PB94-964012
EPA/ROD/R04-94/178
July 1994
EPA Superfund
Record of Decision:
Sangamo Weston/Twelvemile Creek/
Lake Hartwell Site, Pickens, GA
6/28/1994
-------
FINAL RECORD OF DECISION
for the
Sangamo Weston/Twelvemile Creek/Lake Hartwell
PCB Contamination Superfund Site - Operable Unit Two
Pickens, Pickens County, South Carolina
June 1994
-------
Record of Decision Declaration
SITE NAME AND LOCATION
Sangamo Weston/Twelvemile Creek/Lake Hartwell PCB Contamination
Superfund Site - Operable Unit Two
Pickens, Pickens County, South Carolina
STATEMENT OF BASIS AND PURPOSE
This decision document presents the selected remedial action
for the Sangamo Weston/Twelvemile Creek/Lake Hartwell PCB
Contamination Superfund Site - Operable Unit Two ("Sangamo OU2")
in Pickens County, South Carolina, which was chosen in accordance
with CERCLA, as amended by SARA, and, to the extent practicable,
the National Contingency Plan. This decision is based on the
Administrative Record for the Site.
The State of South Carolina concurs with the selected
remedy.
ASSESSMENT OF THE SITE
Actual or threatened releases of hazardous substances from
this Site, if not addressed by implementing the response action
selected in this ROD, may present an imminent and substantial
endangerment to public health, welfare, or the environment.
DESCRIPTION OF THE REMEDY
This Operable Unit is the final action of two Operable Units
for the Site. Operable Unit One addressed the land-based source
areas which included the Sangamo Weston Plant and six satellite
disposal areas. This action, Operable Unit Two, addresses the
sediment, surface water, and biological migration pathways
downstream from the source areas. This action addresses the
primary PCB human exposure pathway by mitigating the consumption
of PCB contaminated fish harvested from Lake Hartwell.
The major components of the selected remedy include:
Continuation of the existing fish consumption advisory on
Lake Hartwell;
Public education program designed to increase awareness of
advisory and methods to prepare/cook fish to reduce quantity
of contaminants consumed;
Continued monitoring of aquatic biota and sediment to
support continuance and/or justify modifications to existing
fish advisory; and
-------
Regulation of Twelvemile Creek Impoundments that will
facilitate burial of contaminated sediment while mitigating
adverse impacts to Lake Hartwell water quality.
STATUTORY DETERMINATIONS
The selected remedy is protective of human health and the
environment, complies with federal and state requirements that
are legally applicable or relevant and appropriate to the
remedial action, and is cost-effective. This remedy utilizes
permanent solutions and alternative treatment technologies to the
maximum extent practicable. This remedy does not satisfy the
preference for treatment as a principal element based upon the
excessive costs associated with removal, treatment and disposal
of approximately 4.7 million cubic yards of PCB contaminated
sediment spread out over approximately 730 acres.
Because this remedy will result in hazardous substances
remaining onsite above health-based levels, a review will be
conducted no less often than every five years after commencement
of remedial action to ensure that the remedy continues to provide
adequate protection of human health and the environment.
John H. Hankinson, Jr. Date
Regional Administrator
EPA - Region IV
-------
TABLE OF CONTENTS
SECTION PAGE
1.0 Site Name, Location, and Description 1
2.0 Site History and Enforcement Activities 1
3 .0 Highlights of Community Participation 6
4.0 Scope and Role of Response Action 7
5.0 Summary of Site Characteristics 8
5.1 Summary of Physical Characteristics 8
5.1.1 Surface Features 8
5.1.2 Surface Water Hydrology 10
5.1.2.1 Surface Water Hydrology of Twelvemile
Creek Watershed 11
5.1.2.2 Surface Water Hydrology of Lake Hartwell...13
5.1.3 Meteorology 14
5.1.4 Demographics 15
5.2 Sediment Investigation Summary 16
5.2.1 Phase I Investigation 17
5.2.2 Phase II Investigation 18
5.2.3 Field Screening-Modified Spittler Method...18
5.2.4 Nature and Extent of Contamination 19
5.2.5 Other Contaminants in Sediment 24
5.2.6 Summary of Surface Water Investigations....28
5.3 Biological Investigation Summary 28
5.3.1 Twelvemile Creek Watershed Investigations..29
5.3.2 Lake Hartwell Investigations 33
5.3.3 Assessment of Resource Use 38
5.4 Contaminant Fate and Transport 40
5.4.1 Future Sedimentation 40
5.4.2 Future Sediment and Water Quality Trends...43
5.4.3 Bioaccumulation and Future Fish
Concentrations 48
6.0 Summary of Site Risks 50
6.1 Contaminants of Concern 50
6.2 Exposure Assessment 51
6.3 Toxicity Assessment 52
6.3.1 Carcinogenic Toxicity of PCBs 52
6.3.2 Noncarcinogenic Toxicity of PCBs 53
6.4 Risk Characterization 53
6.5 Monte Carlo 54
6.6 Ecological Risk Assessment 55
6.7 Uncertainty Assessment 56
6.8 Summary 56
7.0 Description of Alternatives 66
7.1 Final Cleanup Goal for Sediment 67
7.2 Final Cleanup Goal for Fish 68
7.3 Remedial Alternatives 70
-------
SECTION PAGE
8.0 Summary of the Comparative Analysis of Alternatives 82
8.1 Overall Protection of Human Health and
the Environmental 82
8 .2 Compliance with ARARs 83
8.3 Long-Term Effectiveness and Permanence 84
8.4 Reduction of Toxicify, Mobility, and Volume 85
8.5 Short -Term Effectiveness 86
8.6 Implementability 88
8.7 Cost 88
9.0 The Selected Remedy 89
9.1 Continuation of l
-------
LIST OF TABLES
TABLE/TITLE PAGE
5-1 Monthly Mean Temperature and Precipitation for
Clemson and Pickens, SC 15
6-1 PCB Exposure Point Concentrations in Shallow Sediment 57
6-2 Exposure Point Concentrations in Fish 58
6-3 Exposure Parameters for Incidental Ingestion of and Dermal
Contact with Sediment (Cur rent/Future Use Scenario) 60
6-4 Exposure Parameters for Ingestion of Fish by Adult
Recreational Fisherman (Current/Future Use Scenario) 61
6-5 Potential Direct Contact Risks Associated with PCBs
in Sediment 62
6-6 Potential Health Risks Associated with Ingestion of Fish..64
8-1 Time to Achieve Protectiveness 87
9-1 Sediment Sample Locations in Twelvemile Creek Watershed
and Arm of Lake Hartwell 98
9-2 Cost Estimate for EPA's Selected Remedy 100
-------
LIST OF FIGURES
FIGURE/TITLE PAGE
1-1 Hartwell Lake and Watershed 2
1-2 Sangamo OU2 Study Area 3
2-1 Estimate of PCBs Received at the Sangamo Weston Plant 4
2-2 Estimate of PCBs Discharged Into Town Creek 5
5-1 .Twelvemile Creek Drainage Basin 9
5-2 Twelvemile Creek Profile 12
5-3 PCB Distribution for Grab Samples 25
5-4 Vertical Distribution of Mean PCB Concentration per
Transect Interval 26
5-5 Vertical Distribution of Maximum PCB Concentration per
Transect Interval 27
5-6 Twelvemile Creek Drainage Sampling Locations 30
5-7 Lake Hartwell Fish Sampling Stations 34
5-8 PCB Concentrations in Largemouth Bass Fillets (1990-93)...36
5-9 PCB Concentrations in Channel Catfish (1990-93) 36
5-10 PCB Concentrations in Hybrid Bass Fillets (1990-93) 37
5-11 Simulated Bed Profile Twelvemile Creek Arm 41
5-12 Predicted Average Sediment Burial Rates 42
5-13 Initial WASP4 Surface Sediment PCB Concentrations 44
5-14 WASP4 Surface Sediment PCB Concentrations at 10 Years 45
5-15 WASP4 Surface Sediment PCB Concentrations at 20 Years 46
5-16 WASP4 Surface Sediment PCB Concentrations at 30 Years 47
5-17 Initial and Predicted Concentrations of Selected Biota....49
7-1 FGETS Modeling Results for Selected PCB-Cleanup Goals 69
7-2 Area to be Capped in Alternative 3A 74
7-3 Sediment Control Structure Location of Alternative 3B 76
-------
FIGURE/TITLE
7-4 Sediment Control Structure/Area to be Capped for
Alternative 3C , 78
7-5 Confined Disposal Facility for Alternative 4 80
9-1 Current Lake Hartwell Fish Advisory 90
9-2 Fish Sampling Stations in Lake Hartwell 93
9-3 Corbicula Basket Sampling Locations in Twelvemile Creek...95
9-4 Sediment Sampling Stations in Lake Hartwell 96
9-5 Sediment Sampling Stations in Twelvemile Creek 97
-------
Final ROD-Sangamo OU2
June 1994
page l
1.0 SITE NAME, LOCATION, AND DESCRIPTION
The Sangamo Weston/Twelvemile Creek/Lake Hartwell
Polychlorinated Biphenyl (PCB) Contamination Site - Operable Unit
Two (hereinafter referred to as "the Sangamo OU2 Site") is
located in Pickens County, South Carolina. The Sangamo OU2 Site
comprises the sediment, surface water, and biological migration
routes downstream from the Sangamo Weston Plant and satellite
disposal areas that have site related PCB-contamination. The
Sangamo Weston Plant and satellite disposal areas constitute
Operable Unit One of the Site.
Lake Hartwell was constructed by the Savannah District U.S.
Army Corps of Engineers (COB) between 1955 and 1963 by damming
the Savannah, Seneca, and Tugaloo Rivers. Figure 1-1 provides an
illustration of the 56,000 acre Hartwell Reservoir located on the
Georgia-South Carolina border. The OU2 study area includes
approximately 40 stream miles of Twelvemile Creek and its
tributaries, the Twelvemile Creek Arm of Lake Hartwell, and
portions of the Keowee and Seneca River Arms of Lake Hartwell
down to the Route 37 (Rt. 37) bridge just south of Clemson, South
Carolina. Figure 1-2 provides an illustration of the Sangamo OU2
study area. The primary focus of OU2 is centered on this area;
however, samples were collected throughout Lake Hartwell during
the OU2 investigations including that portion of the reservoir
between Rt. 37 and Hartwell Dam.
2.0 SITE HISTORY AND ENFORCEMENT ACTIVITIES
Sangamo Weston, Inc., owned and operated a capacitor
manufacturing plant in Pickens, South Carolina from 1955 to 1978,
near the headwaters of Lake Hartwell. The plant manufactured
several varieties of capacitors, including electrolytic, mica,
and power factor capacitors as well as potentiometers.
Schlumberger Industries, Inc. (SII) is the current owner of the
plant site, as a result of a merger with Sangamo Weston in 1989.
The plant used several varieties of dielectric fluids in its
manufacturing processes, including PCB-containing dielectric
fluids. PCBs reportedly enhanced the performance and durability
of the fluids. The PCBs used for this application were primarily
Aroclors 1242, 1254, and 1016. Waste disposal practices from the
Sangamo Plant included land-burial of off-specification
capacitors and wastewater treatment sludges on the plant site and
at the six satellite disposal areas. PCBs were also discharged
with effluent directly into Town Creek, which is a tributary of
Twelvemile Creek. Twelvemile Creek is a major tributary of Lake
Hartwell. The use of PCBs was terminated by Sangamo Weston in
1977 prior to an EPA ban on PCB use in January 1978.
Estimates of the type and quantities of PCBs received, used,
and discharged by Sangamo Weston were derived by reviewing
Monsanto Corporation shipping records, Sangamo Weston records,
interviews with Sangamo Weston employees, engineering and
analytical reports completed by firms under contract to Sangamo
Weston, and EPA documents. Figures 2-1 and 2-2 provide a
graphical illustration of the PCBs received at the Sangamo Plant
and an estimate of PCBs discharged to Town Creek, respectively.
-------
Final ROD-Sangamo OU2
Jvme 1994
page 2
Hlrtwell
SCALE
o
4 e MILES
I I I , I
Figure 1-1
Hartwell Lake and Watershed
Sanaamo OU-2 Rl
-------
Final ROD-Sangamo OU2
June 1994
page 3
1
N
Sangamo Weston Plant
Oftsite DisposalArea
A Landfill
Dam
SCALE
0
1 ' '
4 MILES
i 1
1 1
Figure 1-2
Study Area
Sangamo OU-2 Rl
-------
Final ROD-Sangamo OU2
June 1994
page 4
Date
Rgure 2-1
Estimate of PCBs Received at the Sangamo Weston Plant
10,000,000 T
1,000,000
£ 100,000
CO
g
Ui
ta
CB
O
HIIII\IIII4
mmcecoceieiocecoieco
0.1
e>o>a>
Date
Figure 2-2
Estimate of PCBs Discharged Into Town Creek
-------
Final ROD-Sangamo OU2
June 1994
page 5
The average quantity of PCBs used by Sangamo Weston ranged from
700,000 to 2,000,000 lbs/yr from 1958 to 1977. Data indicate
that an estimated 3 percent of the quantities received and used
were discharged to Town Creek. This approach leads to an
estimated cumulative discharge of PCBs into Town Creek of over
400,000 Ibs through 1977.
Historical surficial and core sediment studies of the
Twelvemile Creek watershed and Lake Hartwell were conducted by
several entities including the South Carolina Department of
Health and Environmental Control (SCDHEC), COB, EPA, RMT (for
Schlumberger), and several Clemson graduate students. These
studies were initiated in 1976 by SCDHEC and occurred
intermittently through the mid to late 1980s. PCB concentrations
in surficial sediments were highest from samples collected near
the plant's discharge point on Town Creek and generally decreased
with increasing distance downstream from the Sangamo Weston Plant
Site. PCB concentrations in sediment core samples were highest
in samples collected from the Twelvemile Creek Arm of Lake
Hartwell. PCB concentrations generally increased to maximum at a
depth of 10-30 cm.
In the mid-1970s, SCDHEC and EPA discovered that fish from
certain areas of Lake Hartwell were contaminated with PCBs at
levels above the U.S. Food and Drug Administration (FDA) safe
tolerance limit of 5 ppm (5 mg/kg). SCDHEC originally issued a
health advisory in 1976 that warned the public against eating
fish from the Seneca River Arm of Lake Hartwell north of State
Highway 24 (Hwy 24) and Twelvemile Creek (Figure 1-1). FDA
lowered the PCB tolerance level to 2 ppm (2 mg/kg) in 1984. As a
result, SCDHEC modified the original health advisory in 1985 to
state the following:
1) All fish taken from the Senec». River Arm of Lake Hartwell
north of SC Hwy 24 and Twelvemile Creek should be released
and not eaten;
2) All fish larger than three (3) Ibs. taken from the remainder
of Lake Hartwell should be released and not eaten; and
3) Fishing is not prohibited but SCDHEC advises that these fish
not be eaten due to the presence of elevated levels of PCBs.
Swimming, boating, and other related activities are not
restricted by this advisory.
This advisory remains in effect, and signs warning against
eating fish have been posted at the majority of the public boat
launch and recreation areas in South Carolina since 1987.
SCDHEC has conducted studies in Lake Hartwell since 1976 to
evaluate the levels of PCB contamination in fish. Their results
indicate that PCB concentrations in non-migratory fish (i.e
largemouth bass, catfish) collected within the Twelvemile Creek
embayment remain above the FDA 2 mg/kg tolerance level with no
apparent decrease over time. PCB concentrations in these non-
migratory species generally decreased with increasing distance
from the source area. However, migratory species (i.e.
-------
Final ROD-Sangamo OU2
June 1994
page 6
hybrid/striped bass) collected at all stations in the lake had
elevated levels of PCBs which frequently exceeded the 2 mg/kg
tolerance level. There was no apparent decrease in migratory
fish tissue PCB concentrations over time.
In 1987, based upon the EPA Hazard Ranking System, SCDHEC
monitoring programs, and accompanying concerns of citizens in the
area, the Sangamo Site was proposed for inclusion on the National
Priorities List (NPL). The Sangamo Site was finalized on the NPL
in February 1990. As a result, EPA issued special notice to SII
in April 1990 for performance of a Remedial Investigation and
Feasibility Study (RI/FS). Since SII declined to conduct the
RI/FS, EPA assumed the lead-role in performing the RI/FS at the
Sangamo OU2 Site and formally initiated the process in September
1990. The RI/FS process for OU2 was divided into two separate
studies, a Sediment Investigation and Biological Investigation,
which were conducted concurrently. The Sediment Investigation
was conducted by the Oak Ridge, TN office of Bechtel
Environmental under the funding and direction of EPA. The
Biological Investigation was conducted by the Savannah District
of the U.S. Army COE, under funding and direction provided by
EPA.
3.0 HIGHLIGHTS OF COMMUNITY PARTICIPATION
EPA held an OU2 RI Kick-Off Public Meeting at the Liberty
(South Carolina) Senior High School Cafeteria on May 9, 1991 to
present the approach for the Phase I Sediment RI and overall
Biological Investigation strategy. On March 3, 1992, a public
meeting was held at the Ramada Inn in Clemson, SC to discuss the
Phase I Sediment Investigation results, planned approach for the
focused Phase II Sediment Investigation, and provide an update on
the Biological Investigations. The Final Remedial Investigation
Reports (Sediment and Biological components) and a summary fact
sheet were released to the public in May 1993. A public meeting
was held on June 3, 1993 at the Ramada Inn in Clemson to present
the findings from the completed investigations and the Baseline
Risk Assessment.
The Draft Feasibility Study Report and a summary fact sheet
were released to the public in February 1994 in an effort to
facilitate greater public involvement in the remedy selection
process at the Sangamo OU2 Site. On March 28, 1994, a public
forum was held at the Strom Thurmond Institute on the Clemson
University Campus to discuss the various remedial alternatives
under consideration in the Draft FS Report.
The Final FS Report, Proposed Plan Fact Sheet, and all other
site-specific documents EPA relied upon to develop the Proposed
Plan were arranged in an Administrative Record and released to
the public in April 1994. The Administrative Record was
maintained and available for public review at the EPA Records
Center in Region IV, and the following five information
repositories: The Village Library in Pickens, SC; Pickens County
Public Library in Easley, SC; R.M. Cooper Library at Clemson
University, SC; Hart County Library in Hartwell, GA; and the
-------
Final ROD-Sangamo OU2
June 1994
page 7
Corps of Engineers Lake Hartwell Natural Resources Management
Center near the Dam in Hartwell, GA.
Concurrent with the public release of the Proposed Plan, a
notice of availability of the Administrative Record was published
in the Greenville News and Anderson Independent on April 11,
1994; the Athens Daily, Easley Progress, Hartwell Sun, and
Pickens Sentinel on April 13, 1994; and the Clemson Messenger-
Journal Tribune on April 16, 1994. A public comment period was
held from April 11 through May 11, 1994. A Proposed Plan Public
Meeting was held on April 19, 1994 at the Ramada Inn in Clemson,
SC. At this meeting, representatives from EPA, the Savannah
District COE, and SCDHEC discussed the site conditions, the
remedial alternatives under consideration and presented the
rationale behind the preferred alternative. During this meeting,
a request for an extension to the public comment period was made.
As a result, the public comment period was extended until June
10, 1994. The notice of this extended public comment period was
published in the Greenville News on May 2, 1994.
A response to comments received during this 60-day public
comment period is included in the Responsiveness Summary, which
is attached to this Record of Decision (ROD) as Appendix B. This
decision document presents the selected remedial action for the
Sangamo OU2 Site in Pickens County, South Carolina, chosen in
accordance with CERCLA, as amended by SARA, and in accordance
with the National Contingency Plan.. The decision for this Site
is based on materials in the Administrative Record and comments
received during the public comment period.
4.0 SCOPE AND ROLE OF RESPONSE ACTION
Due to the size and complexity of the Sangamo Site, and in
order to simplify the investigation and response activities, EPA
divided the Site into two discrete study areas known as Operable
Units (OUs). Operable Unit One (OU1) consists of the land-based
source areas, including the Sangamo Weston Plant Site and six
satellite disposal areas in Pickens County used for the disposal
of solid wastes containing PCBs (Figure 1-2). The six satellite
disposal areas are located in rural areas within a 3-mile radius
of Pickens and have the following designations: Breazeale, Cross
Roads, Dodgens, John Trotter, Nix and Welborn. EPA selected the
remedy for OU1 in a ROD signed on December 19, 1990. Under a
Consent Decree negotiated with EPA, Schlumberger initiated
remedial action at OU1 in November 1993. This included the
excavation of waste and PCB contaminated soils from the six
satellite disposal areas and transportation of the excavated
materials to temporary storage facilities at the former Sangamo
Plant Site. A treatment system will be built and tested at the
plant at which time the contaminated material will be treated by
means of low temperature thermal desorption. In addition,
groundwater recovery and treatment systems will be installed at
the plant site and at applicable satellite areas.
Operable Unit Two (OU2), the subject of this ROD, addresses
the PCB-contamination in the sediments and aquatic biota of the
study area (Figure 1-2). The exposures associated with ingestion
-------
Final ROD-Sangamo OU2
June 1994
page 8
of fish contaminated with PCBs from all fish sampling stations in
Lake Hartwell exceeded EPA's acceptable human health risk range.
OU2 represents the final response action EPA expects to implement
as part of its overall strategy in remediating the Sangamo Site.
This response action addresses the principal threat posed at the
Site which is ingestion of PCB-contaminated fish.
5.0 SUMMARY OF SITE CHARACTERISTICS
Due to the atypical size and complexity of the Sangamo OU2
Site, the Sediment and Biological Investigations generated an
immense and extensive data base . In the interest of brevity,
this information will not be reiterated in this ROD. This
section presents a brief, comprehensive overview of the Sangamo
OU2 Site characteristics as assessed during the Remedial
Investigation. The reader is referred to the Final Remedial
Investigation Report (May 1993), Volumes I and II and/or the
Final Biological Investigation Report (February 1994) for a more
detailed account of this subject matter.-
5.1 SUMMARY OF PHYSICAL CHARACTERISTICS
This section presents information concerning the physical
characteristics of the study area including its surface features,
surface water hydrology, meteorology, and demographics.
5.1.1 Surface Features
The Sangamo OU2 study area generally consists of sloping to
moderately steep uplands that are dissected by a well-developed
drainage system. There are a number of small towns in the area
in addition to the city of Clemson. Outside of the towns and the
residential developments along the Hartwell lakeshore, the
countryside is predominantly rural. Much of the undeveloped area
is forested, but significant agricultural tracts can also be
found.
The northern portion of the Twelvemile Creek watershed is in
the foothills of the Blue Ridge Mountains, an area characterized
by steep, heavily forested slopes and ridges. Twelvemile Creek
originates near 1,920-ft-high Walnut Cove Mountain, in an
isolated area approximately 7 miles north-northeast of Pickens,
which includes first- and second-order streams with abundant
riffles (Figure 5-1).
The relief is somewhat lower in the vicinity of Pickens and
the OUl source areas. In this portion of the watershed,
Twelvemile Creek occurs as a meandering, relatively slow-moving,
fourth-order stream with abundant pools, sand and gravel bars,
and broad floodplains that in some reaches exceed 1,000 ft.
However, approximately 1 mile south of Pickens, the stream valley
narrows and the creek has a large number of sharp bends. In this
area, Twelvemile Creek resembles a mountain stream with exposed
rocks, swift-moving water, and abundant riffles that in some
locations may be considered rapids. Floodplains are generally
absent as the creek valley narrows to widths of only 100 to 200
-------
Final ROD-Sangamo OU2
June 1994
page 9
UK
JOCASSEE
tAKE
KEDWEE
SCALE
0
1
4 MILES
1
1 1
Twelvemile Creek Drainage Basin
Gaging Station
Figure 5-1
Twelvemile Creek Drainage Basin
-------
Final ROD-Sangamo OU2
June 1994
page 10
ft., bounded by moderate to steep slopes that are generally
covered with trees and other vegetation. Access to the creek is
difficult and generally limited to the few bridges in the area.
These conditions continue downstream (south) to the headwaters of
Hartwell Lake, just upstream of the Maw Bridge (Hwy 337 Bridge).
The Twelvemile Creek Arm of Lake Hartwell is characterized
by a narrow channel with steep, heavily vegetated shoreline with
only limited access. The waters in the upper portion of this arm
have been impounded and thus water velocities are greatly
reduced. There is a sharp bend in the lake near the Madden
Bridge, but otherwise the channel has a linear, northeast-
southwest trend. The lake becomes progressively wider
downstream, reflecting the widening of the former Twelvemile
Creek floodplain, and the number and size of inlet coves
increases. The lakeshore along the Twelvemile Creek Arm is
generally undeveloped due to the relief and development
restrictions imposed by the COE, which has classified much of the
western shoreline and the major inlet coves on the eastern
shoreline as a Protected Lakeshore Area. This designation does
not necessarily prohibit all development, but has served to
greatly limit the type and extent of development within this
portion of the lake.
In the downstream direction, past the Keowee River
confluence, into the Seneca River Arm and finally into Lake
Hartwell proper, the shoreline typically becomes more gradually
sloped and is more developed with less vegetative cover than the
shoreline of the Twelvemile Creek Arm. Relief around the main
portion of the lake is much lower, primarily due to the pool
elevation, which becomes progressively closer to the hilltops
toward the south.
Lake Hartwell was formed by the construction of Hartwell Dam
across the Savannah River approximately 7 miles below the point
where the Tugaloo and Seneca rivers converge to form the Savannah
River. Construction of the dam began in 1955 and was completed
in 1963. The dam is a concrete and earth structure that spans
18,000 ft across the Savannah River and rises 204 ft above the
riverbed at its highest point. At normal pool level, Hartwell
Lake extends for 49 miles up the Tugaloo River and 45 miles up
the Seneca River and its tributaries (i.e., Twelvemile Creek and
the Keowee River).
At a normal pool elevation of 660 ft. mean sea level (MSL),
the lake has an area of 56,000 acres and large expanses of open
water, particularly below the point where Twenty-six Mile Creek
meets the Seneca River. Because of the relief and large number
of coves, the lake has a shoreline of 962 miles. There are a
number of large arms in the main portion of the lake. The major
ones include the Tugaloo Arm, Twenty-six Mile Creek Arm, Coneross
Creek Arm, and Eighteen Mile Creek arm.
5.1.2 Surface Water Hydrology
The Sangamo OU2 study area is defined by hydrologic
features; therefore, the surface water hydrology is one of the
-------
Final ROD-Sangamo OU2
June 1994
page 11
most important physical characteristics of the study area. This
section presents an overview of the hydrologic characteristics of
the Twelvemile Creek watershed and Lake Hartwell, which are the
principal hydrologic units in the study area.
5.1.2.1 Surface Water Hydrology of Twelvemile Creek Watershed
The Twelvemile Creek watershed has an area of 140-mi2 and
includes first-, second-, third- and fourth-order streams. The
tributaries to Twelvemile Creek are predominantly first- and
second-order streams. The major tributaries, Town Creek, Wolf
Creek, Rices Creek, and Golden Creek, are third-order streams.
Twelvemile Creek is a third order stream above the mouth of Town
Creek; below this point, Twelvemile Creek is a fourth-order
stream. Twelvemile Creek is the longest stream segment in the
watershed, which flows southward for approximately 24 miles until
reaching the headwaters of Hartwell Lake. Within this 24-mile
reach, approximately 80 tributaries flow into Twelvemile Creek.
The gradient of the Twelvemile Creek basin is represented by
the longitudinal profile shown in Figure 5-2. The profile
extends from the confluence of the Middle and North forks into
Hartwell Lake as far as the Rt. 37 Bridge (the downstream limit
of the study area in Hartwell Lake). From this figure, it can be
seen that the total elevation drop from the confluence of the
North and Middle forks to the headwaters of the lake is
approximately 200 ft, which over a distance of 17 miles amounts
to an average slope of 11.8 ft per mile.
The bulk of the stream flow is derived from runoff. Flow
data for Twelvemile Creek indicate an average daily flow of 198
cubic feet per second (cfs), with historical daily flow ranging
from 30 to 5360 cfs. Sediment in the creek is composed primarily
of sand and has a generally low total organic carbon content
(TOO, ranging from 0.1 to 3.6 percent. NPDES permit data
indicate that Twelvemile Creek receives discharges from sewage
treatment plants and a small number of industrial facilities.
These discharges average 1 to 2 million gallons per day. During
low flow periods, the permitted discharges may comprise 10-13
percent of the flow in the creek; during normal flows, the
discharges comprise less than 2 percent of the flow and become
negligible during high flow periods.
The three impoundments on the lower section of Twelvemile
Creek are all of masonry construction. The lowermost impoundment
(Woodside II) is the largest of the three, with an upper pool
elevation of 722 ft MSL and a lower pool elevation of 684 ft MSL
(difference of 38 ft) . This impoundment was built in 1905. The
middle impoundment (Woodside I) is located in the community of
Cateechee and was rebuilt in 1937 after it failed in 1934. The
middle impoundment has an upper pool elevation of 760 ft MSL and
a lower pool elevation of 736 ft MSL (difference of 24 ft). Both
the lower and middle impoundments were constructed by Norris
Mills and are currently owned and operated by Consolidated Hydro
of Greenville, South Carolina. These two impoundments were
renovated in 1983-1985 and currently produce a combined
2.5 million kWh/yr. In 1984, up to 20 ft of sediments were
-------
Town Creek
Confluence
UJ
900 r
850
800
750
700
650
600
Upper Impoundment
Middle Impoundment
Lower Impoundment
Headwater Hartwell Lake
Maw Bridge (Rt. 337)
Madden Bridge (Rt. 15)
Hwy 133 Bridge
Normal Pool Elevation:
660UMSL
US 123 Bridge
Hwy 93 Bridge
Rt. 37 Bridge
I
550
500
450
4 6 8 10 12 14 16 18 20 22 24
Distance Downstream from Confluence of Middle and North Forks of Twelvemfle Creek (ml.)
Figure 5-2
Twelvemile Creek Profile
26
28
I
CO
r
« g
S> (D
to o
tO
-------
Final ROD-Sangamo OU2
June 1994
page 13
flushed from the pools behind the lower and middle impoundments.
In September 1993, an estimated 43,000 cubic yards of sediment
were flushed from behind the lowermost impoundment. Currently,
sediments are flushed from these pools approximately biannually.
The third, or uppermost, impoundment was built in 1926 and
is the smallest of the three impoundments, with an upper pool
elevation of 783 ft MSL and a lower pool elevation of 763 ft MSL
(difference of 20 ft). This upper impoundment was formerly used
for generation of hydroelectric power. The impoundment was
purchased by the Easley-Central Water District in 1962 and is
currently used as the intake point for the Easley-Central Water
Plant. Sediments are flushed from behind the upper impoundment
quarterly. Approximately 6 ft of sediments are flushed from
behind the impoundment during each flushing event.
Surface water in the Twelvemile Creek basin is currently
utilized for drinking water supply, fishing, and industrial uses.
Twelvemile Creek is classified as a Class B stream according to
South Carolina Regulations (Regulation 61-68, Water
Classifications and Standards). Under the regulations, Class B
waters are defined as being suitable for secondary-contact
recreation (fishing, boating, wading) and drinking water supply
(assuming conventional treatment methods are used) as well as
both agricultural and industrial uses.
5.1.2.2 Surface Water Hydrology of Lake Hartwell
The surface water hydrology of the Hartwell Lake reservoir
is different from that described for the Twelvemile Creek basin
because it is an impoundment with a drainage basin 2,088 mi2.
Hartwell Lake is managed by the COE for flood control and
electric power generation, both of which are affected by the
storage capacity of the reservoir, which is 2,550,000 acre-feet
of water (equivalent to 830 billion gallons). Since its
construction, the reservoir has become one of the major
recreational lakes in the Southeast. Current management
practices therefore consider recreational benefits as well as
flood control and power generation.
The COE maintains a pool elevation of 660 ft MSL during the
recreation season (summer); in mid-October through mid-December
the lake level is dropped 4 ft to a level of 656 ft MSL. The lake
is drawn down in the fall in anticipation of the increased
rainfall that the area usually receives during the winter and
spring. The annual fluctuations in the lake commonly exceed 4
ft; peak elevations commonly occur in April or May. The lowest
recorded monthly average lake level of 643.3 ft MSL occurred in
December 1981.
Water flows in the Lake Hartwell system were characterized
by the inflow and outflow data provided by the COE. These data
consist of 30 years of daily inflow and outflow data collected
between 1962 and 1992. The average monthly inflow to the lake
for the period 1962-1992 ranged from a low of 2,693 cfs in
September to a high of 6,222 cfs in March. Average monthly
outflow from Hartwell Dam in the same period ranged from a low of
-------
Final ROD-Sangamo OU2
June 1994
page 14
3,03.4 cfs in September to a high of 5,190 cfs in April. On a
yearly basis, the average monthly inflow to Hartwell Lake has
been 4,090 cfs; outflow has been 3,997 cfs. Given the reported
storage capacity of the reservoir (2,550,000 acre-feet or 830
billion gallons), the estimated hydraulic residence time in the
lake is 25 days.
Lake Hartwell is Class A surface water (South Carolina
regulations) suitable for primary contact recreation (swimming,
waterskiing), secondary contact recreation (fishing, boating,
wading), drinking water supply, and agricultural/industrial uses.
The lake currently receives a significant level of point and non-
point source discharges. NPDES-permitted discharges include
industrial facilities, electric power generating stations/ and
various sewage treatment plants. The permitted industrial
discharges include heavy metals, volatile organic compounds,
cyanide, phenol, oil and grease, and textile dyes. Permitted
discharges comprise only 18 million gallons per day
(approximately 200 cfs). Since the reservoir continues to be a
source of potable water for a number of communities, these
discharges apparently have not had an appreciable impact on water
quality in the lake.
5.1.3 Meteorology
Pickens County has a temperate climate characterized by mild
winters and relatively abundant rainfall. The National Weather
Service has two precipitation and temperature gauging stations in
the Twelvemile Creek area, one at Pickens and the other at
Clemson. Table 5-1 summarizes the monthly mean temperature and
precipitation for these two locations.
Generally the mean monthly temperature is higher in the
summer and winter months in Clemson while in the spring and fall
it is higher in Pickens. The average monthly precipitation for
the Pickens and Clemson stations ranges between 3.5 in. and
6.1 in. Annual precipitation in both Clemson and Pickens is more
or less evenly distributed throughout the year. The average
annual precipitation for Pickens from 1951-1992 was 56.04 in.
The maximum and minimum annual rainfall events occurred 1964 and
1981, respectively, when 78.46 in. and 33.40 in. were recorded.
In Clemson, the average annual precipitation from 1911-1992 was
53.23 in. The maximum and minimum annual rainfall events
occurred in 1936 and 1925, respectively, when 73.36 and 34.31 in.
of precipitation were recorded.
Free water surface evaporation data was obtained for the
study area from the Evaporation Atlas for the Contiguous 48
United States (NOAA 1982). These data indicate that annual
evaporation from Hartwell Lake should be in the range of 40 to 41
in. With annual precipitation in the area greater than 53 in.,
Hartwell Lake receives 12 to 13 in. of net precipitation per
year.
-------
Final ROD-Sangamo OU2
June 1994
page 15
TABLE 5-1
Monthly Mean Temperature and Precipitation for Clemson and
Pickens, South Carolina1
Month
January
February
March
April
May
June
July
August
September
October
November
December
Clemson, SC2
Temp.
(°F)
42.8
45.4
51.9
60.6
68.4
75.7
78.4
77.6
72.7
61.8
51.6
44.2
Precipitation
(inches)
5.01
5.01
5.65
4.31
4.12
3.95
4.86
4.57
3.54
3.68
3.49
4.98
Pickens, SC3
Temp.
(°F)
42.0
45.3
52.5
61.2
68.3
74.7
77.8
77.1
71.8
62.0
52.4
44.2
Precipitation
( inches )
5.14
5.12
6.07
4.53
4.60
4.53
4.76
4.40
3.64
4.05
4.02
4.91
1 Data Source: National Climatic Data Center
2 Reporting Period 1911-1992
3 Reporting Period 1951-1992
5.1.4
Demographics
Demographics and land use in the Hartwell Lake area are
variable, with small towns and rural residential development in
the Twelvemile Creek watershed giving way to larger towns and
more concentrated development in the areas surrounding Hartwell
Lake. According to 1990 census data, approximately 356,000
people live in the six counties that border Hartwell Lake:
Anderson, Pickens, and Oconee counties in South Carolina; and
Hart, Franklin, and Stephens counties in Georgia. Of this total,
an estimated 297,000 people live in the South Carolina counties;
93,894 live in Pickens County alone.
The major community in the upper portion of the Twelvemile
Creek watershed is Pickens, which had an estimated population of
3,042 in 1990. Two communities are located along Twelvemile
Creek: Cateechee (estimated 1990 population of 158) and Norris
(estimated 1990 population of 884). The town of Clemson, with an
estimated 1990 population of 11,096, is the only large community
directly on the shoreline of the lake. Outside of the small
towns and communities, the majority of the Twelvemile Creek
-------
Final ROD-Sangamo OU2
June 1994
page 16
watershed (and Pickens County in general) is undeveloped. Most of
the acreage bordering Twelvemile Creek and its tributaries is
either forested or cleared for agricultural purposes.
The entire Hartwell project, both land and water usage, is
managed by the COE Savannah District. Hartwell Dam was
constructed by the COE as part of a flood control and
hydroelectric power project. In addition to these primary uses,
Hartwell Lake has developed into one of the largest and most
popular recreational lakes in the southeastern U.S. According to
a 1989 COE survey, more than 15 million people visited the lake,
making it one of the three most visited COE lakes in the nation.
Data from a recreational area on Twelvemile Creek indicates that,
out of a total of 301,000 visitors, 125,787 visits were made for
fishing, 76,892 for boating, 30,310 for water skiing, and 28,139
for swimming.
Development along the shoreline of Lake Hartwell is at least
partially controlled through the COE Lakeshore Management Plan
(COE 1989). The COE controls 23,566 acres above the normal pool
elevation of 660 ft MSL. The COE acreage in effect establishes a
buffer zone of variable width around the entire shoreline. This
buffer zone enables the COE to restrict access, development, and
the types of activities permitted. These restrictions are
effected through a system of permits and land use allocations as
established in the Shoreline Management Plan. In order to
administer the permit program, the shoreline of Hartwell Lake was
allocated into four categories: Limited Development Areas, Public
Recreation Areas, Protected Shoreline Areas, and Prohibited
Access Areas.
Surface water supplies the bulk of potable water utilized by
the residents of Pickens County and surrounding areas. Currently
there are two active water intakes on Twelvemile Creek; these
intakes provide potable water supply to the Pickens Water Plant
and Easley-Central Water Plant. The Pickens Water Plant intake
point is located near the confluence of the Middle and north
Forks of Twelvemile Creek. Withdrawals from this intake enable
the Pickens Plant to supply 2 to 2.5 mgd of potable water to
approximately 4,000 residents in the Pickens area. The intake
for the Easley-Central Water Plant is located in the uppermost of
the three impoundments on Twelvemile Creek. According to the
Easley-Central Water Plant, withdrawals from this intake supply
an average of 1.2 mgd of potable water to approximately 1,100
residents in the plant's service area.
5.2 SEDIMENT INVESTIGATION SUMMARY
Field investigation activities for the sediment component of
the Sangamo OU2 RI were conducted in two seasonal sampling
events. The initial, Phase I event was conducted during July and
August 1991. The objectives of Phase I were to verify
extent-of-contamination data from previous investigations and
further characterize the magnitude and distribution of PCB
contamination in the Twelvemile Creek watershed and Hartwell
Lake. Results of Phase I sampling were used to define a focused
-------
Final ROD-Sangamo OU2
June 1994
page 17
sampling program for the second, Phase II sampling event. Phase
II was conducted during April and May 1992. The objective of
Phase II was to address specific data gaps regarding the extent
of PCS contamination in sediments of the upper Hartwell Lake
reservoir (i.e., Seneca River, Keowee River, and Twelvemile Creek
Arms) and the Twelvemile Creek watershed.
5.2.1 Phase I Investigation
Phase I field sampling encompassed an area extending some 50
miles, from the Sangamo Weston Plant site and the satellite
disposal areas to the Hartwell Dam. A total of 480 samples,
exclusive of quality assurance/quality control (QA/QC) samples,
were collected during Phase I from Twelvemile Creek, its
tributaries, and Hartwell Lake. Phase I samples included 88
surface water, 341 sediment core, and 51 sediment grab samples,
at 50 locations throughout Lake Hartwell. An onsite field
screening method, the Modified Spittler Method (See Section
5.2:3), and offsite Contract Laboratory Program (CLP)
laboratories were used to analyze the samples.
A stainless steel ponar grab sampler was used for collecting
the surface sediment sample. The depth of sample collected by
the grab sampler was approximately 10 to 15 cm. A representative
sample from this surface zone was important to adequately
characterize the horizontal contamination in the biologically
active zone and to determine the quality of sediment that was
recently deposited. All 51 grab samples were analyzed by CLP
laboratories. Four of these samples were selected for full-
screen Target Compound List (TCL)/Target Analyte List (TCL)
analyses. Offsite analyses were also conducted for grain size,
TOC, and cation exchange capacity, all which are factors that
affect the mobility of PCBs and other contaminants.
Sediment core samples were recovered to a maximum depth of
50 cm to characterize the vertical distribution of PCBs in the
sediment. In shallow water areas, core samples were collected by
driving a stainless steel core tube into the sediment by hand.
In deep-water areas, a gravity corer with a stainless steel liner
and core catcher was used to collect the subsurface sample. In
both cases, sediment was extruded from the top of the core tube
at 5-cm intervals and placed in jars marked according to the
depth interval. All core samples were analyzed for total PCBs.
Split samples were collected from 19 percent of the sediment
cores analyzed onsite and sent for offsite PCB analyses at a CLP
laboratory. This was done to determine the correlation between
field and laboratory analytical results.
Surface water samples were collected at all 50 locations. A
near-surface water sample was collected at all stations by
dipping the mouth of the sample container just beneath the
surface of the water. Where water depths exceeded 4 to 6 ft, a
near-bottom sample was collected approximately l ft above the
sediment interface using a peristaltic pump and teflon tubing.
All 88 water samples were analyzed for PCBs and TOC at offsite
CLP laboratories. Samples from 4 stations received full scan
TCL/TAL analyses. Selected surface water samples were also
-------
Final ROD-Sangamo OU2
June 1994
page 18
analyzed for the following water quality parameters: total
dissolved solids (TDS), total suspended solids (TSS), chloride,
flouride, sulfate, and nitrate-nitrite.
Given the large number of samples collected on the open
water of Hartwell Lake, the technical impracticality of
conducting an accurate location survey utilizing traditional
survey methods, and the importance of precise sample locations in
determining the nature and extent of contamination in OU2, sample
location coordinates (latitude/longitude) were determined
utilizing a hand-held Magellan Global Positioning System (GPS).
GPS is a network of 24 satellites orbiting the earth at an
altitude of 10,900 miles. The satellites and the hand-held
receivers both generate a continuous code which allows the
receiver to calculate how long it takes the signal to travel from
the satellite, and then to extrapolate the satellites's distance.
Once the hand-held receiver gets position information from three
satellites, position coordinates are determined by triangulation.
5.2.2 Phase II Investigation
During the Phase II study, a more extensive sampling program
was conducted in areas determined during the Phase I study to
have the highest PCS concentrations. The Phase II field sampling
effort focused on the Keowee River, Seneca River and Twelvemile
Creek Arms of Hartwell Lake as well as Twelvemile Creek itself.
Not including QA/QC samples, 744 samples consisting of 185
surface sediment, 550 sediment core, and 9 storm-event surface
water samples were collected and analyzed by field screening and
CLP analytical methods. Sample collection and location
techniques were as described under Section 5.2.1.
Twenty transects with a maximum of five sediment coring
stations per transect were established across the lake and creek
bed. The sediment core stations along each transect were
selected following a review of the bottom depth profile collected
with a bathymetric profiler. Sediment grab samples were then
collected between each transect location. Sediment was sampled
from a total of 374 locations during the Phase II study.
Composite surface water samples were collected from the following
bridges, which cross Twelvemile Creek and the upper reaches of
the Hartwell Lake reservoir: the Maw Bridge (Rt. 337), Madden
Bridge (Rt. 15), Hwy 133 Bridge, Hwy 93 Bridge, and Rt. 37
Bridge. Samples were collected from these stations following two
significant storm events to evaluate the potential for
resuspension of PCBs into the water column.
5.2.3 Field Screening - Modified Spittler Method
Both phases of the Sediment Investigation utilized the
Modified Spittler Method, a field screening analytical technique,
to significantly reduce cost and turnaround time on analyses.
Reduced turnaround time (approximately 24 hours) enabled sampling
teams to identify locations where additional sampling was
warranted and to evaluate data quality on a real time basis.
-------
Final ROD-Sangamo OU2
June 1994
page 19
The Modified Spittier Method is an EPA-approved field
screening technique for quantifying total PCB concentrations in a
sediment/soil matrix. The technique employs a gas chromatogram
(GC) with an electron capture detector (BCD), and has
demonstrated reliability in the quantification of PCB
concentrations in excess of 1 ppm (mg/kg). The primary
difference in time and cost between the CLP procedure and
Spittler Method is incurred in the extraction procedure. Using
commonly available laboratory equipment, one analyst can extract
10 to 20 samples in less than 2 hours. The cost of GC/ECD
analysis using the Spittler Method is between $50 and $100,
compared to approximately $300 for conventional CLP analysis.
A total of 122 samples were analyzed for PCBs by both the
modified Spittler and CLP methods during Phase I and II. A
statistical comparison of the 67 Phase II samples analyzed by
both CLP and Spittler was performed by the EPA Environmental
Monitoring Systems Laboratory (EMSL) in Las Vegas, Nevada.
Twenty-eight of these samples were reported as "non-detect" by
both analytical methods. Of the remaining 39 samples, PCB
concentrations reported by Spittler were greater than those
reported by CLP for 35 samples, 3 samples analyzed by Spittler
were less than the concentration reported by CLP, and 1 sample
was not detected by Spittler, but was detected by CLP (i.e. false
negative).
In general, PCBs results obtained by field screening were
conservative. Ninety percent of the results were higher than
those obtained by the CLP method, and only one false negative was
reported. Only one result was an order of magnitude.greater than
the CLP laboratory result. Although correlation coefficients
were low, the data sets were generally in agreement when the
factors affecting comparability are taken into consideration.
The major reasons for poor statistical correlation in results
were attributed to nonhomogeneity of the split samples and the
differences in PCB quantitation.
Based upon this comparison, the Modified Spittler Method was
found to detect PCBs as well as the CLP laboratory. It was
concluded that the Spittler Method produces data of Data Quality
Objective (DQO) Level II/III when compared to DQO Level IV for
CLP analyses. However, the level of data quality generated by
this field screening technique allowed EPA to cost effectively
determine the nature and extent of PCB contamination in a study
area that encompassed approximately 40 stream miles and over
1,000 acres of open water.
5.2.4 Nature and Extent of PCB Contaminated Sediment
Previous investigations and the Sangamo OU2 RI have
delineated in detail the distribution and magnitude of PCB
contamination in sediment within the study area. Most of the
contaminated sediment lies within the upper portion of Hartwell
Lake, specifically the Twelvemile Creek Arm and Seneca River Arm
opposite the city of Clemson. Within the Twelvemile Creek
watershed, minor levels of PCB contamination have persisted in
Town Creek near the Sangamo Weston plant site and in the
-------
Final ROD-Sangamo OU2
June 1994
page 20
impoundments associated with the 3 small dams on Twelvemile
Creek. Within Hartwell Lake, localized accumulations of
contaminated sediment have been detected as far downstream as the
dam, with concentrations generally decreasing with increasing
distance downstream from the Twelvemile Creek Arm.
Concentrations have been decreasing since the mid-1980's,
reflecting a decline in PCB input to the lake as well as the
burial and/or dilution of previously deposited sediment through
mixing with clean sediment.
A summary of the significant findings regarding PCB
contamination in sediment for eight subsections of the study area
is presented below. The sediment investigation results are also
graphically illustrated on Figures 5-3, 5-4, 5-5 at the end of
this section. Bridges and landmarks used in the description of
the study area subsections can be found on these figures or
Figure 5-1 and 1-2.
Tipper Twelvemile Creek Watershed - This section includes
segments of the principal tributaries of- Twelvemile Creek that
collect drainage or other discharges from the Sangamo Weston
Plant Site and the six satellite disposal areas. This section
also includes Hagood Reservoir and tributary reaches associated
with the Midway Dump and Easley-Pickens Landfill. Sampling
conducted for the RI and previous investigations supports the
following conclusions:
Sediment contamination in Twelvemile Creek was limited to
isolated occurrences between the mouth of Town Creek and the
upper (Easley-Central) impoundment; PCB concentrations were
typically
-------
Final ROD-Sangamo OU2
June 1994
page 21
The upper impoundment had the lowest levels of PCB
contamination among the 3 impoundments; concentrations were
typically below detection limits or <0.5 mg/kg.
The middle (Woodside I) impoundment had only low levels of
PCB contamination (<1 mg/kg), most of which was detected in
core intervals below 10 cm.
More highly contaminated sediment has accumulated in the
lower (Woodside II) impoundment; most samples had <1 mg/kg
but concentrations ranged up to 6 mg/kg in core samples.
All 3 of the impoundments are flushed on a periodic basis to
remove sediment and the slightly contaminated sediment
detected in these impoundments will eventually be released
into the Twelvemile Creek Arm of Lake Hartwell.
Upper Twelvemile Creek Arm - This section begins at the Lay
Bridge and extends south, in a downstream direction, to Madden
Bridge (Route 15 Bridge). Historical and RI data indicate that
potentially significant quantities of PCB-contaminated sediment
have accumulated in the upper portion of the Twelvemile Creek Arm
of Hartwell Lake, particularly in the open water section
immediately above the Madden Bridge. Specific conclusions
include:
The most significant accumulations of PCB-contaminated
sediment coincided with depositional areas for fine-grained
materials; in the upper arm, these areas typically occurred
in proximity to the shoreline or in coves.
The uppermost portion of the arm, in the vicinity of Maw
Bridge, has only minor levels of contamination (1-2 mg/kg)
due to the higher water velocities and limited
sedimentation.
The open water area upstream of the Madden Bridge represents
the principal site for deposition of contaminated sediment
in the Twelvemile Creek arm; concentrations have ranged as
high as 153 mg/kg in historical samples and to a maximum of
61 mg/kg in the RI samples.
The magnitude of PCB contamination in the upper arm has
declined significantly since the mid-1980's, reflecting the
decreased input of PCBs to the lake and, in the open water
area, burial of the more highly contaminated sediment
through deposition of several feet of sediment with much
lower levels of contamination. This burial is consistent
with results from the HEC-6 sediment transport modeling
discussed in Section 5.4.
The vertical extent of contamination in the open water area
was not defined. Given that the source of PCB contamination
predated the impoundment of the creek and that releases
continued for many years thereafter, contaminated sediment
probably extends down to the original elevation of the land
surface (and former channel).
-------
Final ROD-Sangamo OU2
June 1994
page 22
Middle Twelvemile Creek Arm - This section begins at Madden
Bridge and extends south to the Hwy 133 Bridge. A variable
distribution of low to moderate levels of contamination was
identified, with higher concentrations generally limited to
depositional areas having finer-grained sediment. Specific
conclusions include:
Sediment PCB concentrations in the middle portion of the arm
have varied over time, but since 1985 have consistently been
greater than 1 mg/kg and in most cases greater than 3 mg/kg.
RI data show surface sediments contaminated with 1-3 mg/kg
of PCBs and slightly higher concentrations in deeper
sediment from core samples.
The more highly contaminated sediment was encountered in
samples from depositional areas such as Moore Bend and the
open water area upstream of the Highway 133 Bridge.
PCB concentrations in samples collected from depositional
areas show a downward trend since 1988, indicating decreased
input of PCBs into the lake and that more highly
contaminated sediments have either been buried, resuspended
and transported away, or diluted through mixing with cleaner
sediments.
Alternating sequences of vertically increasing or decreasing
contamination observed in the majority of sediment cores
from the Twelvemile Creek arm complicate the interpretation
of the vertical extent of contamination. This vertical
extent was not defined in the middle segment of the arm.
Lower Twelvemile Creek Arm - This section begins at the Hwy 133
Bridge and extends south to the U.S. Hwy 123 Bridge adjacent to
Clemson. The lower portion of the Twelvemile Creek arm has also
been a depositional area for contaminated sediment, as evidenced
by the widespread PCB contamination in this area. A summary of
the major findings include:
RI data show PCB concentrations in surface sediments
generally in the range of 1-2 mg/kg; higher concentrations
were detected in core samples, with maximums in the 18-21
mg/kg range. Concentrations generally decreased with
increasing distance downstream but localized accumulations
of more highly contaminated material were identified.
The vertical limits of contamination generally occurred in
the upper 30-40 cm of sediment in most of the historical and
RI sediment cores.
The more highly contaminated sediment in the lower arm was
usually detected in samples collected from within or in
close proximity to the former submerged channel of
Twelvemile Creek.
A decline in sediment PCB concentrations was observed at the
mouth of Twelvemile Creek, reflecting the influence of the
-------
Final ROD-Sangamo OU2
June 1994
page 23
Keowee River arm, which has a much higher average flow than
Twelvemile Creek.
PCB concentrations in the shallow sediment (0-15 cm) of the
lower arm have declined significantly, reflecting the
decreased input of PCBs and/or that the more highly
contaminated sediment detected in previous investigations is
getting buried by or diluted through mixing with cleaner
sediment.
The persistence of relatively high levels in deeper sediment
(down to 40 cm) supports conclusions based on the HEC-6
modeling that only limited amounts of sedimentation occur in
the lower portion of the Twelvemile Creek arm.
Keowee Arm - The Keowee Arm is a branch of Lake Hartwell that
meets the Twelvemile Creek Arm just north of Clemson. Historical
data and the Phase I/Phase II sampling results from the Keowee
River Arm indicate the following:
RI data have shown generally low levels of contamination (<2
mg/kg) in the lower part of the Keowee arm, most likely
related to hydrodynamic interactions with the Twelvemile
Creek or Seneca River arms.
Historical data included higher levels of contamination (6-7
mg/kg) in core samples collected from locations farther
upstream in the Keowee arm, possibly indicating an
additional PCB source in the upstream portion of the Keowee
arm. This contamination is probably not related to the
Twelvemile Creek source because of its occurrence several
miles upstream in the Keowee arm and because flow in the
Keowee is several times greater than flow in Twelvemile
Creek. Data were insufficient to characterize the magnitude
or determine the location of this source.
Seneca River Arm/Upper Section of Hartwell Lake - This section
begins at the Hwy 123 Bridge and extends south to the Route (Rt)
37 Bridge. PCBs were detected throughout this area as concluded
below:
Surficial sediments had contamination in the range of 1-2
mg/kg everywhere except the excavated channel opposite the
Clemson University diversion structures.
Deeper contamination was encountered in the former Seneca
River channel and the open water area just upstream from the
Highway 37/Southern Railroad embankment. Deposition of more
highly contaminated sediment near the Highway 37 embankment
was attributed to the hydrodynamic influences associated
with the constriction of the channel at this location.
Residual PCB contamination likely exists in the former
Seneca River channel located on the western edge of the
Clemson University campus, behind the diversion structures.
-------
Final ROD-Sangamo OU2
June 1994
page 24
The vertical extent of contamination was defined and
generally limited to depths of 20-40 cm. A good correlation
between the depth of contamination in cores collected in
1988 and for the RI (in 1991-1992) indicate that only
limited sedimentation has occurred in this portion of the
lake during this 3 to 4-year period.
Maximum PCB concentrations in sediment core samples declined
from levels of 16 to 18 mg/kg in 1988 to concentrations of 5
to 11 mg/kg in 1991/1992; explanations for the decline, if
predicted low sedimentation rates are accurate, include
spatial variability in the distribution of contamination and
continued mixing of the vertical sediment profile.
Lake Hartwell Proper - This section includes the remainder of
Lake Hartwell, beginning at the Rt 37 Bridge and extending south
to the Hartwell dam. Historical sampling in this area indicate
that PCB contamination has migrated as far downstream as the
Hartwell dam. This was not confirmed during the RI. The
occurrence and magnitude of PCB contamination in Lake Hartwell
declined significantly downstream of the confluence of the Seneca
River Arm and Twenty-six Mile Creek, reflecting the input of
significant quantities of uncontaminated sediment from Twenty-six
Mile Creek and the various other tributaries. The majority of
widespread contamination in the lower part of the lake was
limited to concentrations below 1 mg/kg.
5.2.5 Other Contaminants in Sediment
A limited number of sediment grab and core samples were
collected for full-screen TCL/TAL analyses. No significant
contaminants other than PCBs were detected in these analyses.
Specific conclusions include:
No volatile organic contaminants were detected in any of the
samples other than an estimated concentration of acetone
(2200J ^g/kg) in one duplicate sample.
A number of semivolatile organic compounds were detected in
one of the samples, mostly in low, estimate concentrations
that were not indicative of significant contamination.
Several unidentified semivolatile organic compounds were
reported in sediments at 4 sampling stations, ranging in
concentration from 2 mg/kg to 200 mg/kg. These compounds
were most likely petroleum hydrocarbons related to urban
run-off or boating discharges of fuel and oil.
Several pesticide compounds (DDE, DDD, and DDT) were also
detected, in low, estimated concentrations of 0.0024 to
0.053 mg/kg.
Metals analyses did not identify any sediment with elevated
concentrations of heavy metals; the following metals were
not detected: antimony, arsenic, cadmium, mercury, selenium,
and thorium.
-------
Headwaters
10
8J
£. 8
S
4
2
Maw Bridge
(Rt. 337)
I
Madden Bridge
(Rl. 15)
Hwy133
Bridge
I
o
US 123
Bridge
Hwy93
Bridge
O °0
80°
O
a o°o
00° n%° O C
..o...rt 8 9 ...&..
a
6
o 8 o - a
°
O
° 6°
Op0 o °°
om ° o °
3 ° O O. n ..
o
o
^H 1i-F
M L K ' J
30.000
Gb
.°o
o
o °oo.
o o o o
o ° o
o a
o
-t-
Rt.37
Bridge
I
-i 'B
\ H
R O
10,000
O Phase I Sample - CLP Data
O Phase II Sample - Field Screening Data
1 I '
p 'o
20,000
40,000
B I
50,000
Downstream Distance Along Thalweg from Headwaters of Twelvemlle Creek Arm of Hartwell Lake
Figure 5-3
PCB Distribution for Grab Samples
60,000 ^
8
i
CO
PJ (D 3
tq 5
a> H
u> o
to vo G
1/1 rf* to
-------
HiadwatorLakeHartwell
NOfTTU' HOC
WOftMSL
I Normal Pool Elevation
Mwr Bridge (Rl 337)
Madden Bridge (Rl. 15)
Mean PCB Concentration per Transect Interval
Hwy 133 Bridge
US 123 Bridge Hwy 93 Bridge
Rl 37 Bridge
10,000
NOTE: Man PCB conMntntbnt wen datgrmlMd by ntngtig U raulti tor a given eampto depth
Menal tor in coring ibttom "long tqtttn tnnuct.
20,000 30,000 40,000
Dmmetreim Dlttintt Along Thetwej (ram Hudwaten of Twil»emlle Creek Arm ol rhrhrell Uke (H)
80,000
60,000
H-
S
f
CO
Figure 5-4
Vertical Distribution of Mean PCB Concentration per Transect Interval
0) (D §
U) O
(T) H
VD O
tO VD Cj
-------
Maximum PCB Concentration per Transect Interval
Hwdwater Uk< Hartvnll
I Normal Pool Elivatlon
t 660ftMSL
Maw&ldge(R1.337)
Madden Btldge(Rt. 15)
H*y133Bi1dg»
i
US123Brldg» Hwy 93 Bridge
f.
m.378fMgi
10,000
NOT!: Ttit nwdrrom PCB concaitnUon rtprmntt thi imlmm Mtdn) mul lor umpht
cenict«) from a glm nmpfc diptli Intoml for ill coring ibllont ik»g glvni (nnucl.
20,000 30.000 40.000
OownrtrMm Dlttinci Along Tnilwig from Hiidmtin ot Tfcilnmlli Critt Ann ol Hirtwill Liki (ft)
50.000
80,000
No umilii for glnn Intinril
(--.I PCB Concintntlon <1 mj/)cg
[.'.J PCB Concentration Sl< t mg/kg
PCB Concintntlon J5<10 mg/kg
PCS Contintntloft 210
5
CO
Figure 5-5
Vertical Distribution of Maximum PCB Concentration per Transect Interval
0) 0> 3
(JQ o
n> H
\o o
to vo cj
»J rf* tO
-------
Final ROD-Sangamo OU2
June 1994
page 28
5.2.6 Summary of Surface Water Investigations
Surface water quality investigations have focused on the presence
of PCBs in waters downstream of the Sangamo Weston Plant Site,
the OU1 offsite disposal areas, and/or the Midway dump and
Easley-Pickens landfill. Relevant conclusions from these
historical investigations and the RI studies include:
The most frequent and highest levels of PCB contamination
(1.7-20.3 itg/1) were detected in surface water samples
collected from locations downstream of the Sangamo Weston
outfall on Town Creek.
PCBs were either not detected or present in concentrations
<1 fig/I in surface water samples collected from drainage
ditches and tributaries associated with the OUl offsite
disposal areas (Breazeale, Cross Roads, Dodgens, Trotter,
and Welbora disposal areas). These areas are apparently not
releasing appreciable levels of contamination to Twelvemile
Creek or its tributaries.
Based upon NPDES data, PCB concentrations in discharge from
the Sangamo Weston wastewater treatment plant outfall on
Town Creek during the mid-1970's averaged 13 /*g/l and ranged
as high as 235 jtg/1; NPDES monitoring data from this outfall
for 1992 reported PCB concentrations in the range 0.4 to
40.3 /*g/l. This continued release of PCBs to Town Creek and
the downstream portions of OU2 are expected to terminate
once the OUl remedial action at the Sangamo Weston plant
site has been completed.
PCBs were not detected in any of the surface water samples
collected for Phase I of the RI; detection limits for these
unfiltered samples ranged from 1.2 to 1.3 pg/1.
A limited number of these RI samples were also analyzed for
full TCL analyses; no volatile organics, extractable organic
compounds, pesticides or cyanide were detected, and
inorganic (metals) concentrations were not indicative of
contamination.
PCBs were not detected in any of the storm event samples
collected for Phase II of the RI; detection limits for these
unfiltered samples ranged from 1.2 to 1.3 pg/1.
5.3 BIOLOGICAL INVESTIGATION SUMMARY
Field activities for the Biological Investigations were
conducted by the COE in the Spring 1991, 1992, and 1993. The
field activities were divided into two individual study areas;
the Twelvemile Creek watershed and Lake Hartwell. The specific
methods employed and results from these two investigations are
summarized in the following sections.
The information gathered from the Biological Investigations
conducted in the Twelvemile Creek watershed further support the
-------
Final ROD-Sangamo OU2
June 1994
page 29
conclusions made in the sediment component of the RI that: 1)
the Sangamo Plant Site is the primary source of PCB contamination
in Twelvemile Creek; and 2) the contribution of PCB contamination
in the Twelvemile Creek watershed from the six satellite disposal
areas is negligible. The aquatic biota in Twelvemile Creek
appear to have suffered from the influence of man's activities.
Although PCBs have likely contributed to these perturbations
producing cumulative effects, other factors have no doubt had an
additional impact on these animals at the population and
community levels.
Fish in Lake Hartwell continue to be contaminated with PCBs
often at levels that exceed the FDA safe tolerance limit of 2
mg/kg. PCBs appear to be the only pollutant of concern in Lake
Hartwell fish. The highest concentrations of PCBs in fish are
located in the Twelvemile Creek Arm. PCB concentrations in fish
species that are generally non-migratory decrease both downstream
and farther from the Twelvemile Creek Arm. Migratory fish
species have PCB concentrations that are similar throughout the
reservoir and are high enough to be a concern.
PCBs were detected at all levels of the food chain in both
the Twelvemile Creek watershed and Lake Hartwell. Pathways for
PCB uptake in aquatic biota of Lake Hartwell include ingestion
through the food chain and contact, most likely through the
gills. Exposure results from PCBs being adsorbed to sediments
and transported down Twelvemile Creek into Lake Hartwell.
The health of fish in Lake Hartwell does not appear to be
affected at the population level for fish that have PCB
concentrations at the current levels (around 5 mg/kg). However,
there is evidence that as concentrations increase to greater than
20 mg/kg fish health can be affected. Largemouth bass was the
target species used during these investigations. The sensitivity
or tolerance of other fish species to PCB contamination is not
known.
5.3.1 Twelvemile Creek Watershed Investigations
Field investigations in the Twelvemile Creek watershed were
conducted during Spring 1992 to: 1) determine the concentration
of PCBs in one primary sport fish species and one bottom feeding
species; 2) assess the condition of the fish community by
determining the Index of Biotic Integrity (IBI); 3} assess the
condition of the macroinvertebrate community by using Benthic
Rapid Bioassessment techniques; 4) assess fish health of
redbreast sunfish by using bioindicators and the Health
Assessment Index (HAI); 5) assess composition and fate of
harvest to determine public utilization by using a creel survey;
and 6) determine on-going PCB contamination due to transport of
drifting organic matter and bioaccumulation by a sediment
dwelling, filter feeding organism, the freshwater clam (Corbicula
fluminea).
Twelve sampling stations were established in Twelvemile
Creek and three uncontaminated tributaries. These sample
stations are shown in Figure 5-6. The length of each station was
-------
Final ROD-Sangamo OU2
June 1994
page 30
Figure 5-6
Twelvemile Creek Drainage Sampling Stations
-------
Final ROD-Sangamo OU2
June 1994
page 31
from 0.3 to 0.5 miles to encompass gross habitat types and ensure
adequate sample size. Stations were permanently marked with
"Carsonite" markers located in the creek bank at the upper and
lower boundaries. All 12 stations were not involved in each sub-
element of the overall investigation (i.e. PCB tissue analysis,
IBI, HAD .
Twenty fish were collected for PCB contaminant analysis from
each of the stations in the Twelvemile Creek watershed and one
tributary reference site. Redbreast sunfish (Lepomis auritus)
were used as a sport fish species for assessment of human health
and environmental risk. Redbreast sunfish were chosen based on
abundance and likelihood of use by anglers. Ten redbreast were
collected per site. Ten northern hogsuckers (Hypentelium
nigricans) were collected per site to determine PCB
concentrations in a bottom feeding species to assess
environmental risk. Northern hogsuckers were selected based on
abundance and tropic level. Largemouth bass and bluegill were
substituted for redbreast sunfish at three stations.
Concentrations of PCBs in fish collected in the Twelvemile
Creek watershed were highest in Town Creek and Twelve Mile Creek.
Four of the 12 stations sampled had PCB values exceeding 2.0
mg/kg. Three of these stations (5, 8, and 12} were downstream
from the Sangamo Plant Site. Station 4, the other station having
high PCB values, was located just upstream from the confluence of
Town Creek and Twelvemile Creek and was probably affected by fish
migrating from both creeks. Stations upstream from station 4
could not be affected by migration due to a low head dam located
at the upper boundary of station 4. Mean PCB concentrations
ranged from a high of 12.5 mg/kg for northern hogsuckers in Town
Creek to a low of 0.88 mg/kg in redbreast sunfish in Twelvemile
Creek for stations downstream from the Sangamo Plant Site. The
remaining stations located in tributary streams or upstream from
the Sangamo Plant Site had lower PCB concentrations in fish than
those located downgradient from the plant site.
Several non-diagnostic studies which employed a metric/index
system were conducted to assess the health or condition of
individuals, populations and/or communities in the Twelvemile
Creek watershed. The benthic macroinvertebrate community of
Twelvemile Creek watershed was evaluated using EPA's rapid
bioassessment protocols. Samples were collected at a reference
site and the remaining sites shown on Figure 5-6. Eight of the
nine stations sampled in the Twelvemile Creek drainage were
classified as moderately impaired. Station 11, located upstream
of the Sangamo Plant site on Town Creek, was classified as
non-impaired and had high values for taxa richness and exhibited
an abundance of intolerant taxa. Sites immediately downstream of
the Sangamo facility exhibited evidence of greater impairment
than did upstream or tributary sites.
The Index of Biotic Integrity (IBI) was used to assess the
general health of the fish community in the Twelvemile Creek
watershed. The IBI index is based on a set of metrics determined
from species composition and abundance for a fish community.
These metrics focus on basic ecological characteristics of the
-------
Final ROD-Sangamo OU2
June 1994
page 32
fish fauna and have been used effectively in stream monitoring
programs for the past decade. Nine stations in the Twelvemile
Creek watershed and three reference sites were sampled. The IBI
was effective for ranking the reference sites as expected based
on no identifiable impacts at station 1 (IBI 40), known impacts
of agricultural runoff at station 2 (IBI 30), and a known point
source discharge at station 3 (IBI 12). Therefore, it appeared
appropriate to use the IBI to assess the relative health of fish
populations in the Twelvemile Creek watershed with confidence.
The IBI scores indicated that all stations sampled have been
impaired to some degree. Most stations throughout the watershed
were classified as either "fair" or "good" indicating moderate
negative effects on the fish communities that could have resulted
from habitat degradation from man's activities. Only the two
reference stations 2 and 3 that were identified as having known
impacts and station 5, located in Twelve Mile Creek downstream
from Town Creek, were classified as "poor" or "very poor." The
IBI classification at station 5 ("poor") and a comparison of
relative health between sample stations and reference stations
indicated that station 5 has been severely impacted. Both
habitat degradation and water quality impacts including PCB
contamination likely contributed to this adverse impact.
Both the macroinvertebrate and fish communities have been
impaired throughout the Twelvemile Creek watershed. Assumably,
these impairments have resulted from PCB discharges, habitat
degradation, particularly siltation due to erosion from
development, siviculture and agriculture, and poor water quality
from both point and non-point discharges throughout the
watershed.
The Health Assessment Index (HAD was employed to assess the
health of redbreast sunfish at selected stations in the
Twelvemile Creek drainage and a reference site. HAI is an
autopsy based index that uses divergence from the normal of
various organs, tissues and blood parameters to evaluate fish
health. The HAI scores for redbreast sunfish were variable
throughout the watershed ranging from 39.3 to 62.7. Station 12
located in Town Creek immediately downstream from the Sangamo
site had the lowest HAI score, 39.3, indicating healthy fish at
this site.
It is not evident from evaluating the HAI scores that the
health of redbreast sunfish has been effected by PCB
contamination. HAI scores were variable throughout the drainage.
However, scores in the range seen during this investigation were
not considered high. It is likely that health problems that
would manifest themselves to visual inspection through autopsy do
not occur in fish contaminated at the lower PCB concentrations
seen in redbreast sunfish in the Twelvemile Creek drainage.
Drift net samples were used to collect detritus (floating
organic matter) in Town Creek downstream of the Sangamo site and
in Twelvemile Creek at Lay Bridge. Additionally, cages of
"clean" Corbicula fluminea (fresh water clams) were placed at
-------
Final ROD-Sangamo OU2
June 1994
page 33
these sites for 28 days to determine rates of bioaccumulation of
PCBs by analyzing the muscle tissue. Drift net samples of coarse
particulate matter collected from Town Creek, immediately
downstream of the Sangamo Plant Site had detectable PCB
concentrations ranging from 0.066 to 0.51 mg/kg. PCB
concentrations in corbicula collected at this station ranged from
0.75 to 0.88 mg/kg. Corbicula collected from the station
immediately downstream of the Lay Bridge had PCB concentrations
that ranged from 0.45 to 0.57 mg/kg. From these results, it can
be concluded, that even though PCBs have not been detected in
water samples collected from the study area, transport of PCBs
down Town Creek and Twelvemile Creek is occurring.
5.3.2 Lake Hartwell Investigations
Studies in Lake Hartwell were conducted during Spring 1991,
1992, and 1993 to: 1) determine the concentration of PCBs in
primary sport and forage fish species; 2) assess fish health of
largemouth bass using bioindicators and HAI; 3) determine PCB
concentrations in seston and macroinvertebrates; and 4) assess
composition and fate of harvest to determine public utilization
using a creel survey.
Fish species selected for Lake Hartwell contaminant analysis
included: 1) sport and/or commercial species, 2) fishes that
seasonally occupy different habitats or exhibit seasonal
migrations, and 3} primary forage species. Species selected for
collection and analysis were: largemouth bass, hybrid bass,
channel catfish, threadfin shad (Dorosoma petenense), gizzard
shad (Dorosoma cepedianum), blueback herring (Alosa aestivalis),
and bluegill (Lepomis macrochirus).
A total of 134, 144, and 144 fish were collected from six
locations throughout Lake Hartwell during spring 1991, 1992, and
1993, respectively. These six stations are illustrated on Figure
5-7 and were selected to represent a gradient of PCB
contamination in fish ranging from the most impacted (SV-107 in
the Twelvemile Creek embayment) to a background location (SV-641)
in the Tugaloo Arm. These fish were analyzed for PCB
concentrations in standard U.S. FDA fillets with skin on, de-
scaled with rib cage (except for catfish where the skin was
removed). Additionally, the 60 largemouth bass collected in 1991
were analyzed for priority pollutants. Composite samples of
forage fish, including bluegill, threadfin shad, gizzard shad,
and blueback herring, were collected from SV-107 (highly
impacted), SV-532 (intermediate zone), and SV-641 (background).
PCB concentrations in fish have historically been highest in
the Twelvemile Creek Arm of Lake Hartwell (SV-107) and decrease
in a downstream direction. This is expected since PCBs enter
Lake Hartwell bound to sediments that are transported down
Twelvemile Creek, as discussed in Section 5.2 of this ROD.
Results of PCB analyses collected from 1990-1993 for largemouth
bass, channel catfish, and hybrid bass are graphically
illustrated on Figures 5-8, 5-9, and 5-10, respectively. These
years were comparable since analyses were completed in the same
manner. Although 4 years of data may be insufficient to justify
-------
Figure 5-7
Lake Hartwell Fish Sampling Stations
vl.
SV-107
SV - 642
-
s
p
CO
,-
(D H
VO O
|u> vo q
*> K>
-------
Final ROD-Sangamo OU2
June 1994
page 35
meaningful trend analysis, PCS levels decreased significantly in
largemouth bass at SV-107 (Figure 5-8). This trend was not
significant at other stations. As in past samples, PCB levels in
largemouth bass decreased in a downstream direction. PCB
concentrations in channel catfish have also decreased at stations
farther from the source in Twelvemile Creek (Figure 5-9).
However, there has been little change in the PCB concentrations
in channel catfish over the past 4 years
The distribution of PCB concentrations in hybrid bass
throughout the lake remains a concern. Hybrid bass collected
farther from the source in Twelvemile Creek have higher PCB
concentrations than other species, often exceeding the FDA
tolerance level of 2 mg/kg (Figure 5-10). This is thought to be
a result of their migratory behavior and is also seen in the
samples of walleye, another migratory species having high
concentrations of PCBs (mean 3.48 mg/kg) from SV-642 at Hartwell
Dam.
Contaminants other than PCBs were not identified at
concentrations of concern nor were they widely distributed. A
comparison of the priority pollutants to the established USFDA
action levels showed that no chemical exceeded any of the safe
tolerance standards except PCBs. Priority pollutants having no
standards were compared to SCDHEC's fish tissue trend data base.
Out of 60 largemouth bass there were 17 different compounds
(mostly petroleum and organic solvents) that occurred only once
in a single fish and several others that occurred only twice.
There was no pattern of contamination and PCBs are considered the
only contaminant of concern.
Forage fish species had mean PCB concentrations ranging from
3.00 mg/kg in threadfin shad to 12.43 mg/kg in gizzard shad at
SV-107. PCB concentrations in forage fish decreased at stations
farther from the Twelvemile Creek Arm. This supports the
hypothesis that migratory species bioaccumulate PCBs while
located in the upper Seneca River rather than through the food
chain as forage species migrate out of the more heavily
contaminated area.
The effects of PCB contamination on fish health and
reproductive competence in Lake Hartwell were examined utilizing
the HAI, biological indicator analyses, and age and growth
analysis. Largemouth bass were the target fish species sampled.
Comparisons of fish health using the HAI as the indicator were
made among all six stations sampled in Lake Hartwell in 1992 and
1993. Differences in the HAI were seen between stations.
Station SV-107 consistently had the highest HAI score for each
year sampled. Station SV-106, the closest downstream station,
also had high HAI scores. During 1992, these two stations
exceeded 80 while stations farther from Twelvemile Creek ranged
from 42 to 65. Fish from Twelvemile Creek (SV-107) and Martins
Creek (SV-106) had the highest occurrence of fatty liver
conditions. Accumulation of fat in the liver can be associated
with exposure of fish to PCBs. The remaining stations had a
greater percentage of normal livers with little occurrence of
fatty livers. Although HAI scores only indicate the relative
-------
Final ROD-Sangaxno OU2
June 1994
page 36
SV-107 SV-106 SV-532 SV-535
STATION
SV-642
SV-641
Rgure 5-8
PCB Concentrations In Largemouth Bass Fillets (1990-1993)
8 T
SV-107 SV-106 SV-532 SV-535
STATIONS
SV-642
SV-641
Figure 5-9
PCB Concentrations in Channel Catfish (1990-1993)
-------
Final ROD-Sangamo OU2
June 1994
page 37
SV-107 SV-106 SV-532 SV-535
STATION
SV-642
SV-641
Figure 5-10
PCB Concentrations in Hybrid Bass Fillets (1990-1993)
-------
Final ROD-Sangamo OU2
June 1994
page 38
health of fish without indicating cause, decreases in the HAI
scores with distance from the source was consistent with declines
in PCB concentrations in fish at these stations.
Biomonitoring using biological indicators of fish health and
reproductive competence was conducted at both Twelvemile Creek
and Lake Hartwell study sites. The selected suite of
bioindicators measured in this investigation included, among
others, organ dysfunction, nutritional status, gonadal analyses,
detoxification enzymes, and concentrations of reproductive
steriods. The redbreast sunfish was the target species in the
Twelvemile Creek watershed, while largemouth bass was the target
species in Lake Hartwell. Bioindicator analyses conducted in
Twelvemile Creek indicated that PCB contamination probably
contributes, but is not the sole cause of impacts on fish health.
There was evidence of biochemical and physiological differences
between Lake Hartwell stations with largemouth bass collected
from SV-107 exhibiting abnormal responses to several
bioindicators. The most obvious physiological difference in
largemouth bass between stations related to the functions of the
liver. Additionally, age and growth analysis indicated that
largemouth bass from SV-107 had a slower growth rate and were
less plump having lower weights per unit of length.
Cages of clean corbicula were deployed for 28 days at two
locations in the Twelvemile Creek Arm of Lake Hartwell to
evaluate the uptake of PCBs by this filter feeding organism. The
two locations were established in the "goose-neck bend" area
between the Maw (Hwy 337) and Madden (Hwy 15) Bridges and near
the lower end of Twelvemile Creek Arm, just above the Keowee
River confluence. PCBs were detected in corbicula muscle tissue
at concentrations ranging from 0.48 mg/kg at the lower Twelvemile
Creek Arm station to 0.56 mg/kg at the goose-neck bend station.
Adult mayflies were also collected from the Twelvemile Creek Arm,
just above the Hwy 133 Bridge and were found to have 2.6 mg/kg of
PCBs.
5.3.3 Assessment of Resource Use
Public utilization of the fishery resources was assessed by
creel surveys on Twelvemile Creek and Lake Hartwell. Twelvemile
Creek was surveyed at major access points upstream to the
confluence of the Middle Fork and North Fork. Lake Hartwell was
divided into eight sampling areas that allowed analysis of angler
use relative to PCB concentrations in fish within each area.
Additionally, interviews provided data from which to evaluate the
economic importance of the resource. The interviews also
included questions to assess how fishermen used or consumed their
harvest and fisherman awareness of the health advisory. These
questions were the same for both the Lake Hartwell and the
Twelvemile Creek creel survey.
Angler use of Twelvemile Creek appeared to be minimal. Only
21 interviews were obtained during this investigation. Some
anglers were interviewed more than once indicating there may be a
small localized group of anglers that frequent Twelvemile Creek.
Most anglers interviewed were located at bridge crossings near
-------
Final ROD-Sangamo OU2
June 1994
page 39
the confluence of Twelvemile Creek and Lake Hartwell. They were
encountered from March through August only. There was evidence
of some fishing in more upstream areas from the presence of the
types of trash that is customarily associated with bank
fishermen. However, anglers were not encountered at these
locations indicating low use.
Most anglers (66%) do consume fish caught in Twelvemile
Creek. Channel catfish and sunfish made up most of the harvest.
Consumption estimates for those anglers that indicated they ate
fish were higher for Twelvemile Creek (37 g/day) than for Lake
Hartwell (28 g/day). However, the total estimated harvest for
Twelvemile Creek in 1992 was only 116.2 kg (256 Ibs.).
Public utilization of the fishery resources in Lake Hartwell
was also assessed by creel survey. A total of 415,839 fish were
caught in 728,489 angler-hours of effort. A total of 239,726
(58%) were released, while 176,112 fish weighing 115,320 kg
(253,703 Ibs.) were harvested. The top five species sought
included largemouth bass, anything, striped bass, hybrid bass,
and crappie. Largemouth bass and crappie comprised the largest
percentage by number of fish harvested, while largemouth bass and
hybrid bass comprised the largest percentage by weight.
The number of fish harvested was greatest in areas which
comprised the Twelvemile Creek, Keowee River, and Seneca River
Arms which encompass the most heavily contaminated portion of
Lake Hartwell and much of the area under the most extreme
advisory. Additionally, the Seneca River Arm area had the
highest harvest by weight. By weight, 29% of the largemouth
bass, 48% of the hybrid bass, and 66% of the striped bass
harvested from Lake Hartwell came from the Seneca River Arm.
Despite the advisory concerning fish consumption, anglers
continue to harvest the largest portion of fish from Lake
Hartwell from the areas most affected by PCB contamination.
Eighty-five percent of all anglers responding were aware of
the fish consumption advisory. Sixteen percent had stopped
eating fish because of the advisory. Sixty-seven percent of
anglers interviewed indicated they ate fish from Lake Hartwell.
The mean monthly per capita consumption for Lake Hartwell anglers
that eat fish was estimated to be 0.85 kg (1.86 Ibs). The mean
monthly per capita consumption estimate for the Twelvemile Creek
Arm, the most contaminated area, was 0.9 kg (2 Ibs) near the lake
average.
The health advisory on fish consumption appeared to
influence consumptive habits of less than one-fourth of the
anglers interviewed. Most anglers cited other reasons why they
had stopped eating fish caught from Lake Hartwell. Most anglers
indicated that they practiced catch-and-release, presumably for
sport reasons. Some anglers indicated that a stronger advisory,
a ban on fishing, seeing people get sick from eating fish and/or
documentation of increased contamination would affect their fish
consumption habits.
-------
Final ROD-Sangamo OU2
June 1994
page 40
5.4 CONTAMINANT FATE AND TRANSPORT SUMMARY
An integral component of the Sangamo OU2 Rl was to develop
predictions regarding the continued deposition, migration, and
accumulation of PCB contaminants in the sediment, water, and
aquatic biota of Lake Hartwell. These predictions were
accomplished through the use of specialized numerical-based
models that considered the environmental processes and
contaminant characteristics that control PCB migration and
accumulation as well as the intermedia transfer and degradation
of PCBs.
This comprehensive modeling effort focused on changes to an
approximate 10 mile reach of the Twelvemile Creek Arm of Lake
Hartwell over a 30-year period and consisted of: 1) future
sediment transport/deposition using the HEC-6 computer model,
"Scour and Deposition in Rivers and Reservoirs, Version 4.0",
which was developed by the Hydrologic Engineering Center of the
COE; 2) future PCB fate and transport using the "Water Quality-
Analysis Simulation Program (WASP-4), Version 4.0" developed and
supported by EPA; and 3) future aquatic bioaccumulation modeling
using the "Food and Gill Exchange of Toxic Substances (FGETS),
Version 3.0" developed by EPA. The results of these modeling
efforts are summarized below.
5.4.1 Future Sedimentation
The objective of sediment transport simulation was to
determine rates at which presently contaminated sediments would
be buried by relatively clean sediments from tributary drainages
over the next 30 years. This simulation assumes that the
historic hydrologic flow regime and sediments loading to the
system remains unchanged. An overview of sedimentation regimes
within the Twelvemile Creek Arm and upper Seneca Arm is
summarized below. Figure 5-11 provides a bed profile of the
Twelvemile Creek Arm. Figure 5-12 provides an illustration of
the average sediment burial rates as predicted by the HEC-6
model.
The model identified three distinct sediment transport
regimes. The first regime extends from just below the Woodside
II impoundment (Transect T19) to a point approximately 2,400 ft
upstream of the Maw Bridge (Transect T16). This reach is
upstream of the backwater influence of the Hartwell impoundment
(normal pool elevation of 660 ft MSL) and, therefore, behaves as
a river. Energy gradients, bottom shear stresses, and turbulence
intensities are relatively large in this reach; sediment
transport potential is therefore high and sediments of all sizes,
sand included, are carried by the flow as suspended sediment.
The HEC-6 results predict excess sediment transport potential,
resulting in a small amount of scour along this reach.
The second regime extends from upstream of the Maw Bridge to
just above the Hwy 133 bridge; this reach represents a zone of
transition from a high energy fluid environment (river) to a low
energy environment (lake). As the transition from river to lake
occurs, energy gradients, bottom shear stresses, and turbulence
-------
670
66(
60
50
1992 Initial Profile
(Route 337)
Maw Bridge
(Route 15)
Madden Bridge
Hwy133
Bridge
Highway 93
Route 37
Bridge
40 30 20
Distance Upstream from Transect T1 (1000 ft)
10-year Simulation ' 20-year Simulation
Figure 5-11
Simulated Bed Profile Twelvemile Creek Arm
10
30-year Simulation
CO
r* *
Si,
u> i
H * I
-------
Final ROD-Sangamo OU2
June 1994
page 42
SOUTHERN RAILROAD
Figure 5-12
Predicted Average Sediment Burial Rates
-------
Final ROD-Sangamo OU2
June 1994
page 43
levels all decrease, resulting in high rates of sediment
deposition (1.43 to 13.12 cm/yr) as all of the sand and most of
the silt coming into the reservoir are deposited (clays remain as
suspended sediment and continue downstream). The model predicted
a 30-year accumulation of nearly 10 ft of sediment in some areas.
Sedimentation in this area results in the formation of a delta
that advances into the impoundment over time, moving the
headwater location further into the impoundment as well (See
Figure 5-11). With a continuous supply of sediment from the
watershed, this delta eventually will advance below Madden
Bridge.
The third regime extends from upstream of the Hwy 133 bridge
to the Hwy 37 bridge, which marks the downstream extent of the
system modeled. Hydraulically, this portion of the system
behaves as an impoundment with low-energy gradients and
turbulence levels. This quiescent environment favors sediment
deposition; however, only small deposition depths and burial
rates are predicted because most of the sand and silt load is
deposited upstream, and only slow-settling, clay-sized particles
are available for deposition downstream. Most of the clay
entering this reach is deposited, resulting in predicted
deposition depths ranging from 0 to 54 cm and average burial
rates of 0 to 1.8 cm/yr.
5.4.2 Future Sediment and Water Quality Trends
Future PCB fate and transport in the Twelvemile
Creek/Hartwell Lake system was modeled using the WASP4 code
developed by EPA. The modeling was conducted to predict the fate
and transport of PCBs in the Twelvemile Creek/Hartwell Lake
system over the next 30 years. This work was conducted to assist
in the extrapolation of PCB data collected in the present
investigation to future conditions. An overview and results of
the water quality modeling study are summarized below. Figures
5-13, 5-14, 5-15, and 5-16 provide predicted PCB concentrations
in surface sediment initially, at 10 years, 20 years, and 30
years, respectively.
Results from the WASP4 simulations reflect a relatively
complex cycling of PCBs through the system. Initially, PCB
concentrations are relatively high in upstream surface and bottom
sediment segments. Over time, clean sediment from the upstream
tributary is deposited over the contaminated sediments, which
results in burial and dilution of the contaminated sediments.
Because sediment resuspension and deposition can occur
simultaneously, some contaminated sediments from upstream
segments is resuspended in the water column, where advection and
dispersion transport them further into the impoundment. These
contaminated sediments are then redeposited. Because suspended
sediment concentrations decrease with increasing distance into
the impoundment, the rate at which redeposited sediments are
buried is much smaller than at points upstream where they
originated.
The net result of these transport processes is (1) a
significant reduction in PCB concentrations in the surface and
-------
Final ROD-Sangamo OU2
June 1994
page 44
PCB CONCENTRATIONS (ug/U
io4- io5
10*- IO1
to - io1
I - 10
< I
NOTE: PCB conotntratlons expressed in tins of uo/L
for purposs of USP4 xxHllno. Corontrtrtion
can be conwttd to pp* by dividing by 2650.
TRIBUTAR1
OUMe. CENTEH.IK
CROSS SECTION TRANSECT
LOCATION MODIFIED FROK
SAHCWCOUZ-RI
1.0
SCALE IN MILES
Rgure 5-13
Initial WASP4 Surface Sediment PCB Concentrations
-------
Final ROD-Sangamo OU2
June 1994
page 45
PC8 CONCENTRATIONS (ug/L)
io4 - ic1
io>- io«
10* - 18s
10 - i«l
i -10
< i
NOT!: PCS ccnoan trot ions «xpr«ss8d in tens of ug/L
for purposa of IttSM nxJUinj. Conotntrotion
con EM comtrted to pp> by dividing by 2SSO.
LEGEND
TRIBUTART
QUNO. I
CIBSS SECTION flUNSECT
.. LOOTIW MODIFIED FROM
SANCAM OU2-flI
1.0
SCALE IN MILES
Figure 5-14
WASP4 Surface Sediment PCB Concentrations at 10 Years
-------
Final ROD-Sangamo OU2
June 1994
page 46
PCB CONCENTRATIONS (ug/U
to4 - to*
IOJ - 10«
10* - 10*
10 - I8l
I - 10
NOTE: PCS concintrations express*! in t»ns of ua/l
for purpose of WSN mxHllna. Concentration
can M eorw*rt*d to pp* by atViOUig by 3856.
LEGEND
CHANEL CEMBine
CROSS SECTION TRANSECT
.. LOCATION WOIFIED FflW
S*»6»MD«8-*I
1.0
SCALE IN MILES
Figure 5-15
WASP4 Surface Sediment PCB Concentrations at 20 Years
-------
Final ROD-Sangamo OU2
June 1994
page 47
PCB CONCENTRATIONS (ug/U
10' - ID1
iaj-10*
)8«- io3
10 - 19'
1 - 10
< I
NOTE: PCS conontrotions *>pr*sMd in tms of uoA
for purpos* of tUSM xtelina. Conovrtration
eon bi oomwtid to pp« by olvidino by 2S5D.
OUNCL CEKTERLI*
CROSS SECTION 1BMSEC1
,. Loanw MBIFIED turn
SMBMBOUMU
1.0
I
SCALE IN MILES
Figure 5-16
WASP4 Surface Sediment PCB Concentrations at 30 Years
-------
Final ROD-Sangamo OU2
June 1994
page 48
bed sediments in the upper and middle portions of the system, and
(2) an increase in PCB concentrations in surface and bed
sediments near the lower end of the system. Surface and bottom
sediments upstream of the Hwy 133 bridge, which had initial PCB
concentrations ranging from 1,600 to 10,000 /*g/L, are predicted
to decrease to near zero levels within the first 5 years. This
reduction in concentration is due to burial by clean sediments
derived from upstream reaches along with removal by sediment
resuspension.
In the middle reach of the system between the Hwy 133 bridge
and the Hwy 93 bridge, PCB concentrations also decrease over time
for the same reason but at a much slower rate. Initially,
concentrations in this middle reach ranged between 1,800 and
23,000 ng/li. The range is reduced to 400 to 8,000 /tg/L after
5 years, 50 to 2,800 /tg/L after 10 years, and 0 to 40 /tg/L after
30 years. Surface and bottom sediments lying between the Hwy 93
bridge and the Hwy 37 bridge show PCB concentrations increase
over time. Initially, concentrations in these segments ranged
between 400 and 1,400 /*g/L; after 30 years, concentrations in
these segments are predicted to range up to 8,800 /*g/L. This
predicted concentration increase is attributed to hydrodynamic
influences associated with the Hwy 37 embankment.
5.4.3 Bioaccumulation and Future Fish Concentrations
Aquatic bioaccumulation modeling was conducted to predict
future PCB levels in the fish of the Twelvemile Creek/Hartwell
Lake system using the FGETS model. This work was conducted to
assess: (l) how PCB levels in fish would change over.the next
30 years if the contaminated sediments were left unmanaged, and
(2) how long it would take for levels to drop below the current
FDA tolerance level of 2 mg/kg in several species. The results
of the aquatic bioaccumulation modeling study are summarized
below. Figure 5-17 provides a graphical summary of this effort.
FGETS is a time-dependent model that predicts whole-body
fish residues at specified points in time. The outputs of the
model are averaged residues for each year of the fish's life.
FGETS can predict chemical uptake by fish through two principal
pathways: (1) directly from water via respiration, and (2)
through consumption of contaminated food and sediments. For
hydrophobia organic chemicals, such as PCBs, the second pathway
is known to be the more significant one.
Application of the FGETS model requires the construction of
a conceptual model, which includes selection of a representative
food web. Three species of fish were chosen to be included in
the representative food web, based on the species present in
Hartwell Lake, the availability of data for calibration, and the
potential for human exposure. Largemouth bass was selected as
the top predator for the food web model, since it is the most
frequently harvested species and currently has the highest PCB
residues for game (harvest) fish. Gizzard shad and bluegill
sunfish were identified as the two most important forage fish for
Hartwell Lake largemouth bass. Plankton and benthos composed the
lower levels of the food web. Plankton and benthos were assumed
-------
0 YEARS
5 YEARS
10 YEARS
20 YEARS
30 YEARS
- Largemouth Bass
- Gizzard Shad
- Bluegill Sunfish
- Plankton
- Benthos
NOTE:
Predicted PCB concentrations are based on outputs of the FGETS model and
represent averaged whole body flsh residues per fish lifetime. Plankton and
Benthos concentrations represent equilibrium with predicted surface water and
sediment concentrations, respectively.
PCB CONCENTRATIOH (pi
Figure 5-17
Initial and Predicted PCB Concentrations for Selected Biota
Lake Hartwell
CO
UJ O
n> M
u> o
vo rf* 10
-------
Final ROD-Sangamo OU2
June 1994
page 50
to be in chemical equilibrium with the dissolved-phase PCB
concentrations of the water column and particulate-phase PCB
concentrations of the surficial sediments, respectively, as
predicted by the WASP4 water quality model.
Actual modeling consisted of a long-term simulation to
predict future PCB fish tissue concentrations over the next
30 years. Results show PCB concentrations in all components of
the aquatic bioaccumulation model decreased over the 30-year
simulation period (Figure 5-17). PCB concentrations in plankton
and benthos, assumed to be in chemical equilibrium with the water
column and surface sediments, showed a rapid rate of decline
consistent with the WASP4 predictions. In response to decreasing
water column and surface sediment PCB levels, concentrations in
fish declined as well. Mean concentrations in largemouth bass
fillets (10.75 mg/kg initially) declined to 5.16 mg/kg by year 5
and fell to 1.74 mg/kg and below the 2.0 mg/kg FDA tolerance
level by year 10 (year 2003).
Although the mean largemouth bass concentration fell below
the FDA limit by year 10, a longer period of time is required for
concentrations in older age classes to fall below the FDA limit.
PCB concentrations in 10-year-old fish, representing the highest
age class and largest fish in the model, required 12 years (year
2005) to fall below the 2.0 mg/kg FDA limit. By year 30, mean
concentrations in largemouth bass fillets are expected to be very
low, with a mean value of 0.025 mg/kg predicted.
6.0 SUMMARY OF SITE RISKS
CERCLA directs EPA to protect human health and the environment
from current and potential exposure to hazardous substances at
the Sangamo OU2 Site. The Baseline Risk Assessment provides the
basis for taking action and indicates contaminants and the
exposure pathways that need to be addressed by the Remedial
Action. It serves as an indication of what risks the Site poses
if no action were taken. A Baseline Risk Assessment was
conducted to evaluate the potential current and future human
health and ecological impacts associated with exposure to PCBs in
sediment and fish. This section of the ROD contains a summary of
the results of the Baseline Risk Assessment conducted for this
Site. Tables utilized and referenced in this discussion are
included at the end of this Section to facilitate the readability
of this Section.
6.1 CONTAMINANTS OF CONCERN
The Baseline Risk Assessment focused only on PCBs as the
chemical of potential concern. As described in Section 5.0,
full-screen analyses were conducted on a limited number of
sediment and fish tissue (biological) samples. These analyses
did not detect appreciable quantities of volatile organic
compounds (VOCs), semivolatile organic compounds (SVOCs),
pesticides, and/or inorganics (metals).
-------
Final ROD-Sangamo OU2
June 1994
page 51
The human health exposure pathways that were quantitatively
evaluated under this Baseline Risk Assessment were as follows:
1) Ingestion and dermal absorption of PCBs in shallow sediment
by a child and an adult at:
The upper section of Twelvemile Creek Arm,
The upper Twelvemile Creek Watershed,
The Hagood Reservoir, and
Immediately downgradient of the Sangamo Plant Site.
2) Ingestion of PCB-contaminated fish by a recreational
fisherman.
Sediment data were divided into two overall categories in
order to best define exposure. These included sediment sample
locations which were 2.1 m (7 ft) or less below the water
surface, and sediment sample locations which were between 2.1 m
and 3 m (7-10 ft) below the water surface. As discussed
previously, the lake level fluctuates an average of 4 ft, hence
exposing additional sediment. The lower interval was developed
to evaluate this potential exposure. Table 6-1 summarizes the
information used to calculate sediment exposure point
concentrations at the four locations. Sediment data is from the
Phase I and Phase II Sediment Investigation.
The risks posed by the ingestion of PCB-contaminated fish
were evaluated at all six Lake Hartwell fish sampling locations
(SV-107, SV-106, SV-532, SV-535, SV-642, SV-641). These
locations are depicted on Figure 5-7 in Section 5.3 of this ROD.
In addition, risks posed by the ingestion of fish caught in the
Twelvemile Creek watershed, tributaries of the watershed, and a
background location were quantified. Risks were evaluated
separately by fish species and also for all species combined to
account for both individuals who may eat one specie exclusively
and individuals who consume all types of fish. Table 6-2
summarizes the information used to calculate fish tissue exposure
point concentrations at these locations. Fish tissue data are
from the 1991 and 1992 Biological Investigations prepared in
accordance with standard FDA fillet method (i.e. fillet including
rib cage and belly flap with skin on/scales off - except catfish
where skin is removed).
6.2 EXPOSURE ASSESSMENT
For purposes of this Baseline Risk Assessment, it was
assumed that in the future the Twelvemile Creek/Lake Hartwell
system will continue to exist in its present state. All exposure
pathways that are complete under the current land-use conditions
were also assumed to be potentially complete under the future
land-use scenario. The exposure pathways that were
quantitatively evaluated under the current/future use scenarios
are described in Section 6.1 of this ROD.
EPA employed a reasonable maximum exposure (RME) approach to
estimate the potential exposures and associated risks at the
Sangamo OU2 Site. The RME is the highest exposure that is
-------
Final ROD-Sangamo OU2
June 1994
page 52
reasonably expected to occur at the Site. The intent of the RME
is to estimate a conservative exposure case that is still within
the range of possible exposures. The calculation of risk posed
by exposures to PCBs requires the combination of exposure point
concentrations with assumptions regarding frequency, duration,
and magnitude of receptor contact. RME exposure point
concentrations for sediment and fish tissue are listed in Tables
6-1 and 6-2, respectively. The RME exposure parameter
assumptions for sediment and fish that were used to characterize
risk posed by PCBs are listed in Tables 6-3 and 6-4,
respectively. The exposure parameters for ingestion of fish were
derived from results the site-specific recreational angler survey
discussed in Section 5.3.
6.3 TOXICITY ASSESSMENT
For risk assessment purposes, individual chemicals are
separated into two categories of chemical toxicity depending on
whether they exhibit principally carcinogenic (cancer-causing) or
noncarcinogenic effects. The carcinogenic and noncarcinogenic
toxicity of PCBs are discussed below.
6.3.1
Carcinogenic Toxicity of PCBs
EPA assigns weight of evidence classifications to potential
carcinogens. Under this system, chemicals are classified as
follows:
Group A - Human Carcinogen: Sufficient evidence from human
epidemiological studies to support a causal association
between a chemical and cancer.
Group Bl- Probable Human Carcinogen: Limited evidence from human
epidemiological studies of carcinogenicity to humans.
Group B2- Probable Human Carcinogen: Sufficient evidence of
carcinogenicity to animals.
Group C - Possible Human Carcinogen: Limited evidence of
carcinogenicity in animals and an absence of data on
humans.
Group D - Not Classified: Inadequate evidence of carcinogenicity
in animals.
Group E - No Evidence of Carcinogenicity to Humans
PCBs are classified as a Class B2 carcinogen, which
indicates that there is sufficient evidence from tests on animals
to link exposure of PCBs with carcinogenic effects.
Slope factors {SFs) have been developed by EPA for
estimating excess lifetime cancer risks associated with exposure
to potentially carcinogenic contaminants. SFs, which are
expressed in units of (mg/kg-day)'1 or risk per milligram per
kilogram of dose, are multiplied by the estimated intake of a
potential carcinogen, in mg/kg-day, to provide an upper bound
-------
Final ROD-Sangamo OU2
June 1994
page 53
estimate of the excess lifetime cancer risk associated with
exposure at that intake level. The term "upper-bound" reflects
the conservative estimate of the risks calculated from the SF.
Use of this approach makes underestimation of the actual cancer
risk highly unlikely. Slope factors are derived from the results
of human epidemiological studies or chronic animal bioassay data
to which mathematical extrapolation from high to low dose and
from animal to human dose has been applied, as well as statistics
to account for uncertainty (e.g. to account for the use of animal
data to predict effects on humans). A SF of 7.7 (mg/kg-day)'1 for
the PCB Aroclor 1260 was obtained from the Integrated Risk
Information System (IRIS) for use in this Baseline Risk
Assessment.
6.3.2 Noncarcinogenic Toxicity of PCBs
Reference doses (RfDs) have been developed by EPA for
indicating the potential for adverse health effects from exposure
to the contaminants exhibiting noncarcinogenic effects. RfDs,
which are expressed in units of mg/kg-day, are estimates of daily
exposure levels for humans, including sensitive populations, that
are likely to be without risk of adverse effect. Estimated
intakes of contaminants from environmental media can be compared
to the RfD. RfDs are derived from human epidemiological and
animal studies to which uncertainty factors have been applied. A
chronic RfD of 7 X 10"5 mg/kg-day for the PCB Aroclor 1016 was
used in this Baseline Risk Assessment. This daily exposure
estimate has been documented in studies conducted on monkeys to
produce reproductive effects (reduced birth weights).
6.4 RISK CHARACTERIZATION
For carcinogens, risks are estimated as the incremental
probability of an individual developing cancer over a life-time
as a result of exposure to the contaminants of concern, in this
case PCBs. Excess life-time cancer risk is calculated from the
following equation:
Risk = GDI x SF
where:
Risk = a unitless probability (i.e. 2 x 10'5) of an
individual developing cancer.
GDI = chronic daily intake averaged over 70 years
(mg/kg-day).
SF = slope factor, expressed as (mg/kg-day)"1.
These risks are probabilities that are generally expressed
in scientific notation (i.e. 1 x 10"6) . An excess lifetime cancer
risk of 1 x 10"6 indicates that, as a reasonable maximum estimate,
an individual has a l in 1,000,000 chance of developing cancer as
a result of site-related exposure to a carcinogen over a 70 year
lifetime under the site-specific exposure conditions for the
site. EPA's generally acceptable risk range is 1 x 10~* to 1 x
-------
Final ROD-Sangamo OU2
June 1994
page 54
The potential for noncarcinogenic effects is evaluated by
comparing an exposure level over a specified time period with a
reference dose derived for a similar exposure period. The ratio
of exposure to toxicity is called a Hazard Quotient (HQ). An
HQ<1 indicates that an individual's dose of a single contaminant
is less than the RfD, and that the toxic noncarcinogenic effects
from that chemical are unlikely. The Hazard Index (HI) is
generated by adding the HQs for all chemicals of concern that
affect the same target organ within a medium or across all media
to which a given population may reasonably be exposed. An HI<1
indicates that, based on the sum of all HQs from different
contaminants and exposure routes, toxic noncarcinogenic effects
from all contaminants are unlikely. The HQ is calculated as
follows:
Non-cancer HQ = CDI/RfD
where:
GDI = chronic daily intake
RfD = reference dose
GDI and RfD are expressed in the same units (mg/kg-day) and
represent the same exposure period.
The carcinogenic and noncarcinogenic risks associated with
direct contact with or incidental ingestion of PCB contaminated
sediment and ingestion of PCB-contaminated fish are summarized in
Tables 6-5 and 6-6, respectively. Adverse human health risks
resulting from direct contact or incidental ingestion with the
sediment are unlikely to occur, however, sediments are a
continuing source of contamination in the aquatic biota of the
study area. Exposures associated with the ingestion of fish
caught from all sampling stations resulted in unacceptable risks
ranging from 10"2 to 10"*. The highest cancer risk of 4 x 10~2 was
calculated for anglers exclusively consuming largemouth bass from
the Twelvemile Creek watershed. The highest cancer risk for
ingestion of all species combined, 1 x 1CT2, was calculated for
the Twelvemile Creek Arm. The lake-wide risk associated with
ingestion of all species combined was 5 x 10'3.
6.5 MONTE CARLO
The human health risks presented under Section 6.4 are
referred to as the deterministic risks. That is, the resultant
risk is a single value for a single set of specific exposure
parameters. A refined approach to risk assessment can be
achieved by taking into account the probability distributions of
possible values for each of the exposure parameters rather than
relying on discrete parameter values. Such an approach was taken
in the Sangamo OU2 risk assessment through Monte Carlo analysis.
The advantage of this approach is that it furnishes more
information of the variability of the full range of potential
risks that may occur.
The fish ingestion pathway was chosen for Monte Carlo
analysis evaluation because this pathway is associated with the
highest potential risks. The potential cancer risk associated
-------
Final ROD-Sangamo OU2
June 1994
page 55
with exposure to PCBs was evaluated for ingestion of bass
(largemouth and hybrids) lake-wide and for ingestion of bass
(largemouth and hybrids) caught in the area associated with the
highest PCB concentrations (i.e. SV-107). The mode of the risk
distribution is of most interest when interpreting the final
distribution of risks since it represents the most frequently
occurring risk. The most frequently occurring risk in the
distribution, or mode, was estimated to be 2 x 10"* and 3 x 10"*
for ingestion of lake-wide bass and bass at SV-107, respectively.
The results of the Monte Carlo analysis indicated that the l x
10"2 risk associated with the ingestion of bass from location SV-
107 represented the 95th percentile of high end of the overall
risk distribution, while the 5 x 10"* risk associated with
ingestion of lake-wide bass corresponded to the 90-95th
percentile of the overall risk distribution.
6.6 ECOLOGICAL RISK ASSESSMENT
An ecological risk assessment was conducted to evaluate the
impact PCBs may be having on aquatic receptors (plant,
invertebrates, and fish) and terrestrial receptors (birds and
mammals) of the Sangamo OU2 study area. The Biological
Investigations clearly document PCB contamination at all levels
of the aquatic food web. However, although PCBs appear to be
impacting the fish and macroinvertebrate communities in
Twelvemile Creek, habitat degradation from man's influence is
likely causing additional adverse impacts.
A 1990 investigation conducted independently of these RI
studies provided evidence that fish health and populations can be
affected as concentrations in fish increase to an average of 20
mg/kg and greater. Results from the bioindicator analyses
support the contention that individual largemouth bass may be
impacted at SV-107 and SV-106, areas closest to the Sangamo Plant
Site. However, at the present level of PCB concentrations in
fish tissue (approximately 5 mg/kg), the data from the Biological
Investigations do not support a conclusion that significant
impacts to the overall health and the population/community levels
to Lake Hartwell fish are occurring.
Ecological exposures to surface water were not directly
evaluated in the ecological assessment because PCBs were not
detected in surface water at detection limits ranging from 1.2-
1.3 pg/Ii during the RI. Aquatic plants are not likely to suffer
adverse impacts from PCB concentrations just below the detection
limit. Aquatic invertebrates that are intimately associated with
sediments (primarily benthic organisms) may be adversely impacted
in the study area based on a comparison of PCB concentrations in
sediment to EPA's interim sediment quality criteria for aquatic
organisms. Reliable population estimates for avian species
(eastern phoebe, green-backed heron, belted kingfisher, osprey)
and mammals (mink, otter) that would feed in the study area were
not obtainable. Therefore, the certainty of the extent of any
population impacts caused by PCBs in the study area is unknown.
However, population effects on these terrestrial species is not
likely.
-------
Final ROD-Sangamo OU2
June 1994
^ page 56
6.7 UNCERTAINTY ASSESSMENT
In any risk assessment, there is some degree of uncertainty
associated with the estimates of human and ecological risks. The
primary sources of uncertainty included environmental sampling,
exposure parameter estimation and toxicological data. Typically,
these uncertainties are compensated for by employing conservative
assumptions. Consequently, the qualitative and quantitative risk
estimates for Sangamo OU2 should not be construed as absolute
estimates of risk, but rather as conditional estimates based on a
number of assumptions regarding exposure and toxicity.
6.8 SUMMARY
The primary human health exposure pathway of concern at the
Sangamo OU2 Site is the ingestion of fish contaminated with PCBs.
Exposures associated with ingestion of fish caught from all
sampling locations resulted in unacceptable carcinogenic risks
ranging from 10"2 to 10"* and HQs greater than 1. As discussed
under Section 5.4 (Contaminant Fate and Transport) of this ROD,
PCB concentrations in sediment are predicted to decrease over
time. The impact of decreasing sediment concentrations on
overall risk was evaluated for the ingestion of largemouth bass
at SV-107. The output from the food-chain bioaccumulation model
(F6ETS Section 5.4.3) was utilized in this analysis. It was
found that the total modeled risk over a 30-year time frame (2 x
10"3) was not substantially different from the calculated
deterministic RME risk of 1 x 10"2. This was largely because risk
over this time period is dominated by the very high risks
calculated for the first eight years of exposure. In actuality,
a period of approximately 22 years appears to be required for
one-year risks to drop to the 10"6 risk level. It would require
more than 22 years for the entire 30-year risks to decline to <
10"*. With respect to noncancer risks, a period of approximately
20 years appears.to be required for the HQ to drop below one.
After the next 20-30 years, long-term risk associated with the
consumption of fish harvested from Lake Hartwell should be
substantially lower.
Actual or threatened releases of hazardous substances for
the Sangamo OU2 Site, if not addressed by implementing the
response action in this ROD, may present an imminent and
substantial endangerment to public health, welfare, or the
environment.
-------
Final ROD-Sangamo OU2
June 1994
page 57
TABLE 6-1
PCB Exposure Point Concentrations in Shallow Sediment
Area/
Interval
PCB Concentration (mg/kg)
Mean1
95%
UCL Of
Mean"
Maximum
Detected
Concentration
RME
Exposure
Point
Concentration"
Tipper 12 Mile Creek Arm
Beneath 0-2.1 m
of Lake Water
Beneath 0-3 m
of Lake Water
0.91
1.51
1.65
2.93
7.54
7.88
1.65
2.93
Upper 12 Mile Creek Watershed '
Beneath 0-2.1 m
of Lake Water
Beneath 0-3 m
of Lake Water
0.45
0.45
0.70
0.72
0.76
0.91
0.70
0.72
Hagood Reservoir
Beneath 0-2.1 m
of Lake Water
Beneath 0-3 m
of Lake Water
0.29
0.28
0.36
0.33
0.33
0.33
0.33
0.33
Downgradient of Sangamo Plant
Beneath 0-2.1 m
of Creek Water
1.51
3.52
3.52
a) Arithmetic mean concentration (including one-half detection
limit for non-detects).
b) Value represents 95th percent Upper Confidence Limit (UCL)
on the arithmetic mean.
c) RME=Reasonable Maximum Exposure; value listed is the lower
value of the 95th percent UCL on the arithmetic mean and the
maximum detected concentration.
-------
Final ROD-Sangamo OU2
June 1994
page 58
TABLE 6-2
Exposure Point Concentrations in Fish
Species
PCB Concentrations (mg/kg)
Mean
95%
UCL of
Mean
/ ; ' ' '',,"' "'SV-TC
Largemouth Bass
Hybrid Bass
Largemouth Bass
and Hybrid Bass
Channel Catfish
All Species
Combined
6.8
1.4
5.7
5.6
5.4
8.5
2.0
7.9
7.8
7.0
Maximum
Detected
Concentration
RME Exposure
Point
Concentration
, < ,
>7 ; ' .''/','>",
19.7
2.3
19.7
8.3
19.7
8.5
2.0
7.9
7.8
7.0
SV-106
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species
Combined
3.3
4.7
2.5
3.6
3.8
8.0
3.9
4.3
9.2
9.7
4.0
9.7
3.8
8.0
3.9
4.3
SV-532
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species
Combined
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species
Combined
1.7
3.9
1.7
2.5
3.8
5.9
4.4
4.0
3.5
9.6
3.5
9.6
3.5
5.9
3.5
4.0
SV-535
0.80
1.8
0.60
1.2
1.3
3.3
0.82
1.7
1.7
3.9
0.90
3.9
1.3
3.3
0.82
1.7
SV-642
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species
Combined
0.56
2.3
0.71
1.3
0.98
3.7
2.6
2.1
1.1
7.0
2.8
7.0
0.98
3.7
2.6
2.1
-------
Final ROD-Sangamo OU2
June 1994
page 59
TABLE 6-2
Exposure Point Concentrations in Fish
Species
PCB Concentrations (mg/kg)
Mean
95%
UCL of
Mean
Maximum
Detected
Concentration
RME Exposure
Point
Concent rat ion
- , SV-641 (Lake Background)
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species
Combined
0.22
2.0
0.41
0.74
0.3
2.8
1.4
1.1
1.2
9.3
0.8
9.3
0.3
2.8
0.8
1.1
All Areas of Lake Hartwell + 12 Mile Creek
All Species
Combined
Largemouth Bass
and Hybrid Bass
3.0
3.2
3.9
4.2
19.7
19.7
3.9
4.2
, ' , 12 Mile Creek Watershed
Largemouth Bass
Redbreast
Sunfish
4.0
1.7
25
2.2
13.9
7.1
13.9
. 2.2
Tributaries of Watershed ;
Redbreast
Sunfish
Watershed
Background ,
0.74
ND ;
0.2
ND
1.2
ND
0.2
,ND
-------
Final ROD-Sangamo OU2
June 1994
page 60
TABLE 6-3
Exposure Parameters for Incidental Ingestion of and Dermal
Contact with Sediment (Current /Future Use Conditions)
Parameters
Age Period (years)
Exposure Frequency (days /year)*
Exposure Duration (years)
Soil Ingestion Rate (mg/day)c
Skin Surface Area Available for Contact (cm)2
Sediment to Skin Adherence Factor (mg/cm2)'
Dermal Absorption Fraction (PCBs)f
Body Weight (kg)e
Averaging Time (years)
Carcinogenic11
Nohcarcinogenic
Reasonable
Maximum Exposure
Case
Child
4-12
105
8
138
700d
1.0
0.01
27
70
8
Adult
>18
75
30b
100
l,230d
1.0
0.01
70
70
30
a) For children, assumes that children will swim 7 days/week
from Memorial Day to Labor Day (15 weeks) for a total of 105
exposure events/yr. For adults, assumes that adults will
swim 5 days/week for same 15 weeks for a total of 75
exposure events/yr.
b) Based upon national upper-bound time at one residence (USEPA
1991a, 1989a).
c) Value for children is a weighted-average ingestion rate,
assuming 4, 5, and 6 year olds ingest sediment at a rate of
200 mg/day and older children ingest sediment at 100 nig/day
(USEPA 1989a, 1991a). Value for adults is the standard
default value recommended by USEPA (1991a, 1989a)
d) Surface area based on 50th percentile values from USEPA
(1985) for feet. Values for child and adult are the time-
weighted averages.
e) Based on USEPA Region IV guidance (USEPA I992b).
f) Based on USEPA Region IV guidance (USEPA I992b).
g) Value for 4-12 year old child is the time-weighted average
based on data provided by USEPA (1989b). Value for adult is
the standard default value recommended by USEPA (1991a).
h) Based on USEPA (1991a, 1989a) standard assumption for
lifetime.
-------
Final ROD-Sangamo OU2
June 1994
page 61
TABLE 6-4
Exposure Parameters for Ingestion of Fish by Adult
Recreational Fisherman (Current /Future Use Conditions)
Parameters
Fish Ingestion Rate (grains /meal)1
Exposure Duration (years )b
Exposure Frequency by Sampling Location
(meals per year)'
Twelvemile Creek Watershed
SV-107
SV-106
SV-532
SV-535
SV-642
SV-641 (Lake Background)
Lake -Wide Average
Body Weightd
Averaging Time (years)
Carcinogenic'
Noncarcinogenic
Reasonable Maximum
Exposure Value
357
30
60
30.6
27.6
20.6
22.5
37
22.7
27.6
70
70
30
a) Based on the arithmetic mean of fish ingestion rates
reported for 52 anglers surveyed at Twelvemile Creek on Lake
Hartwell.
b) Value is based on the national upper-bound time at one
residence (USEPA 1991a, 1989a).
c) Values based on January to December 1992 data obtained from
Lake Hartwell Recreational Angler Survey. These values are
the arithmetic mean exposure frequency in meals/month,
multiplied by 12 months/year reported for a total of 677
anglers surveyed at Twelvemile Creek/Lake Hartwell. Final
report of survey stated that some anglers were interviewed
more than once, indicating that the survey results may
include fishing habits of subsistence fisherman as well as
recreational fisherman.
d) Standard default value provided by USEPA (1991a, 1989a).
e) Based on USEPA (1991a, 1989a) standard assumption for
lifetime.
-------
Final ROD-Sangamo OU2
June 1994
page 62
TABLE 6-5
Potential Direct Contact Risks Associated with PCBs in
Sediment
Exposure Pathway
Upper
Bound
Excess
Lifetime
Cancer
Risk
HQ for
Noncar-
cinogenic
Effects1
INCIDENTAL 1NGESTION OP SEDIMENT' '
Upper Twelvemile Creek Arm
Child (4-12 years)
Adult
2 x 10-6
3 x ID"*
<1 (0.03)
<1 (0.01)
Upper Twelvemile Creek Watershed
Child (4-12 years)
Adult
9 x 10'7
7 x ID'7
<1 (0.01)
<1 (0.003)
Immediately Downgradient of Sangamo Plant Site
Child (4-12 years)
Adult
5 x lO^5
3 x 10-6
<1 (0.06)
<1 (0.02)
Hagood Reservoir
Child (4-12 years)
Adult
4 x 10'7
3 x 10'7
<1 (0.006)
<1 (0.001)
DERMAL ABSORPTION FROM SEDIMENT
Upper Twelvemile Creek Arm
Child (4-12 years)
Adult
1 x lO'7
4 x 1(T7
<1 (0.002)
<1 (0.001)
Upper Twelvemile Creek Watershed
Child (4-12 years)
Adult
5 x 10"8
9 x 10-*
<1 (0.0007)
<1 (0.0004)
Immediately Downgradient of Sangamo Plant Site
Child (4-12 years)
Adult
2 x ID'7
5 x 10'7
<1 (0.004)
<1 (0.002)
Hagood Reservoir
Child (4-12 years)
Adult
2 x 10-8
4 x 10-*
<1 (0.0003)
<1 (0.0002)
-------
Final ROD-Sangamo OU2
June 1994
page 63
TABLE 6-5
Potential Direct Contact Risks Associated with PCBs in
Sediment
Exposure Pathway
Upper
Bound
Excess
Lifetime
Cancer
Risk
HQ for
Noncar-
cinogenic
Effects1
DIRECT CONTACT WITH SEDIMENT (TOTAIi)
Upper Twelvemile Creek Arm
Child (4-12 years)
Adult
2 x 10-*
3 x 10-6
<1 (0.03)
<1 (0.01)
Upper Twelvemile Creek Watershed
Child (4-12 years)
Adult
1 x 1CT6
8 x 10'7
<1 (0.01)
<1 (0.003)
Immediately Downgradient of Sangamo Plant Site
Child (4-12 years)
Adult
Hagood Reservoir
Child (4-12 years)
Adult
5 x lO*
4 x ICT6
<1 (0.06)
<1 (0.02)
-
4 x 10'7
3 x lO'7
<1 (0.006)
<1 (0.001)
a) Hazard Quotient >1 indicates exposure to PCBs may result in
adverse health effects. Actual value listed in ().
-------
Final ROD-Sangamo OU2
Jxine 1994
page 64
TABLE 6-6
Potential Health Risks Associated with the Ingestion of Fish
Location/Species
Upper
Bound
Excess
Lifetime
Cancer
Risk
SV-107 , - ,' ' < '
Largemouth Bass
Hybrid Bass
Largemouth Bass and Hybrid Bass
Channel Catfish
All Species Combined
1 x 10'2
3 x 1CT3
1 x 10'2
1 x 10'2
1 x 10'2
HQ for
Noncar-
cinogenic
Effects'
,
>1 (50)
>1 (10)
>1 (50)
>1 (50)
>1 (50)
SV-106
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species Combined
5 x 10'3
1 x 1(T2
5 x 10'3
5 x 1CT3
>1 (20)
>1 (40)
>1 (20)
>1 (20)
SV-532 . . .
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species Combined
3 x 10'3
6 x ID"3
3 x 1CT3
4 x 10'3
>1 (10)
>1 (20)
>1 (10)
>1 (20)
SV-535
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species Combined
i x icr3
3 x 10'3
9 x 10-*
2 x icr3
>1 (6)
>1 (10)
>1 (4)
>1 (8)
SV-642 , , ' '.
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species Combined
2 x 10'3
6 x 10'3
4 x 10'3
4 x 10'3
>1 (7)
>1 (30)
>1 (20)
>1 (20)
-------
Final ROD-Sangamo OU2
June 1994
page 65
TABLE 6-6
Potential Health Risks Associated with the Ingestion of Fish
Location/Species
Upper
Bound
Excess
Lifetime
Cancer
Risk
HQ for
Noncar-
cinogenic
Effects1
SV-641 (Lake Background) - .,*,,>,.
Largemouth Bass
Hybrid Bass
Channel Catfish
All Species Combined
3 x 1CT1
3 x 10'3
8 x ID"4
1 x 10"3
>1 (1)
>1 (10)
>1 (4)
>1 (5)
All Areas ;of Lake Hartwell + 12 Mile Creek
All Species Combined
Largemouth Bass and Hybrid Bass
5 x lO'3
5 x 10'3
>1 (20)
>1 (20)
Twelvemile Creek Watershed
Largemouth Bass
Redbreast Sunfish
4 x 10'2
6 x 10'3
>1 (200)
>1 (30)
Tributaries of Watershed .
Redbreast Sunfish
Watershed Background
6 x 10-*
ND
>1 (20)
ND
a) Hazard Quotient >l indicates exposure to PCBs may result in
adverse health effects. Actual value listed in ().
-------
Final ROD-Sangamo 0172
June 1994
page 66
7.0 DESCRIPTION OF ALTERNATIVES
Based upon the findings of the RI and associated Baseline
Risk Assessment (human health/ecological), EPA developed remedial
action objectives to support the identification, development and
screening of remedial alternatives. These remedial action
objectives were:
Mitigate continued migration of PCS-contaminated sediments
into Lake Hartwell by eliminating releases of PCBs into
Twelvemile Creek.
Control or eliminate the downstream migration of PCB-
contaminated sediment within the Twelvemile Creek Arm of
Hartwell Lake.
Limit, to the extent feasible, the transfer of PCB
contaminants from sediment to biota.
Prevent or minimize exposure to fish with PCB contamination
above target risk (or FDA) levels.
Protection of human health is considered the primary driver
for developing and evaluating remedial action alternatives.
Thorough development and evaluation of feasible remedial
alternatives for the Sangamo OU2 Site required the derivation of
cleanup goals for the media of concern: sediment and fish.
Cleanup goals were not developed for surface water since PCBs
were not detected in any sample collected during the RI above the
detection limits of 1.2 to 1.3 /xg/1. Cleanup goals for sediment
and fish were also necessary to identify areas to be addressed by
the range of remedial alternatives that were evaluated in detail.
The Final Cleanup Goals for sediment and fish, and the supporting
rationale are presented in Sections 7.1 and 7.2, respectively.
The contaminants of concern for the Sangamo OU2 Site are
total PCBs. No Aroclor- or congener-specific distinctions are
factored into the evaluation of remedial action alternatives.
Discharges from the Sangamo Weston Plant into Town Creek included
a variety of PCBs consisting predominantly of Aroclors 1242,
1254, and 1016. Analytical data for the RI included both
Aroclor-specific concentrations (from offsite CLP analyses) and
concentrations of total PCBs (from onsite field screening
analyses). However, all of the results, findings and conclusions
in the RI were discussed in terms of total PCBs. The reasons for
this approach are as follows:
A majority of the health-based and environmental criteria
for PCBs are stated in terms of total PCBs (i.e., FDA
tolerance level).
EPA uses a carcinogenic slope factor for PCBs that is based
on Aroclor 1260; this provides a conservative indicator of
potential health effects for all PCB mixtures.
-------
Final ROD-Sangamo OU2
June 1994
page 67
Evaluation of contamination in terms of total PCBs was cost-
effective; it permitted collection of a large number of
samples and the use of field screening (Modified Spittler
Method) for sediment analyses.
7.1 FINAL CLEANUP GOAL FOR SEDIMENT
Potential cleanup goals for PCB-contaminated sediment at the
Sangamo OU2 Site were determined through an evaluation of
available criteria and accepted techniques for calculating
cleanup goals in sediment. This approach generated a range of
potentially viable cleanup concentrations for the Sangamo OU2
sediment based on the following sources:
Existing (published) criteria for PCBs in sediment,
typically derived from either geographic-based, background
PCB concentrations or biological effects observed or
predicted in aquatic organisms.
Precedents from other NPL sites where EPA has identified
protective sediment cleanup goals for PCBs that have been
finalized in RODs.
Site-specific cleanup goals calculated using EPA's
equilibrium partitioning approach based on contaminant
partitioning between sediment and sediment pore water, based
on protection of aquatic life.
From this evaluation, three representative sediment cleanup
goals (1 mg/kg, 0.4 mg/kg and 0.05 mg/kg) were selected for
further analysis. The analysis consisted of an evaluation of the
long-term impacts of potential sediment cleanup goals on PCB
concentrations in fish, and ultimately human health via ingestion
of contaminated fish, given that the ingestion of fish is the
principal exposure pathway of concern at the site. This
evaluation was conducted utilizing the FGETS bioaccumulation
model discussed in Section 5.4.3 of this ROD. The rationale for
selecting these values is presented below.
1 mg/kg - The most frequently selected sediment cleanup goal
for PCBs at NPL sites based on a review of the EPA's ROD
database. A concentration of 1 mg/kg also represents a
reasonable lower limit considering technical feasibility and
cost.
0.4 mg/kg - The mean value for the site-specific sediment
quality criteria calculated using the EPA's equilibrium
partitioning approach; also equal to the Effects Range-
Median criteria based on an evaluation of published criteria
associated with biological effects on aquatic life as
reported by the National Oceanic and Atmospheric
Administration (NOAA).
0.05 mg/kg - Equal to the Effects Range-Low from NOAA based
on an evaluation of published criteria associated with
biological effects on aquatic life; also representative of
-------
Final ROD-Sangamo OU2
June 1994
page 68
the more commonly reported background-based sediment
criteria for PCBs.
The time required for 2-8 year old largemouth bass in the
Twelvemile Creek Arm of Lake Hartwell to achieve 2 mg/kg for the
range of selected sediment cleanup goals are compared to the
baseline condition in Figure 7-1. As shown in this figure, fish
PCB concentrations decline at about the same rate regardless of
the sediment cleanup goal. Therefore, a final sediment cleanup
goal of 1 mg/kg was selected based on technical feasibility
rather than performance or risk-based considerations. This
concentration identified the entire Twelvemile Creek Arm,
extending from the headwaters of the lake downstream to the
confluence with the Keowee Arm, as an area to be addressed. This
area covers approximately 730 acres with a total estimated volume
of 4,722,000 cubic yards (yd3) of contaminated sediment.
7.2 FINAL CLEANUP GOAL FOR FISH
Fish ingestion was identified as the primary exposure
pathway of concern at the Sangamo OU2 site. Potential
remediation goals include the FDA tolerance level of 2 mg/kg for
PCBs in the edible portions of fish, and risk-based levels that
consider the fish ingestion exposure pathway. Both of these
options are described below.
The FDA criterion was identified as a contaminant-specific
Applicable or Relevant and Appropriate Requirement (ARAR). In
addition, the existing health advisory for Hartwell Lake is based
on the continuing presence of PCBs in fish in concentrations
greater than 2 mg/kg. Selection of risk-based cleanup goals for
fish were considered by determining the concentration levels in
largemouth bass that would result in acceptable risk to anglers
(through ingestion of the fish) based on EPA's target risk range
of 1 x 10"4 to l x 10"*. The acceptable concentrations were
estimated using the same methodology used for the baseline human
health risk assessment and for determining risk levels for the
sediment cleanup goals. A fish tissue concentration of
0.036 mg/kg is associated with a 10"* risk, 0.0036 mg/kg. with a
10"5 risk, and so on.
Using EPA's deterministic, reasonable maximum exposure (RME)
approach and site-specific exposure parameters, the 30-year
carcinogenic risk associated with the FDA criterion of 2 mg/kg
results in an estimated risk of 6 x 10"3. This is well above the
upper end of EPA's target risk threshold of 1 x 10~*. It is
important to note that EPA considers the RME assumptions to be
conservative, since they represent upper confidence limits for a
given range of values for a particular risk input parameter or
variable. It is also important to note that consumption of fish
from Lake Hartwell (or other sources) is a strictly voluntary
activity.
Use of the risk-based concentrations for fish remediation
goals (i.e., 0.036 mg/kg) was determined to be technically
impracticable at SV-107 for several reasons. Fish bioaccumulate
PCBs from both the water column and food chain; thus, PCB
-------
ra
CD
o>
p
2-s
100 T
10 -i
1 -
E 0.1 -'
-------
Final ROD-Sangamo OU2
June 1994
page 70
concentrations in sediment and surface water would likely have to
be reduced to levels in the ranges of parts per billion
(0.001 mg/kg) or parts per trillion (0.001 pg/l>) to achieve risk-
based levels in fish. Reducing surface water and sediment
concentrations to these levels is beyond the capability of proven
treatment technologies, particularly when the scale of the site
is taken into consideration. Moreover, as shown in the FGETS
modeling, PCB cycling among fish, plankton, benthos, sediment,
and surface water greatly complicates the removal of PCBs from a
biological system once the contaminants have been introduced.
Even if concentrations in sediment and surface water could be
reduced to levels commensurate with the risk-based fish
concentrations, it would be many years before the fish
concentrations actually declined to acceptable levels.
Given the existence of an ARAR for PCB concentrations in
fish, the technical impracticability of establishing risk-based
cleanup goals, and the classification of fish consumption as a
voluntary exposure, the FDA tolerance level of 2 mg/kg was
selected as the Final Cleanup Goal for Lake Hartwell fish.
7.3 REMEDIAL ALTERNATIVES
Response actions that were identified and passed the
screening of technologies and process options were assembled into
a range of remedial alternatives that included no action,
institutional controls, containment, collection,
removal/disposal, and removal/treatment/disposal. These eight
remedial alternatives were evaluated in detail in the Final March
1994 Feasibility Study Report. A summary description of these
alternatives follows. The reader should refer to Final FS Report
for a more detailed account of this subject matter.
All alternative cost estimates are expressed in 1993 dollars
and are based upon conceptual engineering and design. Capital
cost consists of direct (construction) and indirect (non-
construction) costs incurred in the first year of operation.
Operation and Maintenance (O&M) cost refers to long-term
postconstruction items necessary to ensure continued
effectiveness of a remedial action. Total present worth cost
represents that sum of money, if invested in the base year and
disbursed as needed, would be sufficient to cover all costs of a
remedy over its planned life.
ALTERNATIVE 1 - NO ACTION
As required by CERCLA, a no further action alternative was
evaluated to serve as the basis for comparison with other active
cleanup alternatives. Under this no-action alternative, no
further remedial actions for the contaminated sediments or fish
at the site would be conducted. The no-action alternative would
not affect the existing health advisory issued by SCDHEC, who
would be expected to continue the advisory until PCB
concentrations in fish tissue decline to levels below 2 mg/kg
(FDA tolerance level). The advisory currently warns against the
consumption of fish from the Seneca River Arm of Hartwell Lake
above the Hwy 24 bridge and fish larger than 3 Ib throughout the
-------
Final ROD-Sangamo OU2
June 1994
page 71
entire lake. The advisory would be modified if warranted by
future trends regarding PCB levels in fish. Maintenance of the
fish advisory is assumed to entail the periodic replacement of
existing signs that advise against fish consumption.
As discussed in Section 5.4.3 of this ROD, the mean
largemouth bass PCB concentration is predicted to fall below the
FDA limit by year 10. However, a longer period of time is
required for concentrations in higher age classes to fall below
the FDA limit. PCB concentrations in 10-year-old fish,
representing the highest age class and largest fish in the model,
required 12 years, or the year 2005, to fall below the 2.0 mg/kg
FDA limit. For these reasons, it was assumed that the fish
consumption health advisory would remain in effect for 12 years.
The no-action alternative would also entail periodic reviews
of site conditions to ensure that the alternative remained
protective of human health and the environment. These reviews
would be conducted at least every 5 years as required by CERCLA.
Capital Cost = $7,000
Total O&M Cost = $123,000
Total Present Worth Cost = $130,000
ALTERNATIVE 2A - INSTITUTIONAL CONTROLS
This alternative is a limited action alternative that relies
on a series of institutional controls to prevent or minimize
ingestion of contaminated fish tissue, which was identified as
the primary exposure pathway of concern for the Site. Generally,
these institutional controls consist of the following components:
Continuance of the Existing Fish Advisory
Public Education Program - A program would be initiated to
inform the public on available methods for reducing the
intake of PCBs through fish consumption. Specific
preparation, handling, and cooking techniques can reduce the
quantity of contamination consumed. Information of this
type would be disseminated to the public through a series of
local public meetings and distribution of an informational
pamphlet.
Fish and Sediment Monitoring - In addition to maintaining
the current fish advisory, annual monitoring of PCB levels
in fish and sediment of Twelvemile Creek and Lake Hartwell
would be conducted. Results from this monitoring program
would be utilized to support modifications to the fish
advisory and to monitor concentrations of PCBs in sediment
and fish over time. Results of this program would be made
available to the public.
Regulation of Twelvemile Creek Impoundments - A routine
schedule would be developed for flushing of sediment
accumulated behind the 3 small impoundments located on
Twelvemile Creek. Periodic flushing (most recently in
September 1993) of a large load of sediment over a short
-------
Final ROD-Sangamo OU2
June 1994
page 72
period of time has been documented to have an adverse impact
on the water quality and aquatic biota of the upper portion
of the Twelvemile Creek Arm. These adverse impacts are
attributed to elevated levels of suspended sediment and not
PCBs. A routine flushing schedule would minimize impacts to
the ecosystem while enhancing burial of more contaminated
sediments with cleaner sediments from the Twelvemile Creek
drainage.
Capital Cost = $366,000
Total O&M Cost = $2,842,000
Total Present Worth Cost = $3,208,000
ALTERNATIVE 2B - FISHERIES ISOLATION
Alternative 2B is a more aggressive approach consisting of
management of the fisheries resource within the lake to minimize
ingestion of PCB-contaminated fish. The primary control measure
includes construction of a barrier, or fish fence, to prevent the
movement of migratory fish (i.e. striped bass, hybrid bass, and
walleye) into or out of the most contaminated portions of the
reservoir. These migratory species represent approximately 50%
of the fish harvested by weight from Lake Hartwell.
Placement of the barrier in the vicinity of the Hwy 37
Bridge, just south of Clemson, would result in isolation of the
Twelvemile Creek, Keowee River, and upper Seneca River Arms,
which represent less than 10 percent of the total area of
Hartwell Lake. Isolation of these upstream areas is expected to
result in an accelerated decline in migratory fish PCB
concentrations in the downstream portions of the reservoir (i.e.,
the remaining 90+ percent of the lake). Reduction of fish PCB
levels would allow for rescinding of the existing health
advisories in these areas, returning the majority of lake areas
to normal use. This approach is consistent with the COE's Best
Management Practices for Hartwell Lake, designed to achieve the
maximum beneficial uses for the reservoir, which include fish and
wildlife management and recreational use.
The fish isolation barrier would be designed/constructed to
meet the following performance standards:
Minimize safety hazards for boaters and other users of the
lake (i.e. waterskiers);
Maximize effectiveness to prevent passage of migratory game
fish through the barrier;
Allow boats to pass unimpeded to minimize disruptions of
boating traffic on the lake (i.e. no gate);
Construct a semi-permanent structure that could be removed
once PCB levels decline to health-based levels;
Incorporate value engineering principles to minimize cost
for design and construction;
-------
Final ROD-Sangamo OU2
June 1994
page 73
Minimize operations and maintenance requirements and costs;
Design and installation of fish fence will not impede normal
lake usage.
This alternative would require an approximate 6 month pre-
design (treatability type) phase to develop a detailed design in
accordance with the above performance standards. If pre-design
studies are successful, full scale construction/installation of
this remedy could be accomplished in approximately 1 year. PCBs
would continue to accumulate in the fish upstream of the fish
fence, therefore, fish are predicted to decline to protective
levels (the 2.0 mg/kg FDA tolerance level) in approximately
12 years. By isolating the impacted upstream areas, PCS
concentrations in migratory fish from the remaining 90+ percent
of the lake are predicted to decline to the 2 mg/kg level in
approximately 3-4 years.
Lake use restrictions, such as no-wake zones and warning
signs would be implemented to prevent damage to the fence and for
boating safety. Alternative 2B also includes the institutional
controls described for Alternative 2A.
Capital Cost = $1,232,000
Total O&M Cost = $3,012,000
Total Present Worth Cost = $4,244,000
ALTERNATIVE 3A - CAPPING
Alternative 3A involves the isolation of PCB-contaminated
sediments by placing an 18 inch clean sediment cap over the areas
with the highest contamination. The cap would be constructed
using fine sand (minimum particle size 0.25 mm) to eliminate the
further downstream migration of contaminated sediment and the
transfer of PCBs from sediment to aquatic biota. The cap would
extend 7 miles from just upstream of the Maw Bridge (Rt. 337) to
the confluence with the Keowee River, just upstream of Hwy 123.
The cap would cover an area of 730 acres and would require
approximately 1,800,000 yd3 of sand to construct the cap. Figure
7-2 provides an illustration of the area to capped under this
alternative.
The cap thickness was designed to minimize the impacts
associated with bioturbation and sediment burrowing biota (i.e.
mayfly). Based upon sediment transport modeling for a cap with a
minimum particle size of 0.25 mm, no significant erosion of cap
material was observed in areas below the Madden Bridge. However
in areas above the Madden Bridge (i.e., the "goose-neck" bend
area), the model results showed erosion for all grain sizes
tested (i.e., 10, 2, and 0.25 mm). These results indicate that
maintenance of the cap would be difficult in the headwaters.
Placement of the cap would be accomplished using a hydraulic
barge unloader equipped with a sand spreader. The barge and sand
spreader would likely be supplied with sand through a slurry
pipeline between the barge and the shore or by a separate hopper
barge. A local sand source has been identified near the
community of Liberty, South Carolina. The cost estimate assumes
-------
Final ROD-Sangamo 0112
June 1994
page 74
TRIBUTARY
CLEAN SEDIMENT CAP
0
I
1.0
SCALE IN MILES
Rgure 7-2
Area to be Capped in Alternative 3A
-------
Final ROD-Sangamo OU2
June 1994
page 75
that the sand would be transported by truck to the staging area,
converted to a slurry using lake water, and pumped out to the
sand spreader barge.
Placement of the cap, including design, procurement, and
construction, would require approximately 2 years. Once the
contaminated sediments were isolated beneath the cap, further
transfer to biota would be significantly reduced, and PCB
concentrations in the fish in Twelvemile Creek Arm would decline
at an accelerated rate. A period of 3 to 4 years after cap
placement would be required for the mean concentration of PCBs in
largemouth bass filets to fall below the FDA limit of 2.0 mg/kg.
As part of cap O&M, detailed hydrographic surveys would need to
be conducted periodically to assess the integrity and overall
performance of the clean sediment cap. These surveys would focus
on the upstream segment of the cap in the area where sediment
deposition and scour processes are most active.
Alternative 3A also consists of lake usage restrictions to
minimize erosion of the cap once it is placed and other
institutional controls described under Alternative 2A.
Capital Cost = $48,296,000
Total O&M Cost = $2,843,000
Total Present Worth Cost = $51,139,000
ALTERNATIVE 3B - SEDIMENT CONTROL STRUCTURE
Alternative 3B involves the construction of a 1,600 ft.
fixed-crest weir near the mouth of the Twelvemile Creek Arm. The
weir would maintain a constant pool elevation of 660 ft MSL in
the Twelvemile Creek Arm, minimizing the scour/erosion and
resuspension of contaminated sediment in the headwaters and
thereby reducing the downstream migration of PCB-contaminated
sediment into Hartwell Lake. The constant pool elevation may
also enhance the burial of these contaminated sediments in the
upstream reaches. In addition, the weir would greatly reduce the
migration of fish into and out of Twelvemile Creek Arm. The
location of the sediment control structure is shown on Figure 7-
3.
The weir, or sediment control structure, would be an
effective barrier to the downstream movement of suspended bed-
load sediments, which are transported just above the sediment-
water interface. The weir would allow water to flow from the
Twelvemile Creek Arm to Hartwell Lake but would otherwise be an
effective physical barrier between the two water bodies. The
sediment control structure would not allow the passage of boats
or other watercraft, and lakeshore property owners in the
Twelvemile Creek Arm would not have direct boating access to or
from Hartwell Lake. Monthly pool elevation data and the
reservoir operating rule curve were used to select the crest
elevation for the weir. Based on the pool elevation data, it is
estimated that the pool elevation for Lake Hartwell is above
660 ft MSL approximately 23 percent of the time. The sediment
control structure would be constructed of concrete by utilizing
proven cofferdam techniques.
-------
Final ROD-Sangamo OU2
June 1994
page 76
>SEDIMENT CONTRO
STRUCTURE
TRIBUTARY
TtClVEMU CREEX LAKE
I
1.0
SCALE IN MILES
Figure 7-3
Sediment Control Structure Location of Alternative 3B
-------
Final ROD-Sangamo OU2
June 1994
page 77
Implementation of Alternative 3B, including design and
construction of the weir and establishment of institutional
controls, would require approximately 2 years. Within the newly
impounded Twelvemile Creek Lake, contaminated sediments would not
be isolated from the biota and future contaminant trends in the
fish would be as described under the Alternatives 1 and 2A. The
time to achieve protectiveness in the main body of Lake Hartwell
(i.e., the approximate 90+ percent of the total lake area
downstream of the sediment control structure) was assumed to be
comparable to that predicted for Alternative 2B (fish fence),
given that both the weir and fish fence would greatly limit the
movement of fish into and out of the area with the most highly
contaminated sediment. Using this approach, the FDA Tolerance
Level in migratory fish downstream of the sediment control
structure would be attained in an estimated 3 to 4 years after
construction of the weir was completed.
To ensure public safety, watercraft would be restricted from
the -immediate vicinity of the sediment control structure.
Fishing restrictions in the isolated area, as well as
institutional controls under Alternative 2A, would be implemented
to limit consumption of contaminated fish until protective levels
were achieved.
Capital Cost = $51,226,000
Total O&M Cost = $2,365,000
Total Present Worth Cost = $53,591,000
ALTERNATIVE 3C - OPTIMAL CAPPING/SEDIMENT CONTROL STRUCTURE
Value engineering techniques were used to combine elements
of Alternatives 3A and 3B to achieve similar results with
significantly lower costs and reduced impact to lakeshore
residents and property owners. Figure 7-4 provides a conceptual
approach to this alternative.
The sediment control structure for Alternative 3C consists
of a fixed-crest weir similar in design to the structure proposed
in Alternative 3B. The proposed location for the weir under
Alternative 3C is approximately 1,000 ft upstream of the Hwy 133
bridge. Placement of the structure at this location would
isolate the areas having the highest levels of PCB contamination
in the sediment while at the same time effectively isolating a
smaller portion of the Twelvemile Creek Arm from the main lake.
Moving the weir upstream also allows for its placement at a
relatively narrow constriction of the channel, resulting in a
much smaller weir (length-wise) and construction cost. The weir
would also maintain an upper pool elevation of 660 ft, minimizing
scour and resuspension of contaminated sediment in the headwaters
of the Twelvemile Creek Arm. This action would minimize the
migration of contaminated sediment into the lower portion of the
Twelvemile Creek Arm and Hartwell Lake. The weir is also
expected to reduce the movement of fish into and out of the areas
of highest sediment PCB contamination.
The area downstream of the sediment control structure to the
mouth of the Twelvemile Creek Arm would be capped under this
-------
Final ROD-Sangamo OU2
June 1994
page 78
SEDIMENT CONTROL
STRUCTURE
TRIBUTARY
LITTLE ttaVEHILE CREEK LAKE
CLEAN SEDIMENT CAP
i.o
SCALE IN MILES
Figure 7-4
Sediment Control Structure/Area to be Capped
Alternative 3C
-------
Final ROD-Sangamo OU2
June 1994
page 79
alternative using the same performance criteria (e.g., prevent
burrowing into underlying sediment), conceptual design parameters
(e.g., 18-in. thick, sand-silt mixture), and construction methods
(i.e., hydraulic sand spreader barge) as described for the
Alternative 3A. The area for capping consists of approximately
285 acres and would require approximately 700,000 yd* of material
to construct. Cap integrity below the weir would be maintained
using riprap and other reinforcing materials to prevent scour
from flow over the top of the weir.
Construction of the sediment control structure and placement
of the cap would require a period of 2 to 3 years. Upstream of
the weir, contaminated sediments are not isolated from the biota
and thus PCB concentration in fish would continue to exceed the
FDA limit of 2.0 mg/kg for the an estimated 12 years. Downstream
of the weir in the Twelvemile Creek Arm, cap placement would
accelerate the reduction of PCB levels in fish to protective
levels because contaminated sediments have been isolated,
limiting further transfer of PCBs to biota. An estimated period
of 3 to 4 years (after cap placement) is required for PCB
concentrations to fall below the FDA limit of 2.0 mg/kg in the
capped area as well as in the main body of Hartwell Lake.
Institutional controls under Alternative 2A and fishing
restrictions upstream of the sediment control structure would be
implemented until protective levels in the fish were achieved.
Capital Cost = $31,684,000
Total O&M Cost = $2,365,000
Total Present Worth Cost = $34,049,000
ALTERNATIVE 4 - CONFINED DISPOSAL FACILITY
Alternative 4 involves a much more rigorous remediation
approach than any of the alternatives discussed above. This
alternative involves removal by hydraulic dredging of
contaminated sediment from the Twelvemile Creek Arm with PCB
concentrations greater than 1 mg/kg followed by disposal of the
dredged material in a Confined Disposal Facility (CDF). The
conceptual approach to this alternative is shown in Figure 7-5.
Implementation of Alternative 4 would be a complex undertaking
controlled through scheduling and rigid conformance to procedural
requirements needed to minimize environmental impacts. The
following discussion provides an overview of the remedial
construction activities associated with Alternative 4:
Excavate channel - Rerouting the Twelvemile Creek Arm would
begin immediately south of the high-tension powerline
crossing and would proceed due south for a distance of
approximately 1,600 ft. where it would reconnect with the
original lake channel. The rerouted channel would be
approximately 50 ft. wide at the base with a maximum
excavation depth for the channel at approximately 55 ft.
Material excavated during the channeling operation would be
used for construction associated with the CDF. Rerouting
the channel in this area would only isolate one shoreline
residence on the goose-neck bend.
-------
Final ROD-Sangamo OU2
June 1994
page 80
SCALE: 1' = 1000'
Figure 7-5
Confined Disposal Facility for Alternative 4
-------
Final ROD-Sangamo OU2
June 1994
page 81
Construct Confined Disposal Facility - The COE commonly uses
CDFs for disposal of dredged sediments. The goose-neck bend
area between Maw and Madden Bridges was selected as the
optimum location for the following reasons: 1) this area has
most highly contaminated sediments avoiding the need to
dredge approximately 1.3 million yd3 of material; 2)
approximately 113 acres is available providing sufficient
capacity for disposal of 5 million yd* of material; and 3)
minimal impacts on existing residents would occur.
Dredge and pump dredged sediment into CDF - Sediments from
entire Twelvemile Creek Arm with PCB concentrations >1 rag/kg
would be removed via hydraulic dredging and transported to
CDF via pipeline. Sediment would settle out in CDF, while
water was returned to the lake.
Compact/grade sediment to promote runoff followed by
placement of soil cap over CDF.
Additional detailed information regarding dredging
techniques, dredged material characteristics, and design of the
CDF would be required prior to implementation of Alternative 4.
Implementation of the dredging alternative is expected to require
a period of 3 to 4 years (including treatability testing, design,
procurement and construction). Once sediments with greater than
l mg/kg of PCBs have been removed from the lake, the FGETS model
predicts that the mean PCB concentration in largemouth bass
fillets would fall below the FDA limit of 2.0 mg/kg in 5 to
6 years. Institutional controls would be used to limit
consumption of contaminated fish until protective levels were
achieved.
Capital Cost = $43,422,000
Total O&M Cost = $3,487,000
Total Present Worth Cost = $46,909,000
ALTERNATIVE 5 - STABILIZATION
Alternative 5 is a very complex and costly alternative that
involves removal of contaminated sediments with PCB
concentrations greater than 1 ppm via a combination of shallow
water excavation and hydraulic dredging. Contaminated sediments
in the middle and lower portions of Twelvemile Creek Arm would be
dredged while the upper portion is dewatered and excavated. The
dredged/excavated sediments would be treated by stabilization
with cement and placed in a CDF as described under Alternative 4.
Fish and other biota in the upstream segment would be collected
and destroyed as part of the dewatering operations. This
alternative was evaluated in an attempt to satisfy the preference
for treatment as a principal element for selected remedial
alternatives at NPL sites.
Implementation of Alternative 5 would likely trigger a
substantial number of ARARs, the most significant being the Clean
Water Act and South Carolina Water Classification Standards.
Based on a volume estimate of 4,722,000 yd3 for excavation and
dredging, expected stabilization production rates, and the rate
-------
Final ROD-Sangamo OU2
June 1994
page 82
at which the confined disposal facility can be filled,
implementation of Alternative 5 would entail a duration of 4 to 5
years. Once the contaminated sediment has been removed, fish PCB
concentrations downstream of the Madden Bridge would decline at
an accelerated rate, achieving the FDA tolerance level in
approximately 5 to 6 years. The time to achieve protectiveness
for this alternative therefore is 9 to 11 years.
Capital Cost = $581,957,000
Total O&M Cost = $3,486,000
Total Present Worth Cost = $585,443,000
8.0 SUMMARY OF THE COMPARATIVE ANALYSIS OF ALTERNATIVES
This section documents the comparative analysis conducted to
evaluate the relative performance of each alternative in relation
to each of the evaluation criteria. The purpose is to identify
and clearly understand the advantages and disadvantages of each
remedial alternative described in Section 7.0 of this ROD. As
stated in NCP [40 CFR 300.430 (f)], the evaluation criterion are
arranged in a hierarchial manner that is then used to select a
remedy for the site based on the following categories:
Threshold Criteria'.
Primary Balancing Criteria:
Modifying Criteria:
Overall Protection of Human
Health and the Environment
Compliance with ARARs
Long-term Effectiveness and
Permanence
Reduction of Toxicity,
Mobility, or Volume
Short-term Effectiveness
Implementability
Cost
State Acceptance
Community Acceptance
The Threshold criteria must be met before an alternative is
eligible for selection as a preferred alternative in the Proposed
Plan and ultimate selection in the ROD. The five Primary
Balancing Criteria provide the basis for determining which
alternative provides the best balance of trade-offs among all
others considered. The State of South Carolina has reviewed this
ROD and concurs with EPA's selected remedy described in Section
9.0. The State of South Carolina concurrence letter is attached
as Appendix A to this ROD. Community Acceptance is addressed in
the Responsiveness Summary attached as Appendix B to this ROD.
The following discussion addresses the Threshold and Primary
Balancing Criteria.
8.1 OVERALL PROTECTION OF HUMAN HEALTH AND THE ENVIRONMENT
Overall protection of human health and the environment
addresses whether each alternative provides adequate protection
of human health and the environment and describes how risks posed
-------
Final ROD-Sangamo OU2
June 1994
page 83
through each exposure pathway are eliminated, reduced, or
controlled, through treatment, engineering controls, and/or
institutional controls.
Complexities related to the cycling of PCBs in the biota of
Lake Hartwell cause residual risk levels to remain above 1 x 10"4
regardless of which remedial action alternative is implemented.
A reduction in sediment PCB concentrations does not result in an
immediate reduction in PCB concentrations in fish. High PCB
concentrations in fish during the initial years of the 30-year
exposure duration result in average risk estimates exceeding
acceptable levels. Therefore, all of the alternatives rely on
institutional controls to minimize exposures to PCB-contaminated
fish tissue (limiting these exposures was identified as one of
the primary remedial action objectives). These controls include
public education and continuation of the current fishing
advisory.
The most protective alternatives are those that rely
principally on engineering rather than institutional controls to
reduce or eliminate exposures, given the uncertainty regarding
the performance/reliability of the institutional controls. The
removal/disposal and removal/treatment/disposal actions under
Alternatives 4 and 5, respectively, provide the highest level of
protectiveness, as the contaminated material is removed from the
lake and isolated in an engineered disposal facility.
Alternative 2B entails construction of a fish fence to be
designed and constructed so as to maximize effectiveness while
minimizing operations and maintenance; however, additional, pre-
design studies would be needed to develop the actual'design for
the structure. Alternatives 3A and 3C involve capping actions
that should permanently isolate the contaminated sediment in all
or a major portion of the Twelvemile Creek Arm. The capping
would be highly protective of human health and the environment;
its reliance on long-term monitoring and maintenance to assure
cap integrity is a disadvantage but does not compromise the
overall protectiveness or long-term effectiveness of the action.
8.2 COMPLIANCE WITH APPLICABLE OR RELEVANT AND APPROPRIATE
REQUIREMENTS (ARARs)
Section 121 (d) of CERCLA requires that remedial actions at
NPL sites at least attain legally applicable.or relevant and
appropriate federal and state requirements, standards, criteria
and limitations which are collectively referred to as "ARARs",
unless such ARARs are waived under CERCLA section 121(d)(4).
Applicable requirements are those substantive environmental
protection requirements, criteria, or limitations promulgated
under federal or state law that specifically address hazardous
substances found at the site ("contaminant-specific ARARs}, the
remedial action to be implemented at the site ("action-specific
ARARs"), the location of the site ("location-specific ARARs"), or
other circumstances at the site. Relevant and appropriate
requirements are those substantive environmental protection
requirements, criteria, or limitations promulgated under federal
or state law which, while not applicable to the hazardous
substances found at the site, the remedial action itself, the
-------
Final ROD-Sangamo OU2
June 1994
page 84
site location or other circumstances at the site, nevertheless
address problems or situations sufficiently similar to those
encountered at the site that their use is well-suited to the
site.
All of the alternatives rely on natural fate and transport
processes to reduce PCB concentrations in Lake Hartwell fish to
the FDA tolerance level of 2 mg/kg. Several alternatives include
actions designed to speed up the rate at which these natural
processes act (i.e. by capping or removing contaminated
sediment) . No contaminant-specific ARARs for PCBs in sediment
were identified, which led to the identification of To-Be-
Considered (TBC) criteria and the selection of a sediment cleanup
goal of 1 mg/kg based on an evaluation of these criteria.
Alternatives 3A (Capping), 4 (Confined Disposal Facility) and
5 (Stabilization) would best achieve the TBC criteria for
sediment. The sediment control structures (Alternatives 3B and
3C) are expected to impact PCB concentrations in the sediment,
but not to the same extent as capping or removal.
Alternatives 4 and 5 would trigger the largest number of and
most complex action- and location-specific ARARs, including the
Clean Water Act dredge-and-fill requirements, South Carolina
requirements for surface water discharges, and, if PCB
concentrations in dredged/excavated sediments exceed 50 ppm TSCA
requirements for handling/storage/treatment/disposal. The
alternatives could be designed and implemented to attain
compliance with these requirements. Capping activities under
alternatives 3A and 3C would also trigger Clean Water Act and
South Carolina requirements related to surface water.discharges;
compliance with these requirements would be factored into the
design and implementation. Action-specific ARARs were not
identified for the institutional controls under any alternative
or for construction of the fish fence under Alternative 2B.
8.3 LONG-TERM EFFECTIVENESS AND PERMANENCE
Long-term effectiveness and permanence refers to the
expected residual risk and the ability of a remedy to maintain
reliable protection of human health and the environment over
time. The criterion includes the consideration of residual risk
and the adequacy and reliability of controls.
Contaminant transport and bioaccumulation modeling conducted
during the RI showed that Lake Hartwell is a dynamic system in
which PCB concentrations in sediment and biota are gradually
declining in response to natural fate and transport processes
such as burial by clean sediment, resuspension and desorption
followed by flushing out of the reservoir, and other processes.
The modeling results indicate that PCB levels in the sediment
will decline below the 1 mg/kg cleanup goal in 5 to 10 years
(depending on location within Twelvemile Creek Arm) and that PCB
concentrations in fish will decline to the FDA tolerance level in
12 years. If the modeling predictions are accurate, even
Alternative 1 (No Action) would eventually result in a permanent
solution for the site in approximately 12 years, given that
consumption of contaminated fish was the primary exposure pathway
of concern.
-------
Final ROD-Sangamo OU2
June 1994
page 85
Alternatives 2 to 5 rely on institutional controls to
prevent exposure to contaminated fish; the performance and
reliability of these controls is dependent on site-specific
factors and therefore difficult to predict. In the context of
technology performance and reliability, the fish fence
(Alternative 2B) has not been proven for the same scale of
application proposed for Sangamo OU2. Capping (Alternatives 3A
and 3C) is a proven technology but is dependent on an extensive
program of monitoring and maintenance to ensure cap integrity and
long-term effectiveness. Capping of only the downstream portion
of the Twelvemile Creek Arm under Alternative 3C would result in
a cap that was easier to maintain, and therefore, it should have
a higher degree of long-term effectiveness. Sediment control
structures (Alternatives 3B/3C) are also widely used and are less
dependent on maintenance.
Although dredging and confined disposal facilities are
widely recognized as reliable technologies, further technology
evaluation, in the form of pilot-scale treatability studies,
would be needed to obtain design and operating data essential to
minimize environmental impacts. Similarly, stabilization
treatability studies under Alternative 5 would require further
technology evaluation to determine material handling and process
requirements (e.g., optimum stabilization mixture). Treatability
studies are also needed to determine effluent treatment
requirements for the dewatering fluids from the confined disposal
facility or material handling/stockpiling areas.
8.4 REDUCTION OF TOXICITY, MOBILITY, OR VOLUME
Reduction of toxicity, mobility, or volume through treatment
refers to the anticipated performance of the treatment
technologies for the remedy. None of the alternatives has an
effect on the toxicity of the contaminants, expect Alternative 5
which involves treatment by stabilization. Alternative 5 would
accomplish the most significant reduction in the mobility of PCBs
through a stabilization treatment process, followed by disposal
in a confined disposal facility. Contaminant mobility would also
be reduced through disposal in a CDF, with no treatment, in
Alternative 4.
The capping alternatives (Alternatives 3A and 3C) also
accomplish a reduction in contaminant mobility, but the effect is
contingent upon proper maintenance of the cap. Reductions in
contaminant mobility are expected with the sediment control
structures (Alternatives 3B and 3C), primarily through
maintaining pool elevation, thereby decreasing scour/erosion in
the headwaters. The capping, sediment control structures and
fish fence alternatives also reduce PCB transfer to migratory
fish by isolating the contaminated sediment (through capping) or
restricting the migration of fish into and out of the areas with
the highest sediment contamination (through the fence or weirs).
The net result is a reduction in the level of PCB contamination
in lake-wide migratory fish.
-------
Final ROD-Sangamo OU2
June 1994
page 86
8.5 SHORT-TERM EFFECTIVENESS
Short-term effectiveness refers to the period of time needed
to complete the remedy and any adverse impacts on human health
and the environment that may be posed during the construction and
implementation of the remedy.
Implementation of Alternatives 1 (No-Action) and 2A
(Institutional Controls) would have the least impact on the
community, workers, and the environment; Alternative 2B has only
a slightly greater impact. Alternatives 3A/3B/3C have greater
impacts on the local community due to construction of sediment
control structures and excessive truck traffic associated with
the transportation of capping material. The greatest impacts
(and least short-term effectiveness) are associated with the more
aggressive alternatives that entail dredging, re-channeling, and
construction of the CDF (Alternatives 4/5).
Table 8-1 summarizes the estimated schedule for each
alternative, presenting the time required to implement, time
required for fish'PCB concentrations to decline to protective
levels (FDA tolerance level), and total time for alternative to
achieve protectiveness. Because the most highly contaminated
sediment is concentrated in a relatively small, upstream portion
of the lake that represents less than 10 percent of the reservoir
area, separate estimates are shown for the time to attain FDA
levels and time to achieve protectiveness in both this upstream
portion (i.e., the Twelvemile Creek/Seneca River Arm) and in the
main body of Hartwell Lake. The impact on a given alternative
with regard to protective levels throughout the lakeis an
important consideration given the significant resource potential
of the reservoir. It is important to note that the time required
for implementation includes design, procurement, and construction
but does not consider the time between issuance of the Proposed
Plan/signing of the ROD and initiation of the Remedial Design.
The table shows that isolation or removal of the most highly
contaminated sediment, through fisheries isolation, capping,
sediment control structures, or dredging often has a much greater
effect on PCS levels in downstream areas than within the
Twelvemile Creek Arm. This response is mostly due to the impact
of limiting or preventing the movement of migratory fish into and
out of the areas with the more highly contaminated sediment
(lake-wide, non-migratory fish have exceeded the FDA levels on an
infrequent and isolated basis while migratory fish have typically
had PCB levels higher than 2 mg/kg). Accelerated declines in PCB
levels would be expected for alternatives involving capping or
removal; otherwise, PCB reductions would be accomplished only
through natural fate and transport processes conditions.
The most favorable alternatives with regard to the estimated
time to achieve protectiveness would be Alternatives 2B and 3A;
fisheries isolation (Alternative 2B) is more quickly implemented
and therefore leads to a slightly faster reduction in PCB levels
in the main body of Hartwell Lake but the capping alternative
would have a much greater effect in the Twelvemile Creek/Seneca
River Arm. If more than 1 to 2 years elapse between the signing
-------
Table 8-1
Estimated Time to Achieve Protectiveness
ALT.
No.
i
2A
2B
3A
3B
3C
4
5
DESCRIPTION
No action
Institutional Controls
Fisheries Isolation
Capping
Sediment Control Structure
Optimal Capping/Sediment
Control Structure
Confined Disposal Facility
Stabilization
TIME To
IMPLEMENT"'
(yrs)
0
0.5
1
2
2
2-3
3-4
4-5
TIME FOR FISH To ATTAIN FDA LEVEL
(yrs)
TWELVEMILE
CREEK/SENECA
RIVERA ARM
12
12
12
3-4
12
3-4 (dovmtran)
12 (>f»tfc«n)
5-6
5-6
MAIN BODY OP
HARTWELL LAKE*"
12
12
3-4
3-4
3-4
3-4
5-6
5-6
TIME TO ACHIEVE PROTECTIVENESSW
(yrs)
TWELVEMILE
CREEK/SENECA
RIVER ARM
12
12
12
5-6
12
5-7 (downturn)
12 (IVMIWH)
8-10
9-11
MAIN BODY OF
HARTWELL LAKE
12
12
4-5
5-6
5-6
5-7
8-10
9-11
NOTE: (1) Includes time needed for pre-design studies, design, procurement, and construction, but does not include time between Proposed Plan/Record of Decision and initiation of
Remedial Design.
(2) Twelvemile Creek Arm only; FGETS modeling results indicate that fish will decline to FDA tolerance level (2 mg/kg) in approximately 12 yean under baseline conditions;
none of the alternatives will extend the period required for this to occur (i.e., none of the alternatives can have a period to achieve protectiveness of greater than 12 yrs).
(3) Downgradient of the Twelvemile Creek Arm (i.e., main body of Harrwell Lake), PCB concentrations will decline more rapidly for actions involving Isolation, capping or removal
of contaminated sediment (Alternatives 2B, 3A-C, 4 and 5). It was assumed that this rate of decline would be comparable to that observed in the Twelvemile Creek Arm for
Alternatives Involving capping (3-4 yrs) or removal of the sediment (5-6 yrs); further modeling is needed to verify this estimate.
(4) Duration only; based on estimated time to implement remedy and duration thereafter needed for fish PCB levels to decline to FDA level, based on the FGETS modeling
results presented in Section 3.3 and assumptions regarding migratory fish PCB concentrations trends.
-
s
f
W
p> m g
IQ Q
(D H
voo
CO VO Cl
^J *> to
-------
Final ROD-Sangamo OU2
June 1994
page 88
of the ROD and the initiation of the Remedial Design, the
dredging alternatives (Alternatives 4/5) will not achieve
protectiveness appreciably faster than that obtained under
baseline (no-action) conditions.
8.6 IMPLEMENTABILITY
Alternatives 1 and 2A would be the most readily
implement able; the main activity associated with the
institutional controls under Alternative 2A (and other
alternatives) involves planning and coordination with COE,
SCDHEC, and other agencies. Alternative 26 (Fish Fence) would be
somewhat more complex to implement, primarily due to the
predesign testing to assess the suitability of various designs.
Alternatives 3A, 3B, and 3C represent an intermediate range
of implementability. Capping and weirs have been successfully
constructed at a wide variety of sites and conditions at Hartwell
Lake and should not present any unusual difficulties. The
reduced scale of the area to be capped and size of weir for
Alternative 3C would simplify both design and construction
activities associated with these individual components. A
limitation to Alternative 3A and, to a lesser extent, 3C,
concerns the availability and transportation logistics of as much
as 1.8 million yd3 of capping material. From a technical and
engineering perspective, construction of the fixed crest weirs
for Alternatives 3B and 3C should be much more implementable than
capping due to the likelihood of encountering problems with the
capping materials.
The most difficult alternatives to implement, in terms of
both technical and administrative feasibility, would be
Alternatives 4 and 5; both involve dredging, which is a widely
used technology, but they would still require considerable
predesign testing .and evaluation to optimize operations and
minimize environmental impacts. Construction of the confined
disposal facility in what is currently a large body of open water
and rechanneling the Twelvemile Creek Arm through a forested
ridge would also present a number of technical challenges, while
not insurmountable, the challenges do make implementation of
Alternatives 4 and 5 considerably more complex than the other
remedial alternatives. These alternatives would also have the
greatest administrative requirements due to the need to fulfill
permit-related requirements related to Clean Water Act dredge-
and-fill permits, South Carolina NPDES permits, and possibly
other permitted activities.
8.7 COST
Costs for each alternative were provided under Section 7.3
of this ROD. Alternative 5 is the most expensive alternative to
implement, and is considered cost prohibitive. The more
aggressive alternatives (Alternatives 3A/3B/3C/4) that involve
engineering controls to satisfy the remedial action objectives
have total present worth costs that generally range from $30-50
million. Alternative 2B achieves protectiveness in the main body
of Lake Hartwell in a shorter amount of time than the other more
-------
Final ROD-Sangamo OU2
June 1994
page 89
aggressive and passive (Alternatives 1/2A) alternatives for only
an additional $1 million over the cost of Alternative 2A
(estimated present worth cost of $3,208,000). The No-Action
alternative is the least expensive.
9.0 THE SELECTED REMEDY
This section of the ROD discusses EPA's selected remedy in
detail for the Sangamo OU2 Site. EPA has selected Alternative 2A
- Institutional Controls to address the contaminated sediments
and fish at the Site. The selected remedial alternative for the
Sangamo OU2 Site consists of the following components: 1)
continuation of existing fish advisory; 2) public education
program; 3) future sediment and aquatic biota monitoring; and 4)
regulation of the Twelvemile Creek Impoundments. These primary
components are discussed below.
9.1 CONTINUATION OF EXISTING FISH ADVISORY
The existing fish advisory on Lake Hartwell shall remain in
effect. Signs warning against consuming fish have been posted at
the majority of the public boat launch and recreation areas since
1987. Figure 9-1 provides an illustration of the posted
advisory. The advisory is currently maintained by SCDHEC, and
SCDHEC will remain responsible for management, supervision, and
administration of the fish consumption advisory in the future.
The advisory will be modified if warranted by future trends
regarding PCB concentrations in fish. Modifications shall be
fully supported and justified by the annual monitoring program
discussed in Section 9.3 of this ROD.
Maintaining the fish advisory will likely entail the posting
and replacement of signs describing the advisory at access points
along the shoreline. For purposes of enhancing the effectiveness
of the advisory, additional signs may be posted. Additional
activities designed to increase public awareness of the existing
fish consumption advisory are discussed in Section 9.2 of this
ROD.
As discussed in the Section 5.4.3, F6ETS bioaccumulation
food-chain modeling, the mean PCB concentration in 2-8 year old
largemouth bass is predicted to fall below the FDA limit by year
10. However, a longer period of time is required for
concentrations in higher age classes to fall below the FDA limit.
PCB concentrations in 10-year-old fish, representing the highest
age class and largest fish in the model, required 12 years to
fall below the 2.0 mg/kg FDA limit. For these reasons, it was
assumed that the fish consumption health advisory would remain in
effect for a minimum of 12 years, or the year 2005.
9.2 PUBLIC EDUCATION PROGRAM
Proper preparation of contaminated fish can reduce the
quantity of contamination consumed. Contaminants (PCBs) are
generally stored in fatty tissue (i.e., belly flap, strip along
the backbone and lateral line, and skin). By removing these
-------
Final ROD-Sangamo OU2
June 1994
page 90
ATTENTION
FISH CONSUMPTION ADVISORY - LAKE HARTWELL
S.C. DEPARTMENT OF HEALTH AND
ENVIRONMENTAL CONTROL (SCDHEC)
ALL RSH TAKEN FROM THE SENECA RIVER ARM
OF LAKE HARTWELL NORTH OF SC HIGHWAY 24
AND 12 MILE CREEK SHOULD BE RELEASED AND
NOT EATEN.
ALL FISH LARGER THAN THREE (3) POUNDS
TAKEN FROM THE REMAINDER OF LAKE HART-
WELL SHOULD BE RELEASED AND NOT EATEN.
FISHING IS NOT PROHIBITED BUT SCDHEC
ADVISES THAT THESE FISH NOT BE EATEN DUE
TO THE PRESENCE OF ELEVATED LEVELS OF
POLYCHLORINATED BIPHENYLS (PCBs). SWIM-
MING, BOATING, AND OTHER WATER RELATED
ACTIVITIES ARE NOT RESTRICTED BY THIS
ADVISORY.
FOR ADDITIONAL INFORMATION,
CONTACT SCDHEC AT:
COLUMBIA GREENVILLE ANDERSON
734-5300 242-9850 225-3731
Figure 9-1
Current Lake Hartwell Fish Advisory
-------
Final ROD-Sangamo OU2
June 1994
page 91
areas when cleaning the fish, individuals can substantially
reduce their intake of PCBs. Broiling, baking, and grilling fish
also provides additional risk reduction. Many states, including
Georgia, provide fish preparation suggestions to the public in
pamphlets and other publications on state fishing regulations;
South Carolina currently does not.
EPA shall develop a pamphlet that outlines methods,
including those described above, for reducing the intake of
contaminants through fish consumption. This pamphlet will be
distributed to the general public and to frequent Lake Hartwell
resource users. The most effective method of distribution to the
general public is to provide the pamphlet to everyone who
purchases a fishing license. Thus, all establishments that sell
fishing licenses in the Hartwell Lake area will be supplied with
the pamphlets, as will marinas and selected retail establishments
(i.e., convenience stores) in the area. In addition to the
pamphlets, advertisements will be placed periodically in
newspapers from all of the counties surrounding the lake. The
advertisements will reiterate the potential risks associated with
consumption of PCB-contaminated fish from Lake Hartwell.
Printed information regarding the fish advisory will also be
supplemented with periodic public meetings hosted by EPA, SCDHEC,
and the COB. Public meetings will begin the first year following
signing of the ROD and will be held at a minimum of 5-year
intervals. The public meetings will serve to update the general
population on PCB concentration trends in sediment and fish as
well as the latest estimates of PCB-related risks resulting from
consumption of contaminated fish. Analytical results from the
annual monitoring program described in Section 9.3 and the five-
year reviews, as required by CERCLA, will be disseminated to
those attending these meetings, local citizens, and frequent lake
users. The meetings would also provide a forum for public input
into the decision-making process regarding continued
institutional controls for the lake. The proposed locations for
these meetings include:
Anderson, S.C.
Cateechee/Norris S.C.
Clemson, S.C.
Hartwell, Ga.
Seneca, S.C.
The above public education program represents a baseline
approach. EPA, COE, and SCDHEC will periodically evaluate the
overall approach and effectiveness of this program by soliciting
public comment and input from Lake Hartwell resource users.
Creel surveys, similar to those conducted as part of the
Biological Investigation, may also be utilized as a mechanism to
obtain information from anglers on Lake Hartwell. The public
education program is expected to evolve while attempting to
identify the most productive method(s) for reaching the targeted
audience. EPA, in close consultation with SCDHEC and the COE
shall modify the public education program as required.
-------
Final ROD-Sangamo OU2
Jxine 1994
page 92
9.3 AQUATIC BIOTA AMD SEDIMENT MONITORING
This section describes the program of annual monitoring of
PCS levels in the fish, corbicula, and sediment of Lake Hartwell
and the Twelvemile Creek watershed.
9.3.1
AQUATIC BIOTA MONITORING
In the interest of consistency and to support trend
analysis, fish samples will be collected from the same 6 Lake
Hartwell stations that have been utilized historically by SCDHEC
and also during the Biological Investigation. These stations are
shown on Figure 9-2 and are generally described as follows:
SV-107 Twelvemile Creek Arm of Lake Hartwell in the vicinity
of County Road 37; Pickens County, SC.
SV-106 Seneca River Arm of Lake Hartwell in the vicinity of
Martin Creek; Anderson County, SC.
SV-532 Seneca River Arm of Lake Hartwell in the vicinity of
Eighteen Mile Creek; Anderson County, SC.
SV-535 Andersonville Island area of Lake Hartwell near the
confluence of the Tugaloo and Seneca Rivers; Anderson
County, SC.
SV-641 Tugaloo River Arm of Lake Hartwell in the vicinity of
Interstate 1-85; Oconee County, SC.
SV-642 Open water area of Lake Hartwell in the vicinity of
Hartwell Dam; Anderson County, SC.
Fish samples shall be collected annually in the spring using
gill nets and/or electrofishing techniques. Hybrid bass (weight
class 3.0-5.0 Ibs.), largemouth bass (weight class 1.5-3.0 Ibs.)
and channel catfish (weight class 2.0-4.0 Ibs.) will be collected
from the above 6 stations in Lake Hartwell. In past sampling
events, 10 hybrid bass, 10 largemouth bass, and 4 catfish were
collected at each station to calculate a representative mean.
The adequacy of these respective numbers in determining a
representative mean concentration in fish tissue is currently
under review by SCDHEC. In any event, sampling conducted
pursuant to this ROD will at a minimum include 10 hybrid bass, 10
largemouth bass, and 4 catfish. Game fish shall be prepared in
accordance with the standard US FDA fillet method (i.e., fillet
including rib cage and belly flap with skin on and scales off -
except for catfish where skin is removed.) . All fish shall be
analyzed for the PCS component of EPA's Target Compound List
(TCL).
Forage fish, consisting of gizzard shad/blueback herring,
threadfin shad and bluegill, shall be collected at stations SV-
107, SV-532, and SV-641. Forage fish will be collected
concurrently with game fish sampling discussed above using
electrofishing techniques. Samples will consist of whole body
composite samples for each species. All forage fish samples
-------
Final ROD-Sangamo OU2
June 1994
page 93
1-
Fish Sampling Station
Tributary
SCALE
4
I I 1 I I
MILES
Figure 9-2
Fish Sampling Stations in Lake Hartwell
-------
Final ROD-Sangamo OU2
June 1994
page 94
shall be analyzed for PCBs. Data generated from the forage fish
collection and analyses shall be utilized to support future food-
chain modeling activities.
In addition to fish sampling in Lake Hartwell, corbicula
(fresh water clams) baskets will be placed at 6 locations in
Twelvemile Creek for 28 days each in the spring of each year to
determine current PCB loading into the Twelvemile Creek system.
The sampling locations for corbicula baskets are shown on Figure
9-3. Corbicula baskets will be deployed concurrent with the fish
sampling activities described above. After 28 days, baskets
shall be recovered and analyzed for PCBs.
9.3.2 SEDIMENT MONITORING
Sediment monitoring locations are shown in Figures 9-4 and
9-5. Sediment samples shall be collected annually from 15
locations in Hartwell Lake and 5 locations in the watershed.
Sediments will be collected using a ponar dredge. Stations SD-
001 through SD-005 in the Twelvemile Creek watershed will be
discrete samples. For the remaining 15 locations in Lake
Hartwell, a transect shall be established at each station and 5
grab samples will be collected along the transect. These 5
samples shall be composited to provide a mean concentration of
PCBs in surface sediments from a given transect. Sample and
transect locations shall be established with sufficient accuracy
to enable collection of future samples from same locations.
Details of the sediment sample locations in Twelvemile Creek and
Lake Hartwell are presented in Table 9-1.
«
All sediment and aquatic biota monitoring shall continue
under the direction of EPA. Data from the annual monitoring will
allow EPA, SCDHEC and COE to continue ongoing efforts to evaluate
contamination trends in the sediment and biota. These trends
will be used to support decisions to modify the advisory, in
response to PCB concentration trends in the fish. Annual
monitoring results shall be summarized and disseminated to the
public in support of the public education program described under
Section 9.2. For purposes of the cost estimate described in
Section 9.6, it was assumed that this annual monitoring program
will continue for a minimum of 15 years from signature of the
ROD.
9.4 REGULATION OF TWELVEMILE CREEK IMPOUNDMENTS
The Federal Energy Regulatory Commission (FERC) regulates
hydroelectric facilities in the U.S. Currently, the middle and
lower impoundments on Twelvemile Creek, Woodside I and II,
respectively, are non-licensed hydroelectric impoundments based
on the date of construction and navigability of Twelvemile Creek.
These impoundments effectively trap a large portion of the
sediment load from Twelvemile Creek, and as a result, are flushed
periodically to remove sediment and to restore storage capacity
within the impoundments. Historically, these trapped sediments
have contained high levels of PCBs. However, the RI and
subsequent sampling in response to a recent flushing event
(September 1993) have shown that more recent sediment
-------
Final ROD-Sangamo OU2
June 1994
page 95
n Cr
A
Easley-Picken
Landfill
Lay Bridge'
Rd.
CBS is located
between the Uiw^
trul Madden Bruges
Coitieuta Basket
Sangamo Weston Plant
OHa'le Disposal Area
AUndlill
Dim
SCALE
MILES
Figure 9-3
Corbicula Basket Sampling Locations in Twelvemile Creek Watershed
-------
Final ROD-Sangamo OU2
June 1994
page 96
1"
Sediment Sampling Location
Tributaty
SCALE
4
. . I i
MILES
Figure 9-4
Sediment Sampling Stations in Lake Hartwell
-------
Final ROD-Sangamo OU2
June 1994
page 97
Sediment Sampling Location
SangamoWeston Plant
Otlsite Disposal Ana
A Landfill
Dam
SCALE
MILES
Figure 9-5
Sediment Sampling Stations in Twelvemile Creek
-------
Final ROD-Sangamo OU2
June 1994
page 98
TABLE 9-1
Sediment Sample Locations in Twelvemile Creek Watershed and
Arm of Lake Hartwell
Station ID
SD-001
SD-002
SD-003
SD-004
SD-005
SD-006
SD-007 - 009
SD-010
SD-011
SD-012
SD-013
SD-014
SD-015
SD-106
SD-532
SD-535
SD-641
SD-642
- location
Town Creek downstream of Sangamo West on
outfall - same location as corbicula basket 2
Twelvemile Creek upstream of Wolf Creek
confluence - same location as corbicula basket 3
Wolf Creek downstream of Breazeale site - same
location as corbicula basket 4
Twelvemile Creek at Easley- Central impoundment -
same location as corbicula basket 5
Twelvemile Creek at Lay Bridge
Maw Bridge
Between Maw and Madden Bridges
Madden Bridge
Between Madden and Hwy 133 Bridges
Hwy 133 Bridge
Mouth of Twelvemile Creek Arm
Between US 123 and Hwy 93 Bridges
Hwy 37 Bridge
SV-106 fish sampling station
SV-532 fish sampling station
SV-535 fish sampling station
SV-641 fish sampling station
SV-642 fish sampling station
accumulating behind the small dams has relatively low levels of
PCB contamination (1 to 3 mg/kg). These low concentrations
confirm that PCB-contaminated sediments continue to flow down
Twelvemile Creek into Hartwell Lake, but the magnitude of this
contaminant flux is small.
Although flushing of these impoundments is not currently
contributing appreciable quantities of PCB contamination to the
Twelvemile Creek Arm of the lake, the release of the equivalent
of a half-year's sediment load (or more) during a short time
period (typically less than 1 week) has been documented to have
potentially significant adverse effects on water quality in the
upper portion of the Twelvemile Creek Arm. The influx of
sediment in the most recent flushing event (September 1993) was
observed to result in a notable fish kill in Twelvemile Creek
below the impoundments and in the upper portion of the Twelvemile
-------
Final ROD-Sangamo OU2
June 1994
page 99
Creek Arm. This was attributed to elevated levels of suspended
sediments and not PCBs. Similar and possibly more damaging
effects to benthic organisms within the upper portion of the
reservoir are also likely to occur.
The frequency at which these impoundments are flushed is
based primarily on an operations schedule established by the
owner/operator rather than hydraulic or other technical
constraints. To minimize the negative impacts on water quality,
and to facilitate burial of relatively more contaminated
sediments downstream, EPA's selected remedy shall include the
development of a routine schedule for sluicing of all three
impoundments (i.e., Woodside I and II, and Easley-Central Water
District). Quarterly flushing of these impoundments is thought
to be adequate to minimize the negative impacts.
Preliminary discussions with Consolidated Hydro, Inc.,
owners/operators of Woodside I and II, have indicated a
willingness to comply with this requirement. It is estimated
that approximately $200,000 in capital improvements to the gating
mechanisms are necessary to fulfill the objectives of this
component of EPA's selected remedy. Compliance with the routine
schedule shall be monitored by SCDHEC. In addition, water
quality and sediment monitoring will be conducted on downstream
areas during the sluicing events to facilitate modification of
the schedule as needed.
9.5 FIVE YEAR REVIEWS
As required by CERCLA, 5-year reviews will be conducted to
ensure that the alternative remains protective of human health
and the environment. The results of the annual monitoring
studies shall be incorporated into the 5-year reviews. The 5-
year reviews shall continue at intervals no less often than 5
years following signature of this ROD until protectiveness has
been achieved.
9.6 COST ESTIMATE
Table 9-2 provides a cost breakdown, by component, of Alternative
2A - Institutional Controls. The estimated total present worth
cost for EPA's selected remedy is $3,208,000, which includes
$366,000 of capital costs.
-------
Final ROD-Sangamo 0172
June 1994
page 100
TABLE 9-2
Cost Estimate for EPA's Selected Remedy, Alternative 2A-
Institutional Controls
'Remedy Component/'
, Cost Iteni
Fish Advisory
Public Education
Biota & Sediment
Monitoring
Twelvemile Creek
Impoundments
5 -Year Reviews
Subtotal
Engineering 10%
Procurement 5%
Construction
Management 15%
Subtotal
Indirects/Profit
Subtotal
Cont ingency 2 0 %
TOTAL
Capital
Cost($)
$4,000
$200,000
$204,000
$20,000
$10,000
$31,000
$265,000
$40,000
$305,000
$61,000
$366,000
, OSM Cost
HS/yx.r
$1,000
$33,000
$132,000
$166,000
$17,000
$8,000
$25,000
$216,000
$32,000
$248,000
$50,000
$298,000
O&M Cost' ,
<$5/yrs.J
$25,000
$28,000
$53,000
$5,000
$3,000
$8,000
$69,000
$10,000
$79,000
$16,000
$95,000
, Present ;
', 4 Worth - '"
Cosfc{$}; -
$12,000
$314,000
$1,202,000
$200,000
$60,000
$1,788,000
$179,000
$89,000
$268,000
$2,324,000
$349,000
. $2,673,000
$535,000
$3,208,000
10.0 STATUTORY DETERMINATIONS
This section of the ROD describes how EPA's selected remedy,
Alternative 2A - Institutional Controls, meets the statutory
requirements as delineated in Section 121 of CERCLA.
10.1 PROTECTION OF HUMAN HEALTH AND THE ENVIRONMENT
The extent to which EPA' selected remedy is protective of
human health and the environment is directly dependent upon the
effectiveness of the institutional controls that are the
principal components of this response action. The reliability of
the institutional controls in limiting exposure is difficult to
predict due to the significance of site-specific demographic
factors governing the likely response of the public to the
advisory and other component of this remedy that are designed to
limit exposures. Risks posed by the primary exposure pathway,
ingestion of PCB contaminated fish, are declining due to natural
fate and transport processes, primarily the burial of
contaminated sediment. Baseline modeling activities predict that
PCB concentrations in largemouth bass fillets would decline to
the FDA tolerance limit in approximately 12 years (year 2005).
-------
Final ROD-Sangamo OU2
June 1994
page 101
Effective implementation of the institutional controls described
in Section 9.0 of this ROD would limit exposures until PCB
contamination dropped to these protective levels. Therefore,
this alternative is considered to be adequately protective of
human health and the environment.
10.2 COMPLIANCE WITH APPLICABLE OR RELEVANT AND APPROPRIATE
REQUIREMENTS
The only applicable requirement identified for the selected
remedy is the contaminant-specific ARAR governing PCB
concentrations is fish. Continuation of the existing fish
advisory in conjunction with the aggressive public education
program addresses the FDA requirement governing consumption of
PCB contaminated fish.
No sediment-related, contaminant-specific ARARs were
identified. TBC and health-based criteria for sediment were
identified in the discussions of ARARs and sediment cleanup
goals. Alternative 2A does not attain the TBC or health-based
levels directly, but these levels would be attained through
natural sedimentation processes in the reservoir. Depending on
the area of the Twelvemile Creek Arm, the selected TBC sediment
cleanup goal of 1 mg/kg will be attained in a period of 5-10
years.
10.3 COST EFFECTIVENESS
EPA believes this remedy will minimize risks posed to human
health and the environment at an estimated cost of $3,208,000.
Complexities related to the cycling of PCBs in the aquatic biota
of Lake Hartwell cause residual risk levels to remain above
acceptable levels regardless of which remedial alternative is
implemented. The most effective means to limit exposure to PCB-
contaminated fish is by the effective implementation of the
primary components of EPA's selected remedy. Alternative 2A -
Institutional Controls. Therefore, in terms of risk-reduction
for money spent, Alternative 2A is the most cost-effective
remedial alternative that was evaluated in detail by EPA.
10.4 UTILIZATION OF PERMANENT SOLUTIONS AND ALTERNATIVE TREATMENT
TECHNOLOGIES TO THE MAXIMUM EXTENT PRACTICABLE
EPA has determined and the State of South Carolina has
concurred with EPA's determination that the selected remedy
represents the maximum extent to which permanent solutions can be
utilized in a practicable manner for Operable Unit Two of the
Sangamo Site. Engineering controls and treatment technologies
did not offer a significant advantage in terms of overall risk
reduction for the cost. EPA and the State have determined that
the selection of Alternative 2A - Institutional Controls,
provides the best balance of trade-offs in terms of the five
balancing criteria, while also considering the statutory
preference for treatment as a principal element and community
acceptance. The selected remedy addresses the primary risk
associated with Sangamo OU2, the ingestion of PCB contaminated
fish, by the most effective means practicable.
-------
Final ROD-Sangamo OU2
June 1994
page 102
10.5 PREFERENCE FOR TREATMENT AS A PRINCIPAL ELEMENT
EPA's selected remedy does not satisfy the preference for
treatment as a principal element due to the excessive costs (>
$500 million) associated with removal, treatment and disposal of
approximately 4.7 million cubic yards of PCB contaminated
sediment spread out over approximately 730 acres. Moreover,
alternatives that involved aggressive engineering controls to
contain, remove, and dispose of PCB contaminated sediment ranged
from $30-$50 million without offering a significant advantage in
terms of overall risk reduction.
11.0 DOCUMENTATION OF SIGNIFICANT CHANGES
Pursuant to the requirements of Section 117(b) of CERCLA,
this section of the ROD documents and discusses the reasons for
significant changes made to the selected remedy from the time the
Proposed Plan was released to the public to the final selection
of the remedy in this ROD. For reference, the Proposed Plan is
attached to this ROD as Appendix C. Specific written and/or oral
comments received during the formal 60-day public comment period
held from April 11 through June 10, 1994 are discussed in the
Responsiveness Summary (Appendix B).
EPA's preferred alternative, as delineated in the Proposed
Plan, was Alternative 2B - Fisheries Isolation. In addition to
the Institutional Controls under Alternative 2A, Alternative 2B
also involved the construction of a fishery isolation barrier
("fish fence") to prohibit the movement of migratory.fish into or
out of the area of Lake Hartwell with the highest PCB
concentrations in sediment. Placement of the fish fence near the
Hwy 37 Bridge, just south of Clemson on the Seneca River Arm,
would result in isolation of less than 10 percent of the total
area of Lake Hartwell. Isolation of these upstream areas was
expected to result in accelerated declines in PCB concentrations
in migratory fish in the downstream portions of Lake Hartwell
(i.e. the remaining 90+ percent of the lake).
The reasons supporting EPA's preference for Alternative 2B
were simply: l) Migratory species represent approximately 50% of
the fish harvested by weight from Lake Hartwell. Accelerated
reductions of PCB concentrations in these fish, for the majority
of Lake Hartwell, would achieve a higher degree of protection by
eliminating, or at the very least, minimizing exposures to this
group of migratory fish; 2) Reduction in fish PCB levels would
allow for rescinding of the existing fish advisories in the areas
downstream of the fish fence, in an accelerated time frame, thus
returning the majority of Lake Hartwell to its highest beneficial
use; 3) The incremental cost to implement Alternative 2B was
only approximately $1 million over that of Alternative 2A.
Moderate, yet unwavering public opposition was expressed to
EPA's preferred alternative at the Proposed Plan public meeting
held in Clemson, SC on April 19, 1994. This and subsequent
comments received during the public comment period suggest that
the general consensus supports Alternative 2A and/or no-further
-------
Final ROD-Sangamo OU2
June 1994
page 103
action. In summary, the reasons supporting public opposition to
Alternative 2B were essentially two-fold. First, the public does
not believe EPA is capable of designing, constructing and
installing a safe fishery isolation barrier that would meet the
specified performance standards (See Section 7.3) at the
estimated cost. Secondly, given the fact that Institutional
Controls provide the most reliable mechanism for reducing
exposures to PCB contaminated fish, the incremental cost for
constructing a fish fence is not justified.
EPA has given serious consideration to the concerns that
were voiced by the public. In response to these concerns, EPA,
in close consultation with SCDHEC and the Savannah District Corps
of Engineers, has selected Alternative 2A - Institutional
Controls as the Final remedial alternative for the Sangamo OU2
Site.
-------
Appendix
Proposed Plan Fact Sheet
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
SUPERFUND PROPOSED PLAN FACT SHEET
SANGAMO WESTON/TWELVE MILE CREEK/LAKE
HARTWELL PCB CONTAMINATION SUPERFUND SITE
OPERABLE UNIT TWO
Pickens County, South Carolina APRIL 1994
Anumberof term* tpecifie to the Superfundproefst (printed in told print) are defined in the giouary at U*e end
of thi* publication.
INTRODUCTION
The Region IV office of the United States
Environmental Protection Agency has
developed this Proposed Plan Fact Sheet for
the Sangamo Weston/Twelvemile
Creek/Lake Hartwell Polychlorinated
Biphenyl (PCB) Contamination Superfund
Site - Operable Unit Two ("Sangamo OU2
Site") in Pickens County, South Carolina.
This Proposed Plan is issued to present the
alternatives that EPA has considered to
address PCB contamination in the sediments
and aquatic biota of the Sangamo OU2 Site.
More specifically, the purpose of this
Proposed Plan is to:
Identify the preferred alternative for
remedial action at the Site and explain the
rationale for that preference;
Describe the other remedial options
considered in detail in the Final Feasibility
Study Report;
Solicit public review of and comment on
all of the alternatives described; and
Provide information on how the public
can be involved in the remedy selection
process at the Sangamo OU2 Site.
PROPOSED PLAN PUBLIC
MEETING
for the
SANGAMO-WESTON
SUPERFUND SITE, OU2
Tuesday, April 19,1994
7:00 P.M.
RAMADAINN
U.S. 76 & HWY. 123
Clemson, South Carolina
EPA, in consultation with the Savannah
District of the U.S. Army Corps of
Engineers (COE) and the South Carolina
Department of Health and Environmental
Control (SCDHEQ, will select a final
remedy for the Sangamo OU2 Site after the
public comment period has ended and the
information submitted during this time has
been reviewed and evaluated. The public is
encouraged to submit comments on all the
alternatives, and on the information that
supports these alternatives to the EPA
Remedial Project Manager listed at the end
of this publication. The final remedy, as
presented in a Record of Decision (ROD),
could differ from the Proposed Plan,
depending upon new information or input
EPA receives during the formal public
comment period.
-------
EPA is issuing this Proposed Plan as part of
its public participation responsibilities under
Section 117(a) of the Comprehensive
Environmental Response, Compensation
and Liability Act (CERCLA, more
commonly known as Superfund). This fact
sheet summarizes information that can be
found in greater detail in the Remedial
Investigation/Feasibility Study (RI/FS)
Reports and other site-related documents
contained in the Administrative Record,
located at the established information
repositories. The reader is referred to the
information repositories listed at the end of
this publication for a more detailed account
of this subject matter.
SUE BACKGROUND
From 1955 to 1987, Sangamo Weston, Lie.
owned and operated a capacitor
manufacturing plant in Pickens, South
Carolina, near the headwaters of Lake
HartwelL In its manufacturing processes,
Sangamo used several varieties of dielectric
fluids which contained PCBs. Waste
disposal practices from the Sangamo Plant
included land burial of off-specification
capacitors and wastewater treatment sludges
on the plant site and at six satellite disposal
areas. PCBs were discharged with untreated
effluent directly into Town Creek, which is
a tributary of Twelvemile Creek.
Twelvemile Creek is a major tributary of
Lake HartwelL It is estimated mat over
400,000 Ibs. of PCBs were released into the
Twelvemile Creek system until the use of
PCBs were discontinued prior to an EPA ban
in January 1978. Schlumberger Industries,
Inc. (SII) is the current owner of the plant
site, as a result of a merger with Sangamo
Weston in December 1989.
Historical sampling events conducted by
SCDHEC, COE, and EPA documented the
presence of widespread PCB contamination
of the sediments, surface water and aquatic
biota from the Sangamo Weston outfall on
Town Creek, through Twelvemile Creek, and
well into the Twelvemile Creek Arm of i^fcf
HartwelL During these sampling events, it
was discovered that fish from certain areas
of Lake Hartwell contained PCBs which
exceeded the recommended limit established
by the U.S. Food and Drug Administration
(U.S. FDA). Subsequent to these findings,
a fish consumption advisory was initially
issued in 1976 and later modified by
SCDHEC in 1985 as a conservative measure
to reduce human exposure. The advisory
states: 1) All fish taken from the Seneca
River Arm of Lake Hartwell north of SC
Highway 24 and Twelvemile Creek should
be released and not eaten; and 2) All fish
larger than 3 Ibs. from the remainder of
Lake Hartwell should be released and not
eaten. This advisory remains in effect and
warning signs against eating fish have been
posted at the majority of the public boat
launch and recreation areas since 1987.
In 1987, based upon the EPA Hazard
Ranking System, SCDHEC monitoring
programs and accompanying concerns of
citizens in the area, the Sangamo Site was
proposed for inclusion on the National
Priorities List (NPL). The Sangamo Site
was finalized on the NPL in February 1990.
A Remedial Investigation, consisting of
separate sediment and biological
components, was then initiated to define the
nature and extent of PCB contamination in
the Sangamo OU2 study area. The OU2
study area extends from the Sangamo
Weston Plant's discharge point on Town
Creek, through the Twelvemile Creek
watershed, and finally into the upper portion
of Lake HartwelL Figure 1 at the end of
this publication provides a general
illustration of Lake Hartwell and the
surrounding area. Figure 2 provides a more
focused description of the OU2 study area..
-------
Field sampling and analysis for the sediment
component of the OU2 RI was performed in
July and August 1991 (Phase I) and April
and May 1992 (Phase It). The sediment
investigation included the collection and
analysis of over 1,100 sediment cores and
grab samples to adequately characterize the
horizontal and vertical distribution of PCB
contamination in the sediments of the study
area. The sediment investigation determined
that most of the PCB-contaminated sediment
lies within the upper portion of Lake
Hartwell, specifically the Twelvemile Creek
Arm. PCB concentrations of surficial
sediments (Lc. grab samples) were typically
less than 5 parts per million (ppm), but
higher concentrations up to 61 ppm were
detected in the deeper sediments (Le. core
samples). Figure 3 illustrates average
sediment PCB concentrations for an
approximate 10 mile reach extending from
Maw Bridge (Rt 337)to the Rt 37 Bridge at
the southernmost boundary of the focused
OU2 study area.
Field sampling and analysis for the
biological component of the OU2 RI was
conducted in the Spring of 1991 and 1992.
Largemouth bass, catfish, and hybrid bass
were collected and analyzed from 6 stations
in Lake HartwelL These stations are
illustrated on Figure 4. The biological
investigation determined that fish in the
Twelvemile Creek drainage and Lake
Hartwell are contaminated with PCBs, often
at levels that exceed FDA's safe tolerance
level of 2 ppm. Furthermore, caged fresh
water clams and forage fish (small, young
fish) accumulated PCBs over a relatively
short period of time which confirmed the
presence of food chain and respiratory
transport mechanisms. PCB concentrations
in sedentary fish (Le. largemouth bass,
catfish) are highest in the Twelvemile Creek
Arm and decrease with increasing distance
downstream, similar to the gradient of
sediment contamination. Migratory fish
species (Le. hybrid bass) have PCB
concentrations that are similar throughout the
entire reservoir with concentrations
frequently exceeding FDA's safe tolerance
level PCB concentrations for largemouth
bass and hybrid bass collected from the six
stations, years 1990-1993, are shown on
Figures 5 and 6, respectively.
SCOPE OF THE PROPOSED ACTION
The Sangamo Site consists of two Operable
Units. Operable Unit One (OU1) consists of
the land-based source areas which include
the Sangamo Weston Plant Site and the six
satellite disposal areas located within a 3-
mile radius of Pickens. Schlumberger
Industries, under the oversight of EPA,
initiated cleanup activities for OU1 in
November 1993. On-going remedial actions
consist of excavation of waste and
contaminated soils from the six satellite
disposal areas and transportation of the
excavated materials to temporary storage
facilities at the former Sangamo Plant Site.
The excavated material will remain in these
storage facilities until such time as a
treatment system can be built at die plant
site, at which time the contaminated
materials will be treated by means of low
temperature thermal desorpu'on.
This publication was specifically developed
to present the preferred remedial alternative
for Operable Unit Two (OU2). OU2
addresses the sediment and biological
migration routes downstream from the OU1
land-based source areas. OU2 represents the
final response action EPA expects to
implement as pan of its overall strategy in
remediating the Sangamo Site.
-------
SUMMARY OF SITE RISKS
Human Risk
CERCLA directs EPA to protect human
health and the environment from current and
potential exposure to hazardous substances at
the Site. A Baseline Risk Assessment was
conducted to evaluate the potential current
and future human health and ecological
impacts associated with exposure to PCBs in
sediment and fish. The human exposure
pathways which were quantitatively
evaluated under current and future land-use
conditions included: 1) Ingestion and dermal
absorption of PCBs in the shallow sediment
.by a child and an adult; and 2) Ingestion of
PCB-contaminated fish by a recreational
fisherman. PCBs were not detected in the
surface water during die RI and therefore
this medium was not evaluated in the risk
assessment
EPA evaluated PCBs according to their
potential to produce either cancer and/or
noncancer health effects. The carcinogenic
risk range EPA has set for Superfund
cleanups to be protective of human health is
1 X HT* to 1 X 10"6. For example, a cancer
risk of 1 X 10* indicates that an individual
his a 1 in 1,000,000 ( or 1 in 10,000 for 1X
10"*) incremental chance of developing
cancer as a result of site-related exposure to
a carcinogen over a 70 year lifetime under
the specific exposure conditions at the Site.
EPA generally uses the cumulative
benchmark risk level of 1 X 10"4 for all
exposures relating to a particular medium to
trigger action for that medium. Noncancer
exposure estimates were developed using
EPA reference doses to calculate a Hazard
Quotient (HQ). A HQ greater than 1
indicates that PCBs are present at
concentrations that could produce harmful
effects.
The risks associated with the ingestion of
fish were calculated for all of the sampling
stations in Twelvemile Creek and i*ir*
Hartwell (Figure 4). Risks were evaluated
separately by fish species and also for all
species combined to account for both
individuals who may eat one specie
exclusively and individuals who consume all
types of fish. Exposures associated with
ingestion of fish caught from all sampling
stations resulted in unacceptable risks
ranging from 10"2 to 10"* and HQs greater
than 1. The highest cancer risk of 4 X 10"2
was calculated for anglers exclusively
consuming largemouth bass caught in die
Twelvemile Creek watershed. The highest
cancer risk for ingestion of all species
combined, 1 X 10~2, was calculated for the
Twelvemile Creek Arm. The lake-wide risk
associated with ingestion of all species
combined was 5 X 10"3. Adverse human
health risks resulting from direct contact or
incidental ingestion with the sediment are
unlikely to occur, however, sediments are a
continuing source of contamination in the
aquatic biota of the study area.
Ecological Risk
An ecological risk assessment was conducted
to evaluate the impact PCBs may be having
on the aquatic receptors of the study area.
The biological investigations clearly
document PCB contamination in all levels of
the aquatic food web. However, although
PCBs appear to be impacting die fish and
macroinvertebrate communities in
Twelvemile Creek, habitat degradation from
man's influence is likely causing greater
adverse impacts at the population and
community levels. The health of fish in
Lake Hartwell does not appear to be effected
-------
at the population level for fish that have
PCB concentrations around 5 ppm.
However, there is evidence that as
concentrations increase to greater than 20
ppm, fish health can be effected.
Based upon the results of the Baseline Risk
Assessment, it was concluded that actual or
threatened releases of hazardous substances
from this Site, if not addressed by the
preferred alternative or one of the other
active measures considered, may present a
current or potential imminent and/or
substantial endangerment to public health,
welfare, or the environment
SUMMARY OF ALTERNATIVES
Thorough development and evaluation of
feasible remedial alternatives for the
Sangamo OU2 Site required the derivation of
cleanup goals for the media of concern:
sediment and fish. Cleanup goals were not
developed for surface water since PCBs were
not detected in any sample collected during
the RI above the detection limits of 1.2 to
1.3 ppb. Cleanup goals were also necessary
to identify areas to be addressed by EPA's
preferred alternative. The final cleanup
goals are discussed below.
Final Sediment Cleanup Goal -
Concentration of 1 ppm was selected based
upon technical feasibility. Remediation goal
of 1 ppm identified the entire Twelvemile
Creek Arm, extending from the headwaters
of the lake downstream to the confluence
with the Keowee Arm, as an area to be
remediated. This area covers approximately
730 acres with a total estimated volume of
4,722,000 cubic yards of contaminated
sediment
Final Fish Cleanup Goal - The U.S. FDA
safe tolerance level of 2 ppm was selected
based upon technical feasibility. A
carcinogenic risk-based approach was
evaluated by determining the concentration
levels in largemouth bass that would result
in acceptable risk to anglers through
ingestion of fish. Utilizing EPA risk
assessment methods, a fish tissue
concentration of 0.036 ppm is associated
with a 10~* risk. The risk-based fish cleanup
goal of 0.036 ppm was determined to be
technically impracticable.
In a similar fact sheet dated February 1994
and in the Final FS Report, EPA identified
the range of remedial alternatives that were
under consideration for the Site. These
alternatives are briefly summarized below.
All alternative cost estimates are expressed
in 1993 dollars and are based upon
conceptual engineering and design. Capital
cost' consists of direct (construction) and
indirect (non-construction) costs incurred in
the first year of operation. Operation and
maintenance (O&M) cost refers to long-term,
postconstruction items necessary to ensure
continued effectiveness of a remedial action.
Total present worth cost represents that sum
of money, if invested in the base year and
disbursed as needed, would be sufficient to
cover all costs of a remedy over its planned
life.
Alternative 1: No Action
As required by CERCLA, a no further action
alternative was evaluated to serve as the
basis for comparison with other active
cleanup alternatives. Under this alternative,
no further remedial actions for the
contaminated sediment or fish at the Site
-------
would be conducted. The existing fishing
advisory would continue and 5-year reviews
would be conducted to assess sediment/fish
PCB concentrations and associated risks to
human health and the environment
Due to the complex, dynamic nature of the
Lake Hartwell system, significant changes in
the nature and extent of contamination are
expected over tune even under the no-action
scenario. A comprehensive modeling effort
consisting of sediment transport, water
quality and contaminant transport, and food-
chain bioaccumulation was conducted during
the RI to predict the future fate and transport
of PCBs in sediments and aquatic biota of
the study area. The predicted results of this
modeling exercise are important
considerations when evaluating the no-action
alternative. For this reason, the predicted
results of the baseline modeling activities are
summarized below.
Future Sedimentation
The objective of sediment modeling was to
determine the rates at which presently
contaminated sediments would be buried by
relatively clean sediments from tributary
drainages over die next 30 years. Average
sediment burial rates for the 30 year
simulation are presented in Figure 7.
Substantial rates of sediment deposition are
predicted from Maw Bridge (Rt 337) to just
above the Hwy 133 Bridge. All of the sand
and most of the silt coming into the
reservoir from the Twelvemile Creek
drainage are deposited in this reach with a
predicted 30-year accumulation of nearly 10
feet of sediment in some areas. This
deposition pattern is favorable since PCB
concentrations have historically been higher
in sediments from this portion of the study
area.
Sediment Quality Trends
The objective of sediment quality modeling
was to predict the fate and transport of PCBs
in the surface and bottom sediments in the
Twelvemile Creek/Lake Hartwell system
over the next 30 years. Surface and bottom
sediments upstream of the Hwy 133 Bridge
are predicted to decrease to near zero levels
within 5 years. Sediments between the Hwy
133 Bridge and the Hwy 93 Bridge are
predicted to fall below the 1 ppm cleanup
goal in approximately 10 years. Surface and
bottom sediments lying between the Hwy 93
Bridge and the Hwy 37 Bridge are predicted
to increase and remain over the 1 ppm
cleanup goal after 30 years. This predicted
increase is attributed to the settling of the
finer, clay sized particles caused by the
constriction of the channel at the Hwy 37
embankment
Bioaccumulation and Future
Fish Concentrations
Aquatic bioaccumulation modeling was
conducted to assess: 1) How PCB levels in
fish would change over the next 30 years if
the contaminated sediments were left
unmanaged, and 2) How long it would take
for levels to drop below the 2 ppm cleanup
goal A representative food web was
developed to evaluate the uptake of PCBs
into fish via respiration and consumption of
contaminated food. Largemouth bass was
selected as the top predator, gizzard
shad/bluegill sunfish as the forage fish, while
plankton and benthos composed the lower
levels of the food web. Initial and predicted
PCB concentrations for the selected food
web are presented in Figure 8. The
bioaccumulation model predicts PCB
concentrations to decrease in all levels of the
food web during the 30-year simulation.
-------
PCS concentrations in older largemouth bass
are predicted to fall below the 2 ppm
cleanup goal by 2005.
Capital Cost = $7,000
Total O&M Cost = $123,000
Total Present Worth Cost = $130,000
Alternative 2A: Institutional Controls
This alternative is a limited action alternative
that relies on a series of institutional controls
to prevent or minimize ingestion of
contaminated fish tissue, which was
identified as die primary exposure pathway
of concern for the Site. Generally, these
institutional controls consist of the
following:
Public Education Program - A program
would be initiated to inform the public on
available methods for reducing the intake
of PCBs through fish consumption.
Specific preparation, handling, and
cooking techniques can reduce the
quantity of contamination consumed.
Information of this type would be
disseminated to the public through a series
of local public meetings and distribution
of an informational pamphlet
Fish and Sediment Monitoring - In
addition to maintaining the current fish
advisory, annual monitoring of PCS levels
in fish and sediment of Twelvemile Creek
and Lake Hartwell would be conducted.
Results from this monitoring program
would be utilized to support modifications
to the fish advisory and to monitor
concentrations of PCBs in sediment and
fish over time. Results of this program
would be made available to the public.
Regulation of Twelvemile Creek
Impoundments - A routine schedule would
be developed for flushing of sediment
accumulated behind the 3 small
impoundments located on Twelvemile
Creek. Periodic flushing (most recently in
September 1993) of a large load of
sediment over a short period of time has
been documented to have an adverse
impact on the water quality and aquatic
biota of the upper portion of the
Twelvemile Creek Arm. These adverse
impacts are attributed to elevated levels of
suspended sediment and not PCBs. A
routine flushing schedule would minimize
impacts to the ecosystem while enhancing
burial of more contaminated sediments
with cleaner sediments from the
Twelvemile Creek drainage.
Capital Cost = $366,000
Total O&M Cost = $2,842,000
Total Present Worth Cost = $3,208,000
Alternative 2B: Fisheries Isolation
Alternative 2B is a more aggressive
approach consisting of management of the
fisheries resource within the lake to
minimize ingestion of PCB-contaminated
fish. The primary control measure includes
construction of a, barrier, or fish fence, to
prevent the movement of migratory fish (Le.
striped bass, hybrid bass, and walleye) into
or out of die most contaminated portions of
the reservoir. These migratory species
represent approximately 50% of die fish
harvested by weight from Lake HartwelL
The fish isolation barrier would be located in
the vicinity of the Hwy 37 Bridge, and
designed/constructed to meet die following
performance standards:
-------
Minimize safety hazards for boaters and
other users of the lake (Le. waterskiers);
Maximize effectiveness to prevent passage
of migratory game fish through the
barrier;
Allow boats to pass unimpeded to
minimize disruptions of boating traffic on
the lake (Le. no gate);
Construct a semi-permanent structure that
could be removed once PCS levels
decline to health-based levels;
Incorporate value engineering principles to
minimize cost for design and construction;
and
Minimize operations and maintenance
requirements and costs.
Lake use restrictions, such as no-wake zones
and warning signs would be implemented to
prevent damage to the fence and for boating
safety. Alternative 2B also includes the
institutional controls described for
Alternative 2A.
Capital Cost = $1,232,000
Total O&M Cost = $3,012,000
Total Present Worth Cost = $4,244,000
Alternative 3A: Capping
Under this alternative, an 18 inch clean
sediment cap would be placed over the
entire Twelveirrile Creek Arm to eliminate
the further downstream migration of
contaminated sediment and the transfer of
PCBs from sediment to aquatic biota. The
cap would extend from just upstream of the
Maw Bridge (Rt 337) to the confluence with
the Keowee River, just upstream of Hwy
123. The cap would cover an area of 730
acres and would require approximately
1,800,000 yd3 of sand to construct the cap.
Alternative 3A also consists of lafr* usage
restrictions to minimize erosion of the cap
once it is placed and other institutional
controls described under Alternative 2A.
Capital Cost = $48,296,000
Total O&M Cost = $2,843,000
Total Present Worth Cost = $51,139,000
Alternative 3B: Sediment Control
Structure
Alternative 3B would consist of constructing
a 1,600-ft long fixed-crest weir at the mouth
of the Twelvemile Creek Arm, just above
the confluence with the Keowee River.
Construction of the sediment control
structure would physically isolate the
Twelvemile Creek Arm from the remainder
of Lake Hartwell resulting in reductions in
downstream migration of contaminated
sediment and movement of fish to/from the
areas with highest sediment PCS levels.
Fishing restrictions in the isolated area, as
wen as institutional controls under
Alternative 2A, would be implemented to
limit consumption of contaminated fish until
protective levels were achieved.
Capital Cost = $51,226,000
Total O&M Cost = $2,365,000
Total Present Worth Cost = $53,591,000
Alternative 3C: Optimal Capping/
Sediment Control Structure
Alternative 3C utilized value engineering
techniques to optimize both the area to be
capped and location of the fixed crest weir.
The weir would be constructed at an
upstream location near the Hwy 133 Bridge,
-------
resulting in the physical isolation of a much
smaller area than Alternative 3B. A similar
clean sediment cap as described under
Alternative 3A would be placed below the
weir and extend downstream to the mouth of
the Twelvemile Creek Arm. The area to be
capped under this alternative is
approximately 285 acres. Institutional
controls under Alternative 2A and fishing
restrictions for the isolated area would be
implemented until protective levels in the
fish were achieved.
Capital Cost = $31,684,000
Total O&M Cost = $2,365,000
Total Present Worth Cost = $34,049,000
Alternative 4: Confined Disposal Facility
Alternative 4 involves a much more rigorous
remediation approach than any of the
alternatives discussed above. Under this
alternative, contaminated sediments from the
Twelvemile Creek Arm with PCB
concentrations greater than 1 ppm would be
removed by hydraulic dredging. All dredged
sediments would be placed in a near-shore
confined disposal facility (CDF) located in
the "goose neck" bend area between the
Maw (RL 337) and Madden "(Rt 15)
Bridges. Twelvemile Creek would then be
diverted around the CDF via a physically
rerouted channel Institutional controls
would be used to limit consumption of
contaminated fish until protective levels
were achieved.
Capital Cost = $43,422,000
Total O&M Cost = $3,487,000
Total Present Worth Cost = $46,909,000
Alternative 5: Stabilization
Alternative 5 is a very complex and costly
alternative that involves removal of
contaminated sediments with PCB
concentrations greater than 1 ppm via a
combination of shallow water excavation and
hydraulic dredging. Contaminated sediments
in the middle and lower portions of
Twelvemile Creek Arm would be dredged
while the upper portion is dewatered and
excavated. The dredged/excavated sediments
would be treated by stabilization with
cement and placed in a CDF as described
under Alternative 4. Fish and other biota in
the upstream segment would be collected
and destroyed as part of the dewatering
operations. Institutional controls would be
used to limit consumption of contaminated
fish until protective levels are achieved.
Capital Cost = $581,957,000
Total O&M Cost = $3,486,000
Total Present Worth Cost = $585,443,000
EVALUATION OF ALTERNATIVES
EPA.has established criteria for use in
comparing the advantages/disadvantages of
each alternative under consideration. The
alternatives discussed above were evaluated
against one another by using the nine criteria
specified in the National Contingency Plan
40 CFR 300.430(e)(9). These nine
evaluation criteria are:
1) Overall Protection of Human Health
and the Environment
2) Compliance with Applicable or
Relevant and Appropriate Requirements
(ARARs)
3) Long-term Effectiveness and
Permanence
4) Reduction of Toxicity, Mobility, or
Volume
5) Short-term Effectiveness
6) Irnplementability
7) Cost
8) State Acceptance
9) Community Acceptance
-------
The first two criteria are referred to as
threshold criteria and must be met for an
alternative to be.eligible for selection. The
next five criteria are referred to as balancing
criteria and are used to weigh major trade-
offs among alternatives. The final two
criteria, state and community acceptance, are
referred to as modifying criteria and are
addressed during the Proposed Plan comment
period. The following discussion briefly
evaluates the relative performance of each
alternative to the evaluation criteria.
Overall Protection of Human Health and the
Environment It is important to note that
complexities related to cycling of PCBs in
the biota of Lake Hartwell cause the residual
risk levels to remain above EPA's threshold
of 1 X 1CT* regardless of which remedial
alternative is implemented. A reduction in
sediment PCB concentrations does not result
in an immediate reduction in PCB
concentrations in fish. High PCB
concentrations in fish during the initial years
of the 30-year exposure duration result in
average risk estimates exceeding acceptable
levels. Therefore, the most effective
measures to minimize exposures to PCB-
contaminated fish tissue are institutional
controls (Lc. public education, continuation
of fish advisory).
However, the most protective alternatives
rely principally on engineering rather than
institutional controls to reduce or eliminate
exposures, given the uncertainty regarding
the performance/reliability of institutional
controls. Alternatives 3A and 3C involve
capping actions that would permanently
isolate the contaminated sediment in all or a
major portion of the Twelvemile Creek Arm.
The more aggressive alternatives 4 and 5
provide an even higher level of
protectiveness, as the contaminated material
is removed from the lake and isolated in an
engineered disposal facility.
Compliance with ARARs - This criterion
assesses the overall compliance with
Federal/State requirements. All the
alternatives rely on natural fate and transport
processes to reduce PCB concentrations in
Lake Hartwell fish to the FDA safe tolerance
level of 2 ppm. Alternatives that involve
capping and isolation (Alternatives
2B/3A/3B/3Q achieve FDA levels in fish
quicker in the main body of the lake.
Alternatives 3A, 4 and 5 would best achieve
the cleanup goal of 1 ppm in sediment
However, Alternatives 4 and 5 would likely
trigger a large number of Federal/State
requirements associated with water quality,
dredging, and disposal
Longrterm Effectiveness and Permanence
With the \ exception of the No-Action
Alternative, all alternatives rely on
institutional controls to prevent exposure to
contaminated fish. The performance and
reliability of these controls is dependent on
site-specific factors and therefore difficult to
predict The fish fence (Alternative 2B)
would require pre-design treatability testing
to demonstrate effectiveness for the scale of
application at Sangamo OU2. Capping
(Alternatives 3 A/3Q is a proven technology
but is dependent on an extensive program of
monitoring and maintenance to ensure cap
integrity and long-term effectiveness.
Sediment control structures (weirs) are also
widely used and are less dependent on
maintenance. Further technology evaluation
and treatability studies would be needed
prior to implementation of Alternatives 4
and S to obtain design and operating data for
the dredging, CDF, and treatment activities.
10
-------
Reduction of Toxicitv, Mobility, or Volume
- Alternative 5 would accomplish the most
significant reduction in the mobility of PCBs
through a stabilization treatment process
followed by disposal in a CDF. The capping
alternatives (Alternatives 3A/3Q also
accomplish a reduction in contaminant
mobility, but the effect is contingent upon
proper maintenance of the cap. The
sediment control structures (Alternatives
3B/3C) and fish fence (Alternative 2B)
would be the most effective in limiting
contaminant mobility related to the migration
of fish into and out of the areas with the
highest sediment contamination.
Short-term Effectiveness - Implementation of
Alternatives 1 (No-Action) and 2A
(Institutional Controls) would have the least
impact on the community, workers, and the
environment; Alternative 2B has only a
slightly greater impact Alternatives
3A/3B/3C have greater impacts on the local
community due to construction of sediment
control structures and excessive truck traffic
associated with transportation of capping
material. The greatest impacts (and least
short-term effectiveness) are the more
aggressive alternatives that entail dredging,
re-channeling and construction of the CDF
(Alternatives 4/5).
Table 1 at the end of this publication
summarizes the estimated schedule for each
alternative, presenting the time required to
implement, time required for fish PCB
concentrations to decline to the FDA safe
tolerance level, and total time to achieve
protectiveness. This information is presented
for the focused Sangamo OU2 study area
(Twelvemile Creek/Seneca River Arm) as
well as the main portion of Lake HartwelL
The table shows that most of the alternatives
rely on natural fate and transport processes
within the reservoir to return conditions to
protective levels (12 years). In the
Twelvemile Creek Arm, the more aggressive
alternatives involving capping, sediment
control structures, or dredging, achieve
protectiveness (in most cases) in 5 to 10
years. Alternative 2B - Fisheries Isolation
achieves protectiveness in the main body of
Lake Hartwell (representing 90% of die total
lake area) in a shorter amount of time (4 to
5 years) than the other more aggressive and
costly alternatives.
Implementability - Alternatives 1 and 2A
would be the most readily implementable.
Alternatives 2B/3A/3B/3C would require
close coordination with the COE, SCDHEC,
and other agencies; however, die
administrative and technical issues associated
with these alternatives are considered
manageable. The most difficult alternatives
to implement, in terms of both administrative
and technical feasibility, would be
Alternatives 4/5.
Cost - Costs for each alternative were
provided under the Summary of Alternatives
section of this publication.
State Acceptance - The State of South
Carolina is currently reviewing this Proposed
Plan. The State feels that at a minimum,
Alternative 2A - Institutional Controls
should be selected The State will make a
final selection of a preferred alternative after
the public comment period is over. The
State will rely on public comment for die
selection of the preferred alternative.
Community Acceptance - The purpose of
this Proposed Plan and the comment period
is to encourage input from the public during
the remedy selection process. Community
acceptance of the preferred alternative will
11
-------
be evaluated after the public comment period
and will be described in the Record of
Decision for the Site.
EPA'S PREFERRED ALTERNATIVE
In summary, based upon the information
available at this time, EPA is proposing
Alternative 2B - Fisheries Isolation, which
also includes those activities described under
Alternative 2A. Placement of the fish fence
in the vicinity of the Hwy 37 Bridge would
result in isolation of the Twelvemile Creek,
Keowee River and upper Seneca River
Arms, which represent less than 10 percent
of the total area of Lake HartwelL Isolation
of these upstream areas is expected to result
in an accelerated decline in migratory fish
PCB concentrations in the downstream
portions of the reservoir (Le. the remaining
90+ percent of the lake). Migratory species
represent approximately 50% of the fish
harvested by weight from Lake HartwelL
Reduction of fish PCB levels would allow
for rescinding of the existing fish advisories
in these areas, returning the majority of lake
areas to normal use. This approach is
consistent in the COE's Best Management
Practices for Lake Hartwell, designed to
achieve the maximum beneficial uses for the
reservoir, which include fish and wildlife
management and recreational use.
Alternative 2B - Fisheries Isolation
represents the best balance among the
criteria used to evaluate die other
alternatives. Alternative 2B is believed to be
protective of human health and the
environment, would attain ARARs, would be
cost effective and would utilize permanent
solutions and resource technologies to the
maximum extent practicable. The selected
remedy does not satisfy the preference for
treatment as a principal element due to the
excessive costs. Based upon comments
received from the public during the comment
period, EPA, in close consultation with the
COE and SCDHEC, may later modify the
preferred alternative or select another
remedial alternative presented in this
Proposed Plan.
COMMUNITY PARTICIPATION
Concurrent with the release of this Proposed
Plan, EPA has initiated a 30-day public
comment period from April 11, 1994
through May 11, 1994 for submission of
written and oral comments on the Proposed
Plan and all supporting documentation
located in the information repositories listed
below. All comments, written or oral,
should be directed to Craig Zeller, EPA
Remedial Project Manager for the Sangamo
OU2 Site, at the address and telephone
number listed below. Upon timely request,
EPA will extend the public comment period
by 30 additional days. EPA will hold a
public meeting at 7:00 PM on April 19,
1994 to present the information contained in
the RI/FS Reports and to explain the
rationale behind the preference for
Alternative 2B - Fisheries Isolation. The
meeting will be held at the Ramada Ian.
located at the intersection of U.S. 76 and
Hwy 123 in Qemson, South Carolina.
Representatives from EPA, the U.S. Army
Corps of Engineers, and the South Carolina
Department of Health and Environmental
Control win be available to answer any
questions the public may have regarding
future activities at the Sangamo OU2 Site.
12
-------
GLOSSARY
Administrative Record - A file which contains all
information used by the EPA to make its decision on
the selection of a response action under CERCLA.
This file is required to be available for public review
and a copy is to be established at or near the site,
usually at the information repository. A duplicate file
is maintained in a central location such as a regional
EPA and/or state office.
Baseline Risk Assessment - An assessment which
provides an evaluation of the potential risk to human
health and the environment in the absence of remedial
action.
Comprehensive Environmental Response,
Compensation and Liability Act (CERCLA)
A federal law passed in 1980 and modified in 1986
by the Superfund Amendments and Reauthorization
Act (SARA). The Act created a trust fund, known as
Superfund to investigate and clean up abandoned or
uncontrolled hazardous waste sites.
Hazard Ranking System A scoring system used by
EPA to evaluate relative risks to public health and the
environment A score is calculated based on actual or
potential release of hazardous substances through all
media present (Le. the air. soils, surface water,
sediments, groundwater). If a site scores above 28.5,
the site is proposed for inclusion of the National
Priorities List
Information Repository - Materials on Superfund
and a specific site located conveniently for local
residents.
Macroinvertebrates - Small animals lacking
backbones found in sediments of water bodies.
National Contingency Plan* The Federal regulation
that guides the Superfund program.
National Priorities List (NPL) - EPA's list of
uncontrolled or abandoned hazardous wastes sites
eligible for long-term cleanup under the Superfund
Remedial Program.
Parts Per Billion (ppbVParts Per Minion (ppm) -
Units commonly used to express low concentrations
of contaminants. For example, 1 ounce of PCBs in 1
million ounces of water is 1 ppm. If one drop of
PCB's are mixed in a competition sized swimming
pool, the water will contain about 1 ppm PCB.
Polychtorinated Biphenyb (PCBs) - A family of
organic compounds used since 1926 in electric
transformers as insulators and coolants, in lubricants,
carbonless copy paper, adhesives and caulking
compounds. Also produced in certain combustion
processes. PCBs are extremely persistent' in the
environment because they do not easily break down.
PCBs are stored in the fatty tissues of hunws and
animals through the ^accumulation process. EPA
banned the use of PCBs in 1978. In general, PCBs
are not as toxic as some other chemicals, githmigh
acute and chronic exposure can cause liver damage.
PCBs «»isft have caused cancer in laboratory nrtirnalt
Remedial Investigation/Feasibility Study (RI/FS)-
Two distinct but related studies, normally remAiir*^
together, intended to define the nature and *****$ of
contamination at a site and to evaluate appropriate.
site-specific remedies.
Record of Decision - A public document that
explains which cleanup alternative wffl be used at a
National Priorities List site and the reasons fat
choosing the cleanup alternative over other
possibilities.
Superfund Amendments and Reauthorfeatioa Act
(SARA) Modifications to CERCLA enacted on
October 17,1986.
13
-------
FOR MORE INFORMATION CONTACT:
Remedial Project Managers
Openble Unit Two Craig Zeller
Operable Unit One - Bemie Hayes
Community Relations Coordinator
Cynthia Peurifoy
AT
U.S. Environmental Protection. Agency Region IV
North Superfund Remedial Branch
345 Courtland Street, HE., Atlanta, GA 30365
H800) 435-9233, or (404) 347-7791
**********************
C. Michael Alexander
Savannah District Fisheries Biologist
U.S. Army Corps of Engineers
Richard B. Russell Lake
4144 Russell Dam Drive
EIberton,GA 30635
(706) 283-8731
**********************
Richard Haynes, District Engineer
South Carolina Department of Health & Environmental Control
2600 Bull Street, Columbia, South Carolina 29201
(803) 734-5487
INFORMATION REPOSITORIES
The Administrative Record for die Sangamo Site is available for review at the following information repositories:
The Village Library Pickens County Public Library
Main Street Easley Branch
Pickens, SC 29671 110 West First Avenue
(803) 859-9679 Easley. SC 29640
(803) 859-9679
R. M, Cooper Library Hart County Library
Clemson University Benson Street
Ckmson, SC 29634-3001 HartweD. GA 30643
(803) 656-5174 (706) 376-4655
U. S. Army Corps of Engineers
Lake Harwell Natural Resources Management Center
Hartwell, Georgia
(706)376-4788
14
-------
SCALE
MILES
Figure 1
Lake Hartwell and Watershed
Sangamo OU2
15
-------
ftl SanginoWextoD Plant
Oftslte Disposal «rn
A Undflll
Dan
SCALE
0 4
| ' ' ' |
MILES
Figure 2
Sangamo OU2 Study Area
16
-------
Mnn PCI Ctncmlntio* pif Tranwd Intern)
1MM
Mmri to a rartv «Wom tk>« I gkwi Innwd.
| I
ft 1C
Omdrum OUbMt *!! Tlulwii htm Nnlwiton if Twiteratlto Gnik Aim ! HirtwtU Ukt (ft)
PCI
Figures
Mean Vertical Dlslribulion of PCBs in Sediments
Twelvemile Creek Arm of Hartwell Lake
Sangamo OU-2 FS
-------
1"
Rsb Sampling Location
Trftmtinr
SCALE
0 4 «
I i i i I I . I
MILES '
Rgure 4
Fish Tissue Sample Stations
Lake Hartwell
18
-------
SV-107 SV-106 SV-S32 SV-535
STATION
SV-642
SV-641
Figure 5: PCB concentrations in largemouth bass fillets collected
from six stations In Lake Hartwell, 1990-1993.
D1990 B1991 B1992 OB 1993
SV-107 SV-106 SV-532 SV-535
STATION
SV-642
SV-«41
Figure 6: PCB concentrations in hybrid bass fillets collected
from six stations in Lake Hartwell, 1990-1993.
19
-------
I I <0cmfyr
|>>>;-] 0-1 cm/yr
1-5cm/yr
IB 10-15 cm/yr
0 2500 3000
i. ..... _i_-..- i - f J
SCALE IN FEET
SOUTHERN RAILROAD
Figure 7
Predicted Average Sediment Burial Rates
Sangamo OU2
20
-------
0 YEARS
5 YEARS
10 YEARS
20 YEARS
30 YEARS
Gizzard Shad ^HC- Blueglll Suntish
PCBCONCENTRATIOK
^J
0.001-0.1 0.1-1
NOTE;
Predicted PCS concentrations are based on outputs of the FQETS model and
repcesent averaged whol« botfy tlsh re»idues p«c Bsh Ketlme. PtanWon and
Benthos concentrations represent tqutbrium with predicted surface water and
sediment concentrations, respectively.
Figure 8
Initial and Predicted PCB -ncentrations for Food Web
Lake Hartweli Janaamn OU2
-------
Table 1
Estimated Time Required to Achieve Protediveness
Sangamo OU-2 FS
ALT.
No.
1
2A
2B
3A
3B
3C
4
5
DESCRIPTION
No action
Institutional Controls
Fisheries Isolation
Capping
Sediment Control Structure
Optimal Capping/Sediment
Control Structure
Confined Disposal Facility
Stabilization
TIME To
IMPLEMENT0'
(yrs)
0
0.5
1
2
2
2-3
3-4
4-5
TIME FOR FISH To ATTAIN FDA LEVEL
(yrs)
TWELVEMILE
CREEK/SENECA
RIVER* ARM
12
12
12
3-4
12
3-4 (domino)
12 (*««»)
5-6
5-6
MAIN BODY OF
HARTWELL LAKE"*
12
12
3-4
3-4
3-4
3-4
5-6
5-6
TIME TO ACHIEVE PROTECTIVENESS"
(yrs)
TWELVEMILE
CREEK/SENECA
RIVER ARM
12
12
12
5-4$
12
5-7**
12 down)
8-10
9-11
MAIN BODY OF
HARTWELL LAKE
12
12
4-5
5-6
5-6
5-7
8-10
9-11
NOTE: (I) Include! lime needed for pre-deiigo studies, design, procurement, and construction, but does not include lime between Proposed rlan/Record of Decision and initiation of Remedial .
Detign. .. '
(2) Twelvemile Creek Ann only; POETS modeling result* indicate that flib will decline to FDA tolerance level (2 ing/kg) in approximately 12 yean under batcllne conditions; none
of the alternatives will extend the period required for this to occur (i.e., none of the alternatives can have a period to achieve protectlveoeaiof freater than 12 yn).
(3) Downgradient of the Twelvemile Creek Arm (i.e., main body of Hartwell Lake), PCB concentration! will decline more rapidly for actions involving isolation, capping or removal of
contaminated sediment (AltemaiivM 2B, 3A-C, 4 and 5). II WM usumed that Ihii rat* of decline would be comparable to (hat observed in the Twelvemile Creek Arm for Alternatives
involving capping (3-4 yn) or removal of (he sediment (5-6 yn); further modeling it needed to verify thii estimate.
(4) Duration only; bated on estimated time to implement remedy and duration thereafter needed for fish PCB levels to decline to FDA level, based on the FOBTS modeling result*
presented in Section 3.3 and assumptions regarding migratory fish PCB concentrations trends.
-------
USE THIS SPACE TO WRITE YOUR COMMENTS
Your input on the Proposed Plan for the Sangamo Superfund Site - OU2 is important In
helping EPA select a final remedy for the site. You may use the space below to write
your comments, then fold and matt, A response to your comment will be included in the
Responsiveness Summary.
REQUEST TO BE PLACED ON TEE SANGAMO WESTON, BVC./TWELVE MILE CREEK/LAKE
HARTWELL SITE MAILING LIST
If you would Wee to be placed on the mailing list for the Sangamo Site, please complete
this/arm and return to: CtynthiaPeurtfoy, Community Relations Coordinator. EPARegion
IF, North Superfund Remedial Branch, 340 Courtland Street, Atlanta, Georgia 30360, or
call 2-800-430-9233.
XELEEHQEEL.
23
------- |