• National Exposure Registry
• Benzene
• Dioxin
• Journal Articles
• Trichloroethylene(TCE)
• Trichloroethane(TCA)
• Registry Data

SECTION 3

TCE SITES: SELECTION AND DESCRIPTION

The site selection process used for developing subregistries in the National Exposure Registry is described in detail in the Policies and Procedures Manual (1).

During the site selection process, a total of 279 potential sites were identified as having TCE as a contaminant of drinking water. After reviewing the sites with respect to certain criteria, the sites described in this section were selected as the most appropriate candidates for inclusion in the TCE Subregistry. The criteria for selection included documentation of drinking water contamination and level(s) of contamination, identification of and estimated size of the exposed population, identification of any existing potential exposure or past potential exposure (including duration of exposure and the time elapsed since exposure), identification of susceptible subpopulations, identification of any existing biomonitoring data, identification of the number and levels of secondary contaminants, and confirmation of documented individual exposures. Also, sites where the state and local officials supported development of a subregistry were given priority.

A total of 13 sites located in 3 states were selected for inclusion in the TCE Subregistry. The exposure period for each of these sites represents estimated beginning dates and confirmed ending dates of contamination. The beginning dates reflect the most appropriate dates based on existing environmental data and other information obtained from local, state, or other federal agencies.

The type, amount, and quality of environmental data varied greatly from site to site. The quantity, number, and source of chemicals released varied considerably across sites. Also, the local soil and weather conditions very likely affected the environmental measurement and sampling schemes and added to the variations from site to site(1).

Table 3-1 provides the name(2) and location of each site. The table also provides a summary of environmental data by site. In most cases, documented contamination for a given household address was based on one environmental sample. Table 3-2 summarizes the number of samples available for each household. As is shown in Table 3-2, for about 80% of the households, only one water sample was available; the remainder of the households had two or more available water samples. Following is a summary of the 13 sites included in the TCE

Table 3-1.--Summary of environmental data by site.

TCE Subregistry Site	Number of Households*	Year Exposure Began	Exposure Period† (Number of Years)	Maximum (Median) Level of TCE Reported (ppb)§
Michigan
Verona Well Field and Dowagiac	66	1970/ 1975	13.0/ 7.0	2,000 (6.0)
McGraw-Edison Corporation	55	1975	7.0	733 (1.0)
Indiana
Superior Street	134	1976	11.0	19,380 (84.0)
Central Area	28	1980	7.0	114 (0.4)
Gemeinhardt Piccolo Company	100	1972	17.0	1,600 (4.0)
Conrail Rail Yard	49	1970	18.0	1,520 (78.0)
Illinois
Acme Solvents Reclamation, Inc.	13	1980	10.5	100 (1.0)
Beloit Corporation	3	1983	7.5	3 (2.0)
Byron Johnson Salvage Yard	25	1980	10.5	249 (9.1)
Frinks Industrial Waste	5	1978	12.5	16 (14.0)
Southeast Rockford Groundwater Contamination	331	1978	12.5	122 (15.0)
Warner Electric Brake and Clutch Company	74	1978	6.5	5,220 (234.0)
Total	883	--	--	--

*Represents each household where at least one water sample was positive for TCE.

†Exposure period is based on best available evidence of when contamination occurred and when exposure ceased following water treatment or switch to an alternative water source.

§ppb = parts per billion.

Table 3-2.--Total number of positive water samples per site address.

Number of Samples	Frequency	Percent
0	7*	0.7
1	688	78.0
2	126	14.3
3	38	4.3
4	24	2.7
Total	883	100.0

*These samples are associated with 83 people in the TCE Subregistry. (Since the time of health outcome data analyses the environmental sampling data has been received.)

Subregistry. Each summary includes a brief description of the source of contamination, the period of documented contamination, and the contaminants detected in sampling activities.

DESCRIPTION OF TCE SUBREGISTRY SITES

Michigan

Verona Well Field Area

Verona Well Field (see Figure 1, Appendix A) supplied potable water to the 35,000 residents of Battle Creek, Michigan. Contamination in the area was first documented in August 1981, when water samples collected from a local church revealed volatile organic compound (VOC) contamination. Subsequent sampling of the municipal wells by the Michigan Department of Public Health indicated that 10 of the 28 wells in the field were contaminated. Private water supply wells located in residential areas around the well field were also found to be contaminated. The contaminated municipal well field was listed by the U.S. Environmental Protection Agency (EPA) as a National Priority List (NPL) site in July 1982. In 1983, an EPA remedial investigation identified the primary sources of contamination.

By the end of 1982, 81 out of the 142 private wells tested were found to be contaminated according to ATSDR records. The maximum contaminant concentrations in the private wells (see Table 3-3) were much greater than the concentrations in the public water supply. The public water supply was subject to a blending process that lowered the concentration

Table 3-3.Verona Well Field and Dowagiac, Michigan, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	77	0.01	2,000	6.0
(cis)-1,2,-Dichloroethene	70	ND**	3,900	6.0
1,1-Dichloroethane	59	ND	150	2.0
Tetrachloroethylene	57	ND	230	1.0
1,1,1-Trichloroethane	55	ND	134	3.0
1,2-Dichloroethane	51	ND	325	3.0
1,1-Dichloroethene	51	ND	60	1.0
(trans)-1,2-Dichloroethane	16	ND	1	ND
Freon	5	ND	<0.01	ND
Methylene chloride	4	ND	3	1.5
Chloroform	3	ND	<0.01	ND

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

Note: The Verona Well Field and Dowagiac data were combined here, and elsewhere through- out this report, for two reasons. First, all registrants associated with these two sites were derived from a listing of participants and nonparticipants of the Battle Creek health study (12). The researchers for that study did not distinguish cases according to site area. Second, Verona Well Field and Dowagiac share the same cost recovery identification number.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

of contamination to below the EPA maximum contaminant limit (MCL) of 5 parts per billion (ppb) for drinking water (8). Only those residences using private well water were included in the Battle Creek Epidemiology Study (12) and subsequently, the TCE Subregistry.

Verona registrants were selected for inclusion in the TCE Subregistry based on their eligibility for inclusion in the 1986 Battle Creek health study (12). All participants associated with the 1986 Battle Creek health study(3) (both the exposed and comparison populations) were asked to take part in the National Exposure Registry. The Battle Creek health study, supported by ATSDR, was conducted by the Centers for Disease Control and Prevention (CDC) in conjunction with the Michigan Department of Public Health. The study included VOC-exposed and comparison (very low or no exposure) participants. People who were eligible but had declined to participate in the health study were asked to participate in the TCE Subregistry; 72 chose to participate. The comparison population from the epidemiology study was not included in the analysis presented in this report.

The beginning exposure date was estimated to be 1970. This date is about 1.5 years after the solvent company identified by EPA as the source of the contamination started its operations. This date was considered to represent the minimum length of time necessary for the contamination to have occurred and for the VOCs to have reached the wells at a detectable level. Most residents with contaminated private wells were changed to a VOC-free water source between 1982 and 1983.

Dowagiac

In 1983, VOCs were discovered in private residential wells in Dowagiac, Michigan, located in Cass County approximately 70 miles southwest of Battle Creek (Figure 2, Appendix A). The source of contamination appeared to be an industrial plant that used solvents to degrease manufactured metal parts. The solvents were stored in underground tanks that are believed to have leaked. Such solvents had been stored in the tanks for decades. Evaluation of the existing environmental data indicated that contamination in Dowagiac started many years prior to 1977.

The contaminants reported in residential wells are shown in Table 3-3. The exposure period considered for the TCE Subregistry was from January 1, 1975, through December 31, 1981.

McGraw-Edison Corporation

The McGraw-Edison Corporation plant of Albion, Michigan (Figure 3, Appendix A), manufactured air conditioners, humidifiers, dehumidifiers, electric heaters, and air cleaners from 1958 through 1980. From 1970 through 1980, TCE-contaminated industrial distillate (still bottoms) was spread over a dirt road and parking lots adjacent to the plant to control dust. Based on existing environmental information, it is estimated that 15,000 gallons of contaminated still bottoms was spread on site. Analysis of the on-site residue revealed TCE concentrations as high as 27,000 parts per million (ppm).

Two distinct aquifers lie below the McGraw-Edison plant in unconsolidated glacial sediments containing a sandy unit interspersed with layers of gravel, coal, and silt, and a discontinuous clay layer. The upper aquifer varies in thickness from 1 to 50 feet. The lower aquifer lies 10 to 75 feet below the ground surface and is a minimum of 290 feet thick. The general direction of groundwater flow is south-southwest.

From 1980 to 1981, TCE was found in the on-site shallow monitoring wells at a maximum level of 946,937 ppb, and in the on-site deep wells at 8,590 ppb. At the time of off- site sampling, the contaminant plume had migrated approximately 0.5 mile. The nearest residence was 0.2 mile away. TCE was the only contaminant detected in most of the well samples (see Table 3-4). Concentrations of TCE as high as 733 ppb were detected in private residential wells.

Table 3-4.--McGraw-Edison Corporation, Michigan, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	135	ND**	733	1.0
Chloroform	4	1	3	2.5
(cis)-1,2,-Dichloroethene	4	5	18	5.5
1,2-Dichloroethane	1	9	9	9

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

People with contaminated private wells were initially supplied with bottled water and were later connected to the municipal water system. The period of exposure for subregistry purposes was determined to be from January 1, 1975, through December 31, 1981.

Indiana

Superior Street

The Superior Street site is located on the east side of Elkhart and includes the East Jackson Street neighborhood (Figure 4, Appendix A). According to Elkhart County Health Department records, TCE contamination was first documented from samples taken in 1976 in response to citizen complaints. However, the results of the sample analyses were not available for review. The timing of the complaints corresponded with a January 1976 fire at an underground storage tank facility south of Superior Street.

Residential well sampling was again conducted in 1985 and revealed TCE contamination; the results of this sampling are presented in Table 3-5. After the 1985 groundwater investigation, further testing was conducted in adjacent areas (see other Elkhart sites reported in the sections that follow) to identify the range of the contamination. After 1986, either an alternative water supply or installation of a filtration system was provided for residences using contaminated water. The exposure period was from January 1, 1976, through December 31, 1986.

Central Area

The Central Area (Figure 5, Appendix A) includes central Elkhart south of the St. Joseph River and east to the Elkhart River, and extends south of the city. There are several possible contaminant sources (many leaking underground storage tanks) throughout this area of the city. TCE contamination was first noticed in 1983 by residents complaining of a strong odor in their water. According to Elkhart County Health Department files, in 1983 the first samples from residential wells were taken; they revealed TCE and other chemical contamination (Table 3-6). Letters were issued to residents with contaminated wells warning them of the problem. Private well sampling repeated by EPA in 1985 showed migration of the contamination to other residences in the vicinity. At that point, Superfund expenditures were approved for bottled water and access to city water.

For purposes of the TCE Subregistry, the exposure period was determined to be from January 1, 1980, through December 31, 1986, when residences were put on an alternative water supply or filtration systems were installed.

Gemeinhardt Piccolo Company

The Gemeinhardt Piccolo Company is located in the southwestern corner of the city of Elkhart (Figure 6, Appendix A). Gemeinhardt has manufactured flutes and piccolos on the site since the 1940s. TCE was used in the process until 1972. From 1972 through 1980, 1,1,1-trichloroethane was used. Tetrachloroethylene has been used since 1980.

A groundwater problem was first recognized by a Gemeinhardt employee who complained to the Elkhart County Health Department about the water quality at the plant. Sampling was conducted by the health department in September 1983 and revealed residential well contamination (Table 3-7). Contaminated soil was removed from an on-site seepage bed and two dry wells in the fall of 1984.

Table 3-5.--Superior Street, Elkhart, Indiana, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	146	0.06	19,380	84.0
Tetrachloroethylene	77	ND**	140	ND
1,1-Dichloroethene	74	ND	1,260	1.0
1,1,1-Trichloroethane	51	ND	2,200	2.0
Trichlorofluoromethane	16	ND	340	1.0
1,1-Dichloroethane	15	ND	1,120	1.0
1,2-Dichloroethane	14	ND	54	1.0
1,2-Dichloropropane	14	ND	40	1.0
Benzene	9	2.99	7.5	4.5
Vinyl chloride	5	ND	60	2.5
Bromodichloromethane	4	ND	0.7	0.1
Chloroethane	4	ND	5.7	ND
Chloroform	4	ND	460	26.6
Methylene chloride	4	ND	21.2	0.5
(cis)-1,2,-Dichloroethene	1	9	9	9
1,2-Dichloroethylene	1	9	9	9

Specific chemical names and table values reported to the 10th or 100th decimal

place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

Table 3-6.--Central Area, Elkhart, Indiana, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	28	<0.10	114	0.4
Tetrachloroethylene	26	ND**	105	ND
1,1-Dichloroethene	25	ND	92.1	ND
1,1,1-Trichloroethane	22	ND	0.7	ND
1,2-Dichloropropane	2	3	3	3
Vinyl chloride	2	15	22.2	18.6
(cis)-1,2,-Dichloroethene	1	240	240	240
1,1-Dichloroethane	1	4	4	4
1,2-Dichloroethylene	1	1.7	1.7	1.7
Toluene	1	0.2	0.2	0.2

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

Table 3-7.--Gemeinhardt Piccolo Company, Elkhart, Indiana, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	121	ND**	1,600	4.0
1,1,1-Trichloroethane	112	ND	7,100	18.2
1,1-Dichloroethane	100	ND	160	1.5
Tetrachloroethylene	68	ND	7,500	ND
1,1-Dichloroethane	49	ND	160	6.0
Carbon tetrachloride	23	ND	39	ND
(trans)-1,2-Dichloroethene	23	ND	40	5.0
Methylene chloride	11	19.90	11,000	20.0
Chloromethane	10	9.99	180	11.5
Chloroform	4	ND	0.5	0.4
Toluene	3	0.20	6.8	1.3
(cis)-1,2,-Dichloroethene	2	3.20	7.1	5.2
Ethylbenzene	2	2.10	5.2	3.7
Xylene	2	15.50	45.2	30.4
Chloroethane	1	9.9	9.9	9.9

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

The contaminant migration was to the north-northwest, with an increase in concentration approximately 2,000 feet from the site. The center of the TCE plume was approximately 1,600 feet to the north-northwest, in the vicinity of another musical instrument manufacturing company using some of the same solvents. The exposure period for TCE Subregistry purposes was from January 1, 1972, through December 31, 1988, when the residences were supplied with an alternative water source or filtration systems were installed.

Conrail Rail Yard

This 675-acre site is owned by the Consolidated Rail Corporation and is situated in the southwest corner of the city of Elkhart, Indiana (Figure 7, Appendix A). At the time of data collection, the site contained 72 rail tracks where cars were electronically separated and switched from one track to another. The yard handled approximately 75 trains per day. Car repair and diesel refueling were also conducted on the site. The Conrail facility has been in operation since about 1956, and was extensively renovated during the period 1979 through 1982. Adjacent to the site are light industrial operations; a residential area on County Road Number 1 is located about 2,000 feet beyond the outlying light industrial areas.

The aquifer in this area is located in fairly coarse, highly permeable glacial outwash. The groundwater flow is northwest, toward the St. Joseph River. Information available about the contaminant plume indicated it was 1.5 to 2.0 miles long and discharged into the St. Joseph River. The groundwater flow rate was calculated to be 0.43 foot per day.

The Elkhart County Health Department requested an investigation by EPA in 1986 to confirm indications that local wells were contaminated with carbon tetrachloride and TCE. Five areas on site were identified as potential sources of the contamination. Residential wells downgradient from these sites were affected by the contaminant plume (Table 3-8).

In 1979, the municipal water system in Elkhart provided service for approximately 41,000 persons. The remaining population obtained its water from private residential wells. One-hundred houses using private wells were located north (downgradient) of the site. Since 1987, 76 of the houses with contaminated wells have used whole-house filters or point-of-use filters. The remaining 24 houses have been supplied with bottled water.

Using the groundwater flow rate, ATSDR estimated that the period of exposure was from January 1, 1970, until December 31, 1987. Carbon filtration systems were installed after 1987.

Illinois

Acme Solvents Reclamation, Inc.

The Acme Solvents Reclamation facility is located in a low-density residential area of Winnebago County in northern Illinois (Figure 8, Appendix A). The 20-acre facility was active from 1960 through 1973. Still-bottom residues, sludge, nonrecoverable solvents, paints, and oils were disposed of on site in unlined lagoons or stockpiles. Adjacent to the site were other potential sources for groundwater contamination, including an asphalt-lined sanitary landfill to the west and a mixing plant and quarry to the northeast. Two aquifers are located in the Acme

Table 3-8.--Conrail Rail Yard, Elkhart, Indiana, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	51	<0.10	1,520	78.0
Carbon tetrachloride	51	ND**	6,860	48.0
1,1-Dichloroethene	51	ND	60	ND
Tetrachloroethylene	51	ND	2.4	ND
Chloroform	49	ND	75	ND

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

site area; one aquifer is in the unconsolidated deposits (loose material) which overlie the area, the other is in the bedrock material. The groundwater flow is from east to west.

In 1981, the Illinois Environmental Protection Agency detected chemicals from the Acme site in nearby residential wells (Table 3-9). The exposure period for TCE Subregistry purposes was from January 1, 1980, through June 2, 1990, when baseline data collection for the TCE Subregistry began. The USEPA supplied carbon-filtration systems to the residences with contaminated wells as part of the remediation efforts.

Beloit Corporation

The Beloit Corporation (Figure 9, Appendix A) manufactures equipment for the pulp and paper industry, and is located on 175 acres along the Rock River north of the city of Rockton, Illinois. Prior to 1986, wastes from its manufacturing and equipment demonstration processes were disposed of in three unlined lagoons located at the northern end of the site near the Rock River. Since 1986, the wastes have been treated in a process waste water treatment system. In 1981, sediments from the lagoons were injected into the subsurface soil at the southern boundary of the site. Approximately 220,000 gallons of sludge were injected over 10 acres. On-site groundwater monitoring has been conducted since 1982 by the Beloit Corporation as part

Table 3-9.Acme Solvents Reclamation, Inc., Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	49	ND**	100	1.0
Tetrachloroethylene	17	ND	86	3.0
1,1-Dichloroethane	16	ND	38	4.4
Methylene chloride	16	ND	22	ND
1,1,1-Trichloroethane	14	ND	8	1.5
Bromodichloromethane	13	ND	20	ND
Carbon tetrachloride	13	ND	30	ND
(trans)-1,2-Dichloroethane	13	ND	200	8.0
Butylbenzyl phthalate	12	ND	15.6	ND
Bis (2-ethylhexl) phthalate	12	ND	56.8	ND
(trans)-1,3-Dichloropropane	12	ND	15.6	ND
1,2-Dichloroethane	4	ND	4.99	0.5
Copper	3	10	20	10.0
Iron	3	ND	100	ND
Magnesium	3	41,000	104,000	54,100.0
Manganese	3	ND	50	20.0
Zinc	3	ND	100	ND
Bromoform	1	0.99	0.99	0.99
Chloroform	1	0.99	0.99	0.99
Dibromochloromethane	1	0.99	0.99	0.99

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

of its groundwater compliance monitoring program. The available sampling results from the on-site monitoring showed TCE contamination of up to 42 ppb.

The predominant flow of groundwater is to the south-southwest. Approximately 300 homes were within a one-mile radius of the facility. All 65 homes in the Blackhawk subdivision, located about 600 feet southeast of the site, used private wells for their water source. Of the 65 private wells, 55 have been sampled by EPA since 1982. Sixteen wells were found to be contaminated with VOCs. Three of the sixteen had TCE contamination (Table 3-10). In addition, Rockton municipal well Number 5 is located one-half mile southeast of the site. Although the municipal well was sampled, contamination was not detected. The estimated exposure period was January 1, 1983, through June 2, 1990, when TCE Subregistry data collection began.

Table 3-10.--Beloit Corporation, Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	5	1	3	2.0
1,1,1-Trichloroethane	5	1	212	27.0
Methylene chloride	4	1	1	1.0
1,1-Dichloroethene	3	7	17	7.0
Tetrachloroethylene	3	45	300	48.0
1,1-Dichloroethane	2	1	2	1.5
Bromoform	1	1	1	1
Dibromochloromethane	1	1	1	1

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

Byron Johnson Salvage Yard

The Byron Johnson Salvage Yard is located in a rural area of Ogle County in northern Illinois, between the cities of Byron and Oregon. In 1990, the population within 1.5 miles of the site was estimated to be 500 to 1,000 people, with the majority living west, northwest, and north of the site. The site (Figure 10, Appendix A) consists of two areas--the 20-acre Byron Johnson Salvage Yard, and the 480-acre former Dirk's Farm property. Waste materials containing VOCs, cyanide, and heavy metals had been improperly disposed of during the 1960s and 1970s in both areas, resulting in contamination of soil or sediment and groundwater. Investigation of the site was initiated by EPA in 1970 when a strong chemical odor and soil discoloration in puddles were noticed throughout the salvage yard. Available data from groundwater monitoring wells in the area indicated that both the upper aquifer (Galena-Plateville Dolomite) and lower aquifer (St. Peter Sandstone) were contaminated with VOCs, heavy metals, and cyanide. The plume extended at least 1.5 miles to the northwest and 0.5 mile to the west and north, and affected the water quality of residential wells (Table 3-11).

In July 1984, households using water contaminated with greater than 200 ppb TCE were supplied with bottled water; however, residents continued to use contaminated water for other household purposes. For TCE Subregistry purposes, the exposure period was from January 1, 1980, through June 2, 1990, when baseline data collection began.

Frinks Industrial Waste

Frinks Industrial Waste, a 4.8-acre site located 2 miles north of Pecatonica, Illinois (Figure 11, Appendix A), began operation under a state-issued permit in 1979 as a septic tank pumping service. In 1980, the business was expanded to include the handling and treatment of industrial wastes. A lagoon used to store wastes was constructed in 1975 and closed in 1980. The wastes included oils, cyanide-containing compounds, solvents, and alkaline- and acid-containing wastes. Disposal methods consisted mainly of gravity separation and disposal in unlined lagoons.

Groundwater is present at the facility in both the overlying unconsolidated (loose) till and the underlying bedrock. The aquifer most used as a domestic water supply source in the area is in the upper bedrock formation (Galena-Plateville Dolomite). This aquifer occurs at depths ranging from 20 to 60 feet. The primary direction of groundwater flow in the aquifer is to the south at 15.0 feet per year. Contamination has migrated off site to nearby residential wells (see Table 3-12). This was a rural area; therefore, few wells were affected. The period of exposure for TCE Subregistry purposes was from January 1, 1978, through June 2, 1990, when TCE Subregistry data collection began.

Southeast Rockford Groundwater Contamination

The Southeast Rockford site (Figure 12, Appendix A) is located in a 2-square-mile residential area. Groundwater contamination in the area was first identified by EPA in 1984 during an investigation of illegal dumping by an electroplating company. A 1984 investigation by the Winnebago County Health Department revealed a contaminant plume extending for 2 miles

Table 3-11.--Byron Johnson Salvage Yard, Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	33	0.90	249	9.1
Tetrachloroethylene	32	ND**	8	0.1
(trans)-1,2-Dichloroethane	22	ND	12	ND
1,1,1-Trichloroethane	18	ND	21	9.5
Carbon tetrachloride	7	1	4	3.0
1,1-Dichloroethane	7	3	20	7.0
1,1-Dichloroethene	4	2	6	4.0
Other volatile organic compounds	3	ND	18	16.0

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

Table 3-12.--Frinks Industrial Waste, Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	5	5	16	14.0
1,1,1-Trichloroethane	4	1	2	2.0
Other volatile organic compounds	1	4	4	4
(trans)-1,2-Dichloroethene	1	12	12	12

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

from 20th Street to the Rock River. The plume is believed to have originated from multiple sources. The principal direction of the groundwater movement in southeast Rockford is from the uplands on the east to the Rock River on the west.

Although municipal water was available, many residents of the Rockford site used private wells. Samples taken in 1984 from houses with shallow wells (50 feet or less) in the sand and gravel aquifer showed elevated levels of VOC contamination (Table 3-13). Rockford municipal well Number 35, located in this area, was drilled much deeper (150 feet) in the sand and gravel aquifer and also had elevated VOC levels. Well Number 35 was closed in 1985 because of this contamination. The period of exposure for TCE Subregistry purposes was January 1, 1978, through June 2, 1990, when baseline data collection began.

Warner Electric Brake and Clutch Company

Warner Electric Brake and Clutch Company (Figure 13, Appendix A) is a manufacturing facility located on 93.3 acres in Roscoe, Illinois (population 1,600). The facility has been in operation since 1957. In the brake, clutch, and drive train component manufacturing process, various organic solvents are used for degreasing and parts cleaning. Two on-site, unlined lagoons were used to dispose of waste solvents as part of the facility's National Pollution Discharge and Elimination System (an EPA program).

Sandy, gravely soils and a shallow aquifer make the groundwater highly susceptible to contamination. Available environmental information indicated that a well-defined contaminant plume approximately 5,000 to 7,200 feet long and 12,000 feet wide originated from the unlined

Table 3-13.--Southeast Rockford Groundwater Contamination, Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶ (ppb)
Trichloroethylene	390	0.45	122	15.0
1,1,1-Trichloroethane	369	ND**	436	41.0
1,1-Dichloroethane	331	ND	133	11.0
(trans)-1,2-Dichloroethane	227	ND	100	ND
Tetrachloroethylene	226	ND	32	ND
1,1-Dichloroethene	219	ND	236	4.0
1,2-Dichloroethane	148	ND	52	ND
(cis)-1,2-Dichloroethene	112	ND	456	7.5
Chloroform	57	ND	140	ND
Methylene chloride	16	ND	9	ND
Bromodichloromethane	2	1.50	3.7	2.6
Benzene	1	6.5	6.5	6.5

Specific chemical names and table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from N samples drawn.

§These values represent the maximum reading reported from N samples drawn.

¶These values represent the median reading reported from N samples drawn.

**ND means nondetectable using available analytical methods.

lagoons. Groundwater in the vicinity of the plume flows from the north and northeast to the south and southwest toward the Rock River.

The Town of Roscoe, Illinois, does not have a public water supply, so all houses rely on private wells. Two subdivisions comprising 300 houses are located adjacent to the site (Figure 14, Appendix A). The contaminant plume is located directly underneath the two subdivisions. In early 1983, concentrations of TCE exceeding 500 ppb were found by the Winnebago County Public Health Department in the residential wells of 30 houses. Subsequent investigations by the Illinois Department of Public Health revealed about 100 residences with

Table 3-14.--Warner Electric Brake and Clutch Company, Roscoe, Illinois, residential well sampling.

Chemical	N*	Minimum† (ppb)	Maximum§ (ppb)	Median¶(ppb)
Trichloroethylene	108	0.50	5,220	234.0
1,1,1-Trichloroethane	105	ND**	255	10.0
Tetrachloroethylene	100	ND	45	1.0
1,1-Dichloroethane	4	ND	8	3.7
Methylene chloride	2	1	3	2.0
Other volatile organic compounds	2	ND	2	1.0
(trans)-1,2-Dichloroethene	2	ND	1	0.5
(cis)-1,2,-Dichloroethene	1	6	6	6

Table values reported to the 10th or 100th decimal place are values that were provided to ATSDR.

*N is the number of samples with reported results for the particular chemical.

†These values represent the minimum reading reported from n samples drawn.

§These values represent the maximum reading reported from n samples drawn.

¶These values represent the median reading reported from n samples drawn.

**ND means nondetectable using available analytical methods.

TCE-contaminated wells (Table 3-14). In December 1983, Warner began supplying all residents of the two subdivisions with bottled water. For TCE Subregistry purposes, the exposure period was from January 1, 1978, through July 31, 1984.

1. 1 The contaminant levels and sampling dates might also vary within a site when households or residences are compared, especially if they are served by individual wells. For this and other reasons (for example, variations in well construction and depth), the range in TCE sampling results within sites might be great.

2. 2 The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (10), amended by the Superfund Amendments and Reauthorization Act of 1986 (11), provides for recovery of costs incurred for registries from the responsible party(ies). ATSDR assigns unique names and identification numbers to all Superfund sites; tracks the costs incurred by ATSDR, both direct and indirect, for those sites; and prepares cost recovery packages.

3. 3 The Battle Creek health study (12) was a retrospective cohort study that included two groups: a cohort of people exposed to VOCs and a reference cohort of very low exposure or unexposed people. The exposed cohort included current and former residents of homes with contaminated wells in the areas known as Verona Park, Springfield, and Dowagiac.