CABOT-WROUGHT PRODUCTS - DIVISION OF CABOT CORPORATION
(a/k/a NGK METALS/CABOT BERYLCO, INCORPORATED)
MUHLENBERG, BERKS COUNTY, PENNSYLVANIA
Contaminants of concern are contaminants found on or off site that are at concentrations that might pose a threat to public health. The mere presence of a contaminant being discovered on-or off-site does not imply that a health threat exists. This petitioned public health assessment will evaluate all the contaminants of concern in an effort to determine if there is any threat to public health. ATSDR selects and discusses these contaminants based upon the following factors:
In this section, comparison values used in ATSDR public health assessments will be compared to contaminant concentrations in specific media (e.g., air, soil, groundwater) used to select contaminants for further evaluation. ATSDR and other agencies developed those values to provide guidelines for estimating the media concentrations of a contaminant that are unlikely to cause adverse health effects, given a standard daily ingestion rate and standard body weight. However, the fact that the concentration of a chemical exceeds a given comparison value does not necessarily imply that it is likely to produce adverse health effects. The conditions of exposure and the individual susceptibility factors will determine whether or not the intrinsic toxic potential of a chemical is likely to be expressed. See Appendix C for a description of the comparison values used in this petitioned public health assessment.
ATSDR examines the Toxic Chemical Release Inventory (TRI) to determine if there is any known information on sources of potential contamination in the vicinity of the site or sites in question. The TRI contains information on estimated annual releases of toxic chemicals to the environment (via air, water, soil or underground injection), which are voluntarily reported by companies to EPA. TRI data can be used to give a general idea of the current environmental emissions occurring at or near a site. TRI data may also be used to determine whether the ongoing emissions from reporting facilities might be contributing additional environmental contamination.
A search of the TRI revealed that other Berks County industries were found to release similar contaminants to those found at the NGK site. Those industries were identified by examination of the Toxic Chemical Release Inventory (TRI) for 1989. A metal processor in the northern portion of the County (near Shoemakersville) releases small amounts of beryllium into the air and into a tributary of the Schuylkill River. Because of the distance from NGK, this release is unlikely to impact the Reading area. Two sources of cadmium releases were found in the City of Reading. Large amounts of chromium (282,202 lbs/yr) were released by 10 industrial facilities within the County, four within the Reading area. Trichloroethene was released by three Reading area industries (a total of 303,300 lbs/yr released into the air and 97 lbs/yr into surface water).
A. On-site Contamination
Surface Soil
No surface soil sampling data were available for ATSDR's review. During the late 1960s, many of the waste areas were covered with a highly organic material referred to as "mushroom soil." The mushroom soil was used to cultivate vegetation as ground cover in an effort to prevent erosion and migration of waste (1). The Disposal Area Drain Field is the only waste area that has not been covered (i.e. by mushroom soil, pavement, or gravel) (2, 3).
Subsurface Soil
Subsurface soil samples were collected from the Southwest Red/Lime Sludge Area, the Southeast Red Mud Filter Cake Disposal Area, Pond 1, Pond 2, Pond 3, Pond 6, Former Pond 6 waste pile, Former Government Ore Stockpile, Retention Basin, Disposal Area Drain Field, and Sludge Settling Tank Area (Figure 2). Subsurface soil borings and samples were done at some waste areas down to a depth of approximately 15 feet. Arsenic, beryllium, cadmium, total chromium, and copper were detected at concentrations that exceeded comparison values in on-site subsurface soil. Although there is no comparison value for lead in soil, it was also detected in on-site subsurface soil and will be retained as a contaminant of concern, requiring further consideration (1). (Table 2) (all tables are in Appendix B) shows the maximum concentrations of specific chemicals of concern in subsurface soil on site.
Groundwater
Thirty-one on-site monitoring wells, located in shallow (0-100 feet) and deep (100-200 feet) aquifers under the NGK facility, have been sampled. Three rounds of samples have been collected (December 1989, May 1990, and June 1991) from many of the wells. The six newest wells have only been sampled once. Samples (both unfiltered and field filtered) were analyzed for metals, priority compounds, and volatile organic compounds (VOCs). Groundwater samples from on-site monitoring wells (Figure 4) revealed contaminant concentrations that exceeded comparison values for a number of contaminants. Both shallow and deep aquifers showed contamination (1, 6). A summary of the maximum concentrations of those contaminants detected in on-site groundwater displayed in Table 3.
Four shallow wells existed prior to the RCRA Facility Investigation and sampling was conducted on a quarterly basis by NGK, dating back to about 1981. Some of the highest contaminant concentrations found in on-site groundwater were detected during the early 1980s (7). However, ATSDR will use only the most current on-site groundwater data to evaluate contamination since on-site groundwater has not been used.
Ambient Air
NGK conducted an on-site air monitoring program from June 25, 1991 to July 30, 1991. Sampling and analysis protocols were approved by EPA for NGK under its beryllium National Emissions Standards for Hazardous Pollutants (NESHAP) sampling program. Sampling was accomplished using 24-hour high volume samplers which were run for 7 days and analyzed based on air flow during the 7-day periods. Filters were analyzed for beryllium and chromium (2, 8).
The on-site ambient air sampling locations (RCRA 01 and RCRA 02) were located east of the Disposal Area Drain Field and Former Pond 6 waste pile, respectively, to monitor the potential impacts of those waste areas. The locations of those monitors can be seen in Figures 5 and 6 (8). Table 10 lists the concentrations of total beryllium and total chromium that were recorded during the program.
During the June 25-July 30, 1991, sampling event, the plant operations were reportedly shut down during the final two weeks of July. That sampling plan was to provide ambient air data during production and non-production periods (8). Ambient air results during the last two weeks (non-production period) of July 1991 (July 16-30, 1991) do not vary greatly from the first three weeks (production period). Furthermore, in comparing the on-site monitors (RCRA 01 and RCRA 02) to the nearest off-site monitor (R-1) for this 5 week sampling event, chromium was always detected at a higher concentration off site and beryllium was always detected at a higher concentration on-site. Based on weather observations from the Reading Airport, located approximately two miles southwest of NGK, wind was predominately out of the west/west-northwest during the 5-week sampling period (9). For each week reviewed, higher beryllium air concentrations appear to correspond to winds coming from the west and north (NGK and the waste field). Chromium concentrations, however, do not correspond as clearly to winds from the west and north. Although NGK may be a potential source of chromium, via fugitive dust emissions, data tends to indicate that other sources, such as those identified by the TRI search, may be impacting air quality at the NGK site.
None of the results of the on-site ambient air sampling exceeded the Occupational Safety and Health Administration's (OSHA) Permissible Exposure Limits (PELs) time-weighted average standards. The OSHA limits are based on time-weighted average concentration for a normal 8-hour workday and a 40-hour workweek, to which all workers may be occupationally exposed, day after day. However, three beryllium samples and all of the chromium samples exceeded the ATSDR comparison values and will therefore be discussed as contaminants of concern in the Pathways Analyses and Public Health Implications sections.
Issues regarding the quality of ambient air data are discussed in the Quality Assurance/Quality Control subsection.
B. Off-site Contamination
Soil
On November 18, 1992, one surface soil sample was collected off-site. That sample was collected from the lawn of a private residence approximately 1 mile southwest of NGK. The soil sampling was initiated by the owners of the home and was analyzed for beryllium, chromium (total and hexavalent), and fluoride, which were contaminants detected in their well (see Table 4a, private well 1). Beryllium was detected at a concentration of 2.12 milligrams per kilograms (mg/kg), which exceeds the Cancer Risk Evaluation Guide (CREG) comparison value of 0.2 mg/kg (a description of comparison values used by ATSDR can be found in Appendix C) (10). None of the other contaminants exceeded comparison values. The surface soil sample results and location are listed under Residence #7 in Table 5 and Figure 7, respectively.
Additional surface soil samples were collected in public and private areas around NGK on July 27, October 25, and October 26, 1994 (11, 12). The July 1994 samples were collected at four public locations around the site. The sampling locations are in Figure 7, listed as PA-1, PA-2, PA-3, and UG-1. Based upon prevailing wind directon, the samples collected at UG-1 and PA-3 are located most directly upwind and downwind, respectively, of NGK. Although the upwind samples were intended to indicate background, the proximity to the site and evidence of beryllium detected in air at that location, do not provide for adequate background data (13).
Three composite soil samples were collected at each of the above mentioned locations and analyzed for beryllium (total and soluble), chromium (total), and fluoride. Eleven samples, ranging from 0.77-6.9 mg/kg, revealed total beryllium concentrations that exceeded the CREG comparison value of 0.2 mg/kg (11). Only one of the soluble beryllium samples revealed a concentration above the detection limit. That sample, which was collected at UG-1, was back calculated to a soil concentration of 0.0054 mg/kg, which does not exceed the CREG comparison value (14). Since chromium samples were not analyzed for both total and hexavalent or trivalent chromium, the samples were screened using the pica child RMEG comparison value of 10 mg/kg, for hexavalent chromium. Pica child refers to a behavior of children who have an excessive habit of ingesting nonfood items. They may intentionally ingest 25 times more soil than a child who does not exhibit pica behavior. Concentrations of total chromium were detected in 11 samples, ranging from 17-31.5 mg/kg, which exceeds the comparison value (pica child RMEG) for hexavalent chromium, but not for trivalent chromium. One sample, for both beryllium and chromium, was below the laboratory's detection limit. None of the fluoride samples exceeded comparison values (11). The surface soil sampling results for the public locations are in Table 5.
During the October 1994 sampling, 38 surface soil samples were collected in residential areas around NGK and analyzed for total beryllium and chromium. Concentrations of beryllium exceeded the CREG comparison value in 37 samples, which ranged from 0.5-4.9 mg/kg. Beryllium was below the laboratory detection limit in one sample. Concentrations of total chromium exceeded the pica child RMEG comparison value for hexavalent chromium in 37 samples, which ranged from 10.2-92.6 mg/kg (12). However, total chromium did not exceed the pica child RMEG (2,000 mg/kg) for trivalent chromium in any of the samples. The surface soil sampling results and locations are listed as "residences," numbering 1-6, in Table 5 and Figure 7, respectively.
Groundwater
Off-site groundwater sampling has been conducted at three private wells, three monitoring wells, one piezometer, the Reading Crest Well and a well at the Berks Products Quarry (see Figure 8) (6). All of the wells, except private well 3, are within the EPA specified well inventory area. Private well 1 was sampled on December 6, 1990; December 26, 1990; and June 1991. Private well 2 was sampled on May 30, 1991; June 1991; and July 27, 1994. The Reading Crest Well was developed as a municipal supply well but has never been used; it has been sampled numerous times in the past (1, 5, 6, 15, 16, 17). The other off-site wells have only been sampled once (June 1991). Private well 3, monitoring well 24, and the piezometer represent the shallow aquifer (<100 feet) and all other off-site wells represent the deep aquifer (100-200) (6).
Arsenic, beryllium, cadmium, chromium (total and hexavalent), selenium, fluoride, 1,1-dichloroethene, tetrachloroethene, and trichloroethene were detected in off-site groundwater at levels exceeding comparison values. Some of the contaminants detected in on-site groundwater were not analyzed in off-site groundwater samples. Also, the detection limits for some contaminants analyzed in off-site groundwater are greater than our comparison values. It is not possible to determine whether or not the analyte concentration exceeds comparison values when the detection limit is greater than comparison values. The maximum concentrations, contaminants analyzed, and the detection limits that were reported for contaminants that were not detected for each location sampled are given in Tables 4a, 4b and 4c.
Surface Water
Through stream surveys and the RCRA Facility Investigation, surface water in Laurel Run has been sampled upstream and downstream of NGK on seven occasions. Water quality data for Laurel run are available for the years 1981, 1989, 1990, and 1991. Data from those sampling events are reported in Tables 6a, 6b, and 6c.
Concentrations of beryllium (2500 micrograms per liter [µg/L]), lead (36 µg/L), manganese (450000 µg/L), fluoride (870 µg/L), dichloromethane (methylene chloride) (45 µg/L), and tetrachloroethene (1 µg/L) exceed comparison values (6, 18, 19). Some of the detection limits for arsenic and beryllium are too high. As discussed above, lower detection limits should be used. The highest concentrations of beryllium and manganese, which were reported in the May 13, 1981 stream survey, are unusually high in comparison to the other surface water data (see the Quality Assurance/Quality Control subsection for more information regarding this data). All other data consistently show lower concentrations.
A number of facilities and activities in the past and present contribute to pollution in Laurel Run. Some of the possible sources of pollution include: mushroom farming, a limestone quarry, a lead storage battery plant, a truck terminal, a beryllium processing plant (NGK and its predecessors), and wet-weather highway runoff.
Based on the surface water data, it is difficult to determine what source or sources are contributing greatest to water quality problems in Laurel Run. The highest concentrations of lead, manganese, fluoride, and dichloromethane detected in Laurel Run are found at locations well downstream from NGK rather than at the downstream location closest to the NGK NPDES outfall. If NGK is the primary source of those contaminants, concentrations would likely be greatest just below the NGK NPDES outfall, with decreasing concentrations further downstream as water volume increases. However, groundwater is recharging Laurel Run further downstream, in which case contaminated groundwater from NGK could be a possible source.
In comparing the samples taken upstream and downstream of the NGK wastewater discharge for each sampling event (i.e., comparing samples taken on the same date), increases in beryllium and copper concentrations are always shown. In the same type of comparison, an increase in lead and manganese concentrations is shown in only part of the samples. No other relationship between upstream and downstream contaminant concentrations is seen, although the limitations of the data (i.e., detection limits greater than concentrations being discharged) may have prevented such relationships from being observed. Table 7 shows the contaminants found in NGK wastewater discharge. No appropriate comparison values are available; therefore, no comparison values are reflected in that table. However, from this information we can determine that detectable levels of beryllium, cadmium, chromium (total and hexavalent), copper, lead, nickel, nitrate, and fluoride are discharged to Laurel Run (20, 21, 22, 23).
Sediments
During the RCRA Facility Investigation sediment samples from Laurel Run were collected in December 1989, May 1990, and June 1991. Sediment was collected from the same locations during each sampling event. One sample was taken upstream from the NGK NPDES outfall and two samples were taken downstream of the NGK NPDES outfall (1, 6).
Arsenic and beryllium concentrations exceeded comparison values for sediment at all the sampling locations. No other contaminants analyzed exceeded comparison values. Arsenic concentrations were generally greater downstream of the NGK NPDES outfall than those upstream. All of the samples downstream of the NGK NPDES outfall showed higher beryllium concentrations than the concentrations detected upstream of the NPDES outfall. The difference between the upstream and downstream concentrations for both beryllium and arsenic were marginal, varying only 1-2 µg/L. Data from all three sampling events are in Table 8.
On September 17, 1991, a PADER Water Quality Specialist reported observing unusual sediment on rocks in the NGK NPDES discharge zone of Laurel Run. The unusual sediment materials were washed from some of the rocks, and the materials that settled out were poured into 500 milliliter sample bottles and fixed with acid. On October 1, 1991, after a couple of weeks of heavy rains, the same sampling process was repeated taking one sample upstream and one sample downstream of the NGK NPDES discharge. In addition, a sample of NGK's wastewater effluent was taken during both sampling events and are included in Table 7 (22).
The sediment was not analyzed on a dry weight basis (mg/kg), but was acid-digested and analyzed as a solution (µg/L). This type of analysis would indicate what contaminants can be aggressively (by acid-digestion) leached out of the sediment, but the concentration cannot be compared to comparison values for exposure purposes.
The upstream sediment sample collected on October 1, 1991, revealed beryllium (5.8 µg/L), copper (766 µg/L), lead (248 µg/L), and nickel (39 µg/L). Total chromium was analyzed but was not detected at a detection limit of 50 µg/L. The samples collected downstream of the NGK NPDES outfall revealed beryllium (7290 & 2440 µg/L), total chromium (276 & 160 µg/L), copper (1845000 & 615000 µg/L), lead (104000 & 11700 µg/L), and nickel (19600 & 7290 µg/L) for the September and October 1991 samples, respectively (22). Given the generally low level of contaminants in the stream sediments, these results need further confirmation. In particular, to evaluate their health significance, ATSDR needs to know the amount of these highly contaminated materials (in mg/kg), their distribution within the stream, and possibly their source. See the Quality Assurance/Quality Control subsection for more information regarding this data set.
Air monitoring data for NGK were reviewed for the years 1979 to 1993. Monthly average concentrations of beryllium were reported in µg/m3, at 8 stations surrounding the NGK facility (location of air monitoring stations, otherwise discussed as R-1 through R-8, can be found in Figure 9 numbered simply 1 through 8). Only air sampling for beryllium has been required under the NESHAP regulatory program. Table 9 shows the maximum weekly beryllium concentrations reported for each year since 1979. The same sampling and analysis protocols described in the On-site Contamination subsection were used for off-site ambient air monitoring.
Cabot Berylco reported abnormal stack emissions due to problems with their collector system between December 30, 1980, and February 6, 1981, to which they attributed a high value of 0.04279 µg/m3 at monitor R-1. High concentrations for the year of 1981 were also reported for monitors R-5 through R-8 during that time period (see Table 9).
NGK reported two violations of the NESHAPs regulatory limit of 0.01 µg/m3. The NESHAPs standard is based on a monthly (30-day) average. Violations occurred in June and August 1989 during the excavation of red mud for the construction of a new building (3, 24). Following those two events, the air monitoring station (Figure 9, station R-1) closest to the site recorded high weekly concentrations of 0.08143 µg/m3 and 0.02972 µg/m3 for beryllium in June and August 1989, respectively (25).
Beryllium concentrations recorded on a weekly basis at the R-1 and R-3 stations have exceeded the environmental comparison value of 0.0004 µg/m3 (CREG) for at least 1 week during each year since this sampling program began in 1979.
Off-site ambient air data are available for chromium. Chromium was sampled for at the R-1 station (the off-site station closest to NGK, see Figure 9) during the June 25-July 30, 1991, RCRA ambient air sampling program. Chromium concentrations at the R-1 station exceeded the CREG comparison value during each week of sampling. The results for the off-site station (designated as R-1) can be found in Table 10; in addition, that table includes on-site monitoring data for the stations designated as RCRA 01 and RCRA 02.
Issues regarding the quality of ambient air data are discussed in the Quality Assurance/Quality Control subsection.
Biota
To date, there are no analytical data available for off-site biota (e.g., fish and garden vegetables). Such information is needed to fully evaluate any potential public health impacts. Biota in the areas surrounding NGK and along Laurel Run might be contaminated with chemicals found in Laurel Run's surface water and sediments, off-site groundwater (possibly used for gardening or agricultural purposes), and in off-site ambient air.
C. Quality Assurance/Quality Control
During preparation of this petitioned public health assessment, ATSDR relied on the information provided in the referenced documents and assumed that adequate quality assurance and quality control (QA/QC) measures were followed with regard to chain-of-custody, laboratory procedures, and data reporting. The validity of the analyses and conclusions in this petitioned public health assessment is determined by the completeness and reliability of the referenced information. However, items or issues of concern regarding data quality that were identified by ATSDR are discussed below.
Groundwater samples taken from monitoring wells represent both unfiltered and field filtered samples. EPA Drinking Water Standards and ATSDR health comparison values are based on unfiltered samples. Field filtered samples are sometimes ten or more times lower in concentrations than unfiltered samples, but should always be lower than unfiltered samples. A few groundwater samples showed filtered samples at higher concentrations than the unfiltered samples. ATSDR has reported the highest concentration in such cases.
Groundwater was analyzed for mercury during on-site sampling in December 1989 and May 1990. Mercury was reported at a concentration of 56,900 µg/L (unfiltered) and 33,900 (filtered) in monitoring well 14A for the December 1989 sampling event. No quality assurance or quality control problems were reported for those samples. However, such results are unlikely since: (1) mercury was not detected in the May 1990 groundwater analysis for monitoring well 14A, (2) mercury was only detected in three other on-site monitoring wells, with a maximum concentration of 0.4 µg/L, during the December 1989 and May 1990 groundwater sampling, (3) mercury was detected at relatively low levels in on-site subsurface soil, with a maximum concentration of 4.1 mg/kg, and (4) mercury is not a component of the NGK waste stream. Laboratory error, sample contamination, or an error in reporting are some of the possible situations which could have occurred. Based on all of the above mentioned factors, ATSDR does not regard those elevated samples as representative and; therefore, they will not be evaluated as such.
Beryllium and manganese reported in surface water samples collected during a stream survey on May 13, 1981, were approximately 1000 times higher for beryllium and 10,000 times higher for manganese than concentrations reported for those contaminants during other sampling events. In the stream survey, the concentrations were reported in milligrams per liter (mg/L) and it is possible that a 1000 fold increase resulted if the laboratory reported those values in micrograms per liter (µg/L), but the numbers were transposed for the report. ATSDR was unable to acquire a copy of the original laboratory analysis to investigate this matter. In giving further consideration to those concentrations, it should be noted that the data were collected in 1981 and the other data were collected 8, 9, and 10 years later when water quality might have improved. In addition, the sample that was described in the Off-site Sediment subsection (although it is not truly representative of surface water) shows that suspended solids found in the stream are capable of producing beryllium concentrations in the 2500 µg/L range. Therefore, ATSDR will accept the beryllium and manganese concentrations as reported. The stream survey reported that samples at Station 1 (listed as DS1 in Table 6b), which was originally indicated as the upstream location and Station 2 (listed as US1 in Table 6b), which was originally listed as the downstream location, may have been switched. The data appear to indicate that this is a possibility based on the higher beryllium, fluoride, and copper concentrations shown at station 1 (DS1). Therefore, in Table 6b of this petitioned public health assessment, Station 1 is reported as the downstream location nearest to NGK, and Station 2 is reported as the upstream location. No other quality assurance or quality control problems were reported.
The three samples of unusual sediment that were collected on September 17, 1991, and October 1, 1991, were not analyzed as dry weight sediment samples. The samples were fixed with acid and analyzed as a water sample would have been. However, ATSDR will consider those samples as being representative of the sediment, since sediment was the basis of collection and the focus of the analysis was to identify contaminants leached from the sediment. The data provided from those analyses would more likely represent a sediment exposure than a surface water exposure. Furthermore, no comparison values are applicable to this type of sample. Therefore, ATSDR has only reported the concentrations for those contaminants which were identified as contaminants of concern in on-site media.
Several issues regarding the quality of ambient air sampling data have been raised through a community concern and discussions with EPA. One issue is that the sampling method may be causing some data quality problems. The sampling method, as described earlier, is to run the 24-hour high volume samplers for 7 days before having the filter analyzed rather than on a daily basis which is typically done. The sampling method utilized by NGK had received EPA approval to operate on a continuous basis and is considered to raise data quality questions only for those samplers which are recorded as having low flow (below 39 cubic feet per minute) or as having shut down at some point between daily flow checks by NGK staff. Severe restriction of flow or stopped flow results in inconsistent particle size sampling. The particle sizes sampled under low-flow conditions are not necessarily those upon which the NESHAPs standard was based. For the data submitted by either NGK or Cabot Berylco since 1979, the occurrence of interrupted flow has been recorded and does not involve the June and August 1989 excursions, when the NESHAPs regulatory limit was exceeded. The other air data issue of concern is that the analytical procedure used may not be revealing the total concentration of beryllium (particularly beryllium oxide) that is present in the sample. The analytical method currently being used is an approved EPA method and is being conducted by an approved laboratory. EPA is currently investigating both of those issues. Until those issues are resolved and conclusions determine otherwise, ATSDR will evaluate the ambient air data as they are reported since no other data are available to evaluate. If the EPA investigation of those issues indicate that the sampling and or the analytical method are unrepresentative of the actual beryllium concentration in off-site air, then ATSDR will reassess the data based on those conclusions.
D. Physical and Other Hazards
The physical hazards on site are common to a metal processing plant, and include materials movement, waste handling, and furnace operations. Other than heavy metal parts, metal cutting machines, acid pickling vats, and high temperature furnaces (all of which were to be shut down in November 1992), no unusual physical hazards were noted. The site is secured by a fence (portions not fenced at the edge of the parking lot are in view of 24-hour security personnel).
To determine whether residents in the Muhlenberg Township are exposed to contaminants migrating from the site, ATSDR evaluates the environmental and human components that lead to human exposure. This pathway analysis identifies five elements: 1) source of contamination, 2) transport through an environmental medium, 3) a point of exposure, 4) routes of human exposure such as ingestion, inhalation, or dermal absorption, and 5) a receptor population. ATSDR classifies exposure pathways as completed, potential, or eliminated. A completed pathway existed in the past, and may exist in the present or future if all five elements of an exposure pathway link the contaminant source to a receptor population. Potential pathways, however, are defined as situations in which at least one of the five elements is missing, but could exist. Potential pathways indicate that exposure to a contaminant could have occurred in the past, could be occurring now, or could occur in the future. Pathways are eliminated when at least one of the five elements is missing and will never be present. Completed and potential pathways may also be eliminated when they are unlikely to exist.
A. Completed Exposure Pathways
A list of completed exposure pathways is in Table 11 of Appendix B.
Off-site Groundwater
During the Phase I RCRA Facility Investigation (November 1990) a well inventory, which consisted of reviewing existing state records of private well locations, was conducted. This inventory identified several wells, including private well 3, west of NGK (1). Another well was identified (private well 1) and sampling of that well was requested by EPA in the Phase I RCRA Facility Investigation Addendum (15). In the spring of 1991 a door-to-door water well survey was conducted, at EPA's request, to identify any additional private wells. An area surrounding and hydraulically downgradient (up to 2-3 miles) of NGK was specified by EPA to be surveyed. During the survey a total of 465 residences and businesses were inventoried. One well (private well 2) was identified as being used for consumption and other domestic purposes (5).
Private wells 1 and 2 revealed contaminants that exceed comparison values. All of the contaminants detected in private wells 1 and 2 were detected at elevated levels in on-site groundwater. The detection limit of some contaminants analyzed for private wells 1 and 2 are above comparison values (see Table 4a).
Private well 1 is no longer in use, but was used from 1983 to May 1991, when the well was disconnected and the house was connected to the Muhlenberg Township water supply (5, 6). Private well 2 was installed around 1950 and is currently in use. Initial sampling (May 1991) of private well 2 did not reveal any contaminants above comparison values; however, a second sample in June 1991 showed total chromium at 52.7 µg/L, slightly above the Reference Dose Media Evaluation Guide (RMEG) comparison value of 50 µg/L (5, 6). Private well 2 was last sampled on July 27, 1994, and no contaminants, including chromium, exceeded comparison values (11).
A past completed pathway exists through ingestion, dermal contact, and inhalation routes for the family of 4 (2 children and 2 adults) that used water from private well 1. A current completed pathway exists through ingestion and dermal contact for the family that is presently using water from private well 2.
On-site Ambient Air
No data are available that indicate that on-site workers were exposed to beryllium or chromium in the ambient air that exceed OSHA limits. However, ATSDR will also consider on-site workers, who also live near the site, that may have received additional exposures of beryllium and chromium for extended periods of time (i.e., greater than 8 hours a day or 40 hours a week) over the course of their lifetimes. See Tables 9 and 10 for on-site and off-site ambient air data.
Past, present, and future completed pathways for beryllium and chromium, via inhalation, exist for on-site workers. An accurate estimate of the receptor population cannot be determined, but as many as 1,000 people have worked at NGK in the past and as few as 100 people are currently working at NGK.
Off-site Ambient Air
Off-site ambient air represents past, present, and future completed pathways for people living and working near the NGK plant. Based on ambient air data in Table 9, concentrations of beryllium have exceeded the CREG comparison value at all of the 8 monitoring stations in the past and in 1994 exceeded the comparison value at each of the eight stations. NGK instituted changes in plant operations in November 1992, which includes shutdown of the melting furnaces and hot rolling operations. Despite these changes, air sampling results continue to exceed comparison values and therefore will be evaluated further. Since beryllium is currently being detected off site, a future completed pathway is expected to exist.
Concentrations of chromium which exceed the CREG comparison value were detected off-site (station R-1) during the July 1991 RCRA ambient air sampling program. Based on that 5-week period of sampling data, which indicates a past completed pathway, ATSDR believes that exposures to chromium in the air are presently occurring and are likely to occur in the future.
Inhalation is the only route of exposure that is considered by this pathway. Past and current exposures to concentrations of beryllium and chromium, which exceed comparison values, have occurred within 2 miles of NGK, so potentially a population of less than 14,686 has been exposed (4).
Off-site Soil
Forty-nine surface soil samples collected on public and private property around the NGK facility revealed total beryllium concentrations that exceed the Cancer Risk Evaluation Guide comparison value (0.2 mg/kg). Total beryllium was detected at a maximum concentration of 6.9 mg/kg and soluble beryllium at 0.0054 mg/kg (11, 14). Concentrations of total chromium, also sampled on public and private property, exceed the pica child's RMEG comparison value for hexavalent chromium (but not for trivalent chromium) in 48 samples, with a maximum concentration of 92.6 mg/kg. None of the samples exceeded ATSDR's hexavalent chromium RMEG for non-pica children or adults. Samples were collected in both upwind and downwind locations of NGK (26).
No samples considered to be appropriate for background reference were collected for the area. Beryllium is a naturally occurring metal in the earth's crust, and beryllium studies of background soil in the eastern United States range up to 7 mg/kg (27, 28, 29). Such background soil studies indicate that beryllium exists in the soil (surface soil and otherwise) and are used by ATSDR throughout this document as representing concentrations that could exist for this particular area. Local soil background information would be needed to more accurately assess off-site soil conditions and potential site-related impacts.
Chromium is also a naturally occurring element. Chromium concentrations have been detected up to 1000 mg/kg in background studies in the eastern United States (28, 29).
Although NGK is an apparent contributing source of beryllium in the soil surrounding the site, from data reviewed it cannot be definitively determined to what degree NGK has contributed to off-site contamination. Fugitive dust and deposition of air emissions are possible pathways of off-site migration. Off-site chromium impacts from NGK, however, are less apparent. Regardless of the source or sources of beryllium and chromium in the soil surrounding NGK, ATSDR will evaluate this off-site exposure pathway and the concentrations detected in surface soil.
A completed pathway exists for people who live, work, and recreate in areas located within two miles of NGK. The greatest opportunity for exposure in surface soil would be at residential lawns, parks, playgrounds, gardens, and worksites where excavation occurs. Since beryllium was detected at the residence where private well 1 is located, ATSDR will consider the possibility of impacts on that family from multiple exposure pathways. Dermal contact and inadvertent ingestion of soil are the likely routes of exposure. Inhalation of beryllium is a less likely route, but could have occurred if excavation of soil and ground cover generated significant amounts of dust. An accurate estimate of this population cannot be determined; however, 14,686 people live within a 2-mile radius of the site.
B. Potential Exposure Pathways
A list of potential exposure pathways is located in Table 12 of Appendix B.
On-site Soil
There are past, present, and future potentially completed pathways through on-site soil. On-site workers could have been exposed to contaminants in the past during removal and disposal of process wastes or when working at waste disposal areas prior to their having been covered. Also, area residents who entered the site property prior to fencing could have come in contact with on-site wastes. On-site workers could presently (or in the future) be exposed to contaminants when working in or around the Disposal Area Drain Field which is currently uncovered. Nearby residents also may have been exposed to contaminants through dust created by excavation and/or wind erosion. Potentially, future exposures could occur to workers and the surrounding community during remediation, when waste areas will be excavated and combined, if protective measures are not taken to prevent migration of site contaminants.
Routes of exposure would include dermal contact, inadvertent ingestion of contaminated soil, and inhalation of airborne contaminants. The number of people potentially exposed to contaminants in the past is not known. Currently 4,927 people reside within 1 mile of the site and NGK employs approximately 100 people. Current and former workers and residents who have been on-site would represent the potential receptor population.
Off-site Groundwater
Private well 3 did not show any contaminants above comparison values. But beryllium, which was detected at elevated levels in other off-site wells, was analyzed using detection limits that are above current comparison values (15). Exposures could potentially occur for people using this well or other wells further west of the EPA specified well inventory area.
Groundwater from the off-site monitoring wells, piezometer, and Reading Crest Well showed contaminants that exceeded comparison values, but the groundwater has not been used for consumption or other domestic purposes. Based on sampling, exposure could result if any of those wells are used in the future for consumptive or domestic purposes. Also, based on off-site groundwater sampling, it appears that any wells placed within the EPA specified well inventory area might be susceptible to contamination. Therefore, the construction of wells or the use of groundwater supplied by wells placed within the EPA specified well inventory area should be restricted to prevent future exposures.
People using private well 3, and current or future users of groundwater within the EPA specified well inventory area and west of that inventory area extending to the Schuylkill River, could potentially be exposed to contaminants. Routes of exposure include ingestion (i.e., drinking water or eating foods cooked in water), dermal contact (i.e., bathing, showering, or washing dishes), and inhalation (volatile organic vapors from showering, laundering, sprinkling the lawn, and washing automobiles). The number of people who could potentially be exposed cannot be determined.
Off-site Sediment
There is no evidence that current exposure to Laurel Run sediment is occurring, although there is a potential for past, present, and future exposures, primarily for children, who would be most inclined to play in Laurel Run. Sediment exposures may also have resulted from fishing and bait collecting in Laurel Run.
Although some residents have reported that Laurel Run has been dredged in the past, there is no information indicating that sediments have ever been used in areas where greater potential for exposure could occur (e.g., sandboxes, volleyball lots, picnic areas).
Exposure could occur through dermal contact and inadvertent ingestion through splashing of water and sediment and hand to mouth activities. ATSDR believes that exposure to contaminated sediments from Laurel Run would be minimal, due to the infrequent opportunities for exposure, low probability of contact, and low doses of contaminated sediment that might be received. Even children with pica behavior (excessive ingestion of nonfood items) do not typically directly ingest stream sediment. The number of persons that might have been exposed to contaminated sediment is not known.
Off-site Surface Water
There is potential for past, present, and future exposures to contaminated surface water. Children living near Laurel Run, fishermen, and other stream users are the most likely receptors. Direct exposure to NGK wastewater discharge is very unlikely. Water from Laurel Run is not used as a public drinking water supply and there are no allegations or documentation of its use by individuals for consumption. Due to dilution of surface water and the settling out of contaminants, drinking water supply obtained from the Schuylkill River, downstream from Laurel Run, is unlikely to be significantly affected by pollution from Laurel Run alone. Other sources may also contribute to water quality problems in the Schuylkill River.
Surface water exposure routes would include incidental ingestion and dermal contact. The number of persons that might have been exposed to contaminated surface water is not known.
Off-site Biota
Biota near NGK and along Laurel Run might potentially have been and continue to be contaminated with those contaminants detected in ambient air, groundwater, surface water, and sediment. Biota may come in contact with and accumulate contaminants from their surrounding media (air, water, sediment), although some metals such as beryllium and copper are not readily bioconcentrated (27, 30). People have reportedly eaten fish (notably Suckers) caught from Laurel Run, downstream of NGK. Humans may also be exposed through the consumption of terrestrial plants or animals that have been grown or raised in contaminated areas (nearby residential yards and along Laurel Run).
There is past, present, and future potential for exposure, via ingestion of biota that may be contaminated with metals from the site. The number of people who may have consumed such contaminated food cannot be determined.
Workers' Clothing
A potential past completed pathway exists for the families of workers who were exposed to beryllium. In the past, plant workers were not required to wear protective clothing and thereby accumulated contaminants on their clothing. Contaminants carried home by workers may have resulted in exposures to family members. Family members, particularly through laundering contaminated work clothes, could have been exposed through inhalation, ingestion, and dermal contact. Current safety measures and protective clothing should reduce the opportunity for any present or future exposures of this type to occur.
There are no data or information to determine what levels of contaminants people may have been exposed. The number of people who may have been exposed through this pathway in the past in not known.
A. Toxicological Evaluation
In this section, ATSDR discusses health effects that could result from exposures to site contaminants. People can only develop health effects from a site contaminant if they come in contact with it; therefore, only contaminants present in completed pathways will be evaluated. In order to understand health effects that may be caused by a specific chemical, three factors affecting how the human body responds to exposure need to be considered. These factors include the exposure concentration (how much), the duration of exposure (how long), and the route of exposure (breathing, eating, drinking or skin contact). Lifestyle can affect exposure duration and likelihood. Individual characteristics of each human such as age, sex, nutritional status, overall health, and genetic predisposition can affect how a contaminant is absorbed, distributed, metabolized or eliminated from the body. Together, these factors determine the individual's response to chemical contaminants and what the health effects may be for that individual. Health effects from dermal absorption of compounds in water are hard to evaluate because they depend on length of exposure, exposed skin area, and frequency of washing, as well as properties of the chemical and how well it is absorbed across the skin.
ATSDR examines scientific studies and reports for individual contaminants (including those collected in the ATSDR Toxicological Profiles series). ATSDR uses those data to evaluate the potential for chemicals to cause harm to human health and determines levels of the chemical that can reasonably (and conservatively) be considered as harmless. That information has been incorporated (with safety factors to ensure protection of especially sensitive populations) into guidelines that can be used to identify chemicals of concern for further evaluation. ATSDR uses two kinds of guidelines as comparison values, environmental guidelines and health guidelines. The environmental guidelines can be used to determine if an environmental concentration of a compound is sufficient to merit further study. Such environmental guidelines include the Cancer Risk Evaluation Guides (CREGs), Environmental Media Evaluation Guides (EMEGs), Reference Dose Media Evaluation Guides (RMEGs), and other guidelines used in the tables in Appendix B of this document. The health guidelines include the Minimal Risk Levels (MRLs) and Reference Doses (RfDs). In this case, an estimate is made of the dose people are likely to receive from contaminants at the site, and this value is compared to the health guidelines. The health guidelines are specific for various segments of the population (adult, child, pica child) and for either cancer or non-cancer effects. In some cases, ATSDR has been unable to determine the values for use as guidelines due to lack of scientific data.
It must be emphasized that ATSDR's comparison values are not thresholds of toxicity. They were specifically designed to be protective of public health rather than predictive of adverse health effects. Thus, if a given concentration of an environmental contaminant is lower than the appropriate comparison value, that concentration of the specific chemical in the specific medium may reasonably (and conservatively) be considered safe. However, the fact that the concentration of a chemical exceeds a given comparison value does not necessarily imply that it is likely to produce adverse health effects of any kind. The conditions of exposure and individual susceptibility factors will determine whether or not the intrinsic toxic potential of a chemical is likely to be expressed. That is why contaminants of concern, identified by application of ATSDR's comparison values, are subsequently subjected to further analysis by ATSDR's toxicologists.
A description of the types of ATSDR and other comparison guidelines used can be found in Appendix C.
Some chemicals were found at elevated levels in various environmental media, but there were no completed pathways of human exposure for these chemicals. These included arsenic, copper, lead, antimony, barium, cadmium, manganese, nickel, selenium, thallium, vanadium, nitrate, and 1,1,1-trichloroethane.
Some chemicals were found at elevated levels within completed pathways, but were not at high enough concentrations to be a health concern when likely human doses were considered. These included fluoride, tetrachloroethene, and trichloroethene. In some cases, ATSDR lacks information to evaluate the potential carcinogenicity of these materials. Further, off-site soil, on-site air, and off-site air were not tested for many of these compounds. Beryllium, chromium, and 1,1-dichloroethene were also found at elevated levels within completed pathways and are discussed below in further detail.
Beryllium
Beryllium was found at elevated levels in on-site and off-site groundwater, on-site and off-site soils, on-site and off-site air, and in the water and sediment of Laurel Run.
Beryllium is used in strengthening alloys, primarily of copper, but also aluminum and nickel; it is used in alloys for fatigue resistance, corrosion resistance, and insulation. The major environmental source is combustion of coal. Beryllium can leach from soils and enter groundwater, but only to a limited extent as it is relatively insoluble and binds tightly to soils. Occupational exposure has occurred in mines and factories making alloys and products. Cigarette smoking can also be a major source of beryllium exposure.
Inhalation Exposure
Beryllium is scientifically recognized as a health problem, primarily by inhalation. Exposure through the inhalation route, via the air pathway, is the primary route for beryllium to cause health effects. Acute exposure by inhalation can cause inflammation of the lungs, with chest tightness, coughing, and fatigue, but eliminating acute exposure should result in restoration of the lungs to a normal state. Long-term exposure by inhalation can result in shortness of breath and some scarring of the lung, leading to chronic beryllium disease (CBD). Symptoms of CBD include shortness of breath, fatigue, weight loss, chest and joint pains, cough, and skin rashes.
Diagnosed cases of both acute and chronic beryllium disease have been extremely rare in recent decades. As of 1983, no cases of occupational berylliosis had been reported among individuals first exposed after 1973. With only one exception, no cases of CBD have been reported from indirect or nonoccupational exposure among individuals whose exposure began after about 1950 (31, 32). However, since CBD mimics the symptoms of sarcoidosis and may readily be confused with the latter disease, it is possible that additional, undiagnosed cases of CBD, masquerading as sarcoidosis, have occurred. CBD involves a cell-mediated immune response to beryllium exposure which appears to express itself primarily in people with a genetically-determined susceptibility for the disease. This fact has led to the development of the beryllium lymphocyte transformation (or proliferation) test which can distinguish between sarcoidosis and CBD (33). This test may be performed on a blood sample, but results are more reliable when performed using bronchiolar/alveolar lavage fluid. Corticosteroids, which are often prescribed for CBD, may interfere with the test.
By the inhalation route, some beryllium compounds (both soluble and insoluble forms) have been shown to increase the incidence of lung cancer in laboratory animals (27). Human evidence is more equivocal, however, since most of the positive studies have been inadequately controlled for confounding factors such as smoking. (Heavy smokers have a 10-20-fold increased risk of getting lung cancer. In addition, cigarettes contain approximately 0.5 - 0.7 µg Be/cigarette, some 4.5 - 10% of which escapes with the smoke. For a 2-pack-a-day smoker, the resulting intake of beryllium could be comparable to the dose that would result from breathing 0.01 µg/m3 for 24 hours a day. Thus, it is essential in an epidemiological study of beryllium and lung cancer that the results be controlled for smoking habits.) Currently, EPA considers the epidemiological evidence for beryllium-induced lung cancer in humans to be "inadequate". Where an excess of lung cancer has been detected, it has been more prevalent among workers with acute beryllium disease (i.e., chemical pneumonitis) than those with CBD (27). Historical health outcome data strongly suggest that implementation of OSHA's Permissible Exposure Limit (PEL) has effectively prevented the occurrence of new cases of beryllium lung disease which could possibly be a precursor of Beryllium-induced lung cancer, i.e., if beryllium is, in fact, carcinogenic in humans (32). In this connection, it is worth noting that OSHA's PEL of 2 µg/m3 is 5,000 times higher than ATSDR's CREG of 0.0004 µg/m3, while the NESHAPs regulatory limit of 0.01 µg/m3 is only 25 times higher than ATSDR's CREG.
On-site and off-site ambient air contained beryllium. Based on ambient air data in Table 9, concentrations of beryllium have exceeded the CREG comparison value at all of the 8 monitoring stations in the past and in 1994 exceeded the comparison value at each of the eight stations. In an earlier draft of this document, it was stated that "a slightly increased risk of cancer might be expected" if someone were exposed for an entire lifetime to beryllium at the highest concentration detected in off-site air. However, considering the low confidence attached to the study on which EPA based its classification of beryllium's carcinogenicity, and the generally low concentrations of beryllium in off-site air at this site, ATSDR does not expect any increased risk of lung cancer in nonoccupationally-exposed residents. As indicated by the preceding discussion, it may be of greater relevance to human health that the NESHAP's limit of 0.01 µg/m3, defined as a 30-day average, has only been exceeded twice (both times in 1989) in the last 16 years. On-site air data are limited in scope, representing five weeks in 1991 (Table 10). At that time, levels were insufficient to pose a health problem. Off-site data were taken for a number of years at various stations around the site. These values represent weekly concentrations. The highest concentration was observed in 1989 at station R-1 (Table 9), which was located very near the plant and downwind. This concentration (0.08143 µg/m3) did exceed ATSDR's CREG of 0.0004 µg/m3. However, this concentration was present for only a short time period (CREGs assume lifetime exposure) and this limited exposure is believed by ATSDR to represent no health hazard. Since levels at other sampling points and during all other time periods were significantly lower and of similarly limited duration, ATSDR considers these levels also to represent no public health hazard. Generally, levels off-site have been decreasing in recent years. In the judgement of ATSDR, if any adverse health effects occurred in response to higher off-site exposures in the past, they would probably be limited to CBD in a sensitive (i.e., immunologically predisposed) subpopulation living near the site. Since any past cases of nonoccupational CBD would likely have been misdiagnosed as sarcoidosis, long-term residents who have been diagnosed as having sarcoidosis and who suspect that they may have been exposed to clinically significant levels of beryllium in the past may want to consider consulting an occupational/environmental medicine specialist who can determine whether specialized testing for beryllium sensitivity is appropriate.
Ingestion Exposure
Beryllium metal does not cause disease by ingestion, because it is unable to cross the gut wall and enter the tissues of the body (27, 34). Some beryllium salts are more soluble, but are still not absorbed well. The more soluble beryllium salts (such as beryllium fluoride) react in the gut and form insoluble complexes with phosphates and proteins that are not well absorbed (35, 36). Beryllium has not been found to cause cancer by the oral route in either animals or humans (27). EPA's oral cancer slope factor and ATSDR's CREGs for beryllium in soil and drinking water were derived from a single laboratory study that showed a statistically non-significant increase in total tumor incidence in rats (males only) exposed chronically to 0.7 mg Be/kg/day as beryllium sulfate in drinking water (27, 37). This study does not, however, provide evidence that beryllium is carcinogenic via the ingestion route. The incidence of cancer did not increase with increasing dose. In another chronic study, rats exposed to 31 mg Be/kg/day as beryllium sulfate in feed exhibited no adverse health effects, including cancer (38). There is no evidence that beryllium can cause cancer in humans by the ingestion route.
Off-site groundwater was contaminated with elevated levels of beryllium. However, based on cancer and other health guidelines, levels in groundwater were not considered sufficient to cause any acute or long term effects, including cancer. The maximum concentration of beryllium in off-site groundwater was 5.3 µg/L, which is only 32% higher than the MCL of 4 µg/L. The MCL is not a strictly health-based criterion, and is mentioned here only for the purpose of comparison.
Beryllium levels in off-site soils were within the range expected for background levels in soil, which range from <1 to 7 mg/kg, with an average of about 0.85 mg/kg (28, 29). Therefore, this result may represent a background level rather than the result of migration of contamination from the site. Further, the beryllium may be tightly associated with the soil particles and may not separate from the soil to enter the body (low bioavailability). Although it is not known what form of beryllium may be predominant in the soil, it is known that insoluble forms, such as beryllium oxide, are most commonly found in ambient air and those forms of beryllium generally remain insoluble and immobile when deposited into the soil and sediment (27). Total beryllium was detected in off-site surface soil at a maximum concentration of 6.9 mg/kg which exceeds ATSDR's CREG of 0.2 mg/kg, but is under the RMEGs of 10, 300, and 4000 mg/kg for pica child, child, and adult, respectively. Soluble beryllium was detected at a maximum concentration of 0.0054 mg/kg, which does not exceed any comparison values. Since beryllium does not readily cause disease by ingestion, the concentrations of beryllium detected in soil and drinking water are not expected to pose any hazard to public health.
Dermal Exposure
Beryllium may also act as a direct irritant on the skin, nasal passages, and in the lung. Symptoms are not specific, but redness of the skin, opacity in the eye, coughing and chest tightness may result. The irritation may invoke an immunological (allergic) response, especially in genetically predisposed individuals, and may be worse in individuals previously exposed to beryllium. High levels of exposure via direct contact with beryllium may even lead to ulcers on the skin (27). However, based upon data reviewed by ATSDR, the levels measured in environmental media do not represent a public health hazard via dermal exposure.
Only the trivalent and hexavalent forms of chromium are of any biological significance. Trivalent chromium (Cr3+), the most common form of chromium, is an essential trace nutrient required for the proper function of several enzyme systems, including the glucose tolerance factor, phosphoglucomutase, and the succinate, cytochrome C reductase system (39). Cr3+ is neither irritating nor corrosive, and chronic inhalation or ingestion of Cr3+ compounds produce no adverse health effects. Trivalent chromium compounds have not been reported as carcinogenic by any route of administration. Normal levels of blood Cr range from 20 to 30 µg/dL (evenly distributed between RBCs and plasma), and urinary excretion is generally less than 10 µg/day (40).
Hexavalent chromium (Cr6+), the most toxic form of chromium, readily crosses cell membranes and is reduced intracellularly to Cr3+ (there is no evidence that the reverse reaction occurs to any significant extent) (39). Long-term inhalation exposure to insoluble Cr6+ compounds is associated with irritation and corrosion of the mucosa and submucosa of the respiratory tract, which may lead to ulceration and perforation of the nasal septum. Other skin surfaces may also become ulcerated. Occupational inhalation exposure to chromium, particularly in the chrome production and chrome pigment industries, is associated with an excess incidence of lung cancer. However, hexavalent chromium compounds have not produced lung tumors in animals by inhalation.
Nonoccupational exposure to chromium is primarily through ingestion of food (especially meat, vegetables, and unrefined sugar) and water (MCL = 100 µg Cr6+/L) (40). Acute renal tubular necrosis is the major acute effect of ingestion exposure to chromium (39). However, tissues can accumulate considerable quantities of chromium before pathological changes result. EPA's chronic oral RfD of 0.005 mg Cr6+/kg/day includes an uncertainty factor of 500, [i.e., it is 500 times lower than the no observed adverse effect level (NOAEL = 2.4 mg/kg/day)] determined in an animal study (41). No adverse health effects were detected by physical examination in a family of four persons who drank for 3 years from a private well containing Cr6+ at approximately 1 mg/L, or 1,000 µg/L (approximately 0.029 mg/kg/day in a 70-kg human) (42). While hexavalent chromium is considered to be a human carcinogen when inhaled, it is not thought to be a carcinogen in animals or humans when ingested in water (43).
Although private well 1 did have elevated quantities of chromium (Table 4a), ATSDR inaccurately concluded, in the previous draft of this document, that consumption of water from that well could cause "potential acute health effects". The maximum level detected was 281 µg/L Cr6+ (estimated value). The corresponding dose for a 70 kg adult drinking 2 liters of water per day would be 0.008 mg/kg/day which is marginally (60%) higher than EPA's RfD of 0.005 mg/kg/day for a lifetime (70 years) of exposure. Private well 1 was only used for eight years; therefore, no adverse health effects would be expected. The residents at that location were provided with an alternative water supply in 1991, thereby eliminating the possibility of excess exposure in the future.
Past and current levels of chromium detected in private well 2 are not expected to result in any adverse health effects; however, monitoring for increasing concentrations is advised.
Volatile Organic Compounds (VOCs)
The volatile organic, 1,1-dichloroethene (DCE), was present in private well 1 at levels that exceeded ATSDR's CREG of 0.06 µg/L. EPA classifies 1,1-DCE as a "possible human carcinogen" based on limited animal data and no data in humans. Based on that fact alone, it was stated in an earlier draft of this document that the levels of DCE in private well #1 "pose a slight increased risk of cancer". However, based on a more complete evaluation of the level and duration of exposure, and on a consideration of the basis for the cancer classification of DCE, ATSDR no longer considers this statement to be appropriate. Both EPA's classification of DCE and ATSDR's CREG are based on an animal study which showed a statistically insignificant increase of cancer (i.e., pheochromocytoma) in treated rats compared to controls. In that experiment, male rats were dosed by gavage with 5 mg DCE/kg/day for 2 years (i.e., most of a rat's lifetime). By comparison, the maximum level of DCE detected in private well 1 (i.e., 2 µg/L) would correspond to a dose (based on consumption of 2 L/day by a 70 kg adult) of 0.000057 mg/kg/day, i.e., 87,500 times lower than that in the animal study mentioned above. Even the maximum level of DCE detected in private well 1 was far below ATSDR's RMEGs of 90 and 300 µg DCE/L for children and adults, respectively. Therefore, no cancer or noncancer health effects from exposure to DCE would be expected at the concentrations found in private well 1. Private well 1 is no longer used (44).
B. Health Outcome Data Evaluation
The Commonwealth of Pennsylvania maintains a cancer registry (45). Three years of the registry (1984, 1985, and 1986) were examined for incidence or mortality rate of cancer in general and for lung cancer. Berks County showed no significant increases in cancer incidence or deaths from cancer in general or from lung cancer. However, the cancer registry is new, containing data collected during the last nine years. This database could reflect the consequences of chronic exposures beginning 20 to 30 years earlier, thereby taking into account the long latency period of some chemically-induced cancers. However, such a narrow "snapshot in time" (i.e., the last 9 years) cannot reveal any trends in cancer incidence in this area over the 60 year life of the NGK facility.
ATSDR also examined cancer records on the Centers for Disease Control WONDER computer database. No increases were found in the rate of any cancer type or in all cancer totals, when compared to the rates for the entire state of Pennsylvania, for white males or females. The data for minority populations were limited by small numbers and could not be analyzed. Besides community health concerns, no other sources of health information relevant to the site were identified.
C. Community Health Concerns Evaluation
1. The chemicals (chromium and fluoride) found in local private drinking water well might cause cancer.
Chromium may cause cancer when inhaled, but not when ingested in water (46). Levels in private well 1 exceeded ATSDR's RMEGs for children and adults. However, RMEGs are based on the assumption of exposure over an entire lifetime. Exposure over a much shorter time period would pose little or no threat to human health. Chromium acts as an irritant, and effects are most likely to be seen in sensitized individuals (possible effects are discussed in the Public Health Implications section). ATSDR does not expect any health problems to develop now that this water is no longer used. Fluoride was found in off-site groundwater at elevated levels but these are not high enough to cause health effects under the exposures that are likely to occur. Fluoride has two major effects: acutely, large amounts can be corrosive and irritating; smaller amounts over long periods of time may cause tooth mottling and skeletal degeneration. Fluoride does not appear to cause cancer. Fluoride, when it reacts with beryllium, forms beryllium fluoride, a more soluble form of beryllium that may make the beryllium more mobile in the environment.
The users of private well 1 were provided bottled water, and more recently, they were connected to the city water supply because of the levels of chromium in their well water. No other private wells contained sufficient chromium to present a health problem.
2. Contaminated On-site groundwater could contaminate local drinking water supplies.
The potential exists for future contamination of private wells within and west of the EPA specified well inventory area. The presence of fluoride may increase the soil and water mobility of beryllium. Groundwater monitoring should be conducted to ensure that contamination does not extend beyond the EPA specified well inventory area and that levels of contaminants in private well 2 do not increase. The Reading Crest Well has shown contamination, but has never been used as a drinking water supply.
3. Airborne dust from NGK's old wastewater treatment lagoons might cause medical problems.
The Disposal Area Drain Field is the only waste area on-site that has not been covered. No surface soil samples have been taken to characterize what contaminants might be present in the top 3 inches of that waste area or other potentially contaminated areas. Air monitoring was conducted beside (see Figure 5 and 6) the Disposal Area Drain Field and the Former Pond 6 waste pile during the July 1991 RCRA ambient air sampling program. Based upon weather observations during that sampling program, it appears that beryllium from waste areas is impacting air quality. Results from those monitors are reported in Table 10 and additional discussion regarding the RCRA ambient air sampling program can be found in On-site and Off-site Contamination sections under "Ambient Air."
The data are insufficient to judge potential effects from the dusts arising at these sources. Overall off-site air has in the past frequently exceeded ATSDR's CREG for beryllium in air. However, based on available information, those concentrations have not been sustained long enough for adverse health effects to be expected.
4. Residents living in the Reading area may develop sarcoidosis from exposure to beryllium oxide.
Beryllium causes chronic beryllium disease (CBD), a sarcoidosis-like condition. Its symptoms mimic those of sarcoidosis, and it may even be misdiagnosed as sarcoidosis. The two diseases should not, however, be equated with one another. CBD is a granulomatous lung disease caused by a hyperactive, cell-mediated, immune response to chronic beryllium exposure. It appears to require a genetically-determined sensitivity that does not obey any predictable dose-response relationship. CBD is seen primarily in factory workers and miners; only rarely is it seen in residents near factories or mines, suggesting that a large dose is necessary for the condition to develop. However, the effects of long term, low dose exposures are unknown and may well result in disease, especially when exposure is punctuated with short term, episodic, high level releases. The amount of beryllium exposure needed to cause CBD is uncertain. It is possible that episodic releases of dust containing considerable amounts of beryllium could have taken place. CBD has a latency of several months to several years (disease usually develops 10-15 years after exposure). Symptoms include shortness of breath, fatigue, weight loss, chest and joint pains, cough, and skin rashes. CBD may or may not progress, but it does not spontaneously clear up. There has been only one documented nonoccupational case reported nationwide among individuals whose exposure began after about 1950 (27,31,34,).
Sarcoidosis is a chronic disease of unknown cause characterized by formation of nodules, especially in the lymph nodes, lungs, bones, and skin. Statistics on the incidence and prevalence of sarcoidosis are highly uncertain due to the relative rarity of this chronic disease, the variable severity of its symptoms, the potential for misdiagnosis, and non-representative nature of most study populations. It is clear, however, that the incidence of sarcoidosis in the U.S. is much higher in blacks than in whites. There is no sex predominance in the incidence of sarcoidosis worldwide, and, in caucasian populations, cases of sarcoidosis are almost equally divided between men and women. Although the disease has been reported to be 2-3 times as common among black females as black males, this finding may only reflect the fact that most of the early studies were done in large urban hospitals where the majority of patients seeking medical attention for any ailment happened to be females, especially black females. The current concensus is that there is no predominance of sarcoidosis among women of any race (47). Of all the potential risk factors studied (i.e., genetic, racial, infectious, environmental, occupational, smoking, and presence of other disease), only genetics and possibly geography are well established risk factors for sarcoidosis.
Sarcoidosis is most common in young adults between the ages of 20-40, who live in rural areas. It is an immunologically based response to environmental contaminants, although the specific agents involved are unknown. Because CBD may easily be misdiagnosed as sarcoidosis, ATSDR has examined sarcoidosis as a potential site-related health effect. However, based on currently available data, ATSDR is unable to conclusively establish any clear relationship between the site and the identified cases of sarcoidosis. Any long-term residents who have been diagnosed as having sarcoidosis and who suspect that they may have been exposed to clinically significant levels of beryllium in the past may want to consider consulting an occupational/environmental medicine specialist to determine whether specialized testing for beryllium sensitivity is appropriate.
5. Untreated storm water runoff and treated wastewater from NGK that are discharged into Laurel Run could be having a detrimental effect on aquatic wildlife. Furthermore, the treated waste and untreated storm water that are discharged into Laurel Run may eventually reach Schuylkill River (via Laurel Run) and contaminate local water supplies down river.
Some surveys suggest there is little wildlife immediately downstream (19, 21). However, more recent studies demonstrate that there is an active juvenile fish population downstream and the junction of the Schuylkill River with Laurel Run, is considered a good fishing spot (2, 48). Laurel Run is contaminated near the site, but further downstream recovery occurs. Other facilities may have also impacted the water quality of Laurel Run. Levels of site contaminants in the water and sediments of Laurel Run were generally low, with respect to ATSDR comparison values. However, contaminant concentrations in water and sediment may be detrimental to aquatic wildlife. ATSDR is recommending fish tissue sampling to evaluate any health threat that may exist for people consuming fish from Laurel Run or its confluence with the Schuylkill River.
Due to dilution of surface water from Laurel Run by the much greater volume of water in the Schuylkill River, and the settling out of contaminants, drinking water supplies obtained from the Schuylkill River downstream from Laurel Run are unlikely to be significantly affected by pollution from Laurel Run alone. Other sources may also contribute to water quality problems in the Schuylkill River.
6. Some people may have contaminated private wells and may be unaware of contaminated groundwater. This may affect people who have summer homes and were not interviewed during the well survey.
Other than the wells identified in this report, no wells are known to be in use within the EPA designated well inventory area. All homes and businesses within the well inventory area were contacted by Dunn Geoscience Corporation and water supply sources were verified by homeowner/occupant or the Muhlenberg Township Authority. ATSDR believes that this well inventory has adequately identified private well users who live within the inventory area (2, 5).
7. The Reading Crest Well water is contaminated, and attempts have been made in the past to bring the well into service. There is concern that the well may be brought on-line in the future.
To date, the Reading Crest Well has not been used as a public water supply. A memo, dated November 30, 1990, from the Muhlenberg Township Authority states that the Authority has no plans to develop the Reading Crest Well (49).
8. Laurel Run is impacted by contaminants in surface water from wastewater discharged from NGK. There is concern for the health of children who play in Laurel Run and persons who may have eaten or may currently be eating fish from Laurel Run or at the confluence of the Schuylkill River and Laurel Run.
There are no health outcome data to show whether the health of children who may be playing in Laurel Run or persons who may have eaten fish caught from Laurel Run are being affected. Although individuals have not been identified, there are potential pathways of exposure for people who may have eaten fish from Laurel Run and for people who may be using/playing in the stream. There are no fish tissue data available that might show contamination or allow ATSDR to evaluate possible health effects. The health effects of the primary contaminant of concern, beryllium, are usually the result of inhalation. In order to receive harmful doses of site-related contaminants from stream sediment, considerable quantities of sediment would have to be ingested, which is unlikely.
9. The present parameters for the NGK National Pollutant Discharge Elimination System (NPDES) permit are too high, and compliance for the new, more stringent, standards is not required until August 1993. The Pennsylvania Department of Environmental Resources (PADER) and EPA have delayed compliance deadlines several times already.
Those concerns are regulatory in nature and are to be addressed by the regulatory agencies, PADER and EPA, responsible for establishing those regulations and standards. ATSDR is an agency of the U.S. Public Health Service that addresses the public health impact at hazardous waste sites. Through this petitioned public health assessment, ATSDR could become indirectly involved through its evaluation of surface water and sediment in Laurel Run. Although ATSDR does not have regulatory authority, the Agency would make recommendations to determine the source and to reduce contaminant levels or restrict access to Laurel Run, if ATSDR concluded that Laurel Run represented a public health threat. To date, ATSDR has not determined Laurel Run surface water to represent a public health threat, but is recommending sampling of sediment and fish for site-related contaminants to more fully evaluate those pathways.
10. Deposition from air emissions from the metal facility over its entire history has accumulated in homes and on lawns throughout the community. Such contamination is believed to have impacted the health of persons in the community in the past and could be a current and future health threat.
There are no health outcome data to show whether the health of people living around NGK have been affected. Based upon off-site soil and air data reviewed by ATSDR, soil concentrations do not appear to represent a past, current, or future public health threat and air concentrations do not represent an immediate past (within the last 15 years) or current public health hazard.
11. The analytical procedure used to analyze the concentration of beryllium from air monitoring conducted since 1979 is not measuring all forms of beryllium present in the sample, therefore, is not revealing the actual concentration of beryllium present in the air.
This concern is being addressed by EPA, as it relates to EPA's analytical procedures. If EPA determines that the analytical procedure is not revealing the actual concentration of beryllium present in ambient air, ATSDR will reassess the health threat based upon new data as it becomes available.
12. There are no air monitors due south of the NGK plant to detect the levels of beryllium in which the nearest residents (along Water Street) might be exposed.
On-site air monitoring was conducted, during July 1991, at two stations (RCRA 01 and RCRA 02) along NGK's southern property line (see Figure 5 and 6). Since the concentrations detected (see Table 10) are representative of on-site ambient air at the southern portion of the site, those concentrations are likely to be similar to concentrations along Water Street, which is just off-site (to the south). While the concentrations detected do not represent a health concern, only five weeks of data from those monitors were available for ATSDR's review.
13. People have developed brain tumors and lung cancers due to exposure from site-related contaminants.
Although their cause is unknown and multiple causes may exist, brain tumors have not been linked to beryllium exposures. Lung cancers may be caused by beryllium exposure, but lung cancer also has many other causes, especially smoking tobacco products. It is not known if a cause and effect relationship exists between a particular site-related contaminant, exposure, and brain tumors or lung cancer.
14. The community could be exposed to contaminants during remediation when the solid waste management units are consolidated.
The community could be exposed to contaminants during remediation, if actions are not taken to prevent off-site migration of contaminants. ATSDR has listed this as a potential pathway of exposure and is recommending that protective actions be taken to prevent exposures to on-site workers and the surrounding community.
15. What is beryllium poisoning and its symptoms? Once absorbed, where does it go?
The nature and signs of beryllium poisoning have been discussed in the Public Health Implications Section of this petitioned public health assessment.
16. Health conditions that were reported by citizens as having occurred or occurring in the community include: children with liver, heart problems, asthma, and allergies; cancers of various types; emphysema and general respiratory illnesses; berylliosis; Parkinson's disease; hodgkin's disease; brain tumors; myopathy; mottling of teeth; brittle bones; hair loss; rashes; irritation at night; lots of colds; and a degenerative condition resulting from the side effects of treatment for beryllium poisoning.
A variety of health complaints have been reported by the community. Those that are possibly related to site contaminants of concern have been discussed in the Public Health Implications section of this petitioned public health assessment.
17. Could a child with asthma be more susceptible to beryllium?
It is possible that beryllium dust, when inhaled, might trigger an asthmatic response from the physical irritation of the dust.
18. Dust carried home, from the beryllium plant, on worker's clothing may have resulted in illnesses.
Based on reports from previous employees, it appears that the potential existed for contaminants to have been carried from the plant on their work clothing. Therefore, potential exposures and illnesses could have occurred. Although this may have been a possibility in the past, it is NGK's current procedure to provide protective garments that are collected and to require showers and clean change of clothing at the end of each work day for those employees that might be exposed to beryllium.
19. Exposures may result from residential gardens, community parks, and playing fields.
As discussed in the Pathways Analyses section, there is the potential for exposures to contaminants that may have migrated to the above mentioned off site locations. Although ATSDR is not able to determine whether migration has resulted in contamination of those areas, concentrations detected in recent soil sampling (at the above mentioned locations) does not represent any public health threat.
20. Could contaminants in groundwater cause respiratory infections from showering or cause mottled teeth, stress fractures, and colic in children?
ATSDR has determined that water from heavily contaminated wells should not be used for drinking. However, none of these effects are likely from the levels of contamination present in the groundwater.
21. Work practices at the plant were very poor and dusty in the past.
A number of previous workers have described poor working conditions. This document will be referred to the National Institute for Occupational Safety and Health for investigation of work related health concerns.
22. Exposures may have occurred from going on site before the site was fenced or from fields or caverns where contamination was dumped.
ATSDR has no information regarding dumping in caverns or fields, with the exception of wastes that were piled on-site in the past. Since we do not know the exact time period that nearby residents were on-site, the areas on-site where people went, the activities that they engaged in while on-site, and the location and extent of contamination during those times, it would be difficult to determine whether exposures occurred or the likelihood of adverse health effects. However, it is possible for such exposures to have occurred in the past, before the site was fenced.
23. There are a lot of illnesses around the site, particularly in the Cherokee Ranch area.
ATSDR does not have any health outcome data that are specific to the Cherokee Ranch area alone. But ATSDR acknowledges that several people who attended ATSDR availability sessions made mention of illnesses in the Cherokee Ranch area. As discussed in the Health Outcome Data subsection, data for Berks County do not show a significant increase in cancer; however, limitations for that data do exist.
24. Possible health hazards such as digging at or around the site, contaminated off-site groundwater, and air violations have not been communicated to the public.
A large portion of the public surrounding the site was contacted about the site through the well survey conducted in 1991. EPA has released a document about the site for public comment and conducted a public meeting in 1992. Other activities have taken place, such as the ATSDR site visit and availability sessions, where communications with the public have taken place. However, ATSDR is not in the position to determine whether any or all of those activities have been adequate to communicate the possible hazards to the public, although ATSDR does believe that this document should aid in accomplishing that task.
25. Lake Ontalaunee, from which the City of Reading gets its water supply, is being contaminated with beryllium.
A private citizen provided ATSDR with a table from a report showing results from four lake sediment samples that were collected on October 7, 1985. The expressed concern was regarding beryllium, chromium, and copper that were detected in sediment. Beryllium ranged from 0.86 to 1.4 mg/kg, chromium from 19 to 25 mg/kg, and copper from 17 to 75.6 mg/kg. Of those contaminants, only beryllium exceeds comparison values. Based upon another report completed in August 1993, sediment was sampled for beryllium at five locations at the Lake, but was not detected at the detection limit of 2.0 mg/kg (50). The source or sources of beryllium in sediment at Lake Ontelaunee is unknown, but there does not appear to be any significant contribution from man-made sources since concentrations are within the range of background concentrations for the eastern United States. Beryllium is highly insoluble and is not expected to pose a public health threat by the ingestion route.
26. Further contamination may have resulted from floods on Laurel Run in the past and dredging conducted by the Army Corp of Engineers.
The spread of contaminants onto creek banks and stream front property may have occurred in the past as a result of flooding. However, based on the limited stream sediment samples taken to date, concentrations of contaminants were relatively low. Therefore, significant contamination is not thought to have occurred. However, further sampling of stream sediment have been recommended. The Army Corp of Engineers has completed various work assignments on Laurel Run. ATSDR has not obtained any information indicating that sediments have been removed from Laurel Run.
27. Orange and green colored smoke that caused individuals to experience a burning sensation was emitted from plant stacks in the past. Also in the past, ash from the plant would deposit on automobiles and seemed to deteriorate paint.
ATSDR has no data to evaluate the health impact and concerns related to plant emissions prior to 1979. Beryllium could potentially cause a burning sensation in the nasal passages. No colored smoke or heavy ash have been associated with the site in recent years.
28. Water in Laurel Run turns strange colors when it rains. Could water from Laurel Run that was used on gardens in the past be a health concern?
ATSDR cannot confirm the reports of Laurel Run turning strange colors during storm events. Aside from loading of sediment in the water column and possibly petroleum products being washed from highways and businesses, ATSDR cannot provide any insight as to the change in color of Laurel Run water. In the past, depending on the extent of contamination, it may have been possible for contaminants in stream water to be transferred via irrigation to soil and food stuffs in gardens. Currently, however, contaminants detected in surface water at Laurel Run have been relatively low in concentration (see Tables 6a, 6b, and 6c). Therefore, significant contamination would not likely result from periodic watering of residential gardens from Laurel Run.
29. There is a concern that the RCRA Facility Investigation conducted by Dunn Geoscience Corporation and air monitoring data collected by NGK are not reliable sources of information.
ATSDR has prepared this document based on all the relevant, available data that ATSDR was able to obtain. Documents used are referenced and the information and data are assumed to be reliable and accurate. ATSDR attempts to indicate (in the Quality Assurance/Quality Control subsection) any questions, problems, or inconsistencies that are recognized when evaluating data. Besides those issues raised in the Quality Assurance/Quality Control subsection, ATSDR has no basis, from a technical standpoint, to disqualify any available data as unreliable and inaccurate. Further, it is not within ATSDR's purview to police data monitoring, collection, or reporting. However, ATSDR would make any necessary revisions to this document, if data used herein are found to be inaccurate.
30. Do elevated levels of CD4+ T cells in the lung or blood make people more susceptible to chronic beryllium disease? What is a normal CD4+ T cell count and what would be abnormally high? Should OSHA or ATSDR check the CD4+ T cell count of people who work at and live around beryllium alloy manufacturing facilities?
Is beryllium the antigen in chronic beryllium disease? Is it possible that sarcoidosis has been analyzed in cases that may actually be chronic beryllium disease? If beryllium shows up in a biopsy of lung tissue (dried) from a person diagnosed with sarcoidosis, at what concentration (ppm) would it be a questionable case of chronic beryllium disease?
CD4+ T cells may accumulate in the lungs in response to beryllium exposure. However, they remain a research tool and their significance, except as evidence of an immune response, is not clear. Therefore, CD4+ T cells would not clearly or absolutely indicate susceptiblity to CBD. A normal CD4+ T cell count should range from 900-2800 cells/microliter (51). Because of the many other factors that can affect the numbers and location of immunological cells, CD4+ T cells would be a poor indicator of beryllium exposure and therefore may be of little use to ATSDR and/or OSHA. ATSDR's Health Activities Recommendation Panel reviews each public health assessment to determine whether any further health follow up (such as biological sampling) is warranted. CBD is known to have an immunological component. Simple compounds of beryllium are too small to elicit an immune response; antibodies are formed in response to foreign proteins which are hundreds or thousands of times larger than simple molecules. Therefore, the ultimate antigen, which has not yet been identified, is probably some proteinaceous product to which beryllium is bound. Sarcoidosis is discussed more fully earlier in this subsection, under Community Health Concern number 4. As we state in that response, sarcoidosis presents similar signs as does beryllium disease, except for the demonstrable presence of beryllium in the lungs. Therefore, based on testing lung tissue for the presence of beryllium, sarcoidosis should not be confused with CBD. It is the demonstrated beryllium-immunosensitivity of alveolar lymphocytes, rather than simply the presence of beryllium in the lung at any particular concentration, that supports a diagnosis of CBD as opposed to sarcoidosis. Beryllium may be present in the lungs of exposed individuals in either the presence or absence of disease. If beryllium is actually found inside granulomas from a patient's lung, it is not unreasonable to infer that the beryllium itself may have been the initial stimulus for granuloma formation. Even then, however, a conclusive diagnosis of CBD would require the demonstration of actual beryllium sensitivity. Currently, the easiest way to diagnose CBD is to test for beryllium sensitivity in white cells from blood or bronchoalveolar lavage fluid.
31. Is anyone doing research on a connection between beryllium and brain tumors or arthritic conditions resulting from absorption and or ingestion of beryllium?
Brain tumors have never been associated with beryllium exposures and research on this topic is
unlikely. Much research is being done to address the causes of arthritis, but we do not know if an
association with beryllium is being examined.
