• Home
• About ATSDR
• Press Room
• A-Z Index
• Glossary
• Employment
• Contact Us
• CDC

ATSDR en Español

Search:

Skip directly to: content | left navigation | search

---

Find a Publication

1. ATSDR > Public Health Assessments & Consultations

PUBLIC HEALTH ASSESSMENT

SANGAMO/TWELVE-MILE CREEK/HARTWELL PCB
PICKENS, PICKENS COUNTY, SOUTH CAROLINA

APPENDIX A. Figures

Figure 1. Sangamo Plant Site

Figure 2. Diagram of the interconnections
and water movement between local, municipal
drinking water supplies

Figure 3. Lake Hartwell Sample Stations

APPENDIX B. Tables

Table 1

Demographics
South Carolina and Anderson, Oconee, and Pickens Counties

DEMOGRAPHICS	SOUTH CAROLINA	ANDERSON	OCONEE	PICKENS
1989 POPULATIONa	3,518,756	144,892	56,892	91,316
% WHITE	68.2	82.2	89.4	94.3
% BLACK AND OTHER	31.8	17.8	10.6	6.6
% MALE	48.8	47.7	49.8	50.4
% FEMALE	51.2	52.3	50.2	49.6
% UNDER 10 YEARS	15.1	14.1	13.6	11.8
% OVER 65 YEARS	11.1	13.1	12.3	10.5
% BELOW POVERTY LEVELb	16.6	12.5	14.0	10.7

(a) Revised Population Projections for July 1989, for Division of Research and Statistical Services of the South Carolina Budget and Control Board.

(b) From 1980 census of Population, General Social & Economic Characteristics, South Carolina

TABLE 2

Maximum PCB concentrations (as Aroclors), all congeners, detected in Public Water Supplies by SCDHEC Monitoring Program from 1983-1991

I. ROUTINE SAMPLING POINTS, 1983-1991

SAMPLING POINT	PUBLIC WATER SUPPLY	MAX CONC ug/L (as Aroclors)	DATE	COMPARISON VALUE
				ug/L	SOURCE
RAW WATER	EASLEY-CENTRAL # I	0.209	7/8/87	0.005	CREG
FINISHED WATER	EASLEY-CENTRAL # I	0.120	5/6/87	0.005	CREG
F.H. AT CLARK'S FURN.	EASLEY-CENTRAL # I	0.820 (a)	9/5/90	0.005	CREG
SHELL STATION	EASLEY-CENTRAL # I	0.141	5/6/87	0.005	CREG
PLEZ-U-STORE	TOWN OF CENTRAL	0.203	8/4/89	0.005	CREG
F.H. ON NORTH STREET	EASLEY-CENTRAL # I	3.000 (b)	9/5/90	0.005	CREG
F.H. ON 22 MAIN STREET	EASLEY-CENTRAL # I	1.500 (c)	9/5/90	0.005	CREG
F.H. AT RIVERSIDE PLANT	EASLEY-CENTRAL # I	1.700 (d)	9/5/90	0.005	CREG

_{(a) Pre-flush sample; PCBs were below levels of detection (<0.1) in Post flush Sample.
(b) Post-flush sample; PCBs were <0.1 ug/L in pre-flush sample.
(c) Pre-flush sample; PCBs were detected at a level of 0.75 ppb in post-flush sample.
(d) Post-flush sample; PCBs were detected at a level of 1.1 ppb in pre-flush samples}

II. ADDED SAMPLING POINTS FOR JUNE 1991 ONLY

SAMPLING POINT	PUBLIC WATER SUPPLY	MAX CONC ug/L	DATE	COMPARISON VALUE
				ug/L	SOURCE
F.H. AT CENTRAL & HOWE RD.	HWY 88	0.449 (e)	6/5/91	0.005	CREG
F.H. AT 122 FABRICA ST.	E. CLEMSON	0.106 (f)	6/5/91	0.005	CREG
F.H. AT CLAYTON & WESLEYAN DR.	EASLEY-CENTRAL # I	0.245 (g)	6/5/91	0.005	CREG
F.H. AT MOORE & TARRANT ST.	TOWN OF CENTRAL	0.881 (h)	6/5/91	0.005	CREG
RESIDENCE A	EASLEY-CENTRAL # I	< 0.100	6/5/91	0.005	CREG
RESIDENCE B	EASLEY-CENTRAL # I	0.263	6/5/91	0.005	CREG
RESIDENCE C	EAST CLEMSON	< 0.100	6/5/91	0.005	CREG
BP STATION	HWY 88	0.101	6/5/91	0.005	CREG

_{(e) Pre-flush sample; PCBs were detected at a level of 0.113 ppb in post-flush sample.
(f) Pre flush sample; PCBs were <0.1 ug/L in post-flush sample.
(g) Pre-flush sample; PCBs were unchanged in the post-flush sample.
(h) Pre-flush sample; PCBs were <0.1 ug/L in post-flush sample.}

TABLE 3

PCB Concentrations (as Aroclors) in Drinking Water Collected from Private Residences Within 10 Water Supply Districts, September - November, 1991.

I. PCB Concentrations (as Aroclors) in Drinking Water Samples Collected from Private Residences by SCDHEC^a

WATER SYSTEM	PCB CONCENTRATIONS ug/L		DATE	COMPARISON VALUE
	# Samples	Range		ug/L	SOURCE
Easley - Central # I	4	< 0.1 - 0.152b	9/17-11/6/91	0.005	CREG
Easley - Central # II	1	< 0.1	9/16/91	0.005	CREG
Highway 88	4	< 0.1 - 0.106	9/16 - 11/16/91	0.005	CREG
Liberty	1	< 0.1	9/16/91	0.005	CREG
Town of Central	3	< 0.1	9/16-11/6/91	0.005	CREG
East Clemson	1	< 0.1	9/16/91	0.005	CREG

^{a Level of Laboratory Detection = 0.1 ug/L.
b Sample was split with EPA; EPA detected PCBs at a concentration of 0.16 ug/L.}

II. PCB Concentrations (as Aroclors) in Drinking Water Samples Collected from Private Residences by EPA^c

WATER SYSTEM	PCB CONCENTRATIONS ug/L		DATE	COMPARISON VALUE
	# Samples	Range		ug/L	SOURCE
Easley - Central # I	6d	0.048 - 0.294	9/17/91	0.005	CREG
Easley - Central # II	2	< 0.05	9/17/91	0.005	CREG
Highway 88	3	0.052 - 0.084	9/17/91	0.005	CREG
Liberty	3	< 0.05	9/17/91	0.005	CREG
Town of Central	4	0.06 - 0.22	9/17/91	0.005	CREG
East Clemson	3	0.052 - 0.06	9/17/91	0.005	CREG

^{c. Level of Laboratory Detection = 0.05 - 0.2 ug/L.
d. One sample was split with SCDHEC; SCDHEC detected PCBs at a concentration of 0.152 ug/L.}

TABLE 4

PCB concentrations, all congeners, in Sediments from Twelve Mile Creek, Eighteen Mile Creek, Andersonville Island, Tugaloo River and Lake Hartwell Dam.

Sediment samples in mg/kg (ppm)

YEAR	SV 107 (a)	SV 137 (a)	SV 015 (a)	SV 282 (a)	SV 532 (b)	SV 535 (c)	SV 641 (d)	SV 642 (e)	COMPARISON VALUE
									mg/kg	SOURCE
1976	1.22			< 0.01					0.09	CREG
1977	0.84			0.14					0.09	CREG
1980	0.32			0.19					0.09	CREG
1983	0.94	35.80	0.04	0.14					0.09	CREG
1984	0.90			0.20					0.09	CREG
1986	64.30	0.05	< 0.01	< 0.01	0.58	2.21	< 0.01	1.20	0.09	CREG
1986	4.06								0.09	CREG
1987	8.86	0.05	< 0.01	0.07					0.09	CREG
1987	4.76			0.05					0.09	CREG
1988	1.45	0.03	0.05	< 0.01					0.09	CREG
1989	3.16	0.31	< 0.01	0.12					0.09	CREG
1990	2.91	0.09	0.01						0.09	CREG

(a) Sampling Station located in Twelve Mile Creek.
(b) Sampling Station located in Eighteen Mile Creek.
(c) Sampling Station located in Lake Hartwell at Andersonville Island.
(d) Sampling Station located in Tugaloo River.
(e) Sampling Station located in Lake Hartwell in vicinity of the Lake Hartwell Dam.

TABLE 5

PCB, Dibenzodioxin, Dibenzofuran concentrations, all congeners, in Sediments
from Twelve Mile Creek, 1985 sampling data (1)

I. Dibenzodioxin and Dibenzofuran Congeners (2)

COMPOUND	MAXIMUM CONCENTRATION	DATE	COMPARISON VALUE
			ng/kg	SOURCE
HEXACHLORODIBENZODIOXIN	0	8/6/85	700	EMEG
HEPTACHLORODIBENZODIOXIN	0	8/6/85	700	EMEG
OCTACHLORODIBENZODIOXIN	0.007100	8/6/85	700	EMEG
2,3,7,8 TCDF (DIBENZOFURAN)	0.004800	8/6/85	700	EMEG
TETRACHLORODIBENZOFURAN	0	8/6/85	700	EMEG
HEXACHLORODIBENZOFURAN	0	8/6/85	700	EMEG
HEPTACHLORODIBENZOFURAN	0	8/6/85	700	EMEG
OCTACHLORODIBENZOFURAN	0.000085	8/6/85	700	EMEG

(1) All data have been converted into TCDD Equivalents using the International Toxicity Equivalent factors, 1988.
(2) Concentrations for dibenzodioxins & dibenzofurans are expressed in ng/kg (ppt).

II. PCBs

COMPOUND	MAXIMUM CONCENTRATION (mg/kg or ppm)	DATE	COMPARISON VALUE
			mg/kg	SOURCE
PCB-1242	35.0	8/6/85	0.09	CREG
PCB-1254	13.0	8/6/85	0.09	CREG
PCBs (TOTAL)	48.0	8/6/85	0.09	CREG

TABLE 6

Mean PCB Concentrations, all congeners, in fish (1) from Twelve Mile Creek, Eighteen Mile Creek, Andersonville Island, Tugaloo River, and Lake Hartwell Dam, from SCDHEC Monitoring data, 1977 - 1991

SPECIES: WHITE / CHANNEL CATFISH (2)         PCB Fish Tissue concentration (mg/kg)

YEAR	SV-107	SV-106	SV-532	SV-535	SV-641	SV-642
1977	58.8		8.3	0.8
1978	23.2		7.1	0.4
1979	8.4		6.5	0.4
1980	7.5		2.4	0.4
1981	3		0.8	0.5
1984		2.71
1985	29.1
1986	26.1		6.5	1.8	0.8	2.1
1987	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1988	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1989			10.5
1990	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING
1991	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING

SPECIES: LARGEMOUTH BASS (3)         PCB FISH TISSUE CONCENTRATION (MG/KG)

YEAR	SV-107	SV-106	SV-532	SV-535	SV-641	SV-642
1977	8.5		1.3	0.6
1978	23.3		1.2	0.6
1979	7		2.6	0.7
1980	9		0.7	0.3
1981	8.7		0.9	0.2
1983		0.67	0.96
1984		1.24
1985	24.1
1986	34.1		4.2	0.6	0.2	0.7
1987	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1988	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1989	23.1		3.6	0.8
1990	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING
1991	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING

SPECIES: WHITE / HYBRID BASS (4)         PCB FISH TISSUE CONCENTRATION (mg/kg)

YEAR	SV-107	SV-106	SV-532	SV-535	SV-641	SV-642
1977	10		5.3	6.1
1978	7.5		6.1	2.7
1979	7.5		4.3	2.2
1980	11.7		3.9	2.3
1981	4.5		2.6	0.9
1983			3.4
1984		2.44
1985			3
1986	4.9		5.5	5.5	2.9	2.5
1987	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1988	DELETED	DELETED	DELETED	DELETED	DELETED	DELETED
1989	3.2		6.7		4.1
1990	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING
1991	PENDING	PENDING	PENDING	PENDING	PENDING	PENDING

SPECIES: BLACK CRAPPIE (5)         PCB fish tissue concentration (mg/kg

YEAR	SV-107	SV-106	SV-532	SV-535	SV-641	SV-642
1986	8.9		2.2		0.1	2.5
1989			3.7

(1) Fish Samples from 1977-1981 were fillet composites from 5 fish. Fish samples from 1983-1985 were both fillets of individual fish (n 3-10 fish/station) and fillet composites from 5 fish. Samples from 1986-1089 were individual whole fish (~4-10 fish/station) that were only eviscerated and deheaded; catfish were also skinned, all others were not.

(2) Weight ranges, catfish composites: 0.6-0.9 lbs (1977-1981), 0.78-5.13 lbs (1983-1985); Weight range, individual fish: 0.75-5.24 lbs (1983-1985), 1.8-3.7 lbs (1986), 2.7-3.2 lbs (1989).

(3) Weight range, Largemouth composites: 0.6-2 lbs (1977-1981), 0.96-4.9 lbs (1983-1985); Weight range, individual fish: 0.75-5.24 lbs (1983-1985, 1.8-3.7 lbs (1986), 2.7-3.2 lbs (1989).

(4) Weight range, White /Hybrid composites: 1-2.3 lbs (1977-1981), 0.91-4.13 lbs (1983-1985); Weight range, individual fish: 2.07-4.88 lbs (1983-1985), 3.1-4.8 lbs (1986), 3.1-4.6 lbs (1989).

(5) Weight range, Black Crappie, individual fish: 0.4-2.01 lbs (1986) 0.44-0.99 lbs (1989).

Table 7.

PCBs, dibenzodioxin, dibenzofran concentrations, all congeners, in Fish from Twelve Mile Creek, Eighteen Mile Creek, Lake Hartwell, and Tugaloo River, 1985 EPA Sampling data:

LOCATION	STATION	TYPE FISH	TOTAL PCDDs	2,3,7,8 TCDF	OTHER PCDFs	TOTAL	PCB- 1242	PCB- 1254	PCB- TOTAL
18 MI CREEK	12.1	HYBRID	<20	4.6	<0.02	4.6	0.56	1.6	2.11
18 MI CREEK	12.2	HYBRID	<10	3.4	<O.01	3.4	0.4	1.4	1.8
18 MI CREEK	12.3	HYBRID	<20	9.9	<O.01	9.9	1.8	3.2	5
12 MI CREEK	12.4	HYBRID	<10	29	<0.01	29	46	51	97
12 MI CREEK	12.5	CATFISH	<10	<1.0	<0.01	<10	4.1	23	27.1
12 MI CREEK	12.6	CATFISH	<10	<1.0	<0.01	<10	2.2	1.6	3.8
12 MI CREEK	12.7	LARGE MOUTH	<10	2.4	<0.01	2.4	5.7	10	15.7
12 MI CREEK	12.8	LARGE MOUTH	0.13	3.5	<0.01	3.5	4.6	13	17.6
12 MI CREEK	12.9	LARGE MOUTH	<10	1	<0.01	1	7.7	8.1	15.8
12 MI CREEK	12.10	LARGE MOUTH	<10	6.3	<0.01	6.3	28	34	62
12 MI CREEK	12.11	LARGE MOUTH	<10	2.8	<0.01	2.8	6.9	11	17.9
12 MI CREEK	12.12	LARGE MOUTH	<10	7.4	<0.01	7.4	7.8	8	15.8
12 MI CREEK	12.13	CATFISH	<10	<1.0	<0.01	<10	4.2	11	15.2
12 MI CREEK	12.14	CATFISH	<20	<1.0	<0.01	<20	18	50	68
12 MI CREEK	12.15	NOT LISTED	<10	<1.0	<0.01	<10	1.8	4.4	6.2
LAKE HARTWELL	HD-1	FISH COMP.	<10	22	15.5	37.5	37	110	147
TUGALOO RIVER	COMP 1	HYBRID COMPOSITE	<10	13	<0.01	13	0.5	4.6	5.8
TUGALOO RIVER	COMP 2	HYBRID COMPOSITE	<10	6	<0.01	6	0.77	4.9	5.67
TUGALOO RIVER	COMP 5	HYBRID COMPOSITE	<20	8.5	<0.01	8.5	1.4	5.6	7
TUGALOO RIVER	C 23	HYBRID COMPOSITE	<10	9.2	<0.01	9.2	0.9	3.2	4.1

All PCDD/PCDF data are converted into TCDD equivalents using the International Equivalent Factors/'88.
Concentrations are expresses in ng/kg or ppt.

Laboratory level of detection < 10 ppt.

Concentrations of PCBs are expressed in mg/kg or ppm.

TABLE 8.

      COMPLETED EXPOSURE PATHWAYS

PATHWAY	TIME	SOURCE	MEDIA & TRANSPORT	POINT OF EXPOSURE	ROUTE OF EXPOSURE	EXPOSED POPULATION
FISH	PAST PRESENT FUTURE	SANGAMO WESTON, INC.	FISH	RESIDENCE	INGESTION	LAKE HARTWELL FISH CONSUMERS
MUNICIPAL DRINKING WATER	PAST PRESENT FUTURE	SANGAMO WESTON, INC.	MUNICIPAL DRINKING WATER	RESIDENTIAL TAPS	INGESTION	RESIDENTIAL CUSTOMERS
SEDIMENTS	PAST PRESENT FUTURE	SANGAMO WESTON, INC.	SEDIMENT	SWIMMING & WADING	DERMAL CONTACT	TWELVE MILE CREEK BATHERS

      POTENTIAL EXPOSURE PATHWAYS

PATHWAY	TIME	SOURCE	MEDIA & TRANSPORT	POINT OF EXPOSURE	ROUTE OF EXPOSURE	EXPOSED POPULATION
SURFACE SOIL	PAST PRESENT FUTURE	SANGAMO WESTON, INC.	SURFACE SOIL	SWIMMING & WADING	INGESTION & DERMAL CONTACT	TWELVE MILE CREEK BATHERS

APPENDIX C.

SCDHEC's FISH MONITORING PROGRAM FOR TWELVE MILE CREEK/LAKE
HARTWELL

The following text discusses the various factors and methodological issues that have had a direct effect on the data generated by this program.

Every April, from 1977 to 1985, SCDHEC collected 2-to-3 year old largemouth bass, white catfish, and white/hybrid bass from stations in Twelve Mile Creek (SV-107), in Eighteen Mile Creek (SV-532) and at Andersonville and (SV-535) (See Figure 3). SCDHEC selected these fish species for study because of their association with sport and/or commercial fisheries and because of their varying migratory patterns and habitat selections. SCDHEC later added Black Crappie to the monitoring program when it was found to be an important Lake Hartwell gamefish.

Largemouth bass are widely distributed predators and sport species that undergo seasonal movements from shallow to deep water and have open-water schooling tendencies in late summer and fall. SCDHEC staff do not believe that the largemouth bass have been moving significantly outside their home ranges. As well as being a highly sought gamefish, the hybrid bass are a migratory fish that have the potential to move in and out of the contaminated areas of Lake Hartwell. They are generally schooling open-water fish with a tendency to concentrate in tributary arms and streams during spawning periods. As a result, the data from samples of Hybrid bass have been marked by much variability. White catfish are bottom-dwelling omnivores of sport and commercial value.

In 1986, the sampling program was modified to include 2 additional sampling stations: SV-641 (in the Tugaloo River) and SV-642 (in vicinity of Lake Hartwell Dam). The Tugaloo River is the other riverine embayment of the lake that was not subject to direct input of PCBs.

1. Sampling Periods 1977-1981:

For the sampling years 1977-1981, only fish fillets (that is, the edible tissue) were analyzed for PCBs. Samples were usually composites of 5, similarly-sized fish of the same species; however, composites occasionally ranged from 2-7 fish, depending on availability. Composite samples were used to reduce the cost of tissue analyses (SCDHEC, 1982).

The study protocol during these years occasionally had to be modified when target species were not available at sampling stations; substitutions with other more abundant species were made although the size ranges had to be adjusted. White bass and hybrid bass were included in the same group. Although white bass were the initial target species, hybrids were also inadvertently collected since their similar body features made them difficult to tell apart. During some collections, white bass were unavailable and were substituted with the hybrids. However, since they are considered to be genetically related and have similar feeding and migratory habits, their bioaccumulation rates were assumed to be the same (SCDHEC, 1982).

Twelve Mile Creek (SV-107): Fish in Twelve Mile Creek tend to have the highest PCB concentrations. Between the years 1977-1981, there has been a statistically significant downward trend in the PCBs concentration of white catfish. No downward trend was found for the largemouth and white/hybrid bass except for 1 year, between 1980-1981, for white/hybrid samples.

Eighteen Mile Creek (SV-532): As was noted in SV-107, there has been a significant downward trend in the PCB concentration of white catfish; however, it did not occur to the same degree. This difference was noted particularly between 1977 and 1980. There was also a significant difference in tissue concentrations between the years 1980 and 1981. There was no significant trend noted in the data for white/hybrid and largemouth bass.

Andersonville Island (SV-535): A downward trend was noted in the PCB concentrations of white/hybrid bass. However, this difference was only found to be significant when data from 1977 was compared to the data from the remaining years. Data from years 1978 to 1981 showed no significant difference and no apparent downward trend since 1978. Data on white catfish and largemouth bass show only moderate declines over the sampling years (SCDHEC, 1982).

2. Sampling Periods 1983-1985:

During these years, hybrid bass, largemouth bass, and channel/white catfish were collected from Twelve Mile Creek, Eighteen Mile Creek and from the Martin Creek area (SV-106), approximately halfway between these stations. No samples were collected at the Andersonville Island station. All 3 species were not collected at each sampling station. Both individual fish and composite samples of 5 fish of a similar weight class were assayed for PCBs. From the sampling results, it appears that there was a sizeable increase in the PCB concentrations of catfish and largemouth bass taken from Twelve Mile Creek. There was no appreciable decline in the PCB concentrations in catfish taken from Martin Creek although there was a significant decline in the PCB concentrations in largemouth and hybrid bass from this Creek. Based on these results, SCDHEC and EPA reaffirmed the fishing advisory.

3. Sampling Period 1986:

The SCDHEC sampling protocol was substantially modified in 1986. Besides adding 2 sampling stations and a fourth fish species to the monitoring program, the method of preparing fish samples for analyses was markedly changed.

Black crappie, largemouth bass and hybrid bass were selected for the 1986 study because they are important Lake Hartwell gamefish. According to the 1982 SCWMRD creel survey, they represented nearly 40, 24, and 23 percent, respectively, of the public harvest in Lake Hartwell (catfish represented only 1.76% of the public harvest). Thus, they could be used to develop a reasonable, worst-case scenario for dose estimates. This creel survey did not gather data on sampling location; thus, we do not know whether the survey respondents represented bank fishermen or fishermen using boats.

Reviewing past sampling data from SV-107, SCDHEC staff noted much variability in the PCBs levels among fish of the same weight class and species. To reduce this variability and to allow for more accurate trend analyses, individual fish were de-headed and eviscerated; no areas of fatty tissue were trimmed except from the body cavity. This technique is referred to as a modified whole fish preparation. This method contrasts with the fillet preparation (the edible portion of the fish trimmed of all fatty tissues) which was used prior to 1986. Except for catfish, all the fish were analyzed with the skin on. SCDHEC staff used the modified whole fish preparation as a means of limiting the variability in data believed to have resulted because of individual filleting techniques. They also believed some people may be preparing fish in a similar manner.

Despite this change in technique, they found the same level of variability in the data. SCDHEC staff now believe that this can primarily be attributed to the variability in the lipid content in fish of the same weight class and species. Other contributing factors may be the variability in the contaminant levels of Twelve Mile Creek sediments and the migratory and feeding patterns of each fish species.

Until this point in time, PCB levels in Twelve Mile Creek fish had fluctuated widely while PCB levels in fish from other locations had remained fairly constant. When the 1986 data is compared to the data from prior years, it appears that the fish are becoming more contaminated. SCDHEC staff mainly attribute this perceived increase to the change in sample preparation because they believe the lipid content of modified whole fish samples is greater than that of fillet samples. Since PCBs selectively accumulate in fat tissues, the samples' PCB concentrations should also be greater. They could not conclusively state this since they did not have the resources to analyze similarly-sized samples by both methods. However, the modified sampling method is not the only possible cause for the increased PCB levels in these fish. First, these fish had similar levels in 1985 before the sampling procedure was changed. Second, the PCB concentrations in the sediments of Twelve Mile Creek collected in 1985 and 1986 were found to be much greater than the sediment data from previous years. Thus, the PCB concentrations in these fish more likely reflect the higher PCB concentrations in these sediments. The cause of the increase in these sediments is not known.

White/channel catfish tended to have higher concentrations of PCBs; they are bottom feeders and can ingest PCBs in sediments. For analysis, white and channel catfish were grouped in 1 category. This was done for 2 reasons: it had been done in earlier sampling rounds and there was some difficulty collecting sufficient numbers of either species at all sampling stations. The channel catfish were readily collected at the riverine stations while the white catfish were readily collected at the open-water stations. The data were still able to be used for trend monitoring since the ratio of each species of catfish was similar between stations and it was believed that both species were equally consumed by the public (the 1982 creel survey did not speciate the catfish).

All fish collected from the Seneca River arm of the lake, north of the SC Highway 24 advisory boundary, had mean PCBs values above the current FDA limit. At the remaining 3 stations, only hybrid bass and catfish had mean PCBs levels greater than this level.

4. Sampling periods 1987-1988:

There are no data available for review. SCDHEC contracted with a private laboratory to analyze fish samples collected during these years. However, the sampling results from this laboratory were found to be considerably out of line with prior sampling data. SCDHEC laboratory staff subsequently found serious problems with this laboratory's quality assurance/quality control practices and procedures. Thus, SCDHEC rejected the data.

5. Sampling period 1989:

A total of 74 fish were collected by SCDHEC from stations in Twelve Mile Creek, Eighteen Mile Creek, Tugaloo River and at Andersonville Island. Largemouth bass, black crappie, hybrid bass and 4 channel catfish of similar weight classes were collected at the Eighteen Mile Creek Station. Financial constraints prevented them from collecting all 4 species at the other stations. All fish were prepared for PCB analyses using the modified whole fish technique. All catfish were skinned; the other 3 species were analyzed with the skin on.

As seen in the data from 1986, fish within the advisory exceeded FDA's limit. There was a noticeable decrease in the PCB concentrations of the largemouth bass from Twelve Mile Creek between the years 1986 and 1989 (data from 1987 and 1988 was deleted as mentioned above). A less pronounced decrease was noted in the largemouth bass samples taken from the Eighteen Mile Creek. There was an increase in the PCB concentrations in fish samples collected near Andersonville Island.

There was an increase in the PCB concentrations of the hybrid samples taken from Eighteen Mile Creek; there was also a similar increase in the concentrations of the fish collected from the Tugaloo River. Not only did these fish exceed the FDA limit, but they were collected outside of the advisory area. Yet, the PCB concentrations in hybrid samples taken from Twelve Mile Creek had decreased. Samples of black crappie and channel catfish caught in Eighteen Mile Creek also seemed to have increased in PCB concentrations when compared to the sampling results of 1986. Additional sampling results for later years is needed to further evaluate this trend.

6. Sampling period 1990-1991:

No sampling data from these years were available for review. The SCDHEC laboratory is completing analysis of the 1990 fish data. Fish collected during 1991 will be analyzed by a contract laboratory funded by the federal Superfund Program.

APPENDIX D

Response to Comments received during the Public Comment Period

The following responses address technical comments submitted by the public during the public comment period of September 21, 1992 to October 23, 1992. Comments received about the report's format, site hazard ranking, and the need for a site visit will not be addressed as they were prepared in accordance with ATSDR's Public Health Guidance Manual, March, 1992.

1. Concerns were raised about the assumptions chosen for the exposure scenario for subsistence fishermen. One reviewer considered the assumptions to be overly conservative while another reviewer believed that fishermen were not following the advisory and were collecting fish from areas covered by the advisory. These uncertainties were discussed under 2 headings in the Public Health Implications Section: 1. Polychlorinated biphenyls: a. Food Chain: Fish (Page 22) and 3. Uncertainty (Page 28). The reviewers are encouraged to review these discussions for further clarification.

The risk to the fish-consuming public is best characterized if actual exposure data were available. Since these data did not exist, assumptions were made to estimate the exposure. While the range of assumptions that can be used is broad, the assumptions that are typically chosen account for individuals with the greatest exposures (in this case, subsistence fishermen). This also accounts for risks to the more sensitive members of the population (that is, children, the elderly, the debilitated and the chronically ill). One reviewer considered the use of the maximally exposed individual (MEI) for estimating risk to the fish-consuming public to be overly conservative. In the scenario presented in the Public Health Assessment, the MEI was not actually considered; for this assumption to be made, we would have to assume that subsistence fishermen are consuming maximally-contaminated fish from the area under that advisory. (Contrary to the opinion of this same reviewer, there have been anecdotal reports that fishermen are collecting fish from Twelve Mile Creek, in areas covered by the advisory). The exposure scenario that seemed more plausible was to consider subsistence fishermen who are following the advisory. If the risk from exposure fell within an acceptable range as defined in the National Contingency Plan, then we would have considered an MEI exposure scenario. We did not need to go further since we did not find this to be the case.

A question was raised about the choice of the consumption rate for local fishermen. Again, without data on local consumption patterns, we could only rely on an assumption. In this case, we chose the annual consumption rate (90th percentile) for recreational fishermen from the U.S. west coast. This value is cited in EPA's, Exposure Factor Handbook, July, 1989. It is recommended that these values be used to represent consumption rates for recreational fishermen in any area where a large water body is present and widespread contamination is evident. The 90th percentile rate was chosen because we believed subsistence fishermen could be consuming fish at that rate. We also assumed that the individual's fish diet consisted solely of Lake Hartwell fish. We certainly agree that the risk of exposure is directly related to the amount of contaminated fish consumed per day. Thus, a lower consumption rate would result in a lower estimate of risk. However, without site-specific consumption data from surveys of local fishermen, another consumption rate would merely be arbitrarily selected. Certainly, these risk estimates will be revised when the SCWMRD completes their survey of the actual consumption patterns of Lake Hartwell fishermen.

2. A concern was raised about the assumption that "the effects from Aroclor 1260 is assumed to be representative of all PCB mixtures;" this assumption was used in calculating cancer risk. The objection to the use of this assumption was that it was an "outdated premise." Several attachments were also provided for reference.

We are well aware of the recent efforts to revisit the cancer potency of PCBs; it follows on the heels of EPA's reassessment of dioxins. However, in the case of dioxins, the EPA has withdrawn its cancer potency factor from their Integrated Resource Information System (IRIS) database until they can finalize their opinion.

Such is not the case, however, for PCBs. The cancer potency factor for PCBs was calculated from data generated by a study of laboratory rats fed Aroclor 1260. In developing this cancer potency factor, the EPA assumed that Aroclor 1260 was representative of all mixtures and that the potency estimate for Aroclor 1260 applies to all PCB mixtures (ATSDR Tox. Profile for Selected PCBs (Aroclor-1260, -1254, -1248, -1242, -1232, -1221, and -1016). This opinion has not yet changed, nor has the EPA begun efforts to reassess PCB toxicity.

We agree that the relative toxicities of PCBs mixtures may vary because of their varying levels of chlorination. However, there is of yet insufficient data to adequately characterize these differences. We agree with the comment that risk calculations should be congener specific; however, the EPA has not yet adopted a weighted approach similar to the toxicity equivalent factor approach it developed for dealing with dioxin congeners (this has been referenced in the Public Health Assessment). We are aware that equivalency factors for polyaromatic hydrocarbon compounds (PAHs) and for PCBs can be found. However, the EPA has not yet recognized their validity and has not recommended their use. There is no indication that this will occur in the near future.

3. There was a concern raised about the use of present available data to make assumptions about cancer risk to consumers of municipal drinking water. The main objection was to the characterization of cancer risk for PCB levels in drinking water that were "nondetectable or below health based standards." Certainly, as sampling has shown, the PCB levels in drinking water have fluctuated widely from tap to tap and at each individual tap. Thus, assuming a PCB level will remain the same for a lifetime (or 70 years) becomes highly unlikely and adds to the level of conservancy in the risk estimate. However, in the scenario presented, the maximum concentration level from a private tap was chosen to estimate the maximum risk to consumers. Given the erratic and unpredictable nature of the sampling data, this choice was a reasonable approach. Additional sampling results will better characterize concentration levels at private taps and help to reduce the level of uncertainty in the risk estimate. Even though these water supplies will continue to be monitored, exact characterization of contaminant concentrations will be quite difficult and costly.

We are aware that some sample results were below detectable levels; however, for the samples where PCBs were identified, there are no "health based standards" which could be used for comparison. The EPA has developed an MCL for PCBs as decachlorobiphenyls in drinking water; however, this standard cannot be compared to the data results because they are reported as Aroclors. There is no acceptable methodology yet to measure PCB concentrations as decachlorobiphenyls nor is there an approved method to express Aroclors relative to the toxicity of decachlorobiphenyls. In addition, the MCL is not necessarily a health-based standard; it is a legally enforceable standard that takes into account available analytical methods of detection and water treatment capabilities as well as the cost of achieving various contration levels of contaminants. We do not know what other "standard" the reviewer is citing.

4. There was a concern that contamination of surface soils along the shoreline should not be considered as a possible health threat as it "represents an unsubstantiated conclusion." We concur with the reviewer that risk of exposure is indeed unsubstantiated and will remain so unless additional surface soil samples are collected from locations frequented by the public. We realize that studies have addressed the sediment of the creek bottoms. However, we did not find any sampling results from the accessible shorelines where people can readily fish, wade, and swim. Additional samples from these locations will help us determine if this pathway is a complete one or if it can be eliminated.

5. There was a concern raised about the accuracy of the TRI data. For the Public Health Assessment, we simply cite the available TRI data; we do not critically review it for accuracy. The TRI database depends on self-reporting by each industry. It has been found to have numerous errors in the past and will continue to do so unless there is some effort to validate the data that is submitted. Presently, such a review process is not a requirement of TRI nor are there plans to do so in the future.

6. The comment about the presence of Aroclor 1260 (not used by Sangamo) in some drinking water sampling results suggests the possibility of other contaminant sources. The possibility that PCBs could be emanating in significant amounts from abandoned electrical transformers and capacitors that remained in areas that were flooded by the U.S. Army Corps of Engineers seems quite unlikely. Firstly, the predominant congener found has historically been Aroclor 1242, with Aroclors 1254 and 1016 found in lesser amounts. The incidental finding of Aroclor 1260 in several drinking water samples may point to contribution from an additional source but it is dwarfed in importance by the contribution of PCBs from Sangamo. Secondly, since 1976, Sangamo has been considered to be the source of PCB contamination in fish, sediments, and drinking water. Information about any other actual sources of PCB contamination to this Lake system should be brought to the attention of the EPA project manager for this site.

7. One reviewer raised issue about the inclusion of the words "Twelve Mile Creek/Lake Hartwell watershed" into the OU-2 designation used by EPA. They considered it misleading to consider this situation a watershed problem. We concur with the reviewer that the situation should not include the watershed but involves PCB contamination of the sediments and fish of this Lake System as well as the municipal drinking water systems that receive raw water from it. We feel that this has been explained in detail in the Public Health Assessment. Since the EPA has artificially separated this site into 2 operable units and officially used the term "watershed" in its description of the OU-2 site, its use in the Public Health Assessment was merely for background information. For further clarification, I would refer the reviewer to the May, 1991 Superfund Fact Sheet for this site.

Table of Contents

Agency for Toxic Substances and Disease Registry, 1825 Century Blvd, Atlanta, GA 30345
Contact CDC: 800-232-4636 / TTY: 888-232-6348 

Department of Health and Human Services