media at BNRY would not be expected to cause miscarriages in Livingston.
Data on specific incidences of lupus and MS in Livingston were not available for review. Lupus and MS are chronic inflammatory diseases of unknown etiology (42,43,44). Although the causes are unknown, both environmental and genetic factors are clearly implicated. In genetically-predisposed individuals, the onset of these autoimmune diseases are often triggered by viral infections or reactions to certain drugs, or other chemicals. The onset of MS and Systemic Lupus Erythematosus has been triggered by bee stings and exposure to sunlight. The environmental data available neither implicates nor exonerates chemical exposures at Livingston as possible triggers of autoimmune disease in genetically-predisposed individuals. However, since natural, non-specific triggers are ubiquitous in the environment, the availability of man-made triggers may not be a determining factor in the incidence of these diseases. In other words, an increased incidence of Lupus and/or MS would most likely reflect a cluster of genetically-predisposed individuals, rather than an increased availability of environmental triggers.
ATSDR examined data on overall mortality from all causes in Montana and in Park County. Overall deaths and death rates were slightly less in Park County than in the state of Montana.
Next Section Table of Contents
