Skip directly to search Skip directly to A to Z list Skip directly to site content




The Shaffer Equipment Company (SEC) site, located in Fayette County, Minden, West Virginiaconstructed electrical substations for area coal mines from the period 1970 to 1984. The site isapproximately one acre and has one building (SEC Equipment Building) that served as both awarehouse and office. Electrical equipment such as transformers, switches, circuit breakers, andcapacitors were stored on the site. Dielectric oils that contained polychlorinated biphenyls (PCBs)and 1,2,4-trichlorobenzene were found in on- and off-site soils and sediments. Because PCBs are onsite and PCB-contaminated oils reportedly were burned as starter fuel in the warehouse/officebuilding, on- and off-site soil samples and on-site sediment samples were analyzed forpolychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo-p-furans (PCDFs). On-and off-site soil and on-site sediment concentrations of PCDDs and PCDFs were not judged by theAgency for Toxic Substances and Disease Registry (ATSDR) to be at levels of public healthconcern. Sampling for PCDD/PCDF and PCB contamination was not conducted for thewarehouse/office building.

There is a great deal of community concern and interest about the SEC site. There have beenseveral studies, surveys and reports on the site regarding excess cancers and other adverse healtheffects. In addition, the Concerned Citizens to Save Fayette County (CCSFC) arranged to haveblood and adipose tissue sampling/analysis for PCBs from residents of the area. ATSDR requestedthat data from the CCSFC. The information has been included in this public health assessment.

A past, completed PCB-exposure pathway was identified for on-site SEC workers and their families. Routes of exposure may have been: 1) dermal contact (contact with contaminated soil, sediments,occupational exposure from handling dielectric oils, secondary exposures of family membersprimarily through washing clothes contaminated with dielectric oils), 2) inadvertent ingestion(ingestion of small amounts of contaminated soils, sediments from soiled hands during eating,smoking, drinking and other activities), and 3) inhalation of contaminated airborne dusts. This past,completed pathway does not currently exist because SEC is no longer an active, operating facility.

Several past, potential PCB-exposure pathways were also identified for on-site workers and off-siteresidences that used PCB-contaminated oils as starter fuel for coal igniting. The routes of exposuremay have been: 1) inhalation (incomplete burning of PCBs and formation of PCDD and PCDFcompounds), 2) dermal contact (handling of PCB-contaminated oils), and 3) inadvertent ingestion(ingestion of contaminants from soiled hands). Those past, potential pathways no longer exist, asPCB-contaminated oils are no longer available from SEC.

Other potential pathways that have been identified as past, present, and future concerns involvetrespassers onto the SEC site, children playing in yards and Arbuckle Creek, on-site workers in theSEC Equipment Building, and persons that eat snapping turtles from the area.

Of special concern are individuals who eat snapping turtles that may have been PCB-contaminatedfrom the SEC site. Snapping turtles may bioaccumulate PCBs in the flesh; therefore, pregnantwomen and nursing mothers are cautioned not to eat snapping turtles caught in the site vicinity untilsuch turtles are caught, bioassayed, and determined safe to eat. Sensitive subpopulations tochemical contamination are fetuses and breast-fed infants who may receive PCB exposure fromacross the placental barrier or through mother's milk.

Surface water from Arbuckle Creek is contaminated with fecal coliform bacteria associated with theimproper disposal of human wastes. The fecal coliform source has not been identified. If analysesdo not find PCB contamination in turtles, aquatic life from Arbuckle Creek should still not be eatenuntil the source of the fecal coliform bacteria is removed or unless the food is thoroughly cooked. Recreational uses of the creek should be restricted until further testing indicates fecal coliformcounts are within public health guidelines.

ATSDR has determined that the SEC site (on-site) continues to pose a public health hazard fortrespassing adults and children, and for on-site workers. These groups are at risk of inadvertently ingesting, inhaling, and or having dermal contact/absorption with on-site surface soils and sedimentsthat were highly contaminated with PCBs in the past and where contamination may still exist. Based on recent off-site PCB sampling in the vicinity of SEC, ATSDR believes the SEC site is anindeterminate public health hazard for the general off-site population under current conditionsbecause complete PCB data were not available for off-site surface water and off-site groundwatercontamination. There is no significant human exposure evident in regard to off-site soils, sedimentsand air pathways.

ATSDR has made recommendations to eliminate or reduce the potential for future exposures. Recommendations have also been made for additional on-site characterization, specifically forsurface soil sampling. ATSDR's Health Activities Recommendation Panel made recommendationsfor follow-up investigation and possible retesting of an infant/child showing levels of PCBs inserum, referral of the final public health assessment to the National Institute for Occupational Safetyand Health for review, and possible investigation of past worker exposures, and environmentalhealth education for local public health professionals and the medical community.

ATSDR released a draft of this public health assessment, for public comment, from January 25 -February 23, 1993. ATSDR addressed those comments in Appendix 4 of this document. A publicmeeting is planned to follow the release of the final Shaffer Equipment Company Public HealthAssessment.


A. Site Description and History

The Shaffer Equipment Company (SEC) site is in Fayette County, Minden, West Virginia (Figure1), off Old Minden Road (Route 17). The facility constructed electrical substations for area coalmining industries from 1970 to 1984. SEC stored unneeded, damaged, or outdated transformers,capacitors, switches, and voltage regulation/distribution devices on site. Leaks from that equipment,possible spills, and dumping practices appear to be responsible for the on-site polychlorinatedbiphenyl (PCB) contamination. Local newspapers reported on a public meeting held by EPA andattended by ATSDR (May 29, 1990) that PCB oils (oils contaminated with PCBs) were burned onsite and may have been given away or sold as fuel. ATSDR interviews (1) with several formerShaffer employees also indicated that this may have occurred.

PCBs are commercial compounds that were widely used for insulation and lubrication in electricalcables, transformers, and other equipment. There are 209 individual PCB compounds. Commercialmanufacture of PCBs was stopped in the United States in 1977. Use of PCBs was halted because ofthe persistence of these compounds in the environment and the potential for adverse health effectsfrom exposure to PCBs (2,3).

The site size is approximately one acre and contains one building (SEC Equipment Building) thatserved as both warehouse and office (Figure 2). The site is in a valley and is surrounded by hills. The immediate site is flat and is bordered to the northwest by Arbuckle Creek. Arbuckle Creekexperiences occasional flooding, which is accompanied by sediment migration. Arbuckle Creekflows to the northeast. The site is bounded on the south by hills, north by Arbuckle Creek, and onthe southwest and northeast by fencing. Although access to the southwest part of the property nearthe warehouse/office was restricted by a chain-link fence and locked chain-link gate, the northeastportion of the property is easily accessed through a cattle gate. Not all of the PCB-contaminatedproperty is fenced. Most of the area that EPA excavated and reclaimed during emergency removaloperations of PCB-site contamination in 1984-1985 is not fenced. That includes some property ofthe Berwind Land Company. There are reports of vandalism and recreational use of the site. Heavybrush and trees cover most of the site, although some areas are cleared for parking and turn-around. No concrete or asphalt pavings are present. There are residences across Arbuckle Creek from thesite, both up- and downstream. The Minden wastewater treatment plant is about a mile upstreamfrom the Shaffer site. The plant is permitted to discharge wastewater into Arbuckle Creek.

The West Virginia Department of Natural Resources (WVDNR) visited the site in September 1984. An inspection of the site revealed several hundred transformers and capacitors. A composite surfacesoil sample and a grab-sample of soil/sediment from a site drainage ditch into Arbuckle Creekindicated PCB concentrations of 26,749 parts per million (ppm) and 1,136 ppm, respectively.

Subsequent sampling of soil and sediment by EPA in late 1984 indicated PCB contaminationranging from less than 1 ppm to greater than 200,000 ppm. There were approximately 150transformers and about 50 capacitors on the site; some showing evidence of oil leakage. PCBs hadeither migrated from the site during occasional flooding or had been spilled into Arbuckle Creekbecause PCB concentrations of up to 190 ppm were found in creek sediments.

The Centers for Disease Control (CDC) certified in December 1984 that an imminent threat topublic health existed because of high levels of PCBs in the environmental media (4). EPA approvedfunds for an emergency removal of PCBs from the environment and the removal ofPCB-contaminated equipment in late December 1984. PCB-contaminated soils were excavated andplaced in a clay-lined staging area for treatment. An innovative process for extracting PCBs fromsoil using freon and methanol was tried without success until November 1985, when the process wasterminated. EPA contracted for the removal of the waste pile (4735 tons) by trucking to ahazardous waste landfill in Emelle, Alabama. Removal activities began in September 1987 andended in November 1987. EPA site operations were declared complete in December 1987.

In April 1989, 19 drums containing paints, solvents, and waste oils were discovered on the site. Those drums were to be previously disposed of by the property owner in an agreement with EPA. The property owner declined responsibility and EPA initiated removal of the drums in May 1989.

On May 24, 1989, ATSDR received a petition from the West Virginia Department of Health(WVDOH) to conduct a public health assessment for the SEC site. ATSDR subsequently acceptedthe petition.

In March 1990, EPA conducted additional on- and off-site sampling for volatile and semivolatileorganic compounds, PCBs, and inorganic chemicals from surface soil, subsurface soil, surfacewater, sediment, and raw water (water used for public water supplies from the Minden Mine intakeand the Rocklick Mine intake). That action was in response to a request from the Agency for ToxicSubstances and Disease Registry (ATSDR) for additional off-site sampling. On-site and off-site soilsamples were also collected for polychlorinated dibenzo-p-dioxin (PCDD) and polychlorinateddibenzo-p-furan (PCDF) analyses. Approximately 19 on-site and 5 off-site surface soil/subsoilsamples, 6 surface water samples, 7 sediment samples, 1 spring surface water sample, and 2 rawwater samples were taken.

In May 1990, EPA resampled for verification purposes, a number of the same locations that testedpositive for PCBs from the March 1990 soil/sediment sampling event.

In June 1990, EPA performed additional sampling of three areas near the Shaffer Equipmentbuilding in response to reports made by the Concerned Citizens to Save Fayette County (CCSFC) atthe May 29, 1990 public meeting.

EPA conducted another emergency removal action from November 1990 to January 1991 toremove soil from the remaining areas containing high levels of PCBs. Sampling was conductedbefore, during, and following the removal actions.

ATSDR's evaluation of the 1990-91 sampling results, which are the most detailed and includes on-and off-site sampling, will be the primary focus of this public health assessment.

B. Site Visits

ATSDR staff (Lynn Wilder and Donald Joe) visited the Shaffer site on August 25, 1989 and onMay 29, 1990. Site conditions were noted including accessibility, demographics, and proximity ofany residences to the site. ATSDR met with representatives from the West Virginia Department ofHealth (WVDOH) and the Fayette County Health Department.

C. Demographics, Land Use and Natural Resource Use

Minden, West Virginia is a small coal mining town of approximately 2,000 inhabitants. DuringATSDR's site visit in August 1989, it was noted that the only elementary school in Minden had beenclosed. There were no schools, parks, playgrounds, nursing homes, or hospitals within a 1-mileradius of Minden. The racial character of Minden was estimated during the site visit to be 90%Caucasian and 10% African-American. There are many elderly people and children in Minden. The average age for the area was estimated to be 30 years. Approximately 65-75 people live within400-500 feet of the site. In the past, coal mining was prevalent, but many of the coal mines haveclosed. Homes and yards are well-kept. The general area is economically depressed. Current landuse in the vicinity of the site is residential.

In the 1980 census, Fayette County had a population of 57,863. The majority (eighty-four percent)of the population lives in rural areas of the county. Sixty-two percent of those in the county havelived in the same house for at least five years. An additional twenty-three percent have lived in thecounty for this period but in several different homes (5,6).

Sixteen percent of the county's 19,889 households had annual incomes of less than $5,000. Seventeen percent of the county residents live below the poverty level compared to twelve percent ofthe residents in the state. The per capita income in the county was $5,295 in 1980. (5,6) Reportsfrom the citizens' group (CCSFC) in Minden indicated that the per capita income in Minden is$4,000.

The racial make-up of the county is predominantly white (93 percent). Eight percent (4,505) of thecounty residents are under age five and nine percent (4,952) are over the age of sixty-five. In 1980,15,858 persons over the age of sixteen were employed. Of those, 3,194 (20 percent) were employedin the mining industry. (5,6)

The population figures from 1986 indicated the county's population was falling with a 4.1 percentnet loss. In 1984, the percent of persons under age five remained relatively stable at seven percentwhile the percentage of those over age 65 rose to 14 percent. The birth rate for the county in 1984was similar to that of the state (12.5 to 12.6 per 1000 population). The crude death rate was 10.6 inthe county and the state death rate was 9.8. The infant mortality rate for the county was 7.0 per1000 live births. This is much lower than the rate of 11.0 per 1000 live births in the state. (7)

D. Health Outcome Data

The evaluation of health outcome data may give a general picture of the health of a community. Those data may confirm the presence of excess disease or illness. However, elevated rates of aparticular disease may not be due to hazardous substances in the environment. Other factors such associo-economic and personal habits may have a tremendous influence in the potential developmentof disease. In contrast, even if elevated rates are not found, a contaminant may still have causedillness or disease. ATSDR must depend on previously gathered data to perform a public healthassessment. There were several sources of health outcome data available for ATSDR review andthese are listed as follows:

ATSDR reviewed the 1980 Census for Minden and the Riggans Cancer Mortality Data for FayetteCounty, West Virginia. (5,6,7,8)

In 1983, the West Virginia Department of Health conducted an examination of cancer mortalitystatistics for the Oak Hill, W.V. area. ATSDR reviewed the report, Oak Hill, W.V. Investigation --Stage I: Initial Assessment. (9)

In 1986, the Appalachian Student Health Coalition of Vanderbilt University assisted the localcitizen's group, CCSFC in conducting a health survey of the residents of Minden. In 1989, CCSFCand the Virginia Student Environmental Health Project conducted a similar survey of the residentsof Page, West Virginia to use as a cohort/control population to compare with the Minden survey. ATSDR reviewed both of these surveys. (10)

In 1989, a Beckley, West Virginia gynecologist, who was a member of the Health DepartmentBoard of Directors, conducted a health survey and provided the results to the ATSDR Division ofHealth Studies for evaluation. (11)

Several former employees were contacted by ATSDR by phone to obtain information about healthconcerns. (1)

At a meeting on May 29, 1990, CCSFC informed ATSDR that blood samples from severalvolunteers had been analyzed for PCBs. Concurrently, ATSDR was also informed that adipose (fat)tissue was in the process of being sampled and analyzed for PCBs under the direction of the CCSFC. ATSDR requested the results of the blood and fat sampling (results were expected to be availablenear the end of June 1990). A spokesperson for the CCSFC indicated that the members of CCSFCwould have to vote to release the sampling results; however, no problems were foreseen in givingATSDR the information. ATSDR received the results in February 1991. They have been includedin this public health assessment.

In December, 1990, CCSFC took another set of blood samples (approximately 43) and analyzedthem for PCBs. ATSDR received the results of those analyses in July 1991. They have beenincluded in this public health assessment.

A review and discussion of the above studies, surveys, and reports are contained in the Public HealthImplications Section of this public health assessment.


ATSDR met with citizens of Minden, West Virginia on May 29, 1990 to discuss community healthconcerns about the SEC site. Additional community concerns were gathered through interviewswith former Shaffer employees (8). The following is a summary of those community concerns:

Residents have repeatedly expressed concerns about the health effects of exposures to PCBs. They are especially concerned about the effects on the liver, kidney, skin and respiratory tract.

Currently and throughout the EPA removal process, CCSFC, the local citizens' group has beenactively seeking performance of a health study on previous workers and downstream residents.

The citizens have also expressed a great deal of frustration at the duration and extent of theclean-up process. They feel that not all sources of PCB contamination have been investigated andthat local water supplies have been contaminated by PCBs.

These community health concerns will be evaluated and discussed further in the Public Health Implications Section of this public health assessment.


The tables in Appendix 2 list contaminants in each medium. Those contaminants are evaluated insubsequent sections of the public health assessment to determine whether exposure to them haspublic health significance. ATSDR selects and discusses contaminants based upon several factors. They include concentrations on and off site, the quality of the field and laboratory data, sampledesign, comparison of on- and off-site concentrations to background concentrations (if available),comparison of on- and off-site concentrations to public health assessment comparison values fornoncarcinogenic and carcinogenic endpoints, and community health concerns.

The listing of a contaminant in the tables does not mean that it will cause adverse health effects if exposure occurs at the specified concentrations. The listing of contaminants in the tables means that those contaminants are further evaluated in this public health assessment. The potential for adverse health effects resulting from exposure to contaminants without a comparison value, those that exceed the comparison value, and those of health concern is discussed in the Public Health Implications Section.

Comparison values for ATSDR public health assessments are contaminant concentrations in specificmedia that are used to select contaminants for further evaluation. ATSDR and other agenciesdeveloped those values to provide guidelines for estimating the media concentrations of acontaminant that are unlikely to cause adverse health effects, given a standard daily ingestion rateand standard body weight. See Appendix 3 for a description of the comparison values used in thispublic health assessment.

A. On-Site Contamination


Sixteen on-site samples and one duplicate on-site sample taken in March 1990 indicated that somePCB contamination remains on site (Table 1). PCB concentrations ranged from not detected to 240mg/kg in soil. The on-site soils are contaminated, to a minor extent, with other volatile organiccompounds (VOCs) and semivolatile organic compounds (SVOCs) (Table 1).

Of the VOCs, no ATSDR soil comparison values were exceeded and therefore VOCs in surface andsubsurface soils will not be considered further.

Of the SVOCs, no soil comparison values were found for, 2-methylnaphthalene, acenaphthylene,dibenzofuran, and di-n-octyl phthalate. In addition, comparison values were exceeded forbenzo(a)anthracene, benzo(b)fluoranthene, benzo(a)pyrene, and dibenz(a,h)anthracene. Thesecompounds will be retained for further evaluation in the Public Health Implications Section.

In March 1990, thirteen on-site samples and one duplicate on-site sample were analyzed for PCDDs and PCDFs. The analyses included PCDD and PCDF congeners that were assigned a toxicity equivalency factor (TEF) associated with the most toxic dioxin--2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The TEF was used to calculate the total toxicity associated with all PCDD and PCDF congeners found. ATSDR did not find any PCDD or PCDF levels in the soil samples, except for one sample and its' duplicate sample, that exceeded the ATSDR comparison value EMEG (non-pica child). The sample and its' duplicate were taken from a depression east of the Shaffer Equipment Building. A TEF of 0.128 ug/kg was calculated for the sample and a TEF of 0.801 was calculated for the duplicate sample.

In May 1990, EPA took 14 soil samples (Table 8), four of which were random samples at locationsdifferent than those in March 1990. Those random samples showed PCB levels ranging from 0.148mg/kg to 374 mg/kg. (12)

During a meeting held on May 29, 1990, the CCSFC reported PCB contamination in additionalareas of the site. Thirty additional samples were taken near the Shaffer Equipment building andnear a shale pile in June 1990. PCB concentrations ranged from "not detected" to 40,300 mg/kg(Table 1A). Four of the 30 soil samples were also analyzed for SVOCs (Table 1B). All of thecompounds for which a comparison value was available did not exceed the comparison values andwill not be considered further. A comparison value was not available for 2-methylnapthalene;therefore, this compound will be retained for further evaluation in the Public Health Implications Section.

In November 1990, EPA conducted 58 soil samples which were analyzed for PCBs along with anadditional three confirmation and four duplicate samples. These samples were taken prior to andduring EPA removal actions. The PCB concentrations ranged from 0.3 to 10500 mg/kg (Table 10),with no PCBs being detected in four of the samples using a detection limit of 0.1 mg/kg.

Post sampling was conducted on January 24, 1991, at the excavated areas. The PCB concentrationsranged from 0.44 to 1020 mg/kg (Table 11). The sampling consisted of 53 samples that wereseparated into 9 related areas and analyzed as composites from the original samples. Fouradditional samples were taken from areas on site that were believed to be unimpacted by PCBcontamination. One of those four samples revealed a detectable PCB concentration at 0.33 mg/kg. A 1-3 foot cover of soil, taken from the unimpacted areas, was used to fill the excavated areas. Themajor areas of PCB contaminated soil is believed to have been removed and backfilled; however, ithas not been determined that all areas of the site are free from surface soil contamination.


Two on-site sediment samples collected in March 1990 indicated some PCB, VOC, and SVOCcontamination at the site (Table 2). The highest PCB level of 660 mg/kg was found in thesediments of the excavated area drainage ditch, located in the area where EPA had previouslyinitiated and completed the 1984-1985 emergency removal of PCBs. Sediment comparison valueswere not available for the compounds 2-methylnapthalene and dibenzofuran. Sediment comparisonvalues were exceeded for the compounds benzo(a)anthracene, benzo(a)pyrene, and PCBs. Thosecompounds with no available comparison values and those that exceed the comparison values willbe analyzed further in the Public Health Implications Section.

Two on-site sediment samples (March 1990) were also analyzed for PCDDs and PCDFs, includingcongeners. ATSDR determined that PCDD or PCDF levels associated with the sediment samplesdid not exceed the ATSDR EMEG comparison value.

In May 1990, two sediment samples were taken from the same locations as those collected in Marchof 1990 and analyzed for PCBs only. The May 1990 sampling revealed PCBs at a concentration of112.1 mg/kg at the building area drainage ditch and 39.2 mg/kg at the excavated area drainageditch.


No air sampling data were available for ATSDR's evaluation.

PCB contamination at the Shaffer site are found primarily in soil. The rate of PCB volatilizationfrom soils (i.e., the migration of PCBs by evaporation from the soil into the air) is very smallbecause of soil binding of PCBs (2). The rate of PCB volatilization from surface water into the airis much higher (2); however, PCBs were not detected in surface water samples at the Shaffer site. Because of these factors, PCBs are not expected to be present in ambient air at levels of healthconcern. When additional site monitoring was conducted in March 1990, EPA and ATSDR decidednot to perform air monitoring for PCBs at the Shaffer site (13). However, EPA scanned the site witha field instrument (HNu) in March 1990 to investigate VOCs and SVOCs in the air at the site. Thatinstrument did not detect any organic vapors (VOCs, SVOCs).


No groundwater samples (other than from an on-site spring discussed below) were taken by EPA orwere available for ATSDR's evaluation.

An on-site spring (no known uses) that flows into Arbuckle Creek was sampled for VOCs, SVOCs,and PCBs. Nothing was found above the laboratory measuring instrument's limits of detection. However, the PCB detection limits (0.5 g/l and 1.0 ug/l) of the measuring instrument was higherthan the ATSDR CREG comparison value of 0.0045 ug/l, therefore it cannot be stated withcertainty that PCBs are or are not a problem. The spring was not analyzed for PCDDs or PCDFs;however, because PCDDs and PCDFs have a high affinity for soils, sediments and have lowsolubility in water, they are not expected to be present in groundwater or springwater.


No on-site surface water (except for a spring discussed under on-site groundwater) exists.

B. Off-Site Contamination


Off-site (residential) surface soils were found to be PCB-contaminated at a maximum concentrationof 15 mg/kg in December 1984. In March 1990, five off-site surface soil (but no subsurface soil)samples (Table 3) were taken. Those samples consisted of one background soil sample, three soilsamples from residential yards in the area (from two homes that yielded levels of PCBs during anearlier sampling event and one home that had a potential for PCB contamination), and one samplefrom the mouth of the old Minden Mine. Again, there was some contamination from PCBs, VOCsand SVOCs. The PCB contamination in the samples taken from the yards of three homes and theMinden Mine area ranged from 0 to 2.1 mg/kg.

Comparison values were not available for the compounds 2-methylnapthalene, acenaphthylene, anddibenzofuran. Comparison values were exceeded for benzo(a)pyrene and PCBs. Those compoundsthat did not exceed the comparison values will not be considered further. The others with nocomparison values available or that exceed comparison values will be further evaluated in the Public Health Implications Section.

Four off-site soil samples taken in March were analyzed for PCDDs and PCDFs and their congeners. No sample exceeded ATSDR's PCDD or PCDF EMEG comparison value.


Five surface water samples (two upstream and three downstream from the site) were collected andanalyzed for PCBs, VOCs and SVOCs (Table 4). PCBs were not found in the surface watersamples above the level of the detection limit of the laboratory's measuring instruments. However,as under on-site groundwater, the PCB detection limits of the measuring instrument was higher thanATSDR's comparison value. Some VOC and SVOC surface water contamination was found bothupstream and downstream of the site. Some of these chemicals (eg., methylene chloride, acetone)are common laboratory artifacts or contaminants introduced by laboratory handling, as evidenced bytheir presence in laboratory blank samples. The reported levels of all contaminants exceptmethylene chloride were not above comparison values; therefore these surface water contaminantswill not be considered further in this health assessment. Methylene chloride is will also not beconsidered further. ATSDR believes the elevated concentrations were the result of laboratorycontamination.

No surface water sampling for PCDD or PCDF was performed during March sampling. Althoughno PCDD and PCDF analyses were conducted for surface water, those compounds are not expectedto be found in surface water in any appreciable amounts due to their physical and chemicalcharacteristics. PCDDs and PCDFs will normally leave surface water and adsorb onto sediment thattightly bind the compounds (14). Because PCDD and PCDF were not found above ATSDR'scomparison values in on-site soil or sediment, levels are not expected to be found in surface water.


Five off-site sediment samples were collected in March 1990 and analyzed for PCBs, VOCs, andSVOCs (Table 5). PCBs were found above the comparison value in the off-site sediments at amaximum of 5.2 mg/kg. Benzo(a)anthracene, benzo(b)fluoranthene, and benzo(a)pyrene were alsofound above the sediment comparison value. Comparison values were not available for 2-methylnaphthalene, dibenzofuran, and di-n-octyl phthalate. Those compounds that did not exceedthe comparison values will not be considered further.

No off-site sediment sampling for PCDD or PCDF was performed during any sampling. However, because on-site soils and sediments did not contain PCDDs and PCDFs at levels above comparison values, off-site sediments are not expected to contain levels of PCDDs/PCDFs that came from the site above comparison values.


No off-site air samples were taken during any sampling period.


Two raw water supplies at the Minden Mine intake (West Virginia American Water Company) andthe Rocklick Mine intake (Arbuckle Public Service District) were sampled for PCBs, VOCs, andSVOCs (Table 6). This action was taken by EPA in response to concerns by the CCSFC and thecommunity that PCBs were disposed in the mine shafts. All results were below the detection limitsexcept for levels of methylene chloride, which ATSDR considered a result of laboratorycontamination ("B" identifier in the Table indicates that methylene chloride was found in thelaboratory sample blank). However, the PCB detection limits (0.5 ug/l and 1.0 ug/l) of themeasuring instrument was higher than the ATSDR CREG comparison value of 0.0045 ug/l,therefore it cannot be stated with certainty that PCBs are or are not a health concern.

PCBs are rarely found in groundwater because PCBs have an affinity for soils, sediments, andorganic matter. PCBs also have low leaching potential under normal soil conditions (2).

No PCDD or PCDF analyses were performed on the raw water samples above. No othergroundwater samples were taken during the March sampling.

As discussed in surface water, PCDDs and PCDFs are not expected in any appreciable amounts ingroundwater due to their physical and chemical characteristics. These compounds will separate outof groundwater and adsorb onto soil particles that tightly bind the compounds. Because PCDD andPCDF were not found above comparison values in the soil, high levels are not expected to be foundin the groundwater.


No data are available on PCB levels in aquatic biota (snapping turtles). It is not certain whetherthere are snapping turtles in the site vicinity. No PCB sampling of fish has been performed. EPAhas stated that there are no edible species of fish in Arbuckle Creek.

Some PCBs are highly resistant to degradation and may bioaccumulate in aquatic species. Theconcentration of PCBs in fish, shrimp, and oysters can reach 26,000 to 660,000 times theirconcentrations in water. Area residents have reportedly eaten snapping turtles taken from the area(15,16). There is evidence in the literature that PCBs bioconcentrate in snapping turtles(reptiles)(14).


ATSDR received a copy of a report entitled Community Health and PCB Exposure in Minden, West Virginia dated September 1989 (10) that contained additional PCB sampling data. The data are in Table 7 and the sampling locations are stated in Figure 3. Those results agreed closely with the EPA sampling results from March 1990. Reported on-site PCB levels were lower than discovered by EPA (64.4 mg/kg [soil] versus 240 [soil] and 660 mg/kg [sediment]). Off-site PCB concentrations (soil and sediment) in the report were similar to EPA values.

C. Quality Assurance and Quality Control (QA/QC)

The results of this Petitioned Health Assessment are based on an evaluation of the sampling dataobtained from the March 1990, May 1990, and June 1990 sampling conducted by EPA and the dataobtained from the Community Health and PCB Exposure in Minden, West Virginia report datedSeptember 1989 (10). Some of the data that ATSDR reviewed, showed methylene chloride (Tables4, 6) and acetone (Table 4) contamination. ATSDR believes detection of methylene chloride andacetone was the result of contamination by the laboratory of samples with common solvents used inpreparation of the samples for analysis. The validity of the analysis and conclusions drawn for thisPetitioned Health Assessment are determined by the availability and reliability of the referencedinformation.

In May 1990, EPA resampled several on-site locations previously sampled in March 1990 to verifyPCB concentrations in the soil/sediment (Tables 8, 9). Data from each sampling locations are notlisted in Tables 8 and 9. Only the minimum, lowest positive, and maximum PCB concentration ofall the samples are given.

In comparing the raw PCB data taken in May and March 1990, several discrepancies in PCBconcentrations were noted for samples that were taken from the same location. Ideally, samplesfrom the same location should indicate the same concentration of contaminant. However, when thesame locations were resampled, anomalies in the sampling results occurred between the two dates. For comparison, on-site soil PCB concentrations (110 mg/kg) identified in the March 1990sampling was 9.6 mg/kg in the May 1990 sampling. In another comparison, an on-site soil PCBconcentration of 17 mg/kg identified in March 1990 was 538.9 mg/kg in May 1990 for the samesampling location. The reasons for the discrepancies are unknown. It is very probable thatdifferences in sampling technique, sampling depth, physical changes because of weather and findingthe exact same sampling spot would have had a profound effect on sampling outcome.

EPA, in May 1990, also resampled several on-site locations sampled in March 1990 to verify PCBconcentrations in the sediments (Table 9--note that only the minimum, lowest positive andmaximum concentrations are given). Again, several discrepancies were noted. A PCBconcentration (660 mg/kg) identified in the March 1990 sampling was 39.2 mg/kg in May 1990. Again, the reasons for the discrepancies are unknown.

D. Physical and Other Hazards

ATSDR health assessors saw no apparent physical hazards at the site.


In order to identify other possible facilities that might contribute to the release of contaminants intothe environment near Minden, ATSDR searched the most recent data in the Toxic Chemical ReleaseInventory (TRI). EPA developed TRI from the chemical release (air, water, and soil) informationprovided by specified industries. For any of the 300 - plus toxic chemicals in TRI, EPA requiresthat the manufacturing industry report annual estimated releases to the environment. TRI data arethose releases from manufacturing, processing, and otherwise use of toxic chemicals. The databasedoes not distinguish spills or unintentional releases from routine releases. TRI showed no facilities in the Minden area that released PCBs to the environment.


To determine whether people are exposed to contaminants released from the SEC site, ATSDRevaluated the environmental and human components that lead to human exposure. This pathwaysanalysis consists of five elements: 1) source of contamination, 2) environmental medium in whichthe contaminants may be present or may migrate, 3) points of human exposure, 4) routes of humanexposure such as ingestion, inhalation, or dermal absorption, and 5) receptor population. ATSDRidentifies exposure pathways as completed, potential, or eliminated. A completed exposure pathwayexists in the past, present, or future if all five elements of an exposure pathway link the contaminantsource to a receptor population. Potential pathways, however, are defined as situations in which atleast one of the five elements is missing, but could exist. Potential pathways indicate that exposureto a contaminant could have occurred in the past, could be occurring now, or could occur in thefuture. Pathways are eliminated when at least one of the five elements is missing and will never bepresent. Completed and potential pathways may also be eliminated when they are unlikely to exist,or to be significant. All completed, potential, and eliminated exposure pathways at SEC arepresented in Table 14 of Appendix 2.

A. Completed Exposure Pathways

A past, completed exposure pathway for the on-site workers was identified at the SEC site. WhileSEC was operating, plant workers could have been exposed to PCBs during the handling, recycling,refurbishing, and removal of liquid and filling of PCB-type transformers, capacitors, and switchgear. However, the concentration of PCBs reaching the workers is unknown because no environmentalsampling was done during active facility operations. Indirectly, the workers may have exposed theirfamilies to PCBs by returning home with PCB-soiled clothing.


SEC employees were probably exposed to PCBs during plant operation via inhalation, dermalcontact, and inadvertent ingestion. Interviews by ATSDR with several former SEC employeesindicated that safe handling practices of PCB oils were lax (1). Some employees' hands were soakedin PCB oils while they changed transformer taps. Other employees stated that by the end of the daytheir clothes were soaked with PCB oils and that washing did not always remove the oils. WashingPCB-soaked clothes with other family members' clothes would contaminate those clothes, exposingother family members to PCB oils primarily through dermal contact and inadvertent ingestion. Those employees worked at the SEC site for an estimated average of 5-7 years. ATSDR believesemployees and their families are the people most likely to have been exposed to significant levels ofPCBs, although the actual duration and magnitude of exposure are not known.


Off-site (residential) surface soils were PCB-contaminated at a maximum concentration of 15 mg/kgin December 1984. Later sampling results in March 1990 indicate a maximum of 2.1 mg/kg ofPCBs in residential soils. PCBs adhering to surface soils can migrate or be transported by surface-water runoff and erosion from on-site sources. PCBs found in the sediments of Arbuckle Creek (190ppm, November 1984; 5.2 ppm, March 1990) and in the off-site soil of flood-prone areasdownstream (2.1 ppm, March 1990) of the Shaffer site indicate that migration of PCBs hasoccurred. The concentrations of PCBs in sediments appeared to have decreased over time astransport mechanisms took effect. Arbuckle Creek flows at an average rate of 3000 gallons perminute. Heavily contaminated sediments were probably washed downstream to the New River,reducing PCB concentrations in the remaining sediments. The decrease in PCB concentrations fromthe November 1984 samples to the March 1990 samples indicate this possibility.

A potential exists for PCB exposure to children playing in contaminated residential soils and thecontaminated sediments of Arbuckle Creek. PCB concentrations in both the residential soil andsediments appear to have decreased with time. Children playing in PCB-contaminated soils orsediments may be exposed through inhalation of airborne dusts (soil), dermal contact (soil andsediment), or inadvertent ingestion from soiled hands (soil and sediment).

B. Potential Exposure Pathways


PCB-contaminated oils were also reported to be given away or sold as starter fuel to the generalpopulation (1). If so, exposure could occur from handling PCB-contaminated oils and possiblyinhaling the PCB, PCDF, and PCDDs generated from incomplete combustion of the PCBs. Inadvertent ingestion might also occur if PCB-soiled hands were used in eating, smoking, ordrinking. No data exist on the amounts of PCB-contaminated oils that were sold or given away,who burned the PCBs, who were possibly exposed to PCBs, PCDFs and PCDDs, or the severity ofthe exposure. Because these events took place in the past, it would be difficult, if not impossible, toquantify PCB, PCDF and PCDD exposure.


PCB-contaminated oils were also reported by former employees to be used in the past as a fuel tostart coal fires in the Shaffer Equipment Building (1). As above, exposure to on-site workers in theShaffer Equipment Building could have occurred from handling PCB-contaminated oils andpossibly inhalation of the PCB, PCDD, and PCDDs from incomplete combustion of the oils. Inadvertent ingestion could occur if PCB-soiled hands were used in eating, smoking, or drinking. No documentation exists on the quantity of PCB-contaminated oils used as starter fuel, the durationof use, or the amount of worker exposure that may have occurred.


Early reports from the community stated that residents might have consumed snapping turtles in thearea (15,16). Snapping turtles have not been analyzed for PCBs, PCDDs, and PCDFs.

The EPA does not believe that snapping turtles are consumed from this area. At a public meeting onMay 29, 1990, EPA asked the audience if snapping turtles were being eaten; there was no response. However, because of limited attendance, a lack of response did not necessarily indicate that turtleswere not eaten. Because of the uncertainty surrounding the use of turtles for food and the lack ofdata regarding turtle contamination, ATSDR could not evaluate if exposure to these contaminantsare actually occurring by ingestion. If turtles are being eaten and without data to negatecontamination, their ingestion would be considered a potential human exposure pathway.


Although a fence (part chain-link, part cattle) surrounds part of the site, children and adults mayeasily gain access to the site. Some parts of the fence prevent only large animals from entry. On-siteworkers and trespassers may be exposed to PCBs from dermal contact with on-site soil andsediment, inhalation of contaminated airborne particulates from soil, and from inadvertent ingestionof on-site soil and sediment from soiled hands by playing, eating, smoking, and other activities.


It is not known if the SEC Equipment Building is currently used for a business or commercialactivity, although some activity has been noted at the site. Both PCDD/PCDF and PCB samplinghave not been conducted inside the building. On-site workers, if any, may be exposed to PCBs viainhalation, dermal contact, or inadvertent ingestion from soiled hands by working in the SECEquipment Building, if the building is contaminated. Because PCB-contaminated oils werereportedly burned as a starter fuel in the building (1), incomplete combustion may have generatedPCDD and PCDF compounds.

C. Eliminated Exposure Pathways


Remedial workers could have been, are, or will be exposed to contaminants in a variety ofenvironmental media while conducting on-site activities in the past, present, and future, respectively. It is unlikely that such exposures would be at levels of concern however, provided appropriate workpractices, as defined by the state or federal regulatory or permitting authorities, such as theOccupational Safety and Health Administration (OSHA), are followed. Those include workereducation, certification, supervision and training, and use of personal protective equipment.


Because no farm animals or vegetable gardens were seen in the site vicinity, human exposure shouldnot be occurring through consumption of contaminated plants or farm animals. EPA hasdocumentation stating there are no game and edible species of fish in Arbuckle Creek. Except forsnapping turtles, the food chain (fish, vegetable gardens, and farm animals) was eliminated as an exposure pathway.


A. Toxicological Evaluations

Adults and children who trespass can gain access to the Shaffer site, and may be exposed topolychlorinated biphenyl compounds (PCB) through dermal contact, dust inhalation, andinadvertent ingestion of contaminated soils. Exposure to contaminated on-site soils would beintermittent, and therefore, represents a minimal potential exposure. Children who played in off-siteresidential yards and those who played in Arbuckle Creek may have been exposed to contaminatedsoils and sediments. Past PCB sediment concentrations were high. Nothing is known about thefrequency and duration of exposure. Therefore, ATSDR could not evaluate the public healthimplications of this exposure. Current concentrations of PCBs in sediment and residential soilappear to be low and decreasing with time; therefore exposures are thought to have minimal publichealth consequences. Remedial and other workers on the site have the same potential routes forexposure; however, remedial workers normally follow appropriate work practices and use personalprotective equipment. There are some concerns regarding possible exposure through the ingestion ofsnapping turtles from Arbuckle Creek. Past on-site workers and their families were probablyexposed to levels of PCBs; however, no data exist to quantify exposure. The public healthimplications of this exposure cannot be assessed.

There was vandalism and recreational use of the site prior to EPA emergency removal operations. Activities may have resulted in children and adults being exposed to high levels of PCBs; however,the extent, duration, and magnitude of exposure cannot be quantified. It would be difficult forATSDR to assess public health implications of this past exposure without quantitative data.

ATSDR has received reports that PCB-contaminated oils were used in the past as a starter fuel toignite coal for heat at the SEC Equipment Building and for off-site residential use. Therefore, theequipment building and residences that used the PCB-contaminated oil may be contaminated withPCBs, PCDDs, and PCDFs. Sampling inside the building was not performed. Contamination insidethe building would most likely affect employees and would best be assessed by the OccupationalSafety and Health Administration (OSHA) and the National Institute of Occupational Safety andHealth (NIOSH). For residences that used the PCB-contaminated oils, no environmental data wereavailable and the extent, duration, and magnitude of exposure could not be quantified. Without thisinformation, these past exposures cannot be assessed by ATSDR.

At this time, there is not enough evidence to state that PCBs are carcinogenic in humans. EPA hasclassified PCBs as probable human carcinogens and recommends that all commercial PCB mixturesbe considered to have similar carcinogenic potential. The International Agency for Research onCancer (IARC) has classified PCBs in Group 2B based on sufficient evidence in animals,inadequate evidence in humans, and inadequate evidence for mutagenicity. IARC places chemicalsinto groups based on their evaluation of the cancer risk to humans from exposure to the chemical. Group 2 are those chemicals felt to be probably carcinogenic for humans. Group 2 is divided intotwo subgroups A and B with A having stronger evidence to indicate carcinogenicity than those ingroup B. In addition, NIOSH has recommended that PCBs be regarded as potential humancarcinogens in the workplace (2,17).

Studies in animals and occupationally exposed groups have indicated that the liver and skin are themajor target organs. Increased serum levels of liver enzymes have been seen in some studies ofworkers (2,18,19). Other studies have not shown a large number of workers with high enzymes(20,21,22). Increased levels of liver enzymes detectable in the serum are an easily measured markerof liver damage. Dermatological effects such as chloracne, pigmentary changes, rashes, andswelling or thickening have been described (2,20,21,23,24,25). Chloracne is a chronic skincondition produced by environmental exposures to certain compounds containing chemicals such aschlorine or bromine. There are two predominant skin lesions seen in chloracne cases: the chloracnecyst and the comedo. The chloracne cyst is a skin colored sac with a central opening. The comedois a blackhead. Most studies in workers indicated that those with chloracne also have some evidenceof liver injury (22,26).

PCBs have been shown to cross the placenta, thus there may be PCB exposure to fetuses of womenwho have increased PCB exposure. There have been several reports of effects on development inchildren of mothers who were occupationally exposed to PCBs during pregnancy or were consumersof PCB-contaminated fish during pregnancy. Those effects have included decreased birth weight,gestational age, and cognitive functioning (2,27,28). One subpopulation sensitive to PCBs arenursing infants. PCBs are fat soluble so they tend to concentrate in the areas of the body with highfat content. Breast milk has a high fat content so PCBs are excreted through the milk. Infants maymaintain higher levels of PCBs in their bodies. There have not been many studies to determine theeffects of exposure from PCB-contaminated breast milk. One study that was conducted by Gladenet. al. did not demonstrate any effect on infant psychomotor responses associated with exposurethrough breast feeding (29). During ATSDR's site visit, it was noted that there were many childrenin the vicinity of the site. The number of women in the area who breast-feed their children is notknown. It must be emphasized that ATSDR does not know if fetal exposure to PCBs through breastmilk or during pregnancy is occurring and that such exposures are speculative. For those exposuresto occur, mothers must eat contaminated turtles or breast-feed after eating contaminated turtles. ATSDR has not been able to confirm or deny the hypothesis that contaminated turtles have beeneaten from the site vicinity.

There have not been any studies of PCB levels in the general population, but there have been severalsmall studies of specific groups (see references 2 and 22 for reviews). While differences inmethodology make it difficult to directly compare these studies, they do give an indication of theranges of PCB levels that have been seen. Tables 10 and 11 give a summary of some of these datawhich will be used in the discussion of health outcome data evaluation in this public healthassessment.

The polynuclear aromatic hydrocarbons (PAHs) benzo(a)anthracene, benzo(b)fluorene,benzo(a)pyrene and dibenzo(a,h)anthracene were found at levels exceeding comparison values. PAHs are common contaminants at many sites, often associated with burning of combustiblematerials. These compounds have little acute effect but have been shown to cause cancer. Long-term exposure to levels exceeding comparison values may result in a slight increased risk of cancer. However, exposure to these contaminants at SEC is mostly infrequent and sporadic, and is probablyinsufficient to have any effect.

Appropriate comparison values could not be determined for a 2-methylnapthalene andacenaphthylene, dibenzofuran, and di-n-octyl phthalate. Therefore, we are unable to determine if theconcentrations present are of public health concern. Comparison values may be unavailable for anumber of reasons, including uncertainties and conflicting results in animal studies, lack of humanstudies, and scientific controversies. ATSDR is working to establish comparison values forcompounds commonly found at contaminated sites.

Acenaphthylene and 2-methylnaphthalene are PAHs. No comparison values are available for thesetwo PAHs. However, PAHs, in general, have little acute toxicity especially at the relatively lowconcentrations found at SEC. These two PAHs are not thought to cause cancer.

Dibenzofuran and di-n-octyl phthalate also have no comparison values, largely because the toxicityof these compounds has been little studied. Phthalates as a group have low acute toxicity, but somehave been shown to be carcinogenic by non-genetic mechanisms. Di-n-octyl phthalate itself has notbeen adequately evaluated for carcinogenicity. Dibenzofuran is of low acute toxicity, while its longterm effects are uncertain.

B. Health Outcome Data Evaluation

ATSDR has reviewed the results of PCB analysis of adipose (fat) tissue in 20 residents and formeremployees. In addition, the results from the 1986-1987 serum (blood) sampling was also evaluated(30,31). Those results are provided in the following table:

PCBs in Minden Residents



0.564 ppm

0.100 - 0.970 ppm


7.7 ppb

5 - 14 ppb

ppm = parts per million
ppb = parts per billion

The range and mean values seen in the Minden group are consistent with the levels seen in groupswithout occupational exposure (Table 12). A comparison with the above table and Table 12 does not indicate increased exposure in Minden residents compared to other populations.

Regarding Table 12, the median PCB levels in blood serum of the general population of the UnitedStates without occupational exposure has been found in most studies to be less than 10 ng/mL (ppb),but sex and age of the population must be taken into account (32). Mean serum levels were usuallybetween 4 and 8 ng/mL (ppb), and 95 percent of individuals had levels less than 20 ng/mL (ppb)(2). There have been fewer studies utilizing adipose samples in the population. While there isvariation in groups studied, PCB levels in adipose tissue are between 100 and 200 times the levels inblood serum (22). Studies attempting to correlate serum and adipose levels with health effects havehad inconsistent results (2).

Follow-up serum samples were drawn on 43 individuals in Minden in 1990. The results from thosesamples are consistent with other results. There were 30 persons with no detectable levels of PCBsin their sera (Detection limit was 3 ppb). The range of levels seen in persons with detectable PCBswas 3.2 - 11.0 ppb (Mean = 5.0 ppb) (33). The levels are consistent with those seen in the generalpopulation.

One infant/child (age unknown, but listed as "0" years when the blood sample was taken) in the1990 sampling indicate a sera PCB level of 11.0 ppb. This PCB level is unusual for an infant andthe accuracy of the test is suspect and should be verified. If the results are verified, further tests andinvestigation may be necessary to determine and eliminate the source of the PCB exposure.

There have been a few studies of populations surrounding hazardous waste sites in which exposureassessment has been done (Table 13). PCBs in Minden sera are not significantly higher than theranges of PCBs seen in comparison populations.

A review of the census data for the area demonstrate the factors that may influence exposure and therisk from exposure. The census data indicate that the population is stable and the majority lives inrural areas. A large portion of the population are employed in the mining industry. Industryexposures must be considered when conducting health studies in the area. The per capita income forthe Minden area is less than $4,000. Income influences access to health care and the types ofillnesses that the population experiences.

Several health studies were conducted in the Minden area by individuals or organizations. The WestVirginia Department of Health (WVDOH) conducted a cancer mortality study of Oak Hill (Mindenis on the outskirts of Oak Hill) for the period 1979-1981 (9). The study concluded that there was agreater rate of respiratory cancers in males in the Oak Hill population than in the United States. Anevaluation of the study by ATSDR indicated several weaknesses that may have affected theconclusions of the study. One of the concerns is that the study chose to use the U.S. standardizedmortality rates rather than West Virginia mortality statistics. Because many males in the area havea history of working in the coal mines, they have a higher risk for developing respiratory cancers. That additional exposure would not be reflected in the rates for the United States. The results ofATSDR's evaluation of the study are found in Appendix 3. Another aspect of the cancer mortalityinformation comes from data on the Riggans mortality tapes produced by EPA and National CancerInstitute (NCI). Review of the data indicates that Fayette county has had an increased rate ofrespiratory cancer deaths when compared to the U.S. The increase has been present since the decadeof the sixties so factors such as age and occupation that may be associated with the increase wouldhave existed at that time. This increase existed before SEC was operating; therefore, other factors(e.g., age, smoking status, socioeconomics) that could lead to this increase must be examined (8).

In the summer of 1986, the Concerned Citizens to Save Fayette County (CCSFC) sponsored a health registry of random sampling of Minden and Rock Lick residents. The study was conducted by Vanderbilt University's Appalachian Student Health Coalition and is commonly known as the Vanderbilt Study. In 1989, there was a survey of residents in Page, WV that was intended to provide a control group for the 1986 Vanderbilt Study. The survey in Page was part of a Virginia Student Environmental Health Project Intern Report (10). The reports indicated that certain health conditions such as shortness of breath, unexplained weight loss, persistent cough, and urinary tract infections occur at a higher rate in Minden than in Page, WV. The increased incidence of health problems described in the report for the Minden community does not correlate well with current PCB literature regarding symptoms and health effects associated with PCB exposure. ATSDR's Division of Health Studies reviewed the Vanderbilt study and noted the strengths and weaknesses of the surveys conducted in Minden (1986) and Page (1989). The full results of that review are in Appendix 3. ATSDR concluded that a clear relationship could not be established between PCB contamination and the significantly different observed rates of the four symptoms (described above). In addition there were variables such as age, smoking history, and history of working in the mines that may affect symptom prevalence. Those variables were not addressed in the two surveys. Socioeconomic and nutritional status should also be considered. When such variables are not controlled for, one cannot determine what accounted for the findings (Appendix 3). Although the surveys could not establish a definitive link between PCB contamination and specific health conditions, the surveys did serve to identify diseases of concern for the community such as those described earlier (shortness of breath, unexplained weight loss, persistent cough and urinary tract infections).

Another study was conducted by a Beckley, West Virginia gynecologist (11) who submitted a healthsurvey to ATSDR in 1989 for an evaluation. Although the survey did indicate areas of healthconcern similar to those mentioned previously, the survey did not encompass the Minden area. Thedesign and methodology of the study were obscure. Therefore, ATSDR was not able to draw anyconclusions for this public health assessment.

Past occupational exposures to high concentrations of PCBs have been reported at the Shaffer site. A number of former Shaffer employees were interviewed by ATSDR by telephone to evaluate pastwork exposure to PCBs, and past and current health (1). One of six employees interviewed reportedgetting a skin rash while working with PCB oils. Most employees rated their current health asgenerally good and could not relate any adverse health effects to working with PCBs. One of theemployees participated in the CCSFC serum and adipose (fat) sampling in May 1990.

The area around Minden, West Virginia is economically depressed. Large portions of thepopulation are employed in mining. Such factors must be taken into account when attempting todetermine health consequences from possible exposure to PCBs. Previous health studies conductedin the area have identified health outcomes of concern (shortness of breath, unexplained weight loss,persistent cough, and urinary tract infections), but have not found outcomes that have beenconsistently linked to PCB exposure. The results of the serum and adipose tissue did not indicatebody burdens of PCBs that are significantly different than those seen in other non-occupationallyexposed populations. The review of the toxicological information and the health information do notindicate high levels of exposure to PCBs at the present time or increases in health outcomes that canbe linked to PCB exposure.

C. Community Health Concerns Evaluation

  1. What are the effects of PCB exposure on the liver, kidney, skin, and respiratorysystem?
  2. A discussion of PCB effects on the liver and skin are contained in the ToxicologicalEvaluations portion of the Public Health Implication Section of this document. The effects ofPCBs on the kidney and respiratory system occur mostly at exposures considerably higher thanfound at this site. PCBs may cause kidney damage (hydronephrosis) including swelling andstructural damage. Little effect is seen in the lung, although the lung may be a major route ofabsorption. PCBs can damage lung membranes, and may lower tolerance to infection, butthese effects are unlikely at the exposures seen at this site.

  3. The Concerned Citizens to Save Fayette County, a local citizen's group representing anumber of residents, has actively sought a health study to be conducted on previousworkers and residents downstream (referring to Arbuckle Creek) from the SEC site.
  4. As part of ATSDR's Planned Public Health Actions, ATSDR will contact the NationalInstitute for Occupational Safety and Health (NIOSH) regarding concerns for worker healthand provide NIOSH with a copy of the Shaffer Equipment Company Public HealthAssessment in order that they may be advised of the potential for past PCB occupationalexposures at Shaffer. Follow-up worker health studies, if deemed necessary, are under thepurview of NIOSH. Because the latest sampling results indicate low levels of PCBs inresidential soils (2.1 mg/kg, March 1990) that are not considered to be a public health hazard,and the results of the two PCB serum tests indicate levels of PCBs in Minden residents notsignificantly higher than those seen in comparison populations, a health study of downstreamresidents does not appear to be beneficial or justified at the present time. Contaminants otherthan PCBs are not likely to present a health concern at the exposures likely to occur in the future or to have occurred in the past.

  5. Residents believe that all sources of PCB contamination have not been investigated and that local water supplies have been contaminated with PCBs.
  6. Additional on- and off-site sampling was conducted by EPA in March 1990, May 1990, andagain in June 1990 in response to reports made by the CCSFC. Further on-site sampling andremoval actions were conducted between November 1990 and January 1991. EPA is alsoinvestigating other sites that were allegedly used to dispose of PCBs from the SEC. Theseother sites are distinct and separate from the SEC site and will not be addressed in this publichealth assessment. Local raw water supplies from the Minden Mine intake and the RocklickMine intake were sampled in March 1990 by EPA in response to reports by the CCSFC thatPCBs were dumped into the mine shafts. All results were below detection limits except forlevels of methylene chloride, which ATSDR considered a result of laboratory contamination. However, the PCB detection limits of the measuring instruments were higher than ATSDR'scomparison values. Therefore, additional sampling is needed to determine if these raw water sources could affect public health.


  1. The Shaffer Equipment Company site poses a public health hazard because of the on-site riskto human health resulting from possible exposure to hazardous substances at concentrationsthat may result in adverse health effects. Human exposure to on-site PCBs may occur fortrespassing adults and children, and on-site workers via inadvertent ingestion, inhalation, anddermal contact/absorption with on-site soils and sediments.

  2. The Shaffer Equipment Company site is an indeterminate public health hazard for the generaloff-site population under current conditions because complete PCB data and information arenot available for off-site groundwater and off-site surface water contamination. There is nosignificant human exposure evident in regard to off-site soils, sediments, or air pathways.

  3. A potential health hazard exists for sensitive subpopulations such as fetuses and breast-fedinfants if their mothers eat PCB-contaminated snapping turtles from the area. Currently, thereare uncertainties as to the existence of snapping turtles at the SEC site.

  4. Surface water from Arbuckle Creek is contaminated with fecal coliform bacteria. WhileATSDR does not believe the contamination is related to the site, surface water should not beused until further testing indicates that coliform bacteria are within public health guidelines.

  5. Based upon the EPA post removal sampling, on-site subsurface soils are still contaminatedwith PCBs. Since no post removal surface soil samples were conducted, some areas on-sitemight also still be contaminated. Former workers at the SEC site would normally be the mostlikely population highly exposed to PCBs. Because PCB-contaminated oils were reportedlyburned, on- and off-site soil and on-site sediment sampling and analysis for PCDD and PCDFwere conducted. The levels of PCDDs and PCDFs that were found do not represent a publichealth concern.

  6. Serum samples from 1986-1987 and in 1990 on Minden residents that were analyzed forPCBs did not indicate an increased exposure to PCBs when compared to other populations. The levels are consistent with those seen in the general population.

  7. Cancer mortality data indicates that Fayette County has had an increased rate of respiratorydeaths when compared to the U.S. However, this increase existed before Shaffer wasoperating; therefore, other factors (e.g., age, smoking status, and socioeconomics) that could lead to this increase should be examined.

  8. The PCB detection limit for the off-site raw water supplies, surface water and for the on-sitespring were higher than ATSDR's comparison value. ATSDR cannot state with any certaintythat the off-site groundwater and surface water pathways are not of public health concern until additional sampling is accomplished.

  9. ATSDR evaluated the Vanderbilt Study and concluded that a clear relationship could not be established between PCB contamination and the significantly different observed rates of the four symptoms (shortness of breath, unexplained weight loss, persistent cough, and urinary tract infections).

  10. ATSDR's review of the toxicological information and the health information do not currentlyindicate high levels of exposure to PCBs or increases in health outcomes that can be linked toPCB exposure. Contaminants other than PCBs are not likely to present a health concern at the exposures likely to occur in the future or to have occurred in the past.


A. Recommendations and HARP Statement


  1. Because it is not clear whether snapping turtles are being eaten from the area of the site, thegeneral population and especially pregnant women and women who breast feed infants shouldbe cautioned not to eat aquatic life (snapping turtles) from the site vicinity until a bioassay(chemical analysis) is performed to determine if the PCB levels in edible tissues are safe(below regulatory standards). ATSDR could not conclude with any certainty that snappingturtles are not caught and eaten from the site vicinity.

  2. Even if turtles are not contaminated with PCBs, no aquatic life from Arbuckle Creek should beeaten unless thoroughly cooked, because of the fecal coliform bacteria contamination ofArbuckle Creek. Arbuckle Creek should be posted. Recreational uses of Arbuckle Creekshould be restricted and surface water from Arbuckle Creek should be sterilized before beingused. Such actions are recommended until the source of the elevated fecal coliform bacteria counts is identified and levels are brought below regulatory standards.

  3. Due to the unknown condition of the on-site surface soil and sediment following the 1990-91EPA soil removal action, additional fencing to restrict access is recommended for areas of thesite yet remaining unfenced. Surface soil sampling for PCBs would be necessary in order to characterize the site and determine appropriate future uses and restrictions.

  4. Based on PCB concentrations in subsurface soil, restriction of future site use and activitiesparticularly regarding the excavation of subsurface soil should be considered in an effort to reduce the potential of human exposure to PCBs.

  5. Existing PCB data for on-site groundwater (on-site spring), off-site groundwater (raw watersupplies) and off-site surface water, does not indicate contamination with PCBs and theconclusions of this public health assessment are based on those results. However, the PCBdetection limits of the measuring instruments were higher than ATSDR's comparison values. Therefore, ATSDR cannot state with any certainty that the groundwater and surface waterpathways are not of public health concern until the spring and the raw water supplies areresampled and analyzed with a PCB detection limit of 0.001 ug/l or less.

  6. Remedial workers should be provided with adequate personal protective equipment, asrequired by the Occupational Safety and Health Administration (OSHA), to prevent exposuresto contaminants during remedial activities. Also, workers should be required to follow allother applicable National Institute for Occupational Safety and Health and OccupationalSafety and Health Administration guidelines, advisories, and regulations.

  7. One child was reported with 11.0 ppb PCBs in the serum. This may have been alaboratory error. The child, if identifiable, should be retested. If the serum PCB levels remain high, intervention to reduce PCB levels and exposure should be taken.

  8. When indicated by public health needs, and as resources permit, ATSDR will evaluateadditional relevant health outcome data and community health concerns when available.


In accordance with the Comprehensive Environmental Response, Compensation, and Liability Act(CERCLA) of 1980, as amended, the data and information associated with the Shaffer EquipmentCompany Site have been evaluated for appropriate public health actions. There is currently noindication that people are being exposed to contaminants associated with the SEC site at levels ofpublic health concern. Although exposures may have occurred in the past, there is insufficientinformation to document exposure duration or levels. ATSDR's Health Activities RecommendationPanel (HARP) has determined that an environmental health education program is recommended toadvise public health professionals and the local medical community of the nature and possibleconsequences of exposure to contaminants at the Shaffer Equipment site. The value of obtaining acomplete and accurate exposure history will be stressed as part of the program. In addition,information that is provided on PCBs may include, but not be limited to, the physical nature of thecontaminant, potential exposure pathways (i.e., soil, water, air, food) and exposure routes (i.e,inhalation, ingestion, dermal), potential health effects, symptoms of exposure and testing andtreatment.

In addition, HARP recommends that the public health assessment be forwarded to the NationalInstitute for Occupational Safety and Health for their review and possible investigation of pastworker exposure to PCBs at the Shaffer Site.

The Shaffer Equipment Company data that has been evaluated for follow-up health activitiesincluded copies of laboratory results from the Pacific Toxicology Laboratories and the NationalHealth Laboratories, Inc. Based on this information, HARP recommends that the results of the PCBlevel of the infant/child with a PCB level of 11 mcg/l be evaluated for accuracy. If this result iscorrect, the child should have a repeat test and a case investigation to determine possible exposureroutes to PCB materials, including in-utero exposures.

B. Public Health Actions


  1. The Division of Health Education in conjunction with the local medical community willprovide an environmental health education program to advise health professionals and the localmedical community of the nature and possible consequences of exposure to contaminants at the Shaffer Equipment Site.

  2. The Division of Health Assessment and Consultation will contact the National Institute forOccupational Safety and Health (NIOSH) regarding concerns for worker populations/familiesand will provide NIOSH with a copy of the Shaffer Equipment Company Public HealthAssessment in order that they be advised of the potential for past occupational exposure to PCBs at Shaffer.

  3. The Division of Health Assessment and Consultation will review any new data when it isavailable to ATSDR and will amend the Shaffer Equipment Company Public HealthAssessment if new findings are significant. ATSDR will reevaluate and expand the PublicHealth Actions when needed. New environmental, toxicological, or health outcome data, orthe results of implementing the above proposed actions determine the need for additionalactions at this site.

  4. The Division of Health Assessment and Consultation will notify the CCSFC and the WVDOHof the HARP recommendation that the results of the PCB level of the infant/child with a PCBlevel of 11 mcg/l be evaluated for accuracy. If the result is correct, the test should be repeatedand a case investigation initiated to determine possible exposure routes to PCB materials, including in-utero exposures.

  5. The Division of Health Assessment and Consultation will hold a public meeting, as requestedby the CCSFC, following the release of the final Shaffer Equipment Company Public HealthAssessment.


(listed alphabetically)

Jeffrey A. Church, REHS
Environmental Health Scientist
Community Health Branch

Donald Y. Joe, P.E.
Environmental Engineer
Community Health Branch

Virginia Lee, M.D., M.P.H.
Medical Officer
Federal Facilities Branch

Thomas Umbreit, Ph.D.
Community Health Branch

ATSDR Regional Representative:

Charles J. Walters
Public Health Advisor
Region III


1. Telephone interviews with former Shaffer Employees. Donald Joe, ATSDR. June 1, throughJune 21, 1990.

2. ATSDR. Toxicological Profile for Selected PCBs (Aroclor -1260, -1254, -1248, -1242, -1232,-1221, and -1016). June 1989.

3. Sittig M. Handbook of Toxic and Hazardous Chemicals and Carcinogens. 2nd edition. ParkRidge, NJ: Noyes Publications, 1985.

4. Superfund Record of Communications from C. J. Walters to R. Caron (EPA) dated December 11,1984.

5. U.S. Bureau of the Census. 1980 Census of Population: Volume 1: Characteristics of thePopulation: Chapter B: General Population Characteristics. Department of Commerce. IssuedAugust 1982.

6. U.S. Bureau of the Census. 1980 Census of Population: Volume 1: Characterisitcs of thePopulation: Chapter C: General Social and Economic Characteristics. Department of Commerce. Issued August 1983.

7. U.S. Bureau of the Census. County and City Data Book 1988. Pages 562-587.

8. Riggans WB, Van Bruggen J, Acquavella JF, et. al.. U.S. Cancer Mortality Rates and Trends,1950-1979: Vols. I, II, III. Research Triangle Park, N.C.: U.S. Environmental Protection Agencyand National Cancer Institute, 1983. EPA publication number (EPA) 600/1-83-015a.

9. Environmental Epidemiology Task Force, West Virginia Department of Health, Oak Hill, W.V.,Investigation--Stage One: Initial Assessment, January 11, 1983.

10. Hirschman D and S Walsh. A Virginia Student Environmental Health Project Intern Report: Community Health and PCB Exposure in Minden, West Virginia. September 1989.

11. Letter from Charles W. Merritt, M.D. to J.A. Lybarger, M.D.. March 16, 1989.

12. Eastern Scientific Laboratories, PCB data analysis sheets of 16 soil samples collected at SECsite on May 22, 23, 1990., Purchase order number #0454 dated June 7, 1990.

13. Joe, Don. Comments on the NUS Sampling Plan to the site investigation officer of the USEPA.

14. ATSDR. Toxicological Profile for 2,3,7,8-Tetrachlorodibenzo-p-dioxin. June 1989.

15. Superfund Record of Communications. Telephone call from L.Haddy to C.J. Walters, June 25,1990.

16. Environmental Protection Agency Region III. Federal On-Scene Co-ordinator's Report; ShafferEquipment Site, Minden, West Virginia.

17. Shalat SL, True LD, Fleming LE, and PE Pace. "Kidney Cancer in Utility Workers Exposed toPolychlorinated Biphenyls (PCBs)." British Journal of Industrial Medicine 1989; 46:823-824.

18. Lawton RW, Ross MR, Feingold J, and JF Brown. "Effects of PCB Exposure on Biochemicaland Hematological Findings in Capacitor Workers." Environmental Health Perspectives 1985;60:165-184.

19. Fischbein A. "Liver Function Tests in Workers with Occupational Exposure to PolychlorinatedBiphenyls (PCBs): Comparison with Yusho and Yu-Cheng." Environmental Health Perspectives1985; 60:145-150.

20. Fischbein A, Wolff MS, Lilis R, Thorton J, and IJ Selikoff. "Clinical Findings amongPCB-Exposed Capacitor Manufacturing Workers." Annals of the New York Academy of Sciences1979; 320:703-715.

21. Reggiani G and R Bruppacher. "Symptoms, Signs, and Findings in Humans Exposed to PCBsand Their Derivatives." Environmental Health Perspectives 1985; 60:225-232.

22. Kimbrough RD. "Human Health Effects of Polychlorinated Biphenyls (PCBs) andPolybrominated Biphenyls (PBBs)." Annual Review of Pharmacology and Toxicology 1987;27:87-111.

23. Emmett EA, Maroni M, Schmith JM, Levin BK, and J Jeffreys. "Studies of Transformer RepairWorkers Exposed to PCBs: I. Study Design, PCB Concentrations, Questionaire, and ClinicalExamination Results." American Journal of Industrial Medicine 1988; 13:415-427.

24. Fischbein A, Wolff MS, Berstein J, Selikoff IJ, and J Thornton. "Dermatological Findings inCapacitor Manufacturing Workers Exposed to Dielectric Fluids Containing PolychlorinatedBiphenyls (PCBs)." Archives of Environmental Health 1982; 37(2):69-74.

25. Fischbein A, Rizzo JN, Soloman SJ, and MS Wolff. "Oculodermatological Findings in Workerswith Occupational Exposure to Polychlorinated Biphenyls (PCBs)." British Journal of IndustrialMedicine 1985; 42:426-430.

26. Taylor JS. "Environmental Chloracne: Update and Overview." Annals of the New Academy ofSciences 1979; 320:295-307.

27. Taylor PR, Stelma JM, and CE Lawrence. "The Relation of Polychlorinated Biphenyls to BirthWeight and Gestational Age in the Offspring of Occupationally Exposed Mothers." AmericanJournal of Epidemioloy 1989; 129(2):395-406.

28. Jacobson JL, Jacobson SW, and HEB Humphrey. "Effects of In Utero Exposure toPolychlorinated Biphenyls and Related Contaminants on Cognitive Functioning in YoungChildren." Journal of Pediatrics 1990; 116:38-45.

29. Gladen BC, Rogan WJ, Hardy P, Thullen J, Tingelstad J, and M Tully."Development afterExposure to Polychlorinated Biphenyls and Dichlorodiphenyl Dichloroethene Transplacentally andThrough Human Milk." Journal of Pediatrics 1988; 113:991-995.

30. Pacific Toxicology Laboratories. Results of the Adipose Tissue Analysis for SelectedCongeners of PCBs. June 1990.

31. National Health Laboratories. Results of Serum Analysis of Minden Residents for Total PCBs. 1986.

32. Kreiss K. "Studies on Populations Exposed to Polychlorinated Biphenyls." EnvironmentalHealth Perspectives 1985; 60:193-199.

33. Pacific Toxicology Laboratories. Results of Serum PCB Analysis of Minden Residents. 1990.

34. Fait A, Grossman E, Self S, Jeffries J, Pellizzari ED, and EA Emmett."PolychlorinatedBiphenyl Congeners in Adipose Tissue Lipid and Serum of Past and Present Transformer RepairWorkers and a Comparison Group." Fundamental and Applied Toxicology 1989; 12:42-55.

35. Emmett EA, Maroni M, Jeffreys J, Schmith J, Levin BK, and A Alvares. "Studies ofTransformer Repair Workers Exposed to PCBs: II. Results of Clinical Laboratory Investigations." American Journal of Industrial Medicine 1988; 14:47-62.

36. Kreiss K, Roberts C, and HEB Humphrey. "Serial PBB Levels, PCB Levels, and ClinicalChemistries in Michigan's PBB Cohort." Archives of Environmental Health 1982; 37(3):141-147.

37. Jacobson JL, Humphrey HEB, Jacobson SW, Schantz SL, Mullin MD, and R Welch. "Determinants of Polychlorinated Biphenyls (PCBs), Polybrominated Biphenyls (PBBs), andDichlorodiphenyl Trichloroethane (DDT) Levels in the Sera of Young Children." AmericanJournal of Public Health 1989; 79(10):1401-1404.

38. Jensen AA. "Background Levels in Humans." In: Kimbrough RD and AA Jensen, eds. Halogenated biphenyls, terphenyls, naphthalenes, dibenzodioxins, and related products. New York: Elsevier, 1989:345-380.

39. Stehr-Green PA, Burse VW, and E Welty. "Human Exposure to Polychlorinated Biphenyls atToxic Waste Sites: Investigations in the United States." Archives of Environmental Health 1988;43(6):420-424.

40. CEIMIC Corporation, Data analysis sheets, Contract number 68D90024, 1990.

41. Environmental Health Research and Testing, Inc., Data analysis of 30 samples received June18, 1990., dated June 29, 1990.

42. MDS Laboratories, Data analysis of samples received on November 6, 1990, November 30,1990, and January 28, 1991.


1. Swackhamer DL, Armstrong, DE. 1987. Distribution and characterization of PCBs in Lake Michigan water. J Great Lakes Res 13:24-26.

Next Section     Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #