[Federal Register Volume 63, Number 18 (Wednesday, January 28, 1998)]
[Notices]
[Pages 4236-4243]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-1947]


-----------------------------------------------------------------------

DEPARTMENT OF ENERGY


Record of Decision, Shutdown of the River Water System at the 
Savannah River Site, Savannah River Operations Office, Aiken, South 
Carolina

AGENCY: U.S. Department of Energy.

ACTION: Record of Decision.

-----------------------------------------------------------------------

SUMMARY: The U.S. DOE has decided to implement the No Action 
alternative identified in the Final Environmental Impact Statement for 
the Shutdown of the River Water System (RWEIS) at the Savannah River 
Site (SRS). Under this alternative, DOE will continue to operate and 
maintain the system and maintain the water level of L-Lake.
    DOE will assess the need for future environmental remediation 
alternatives for L-Lake under existing Comprehensive Environmental 
Response, Compensation, and Liability Act (CERCLA) commitments. 
Characterization activities associated with CERCLA closure are expected 
to begin in the year 2000 and be completed in several years. This 
characterization will inform any required remedial action. Pending 
these activities, DOE will continue to operate the RWS. If during 
continued operation of the RWS a system component fails, DOE will take 
appropriate emergency actions. DOE will then determine if the system is 
too costly to repair (by comparing this cost to estimated shutdown 
costs and future possible remediation costs under the CERCLA). If DOE 
determines that the RWS is too costly to repair, it will reevaluate all 
relevant commitments and the information in the RWEIS, to determine 
necessary actions to shut down the RWS. However, the RWS is in good 
condition and not expected to fail over this period of time.
    This RWEIS evaluates three alternatives for the disposition of the 
RWS at the SRS. The RWS is a 50-mile underground concrete piping 
structure and pumping system that was built in the early 1950s to 
provide cooling water for the SRS' five nuclear production reactors. 
The RWEIS alternatives cover the spectrum of reasonable options as 
follows:
    (1) Continue operation of the RWS (No Action Alternative);
    (2) Shut down and maintain the RWS for potential restart (Preferred 
Alternative); and
    (3) Shut down and deactivate the RWS with no maintenance for 
potential restart.
    Based on the RWEIS evaluation of the potential environmental 
impacts, as well as the costs, energy consumption, and regulatory 
implications of the alternatives, DOE has selected the No Action 
alternative and will continue to operate the RWS. Other than potential 
CERCLA remediation activities, if DOE continued to operate and maintain 
the RWS indefinitely the No-Action Alternative would require the 
greatest commitment of money and energy resources. The RWS would 
continue to supply 5,000 gpm to L-Lake from the Savannah River. To do 
so, DOE would spend approximately $1,084,000 annually to provide RWS 
surveillance and maintenance and $494,000 annually for electrical 
energy to pump the water uphill from the river. Finally, DOE would 
continue to dredge the RWS intake canal to keep it clear of debris. 
However, there is great uncertainty regarding the cost of remedial 
action under CERCLA. Therefore, until characterization is completed, it 
will not be evident whether shutting down or continuing to operate and 
maintain the RWS is economically the most prudent course of action.
    In its present configuration, the RWS circulates water from the 
Savannah River to a 1000 acre man-made lake known as L-Lake. L-Lake no 
longer serves to mitigate thermal effluents from L-Reactor because it 
no longer operates. RWS flow is necessary to maintain the full pool 
water level of L-Lake.

[[Page 4237]]

    Low-levels of radionuclides were released to Steel Creek before L-
Lake was constructed. As a result, contaminated sediments are largely 
confined to the former Steel Creek stream bed and floodplain that 
exists under L-Lake. The methods for any needed environmental 
remediation of these low-level radionuclide releases to Steel Creek, as 
well as those to other SRS streams, will be determined under the 
Federal Facility Agreement (FFA). This agreement, between DOE, the U.S. 
Environmental Protection Agency (EPA), and the South Carolina 
Department of Health and Environmental Control (SCDHEC), provides a 
commitment and schedule for the comprehensive remediation of 
contamination at the SRS, including SRS streams and lakes.
    In accordance with the present FFA schedule, DOE will begin 
characterization of the L-Lake CERCLA unit in fiscal year 2000. DOE 
anticipates that this process will lead to an Interim Record of 
Decision (IROD) in Fiscal Year 2001. At that time DOE will decide 
whether L-Lake should be drawn down to facilitate characterization of 
future risks to human health and the environment. The characterization 
process and risk evaluation will lead to the selection of a preferred 
remedial alternative.
    During these future draw down and characterization activities, DOE 
expects to stabilize exposed sediments and address the ``reasonable and 
prudent'' measures for protection of threatened and endangered species 
that the U. S. Fish and Wildlife Service (USFWS) has recommended as a 
result of the Endangered Species Act Section 7 consultation process.

FOR FURTHER INFORMATION CONTACT: For RWEIS information: Andrew R. 
Grainger, NEPA Compliance Officer, U.S. Department of Energy, Savannah 
River Operations Office, Building 773-42A, Rm. 212, Aiken, South 
Carolina 29802, Telephone: (800) 881-7292, Attention: RWEIS, E-mail: 
[email protected]
    For general information on the DOE National Environmental Policy 
Act (NEPA) process: Ms. Carol M. Borgstrom, Director, Office of NEPA 
Policy and Assistance (EH-42), U.S. Department of Energy, 1000 
Independence Avenue, S.W., Washington, D.C. 20585, Telephone: (202) 
586-4600, or leave a message at (800) 472-2756.

SUPPLEMENTARY INFORMATION:

Background

    Located in southwest South Carolina, the SRS occupies an area of 
approximately 300 square miles (800 square kilometers). The Savannah 
River forms the SRS's southwestern boundary for approximately twenty-
seven miles (forty-three kilometers) on the South Carolina-Georgia 
border. The SRS is approximately twenty-five miles (forty kilometers) 
southeast of Augusta, Georgia and twenty miles (thirty-two kilometers) 
south of Aiken, South Carolina, the nearest major population centers. 
The U.S. Atomic Energy Commission (AEC), a DOE predecessor agency, 
established the SRS in the early 1950s for the production of nuclear 
materials to support the United States' national defense, research, and 
medical programs.
    SRS produced these materials by irradiating nuclear fuel and 
targets in SRS' five production reactors (C-, K-, L-, P-, and R-
Reactors). In the reactors, closed pipe loops contained water to cool 
the fuel assemblies by passing water directly across them. The water in 
this closed loop was then pumped to heat exchangers where heat was 
transferred from the closed system to a secondary-cooling system. This 
arrangement of closed loops minimized contamination of the environment.
    The water for the secondary-cooling system was provided by the RWS. 
The RWS pumped river water from the Savannah River using intake canals 
and pumps to the heat exchangers in the reactor areas (C-, K-, L-, P-, 
and R-Reactor areas) by way of distribution piping and water-storage 
basins.\1\ The RWS also pumped water to Par Pond, which was used to 
store additional secondary-cooling water for P- and R-Reactors.
---------------------------------------------------------------------------

    \1\ The water-storage basins are also known as 186-Basins.
---------------------------------------------------------------------------

    After passing through the heat exchangers and absorbing the heat 
from the primary closed-loop cooling system, the heated water in the 
secondary-cooling system was returned to the Savannah River by way of 
several discharge canals and streams. After 1985, when construction of 
L-Lake was completed, heated secondary-cooling water was also returned 
to the Savannah River by way of L-Lake which dissipated the heat from 
the thermal effluent (hot water) from L-Reactor. Thus, in all, the RWS 
is composed of river water intake canals, intake pumps, distribution 
piping to the reactor areas, 186-Basins, discharge canals, receiving 
streams, and lakes (Par Pond and L-Lake).
    At the end of the Cold War in 1992, SRS' mission emphasis shifted 
from the production of nuclear materials to cleanup and environmental 
restoration. Consequently, SRS' reactors were shut down, thereby 
decreasing the demand for RWS flow. From 1988 to 1996, demand for RWS 
flow and flow discharged to onsite streams decreased from 380,000 
gallons per minute (gpm) to 5,000 gpm. Therefore, reflecting decreased 
water needs and DOE's mandate to reduce operating costs, a 5,000 gpm 
pump was installed in 1997 to replace a larger, and more costly to 
operate, 28,000 gpm pump.\2\ Further, because of reduced RWS demand, 
and because SRS' reactors will not operate again, DOE identified the 
RWS as excess infrastructure, costly to operate and maintain, but with 
limited application. Accordingly, DOE prepared the RWEIS to examine the 
environmental impacts of RWS shutdown with the preference of 
eliminating the operational costs of this infrastructure, now only 
marginally useful.
---------------------------------------------------------------------------

    \2\ In the RWEIS, the No-Action Alternative impacts are assessed 
against the baseline provided by operation of the 5,000 gpm pump. 
DOE reviewed installation of the 5,000 gpm pump as a categorical 
exclusion (EEC-SS-G-96-003) in accordance with DOE's NEPA 
regulations. 10 CFR 1021. During assessment of the categorical 
exclusion, DOE determined that a 5,000 gpm pump would be sufficient 
to maintain L-Lake at 190 feet MSL and to provide the minimum 
operating needs of K- and L-Reactor areas without violating any SRS 
permits. Accordingly, the categorical exclusion was approved on June 
6, 1996, and the pump installed thereafter.
---------------------------------------------------------------------------

NEPA Process

    DOE prepared this Record of Decision pursuant to the regulations of 
the Council on Environmental Quality for implementing the National 
Environmental Policy Act (NEPA), 40 CFR Parts 1500-1508, and DOE's NEPA 
Implementing Procedures, 10 CFR Part 1021. This Record of Decision is 
based on DOE's Final RWEIS for the Shutdown of the River Water System 
at the Savannah River Site, Aiken, South Carolina (DOE/EIS-0268), for 
which DOE published a Notice of Intent to prepare on June 12, 1996, in 
the Federal Register, 61 Fed. Reg. 29744. The notice announced a public 
scoping period, ending on July 12, 1996, and solicited comments and 
suggestions on the EIS' scope. DOE held scoping meetings in North 
Augusta, South Carolina on June 27, 1996. Comments received during the 
scoping period and DOE's responses thereto were used to prepare an 
action plan, issued in August 1996, defining the scope and approach of 
the RWEIS. The action plan and reference materials cited in the RWEIS 
were made available for review in the DOE Public Reading Room, located 
at the University of South Carolina-Aiken Campus, Gregg-Graniteville 
Library, 2nd Floor,

[[Page 4238]]

University Parkway, Aiken, South Carolina at (803) 641-3320.
    DOE completed the draft RWEIS in November 1996, and on November 15, 
1996, EPA published a Notice of Availability for the document in the 
Federal Register, 61 Fed. Reg. 58548. This notice started the public 
comment period for the draft RWEIS, which extended through December 30, 
1996. DOE received comments by letter, electronic mail, and statements 
made during public hearings held in North Augusta, South Carolina on 
December 4, 1996, all of which were considered in preparing the final 
RWEIS. DOE completed distribution of the final RWEIS in May 1997, and 
on May 16, 1997, EPA published a Notice of Availability in the Federal 
Register, 61 Fed. Reg. 27024. This ROD is the culmination of and final 
step in the NEPA process for action on the RWS and announces DOE's 
selection of an alternative.

Alternatives Considered in Final RWEIS

No-Action Alternative

    The No-Action Alternative does not change the current status quo 
and involves continued operation of the RWS using a 5,000 gpm pump. 
Under this alternative, L-Lake would maintain its water level at 190 
feet MSL with makeup water provided by the RWS. Par Pond water level 
would continue to fluctuate naturally between 195 feet and 200 feet 
MSL. Under severe drought conditions, and if necessary, the RWS could 
be used to maintain Par Pond water level.

Proposed Action--Shut Down and Maintain Alternative

    The Proposed Action--and Preferred Alternative, the Shut Down and 
Maintain Alternative--provides for shutdown and maintenance of the RWS 
in a standby condition that would allow restart. RWS shutdown would 
result in the L-Lake water level returning to the original Steel Creek 
stream bed over a ten year period. RWS shutdown would not change the 
status quo regarding Par Pond's water level; it would continue to 
fluctuate naturally between 195 feet and 200 feet MSL.
    Under the Shut Down and Maintain Alternative, the RWS operational 
capacity would be preserved in a standby mode to account for unforeseen 
events, mission changes, or remedial action decisions. Maintaining the 
RWS in a standby condition \3\ requires draining the system of water 
and placing the equipment in a protective state minimizing 
degradation.\4\ Under this alternative, the RWS operation could be 
restored to provide water for future missions or, if necessary, to 
maintain Par Pond water level above 195 feet MSL in the unlikely event 
of a severe drought. In addition, the RWS could be restarted if the 
final outcome of the FFA process recommends refilling L-Lake with water 
to manage risk from contaminated sediments in the Steel Creek stream 
bed. During the interim, or in the event none of these potentialities 
are realized, shutdown of the RWS would eliminate operational costs 
associated with this system.
---------------------------------------------------------------------------

    \3\ This standby condition is also referred to as ``lay-up.''
    \4\ Placing the RWS in lay-up also allows maintaining portions 
of the system in a higher state of readiness in order to restore 
pumping capability more rapidly. Maintenance of certain portions of 
the RWS in such a condition might be warranted (1) where those 
portions are likely to be needed for future missions; (2) where they 
might be necessary to maintain Par Pond water levels in the event of 
a severe drought; or (3) where they might be necessary to refill L-
Lake in the event of determination to do so as a result of the FFA 
process.
---------------------------------------------------------------------------

Shut Down and Deactivate Alternative

    The Shut Down and Deactivate Alternative provides for the permanent 
cessation of RWS operation and does not preserve system capabilities, 
even in the most marginal state, for restart. DOE would shut down and 
deactivate the system in a secure, environmentally satisfactory 
condition and isolate all the intake pipes to prevent river water 
intrusion into the RWS. DOE would conduct no maintenance or 
surveillance on the RWS, with the exception of the L-Lake dam, which 
would be maintained until the Lake's water level returned to the 
original Steel Creek stream bed in approximately 10 years.
    Under this alternative, L-Lake water level would return to the 
original Steel Creek stream bed. Par Pond water level would continue to 
fluctuate naturally between 195 feet and 200 feet MSL. Under severe 
drought conditions, the RWS could not be used to maintain Par Pond 
water level, even if necessary. Furthermore, the RWS could not be 
restarted if the final outcome of the FFA process recommended refilling 
L-Lake with water to manage risk from contaminated sediments in the 
Steel Creek stream bed.

Environmental Impacts of the Alternatives

Environmental Impacts of No Action Alternative

    The No Action alternative would preserve the status quo and 
continue current operation of the RWS through a 5,000 gpm pump.\5\ 
Under the No Action Alternative, L-Lake would remain at its normal 
water level of 190 feet MSL. Par Pond water level would continue to 
fluctuate naturally between 195 feet and 200 feet MSL.
---------------------------------------------------------------------------

    \5\ As previously noted, the environmental impacts of the 5,000 
gpm pump were evaluated under a categorical exclusion. On December 
30, 1996, EPA provided comments on the Draft RWEIS and questioned 
the appropriateness of this categorical exclusion. EPA requested DOE 
to describe more thoroughly the impacts associated with the 5,000 
gpm pump. In response to EPA's comment, those impacts were included 
in the Final RWEIS. A discussion demonstrating the appropriateness 
of the categorical exclusion may be found in the RWEIS at page E-61.
---------------------------------------------------------------------------

Environmental Impacts of Shut Down and Maintain & Shut Down and 
Deactivate Alternatives

    The environmental impacts of the Proposed Action--and Preferred 
Alternative, the Shut Down and Maintain Alternative--are the same as 
those of the Shut Down and Deactivate Alternative. Both alternatives 
call for DOE to shutdown the RWS. While the Proposed Action calls for 
DOE to preserve the RWS in a standby condition, the actions necessary 
to accomplish that goal do not entail environmental impacts beyond 
those associated with the shutdown action. Accordingly, the 
environmental impacts of either alternative are the same and DOE 
considers them together in the following paragraphs.
L-Lake
    Under either the Shut Down and Maintain Alternative or the Shut 
Down and Deactivate Alternative, DOE would not augment water flow to L-
Lake. L-Lake cannot maintain a water level of 190 feet MSL, its normal 
full pool water level, without flow augmentation from the RWS. 
Consequently, it would recede to the original Steel Creek stream bed 
conditions over a ten-year period.
    As L-Lake recedes to the Steel Creek stream bed as a consequence of 
either shutdown alternative, habitat for amphibians, reptiles, semi-
aquatic mammals, wading birds, and waterfowl would be gradually reduced 
and eliminated. Consequently, these species would be more vulnerable to 
predation. Eventually, alligators would be displaced due to the loss of 
habitat. Drawdown of L-Lake would result in the loss of nests, eggs, or 
hatchlings.
    In addition, the reversion of L-Lake water level to the former 
Steel Creek stream bed would uncover lake bed sediments. As a result, 
these sediments could be susceptible to the forces of erosion, 
especially during storm events. In addition, the reversion of L-Lake

[[Page 4239]]

water level to the original Steel Creek stream bed could expose some 
sediments, primarily in the Steel Creek stream bed, that could contain 
low levels of contamination, primarily cesium-137.\6\ Animals foraging 
in the L-Lake lake bed or Steel Creek stream bed could be exposed to 
these sediments via inhalation, ingestion, direct radiation exposure, 
and/or skin contact. Similarly, an on-site human working in the L-Lake 
bed lake could be exposed to the contaminants in sediments via 
inhalation, incidental ingestion, direct radiation exposure and/or skin 
contact. An off-site human could be exposed to contaminants in 
sediments through atmospheric or aqueous pathways via inhalation or 
ingestion from sediments that have been re-suspended in air or water. 
The off-site human would not be exposed to direct radiation.
---------------------------------------------------------------------------

    \6\ Cesium-137 is an external radiation hazard from direct 
exposure to gamma radiation which penetrates clothing and skin. 
Measurements taken under the surface of L-Lake show that cesium in 
the Lake sediments is largely concentrated in the original Steel 
Creek floodplain, currently beneath the surface of L-Lake. These 
measurements also show that a maximum (24-hour per day) radiation 
dose received by a human would be approximately 180 mrem per year 
above the typical radiation dose to which Americans are routinely 
exposed (approximately 360 mrem per year). The occupational dose 
limit for adults is 5,000 mrem per year, and this additional dose 
would not exceed that limit. 10 CFR Sec. 20.1201.
---------------------------------------------------------------------------

    Exposure to L-Lake lake bed contaminants is unlikely to pose a 
significant risk to SRS workers, the public, or the environment. The L-
Lake lake bed contaminants would be unlikely to pose a significant risk 
to SRS workers because the concentration of the contaminants in the 
sediments is low and the amount of time that an SRS worker would be 
expected to spend in the lake bed would not yield an annual dose above 
DOE administrative limits (700 mrem). For example, an SRS worker 
spending eight hours per days for 250 days a year over a twenty-five 
year period would receive an annual dose of 41 mrem. The 41 mrem dose 
is well below DOE's 700 mrem administrative limit.
    The L-Lake lake bed contaminants would be unlikely to pose a 
significant risk to the public because the public would not be exposed 
to direct radiation, which is the primary hazard associated with 
cesium-137. The probability of the maximally-exposed individual, 
located at the SRS site boundary, developing a fatal cancer as a result 
of 70 years exposure would be less than one in a million (5.6 X 10E-7). 
In DOE's judgment this risk is extremely small.
    Finally, the L-Lake lake bed contaminants would be unlikely to pose 
a significant risk to the environment because erosion would be 
controlled and contaminated sediments would not pose a significant risk 
to foraging animals. Erosion would be controlled because, based on 
DOE's historic hydrologic data and models, L-Lake would probably recede 
during the growing season. As the Lake's water level slowly receded, 
wetland plants growing in the shore zone would recede down slope with 
the water. Seed banks along the shoreline would germinate and stabilize 
sediments in portions of the newly exposed shoreline. In addition, DOE 
would artificially seed the exposed L-Lake lake bed with appropriate 
vegetation in order to further stabilize the sediment. Thus, exposure 
to contaminants in the L-Lake due to erosion or resuspension of lake 
bed sediment would be minimized because the sediments would be 
protected from wave or wind agitation.
    Furthermore, erosion and transport of contaminated L-Lake lake bed 
sediments would be reduced by the slow drawdown of the water level in 
the Lake, occurring over a ten year period, and by the resulting growth 
of stabilizing vegetation. Because of this slow drawdown and growth of 
stabilizing vegetation, suspension of sediments in the water column 
would be minimized. Further, DOE would maintain the Steel Creek dam 
during the drawdown to impede the transport of those sediments that 
became suspended in the water column. The Steel Creek dam would 
minimize the movement of contaminants suspended in the water column by 
creating a stilling basin to still the water and allow sediments to 
settle out of the water column.
    Contaminants in L-Lake lake bed would not pose a significant risk 
to foraging animals either from radiological or non-radiological 
sources. As described in the RWEIS, radiological contaminants were 
screened against known background contaminant concentrations yielding 
estimated radiation dose rates, which were then compared to applicable 
standards. This comparison indicated that two radionuclides exceeded 
twice the background level, namely cesium-137 and Co-60. Cs-137 and Co-
60 in L-Lake lake bed sediments would primarily cause risk from the 
direct exposure of penetrating gamma radiation. The concentrations of 
these radiological sources was used to estimate a dose rate to selected 
receptor species. The estimated radiation dose rates to selected 
receptor species are well below the applicable standards.
    For non-radiological sources, it was recognized that L-Lake lake 
bed sediments would be exposed and that the sediments could become 
surface soil or facilitate vegetative growth. All samples detected in 
sediments were compared to screening levels for sediments, surface 
soils, and terrestrial plants. No sediment contaminants were present in 
average concentrations that exceeded available screening levels. 
Screening levels were not available for four non-radiological samples 
found in L-Lake sediment--beryllium, cobalt, thallium, and vanadium--
which had an average concentration between two and three times their 
background levels. The potential risk of these contaminants (as well as 
all contaminants that were detected) were assessed by screening their 
respective concentrations against surface soil screening levels.
    Assuming the sediments became surface soils, the average 
concentrations of beryllium, thallium, chromium, and vanadium were 
between two and three times their average background levels. Thallium 
was detected in five of forty-four samples and beryllium, chromium, and 
vanadium slightly exceeded twice its background screening level. This 
indicated that these contaminants either are not present in high 
concentrations or are not widespread because of the few occurrences in 
samples. Accordingly, they would not represent an unacceptable risk. 
The average concentration of cobalt was below its associated screening 
level.
    Assuming the detected sediment concentrations were found in 
terrestrial plants, chromium, thallium and vanadium were between two 
and three times their background screening levels. Again, thallium was 
detected in five of 44 samples. It should also be noted that plants 
absorb chromium, thallium, and vanadium minimally from soils. Because 
of this fact and that L-Lake currently supports a healthy, diverse 
ecological community, it does not appear that effects to L-Lake plants 
from contaminants are occurring or would occur under the Proposed 
Action.
    The screening process is discussed in greater detail in Appendix B 
of the RWEIS.
Par Pond
    Par Pond water level would not be impacted by any of the 
alternatives considered in the RWEIS. Par Pond has not received makeup 
water from the RWS since January 1996, and has been allowed to 
fluctuate naturally between 195 feet and 200 feet MSL. Accordingly, 
ceasing operation of the RWS could not effect Par Pond water level 
because it is not currently receiving make-up water from the RWS.

[[Page 4240]]

    Allowing Par Pond to fluctuate naturally was the product of prior 
analysis and decisions conducted under authority of NEPA and the 
Comprehensive Environmental Response, Compensation, and Liability Act 
(CERCLA). In response to safety concerns presented by a 1992 leak in 
the Par Pond dam, DOE prepared a NEPA document called a Special 
Environmental Analysis (SEA) to assess the environmental impacts of an 
emergency drawdown of the Par Pond water level from 200 feet MSL, its 
normal water level, to 181 feet MSL. The SEA reviewed the anticipated 
and observed environmental impacts of drawing down, repairing, and 
refilling Par Pond, including potential health and ecological impacts 
resulting from the possible exposure to radiocesium contaminated 
sediment. It concluded that ``consideration would be given to begin 
refilling operations as soon as possible, perhaps before dam repairs 
are completed, to minimize impacts on the Par Pond ecosystem.'' DOE, 
Special Environmental Analysis for Par Pond at the Savannah River Site, 
page 4, (April, 1992). In addition, the SEA summarized consequences of 
possible future repair, remedial, and refilling actions, and developed 
a Mitigation Action Plan (MAP) to reduce the impacts of the repair 
activity. Under authority of the SEA, DOE proceeded with the Par Pond 
drawdown and repair project.
    Following the repair action, DOE prepared a Par Pond Interim Record 
of Decision (IROD) to provide a CERCLA remedial action to address the 
interim period following the dam repair until a final remedial action 
could be studied and selected. The IROD's selected remedy consisted of 
refilling and maintaining Par Pond to 200 +/-1 feet until a NEPA 
evaluation could be accomplished to evaluate the environmental impacts 
from reduced flow to Lower Three Runs Creek (the creek below Par Pond 
dam), fluctuating reservoir water level, and discontinuing of river 
water pumping to the reservoir. DOE, Interim Action Record of Decision, 
Remedial Alternative Selection, Par Pond Unit, WSRC-RP-93-1549 (January 
26, 1995). EPA and SCDHEC approved the IROD in February 1995, and Par 
Pond was completely refilled by March 15, 1995.
    In 1995, DOE prepared a NEPA Environmental Assessment (EA), Natural 
Fluctuation of Water Level in Par Pond and Reduced Water Flow in Steel 
Creek Below L-Lake at the SRS (DOE/EA-1070). The EA analysis showed 
that no significant impacts would likely result to the Par Pond 
ecosystem if the Pond's water level were maintained at 195 feet MSL or 
above. Hydrological models analyzed in the EA showed that even without 
RWS flow augmentation, the Par Pond water level is not likely to 
decrease below 196 feet MSL, even in drought conditions. Further, the 
EA analysis showed that elimination of RWS water flow and the 
accompanying reduction of Savannah River nutrients flowing through Par 
Pond would cause the Pond's ecosystem to revert to that typically found 
in reservoirs in the southeastern United States. Based on the analysis 
in the EA, DOE issued a Finding of No Significant Impact (FONSI) in 
August 1995 under which RWS water flow to Par Pond was eliminated. 
Under the FONSI, if Par Pond water level decreases to 195 feet MSL or 
below, DOE will resume water flow augmentation through the RWS. Since 
issuance of the FONSI, natural water flow into Par Pond has maintained 
the Pond above 199 feet MSL.
    Under either of the RWEIS shutdown alternatives, the status quo 
would not be changed and RWS water would not augment natural water flow 
into Par Pond. Under the Shut Down and Maintain Alternative DOE would 
be able to restart the RWS and resume pumping to Par Pond if the water 
level drops below 195 feet MSL, as called for in the FONSI and 
consistent with the IROD. However, under the Shut Down and Deactivate 
Alternative, DOE would not have the capability to restart the RWS to 
augment water flow to Par Pond in the event a severe drought demanded 
such an action.
Other Impacts
    Under either of the shutdown alternatives, DOE would need to find 
an alternate water supply for auxiliary equipment cooling and for fire 
protection water reserves. The alternate water supply would be 
approximately 400 gpm and be drawn from groundwater. This groundwater 
would be provided by existing wells at rates much less than was 
historically provided by these same wells during reactor operations.
    RWS shutdown would result in increased survival of Savannah River 
larval fish and fish eggs because they would no longer be entrained at 
the RWS intake structures. In addition, RWS shutdown would return 225 
acres of original wetlands inundated by the damming of Steel Creek and 
creation of L-Lake. This acreage is approximately the same amount of 
wetlands that exists along the present shoreline of L-Lake that would 
be lost as L-Lake water level recedes to the original Steel Creek 
stream bed.

Environmentally Preferable Alternative

    The proposed action in this instance presents a situation where the 
environmentally preferable action is different based on whether a 
short-term or long-term period of reference is used. Based on the 
analysis in the RWEIS, DOE finds that in the short-term, the 
environmentally preferable alternative is the No-Action Alternative to 
preserve the status quo and continue current operation of the RWS 
through a 5,000 gpm pump. The No-Action Alternative would preserve L-
Lake and prevent the return of the Lake's water level to the original 
Steel Creek stream bed. The preservation of the Lake would, in turn, 
preserve up to 1000 acres of aquatic habitat formed by it, and 
forestall the transition of this habitat to uplands and wetlands 
habitat.
    However, in the long-term, the Shut Down and Maintain Alternative 
is the environmentally preferable alternative. Under this alternative, 
L-Lake water level would return to the original Steel Creek stream bed 
over a ten year period, which would allow for the gradual restoration 
of a more stable ecosystem, such as that in existence prior to 
construction of L-Lake. The pre-Lake ecosystem would be more stable 
because it is the indigenous ecosystem, and because it would not be 
susceptible to potential imbalances, such as those introduced by 
changes in L-Lake water level associated with repair or renovation of 
the Steel Creek dam.
    Furthermore, the restored L-Lake ecosystem would benefit from the 
reemergence of 225 acres of wetlands inundated by the creation of L-
Lake, an amount that approximately equals the amount of wetland acreage 
that would be lost along the shoreline of L-Lake as it gradually 
recedes. After these areas are exposed, they would naturally 
reestablish wetland characteristics with cycles of drying and flooding 
typical of other hardwood swamps on the SRS and in the southeast. As 
typical wetlands they would support diverse ecological communities.
    In addition, while a decrease in aquatic productivity would be 
expected as a consequence of the return of L-Lake water level to the 
original Steel Creek stream bed, an increase in terrestrial 
productivity would occur concomitantly. As the L-Lake water level 
receded, grasses, forbs, shrubs, and trees indigenous to the ecosystem 
would re-colonize the L-Lake lake bed over time. In addition to flora, 
indigenous fauna would return to the ecosystem, and a variety of 
terrestrial and semi-aquatic animal species would inhabit

[[Page 4241]]

the area as L-Lake gradually receded to the original Steel Creek stream 
bed.
    Other than potential CERCLA remediation activities, if DOE 
continued to operate and maintain the RWS indefinitely the No-Action 
Alternative would require the greatest commitment of money and energy 
resources. The RWS would continue to supply 5,000 gpm to L-Lake from 
the Savannah River. To do so, DOE would spend approximately $1,084,000 
annually to provide RWS surveillance and maintenance and $494,000 
annually for electrical energy to pump the water uphill from the river. 
Finally, DOE would continue to dredge the RWS intake canal to keep it 
clear of debris. However, there is great uncertainty regarding the cost 
of remedial action under CERCLA. Therefore, until characterization is 
completed, it will not be evident whether shutting down or continuing 
to operate and maintain the RWS is economically the most prudent course 
of action.

Associated Actions

    DOE considered a number of actions that affect the selection of an 
alternative for the RWS, as well as the timing of implementing a 
selected alternative. The actions are described in the following 
paragraphs.

Remedial Action Process for L-Lake

    Through the FFA, DOE, EPA, and SCDHEC established the procedure for 
environmental restoration activities at the SRS. The FFA integrates DOE 
responsibilities under the Resource Conservation and Recovery Act 
(RCRA) and CERCLA. In response to EPA and SCDHEC comments on the Draft 
RWEIS, DOE recommends further assessment of L-Lake under the FFA, 
possibly resulting in a Baseline Risk Assessment (BRA) and a Remedial 
Investigation/ Feasibility Study (RI/FS).
    A BRA will assess the risk associated with the contaminants 
identified in the L-Lake sediment, primarily located in the original 
Steel Creek stream bed, and it will provide a quantified expression of 
risk for key receptors, such as humans or wildlife, which may be 
exposed to the contaminants. An RI/FS will gather data necessary to 
determine more exactly the nature and extent of contamination in L-Lake 
sediment, establish criteria for remediating the Lake, identify the 
preliminary alternatives for remedial actions, and support the 
technical and cost analyses of the remedial alternatives.
    DOE believes that the analysis and data collection necessary to 
prepare a BRA and RI/FS is more accurately, easily, and economically 
obtained once L-Lake has returned to its original Steel Creek stream 
bed. This is because there are inherent difficulties in taking sediment 
samples while L-Lake is filled if additional samples are needed. 
Because shutdown of the RWS will present no unreasonable risk to human 
health or the environment, and because analysis of L-Lake sediment is 
more appropriate after the Lake has returned to the original Steel 
Creek stream bed, DOE anticipates that this process will be 
accomplished under an Interim Record of Decision (IROD). In accordance 
with the present FFA schedule, DOE will begin characterization of the 
L-Lake CERCLA unit in fiscal year 2000 and begin L-Lake draw down in 
Fiscal Year 2001. The characterization process and risk evaluation will 
lead to the selection of a preferred remedial alternative. It is DOE's 
intention to incorporate National Environmental Policy Act values in 
the IROD and supporting documents.

Remedial Action Process for Onsite Streams

    Steel Creek, Four Mile Branch, Pen Branch, and Lower Three Runs, 
are listed in the FFA as RCRA/CERCLA units because each stream received 
contaminants from past operations. EPA and SCDHEC expressed concern 
about the effect on these units due to the installation of the 5,000 
gpm pump because the installation of that pump reduced water flow 
capacity through the streams from 28,000 gpm to 5,000 gpm. The 
reduction in water flow through the SRS streams has increased the 
concentration of tritium transported to SRS streams from the seepage 
basins.
    Increased tritium concentrations in site streams are the 
consequence of two factors. First, tritium from the seepage basins is 
carried with rainwater to site streams as it percolates through the 
soil at a fairly constant rate. Second, the RWS flow formerly diluted 
the tritium-containing rainwater as it percolated to the streams. 
Installation of the 5,000 gpm pump reduced RWS flow contribution to the 
streams and removed this dilution water. Consequently, installation of 
the 5,000 gpm pump had the effect of increasing tritium concentrations.
    To respond to comments on the draft RWEIS, the Final RWEIS 
evaluated the impacts to workers, ecosystems, and the public due to the 
installation of the 5,000 gpm pump from the resulting increase in 
tritium concentrations. Workers and ecological receptors would be at 
risk due to increased tritium exposure through incidental ingestion and 
skin contact. The public could be at risk due to ingestion of increased 
concentrations of tritium in drinking water. However, the RWEIS risk 
assessment showed that a hypothetical future worker's annual dose would 
be below 1 mrem. The RWEIS environmental risk assessment showed that 
the highest annual dose to an ecological receptor would be 92 mrem. 
Both of these dose rates are well below accepted standards.
    Drinking water taken from the Savannah River would not be impacted 
because installation of the 5,000 gpm did not increase the total amount 
of tritium released to the River. Because the flow rate of water in the 
Savannah River is typically over 10,000 cubic feet per second (compared 
to the 45 cubic feet per second reduction of flow from installation of 
the 5,000 gpm pump) and because the nearest domestic water plant intake 
is approximately 40 miles downstream from SRS, the on-site increased 
concentrations have an insignificant health impact to the public. In 
summary, the increased concentrations of tritium in site streams were 
determined to be acceptable because these concentrations did not pose 
an unacceptable risk to workers, the ecosystems or the public.
    Steel Creek, Four Mile Branch, Pen Branch, and Lower Three Runs, as 
well as other SRS streams have received low-levels of radionuclides, 
including tritium, from past SRS operations. Therefore, all of them 
will be evaluated in accordance with the FFA and be the subject of a 
risk analysis based on hypothetical future residents and industrial 
workers.7 DOE is scheduled to provide information to EPA and 
SCDHEC, which will assist in the characterization of each stream and 
the selection of a remedial alternative.
---------------------------------------------------------------------------

    \7\  These two hypothetical future groups could be at risk 
because the risk analysis assumes that they drink water from SRS 
streams before it is mixed with the Savannah River. A down stream 
receptor is not at risk because the Savannah River flow rate is 
significantly higher than SRS streams and, in effect, dilutes 
tritium to concentrations which do not pose a risk to human health.
---------------------------------------------------------------------------

Water Requirements for Alternatives

    Under the No-Action Alternative, the RWS would continue to supply 
existing operational cooling and make-up water requirements for the 
reactor areas and maintain L-Lake at 190 feet MSL. For either of the 
shutdown alternatives, DOE must supply 400 gpm of groundwater to 
replace that provided by the RWS to cool auxiliary equipment and to 
provide make-up water for fire protection reserves.

[[Page 4242]]

L-Area Sanitary Wastewater Treatment Plant

    The L-Area sanitary wastewater effluent mixes with RWS flow before 
reaching L-Lake. The L-Area Sanitary Wastewater Treatment Plant 
wastewater permit took credit for RWS blending flow in determining the 
extent of treatment necessary before the wastewater was discharged to 
L-Lake. To stop RWS flow, DOE must implement an alternate compliance 
method to manage L-Area sanitary wastewater.

Reactor 186-Basins Alternative Uses Study

    In 1994, DOE analyzed the feasibility of using the SRS C-, L-, P-, 
and R-Reactor 186-Basins 8 and 904-Retention Basins 
9 for aquacultural purposes. In March 1995 DOE advertised 
the availability of the Reactor 186-Basins for commercial use. DOE 
accepted one fish farming proposal that would have relied on ground 
water for make-up water, although this proposal was later withdrawn. At 
the present time, no future uses of the 186-Basins or the 904-Retention 
Basins are planned. DOE could accept similar proposals in the future 
regardless of the RWEIS alternative selected because the basins do not 
rely on the RWS for make-up water.
---------------------------------------------------------------------------

    \8\ 186-Basins are water storage basins which are not 
contaminated.
    \9\ 904-Retention Basins are 50 million gallon basins that were 
designed to receive emergency cooling water in the event of a 
reactor accident.
---------------------------------------------------------------------------

Decision

    DOE selects the No Action Alternative of the RWEIS--Continue to 
Operate the RWS. Under this alternative, DOE will continue to operate 
and maintain the system as well as maintain the water level of L-Lake 
at its full pool water level of 190 feet MSL.
    DOE will assess the need for future environmental remediation 
alternatives for L-Lake under existing CERCLA commitments. 
Characterization activities associated with CERCLA closure are expected 
to begin in the year 2000 and be completed in several years. This 
characterization will inform any required remedial action. Pending 
these activities, DOE will continue to operate the RWS. If during 
continued operation of the RWS a system component fails, DOE will take 
appropriate emergency actions. DOE will then determine if the system is 
too costly to repair (by comparing this cost to estimated shutdown 
costs and future possible remediation costs under CERCLA). If DOE 
determines that the RWS is too costly to repair, it will reevaluate all 
relevant commitments and the information in the RWEIS, to determine 
necessary actions to shut down the RWS. However, the RWS is in good 
condition and not expected to fail over this period of time.
    In accordance with the present FFA schedule, DOE will begin 
characterization of the L-Lake CERCLA unit in fiscal year 2000. DOE 
anticipates that this process will lead to an Interim Record of 
Decision (IROD) in Fiscal Year 2001. At that time DOE will decide 
whether L-Lake should be drawn down to facilitate characterization of 
future risks to human health and the environment. The characterization 
process and risk evaluation will lead to the selection of a preferred 
remedial alternative. Not withstanding a major system failure, DOE has 
decided to operate and maintain the RWS until a preferred remedial 
alternative is selected.
    DOE made this decision after considering the most recent operating 
and maintenance costs and estimated shutdown implementation costs. DOE 
has concluded the amount and uncertainty in shutdown implementation 
costs suggest the RWS should continue to be operated while DOE monitors 
RWS operating and maintenance costs to determine when continued 
operation becomes too costly. For example, if a portion of the system 
failed, DOE may elect to take an emergency action in accordance with 40 
CFR 1506.11, if appropriate, and then determine if repair was too 
costly. If DOE determines that repair is too costly it will announce 
this decision in a future Record of Decision under the requirements of 
NEPA and CERCLA, as necessary.
    Prior to drawdown, DOE will notify the USFWS to ensure all 
``reasonable and prudent'' measures, which were recommended by USFWS 
during the Endangered Species Action consultation process to protection 
of threatened and endangered species, are still adequate and 
appropriate. DOE will also negotiate a schedule with USFWS for the 
review and completion of these measures.
    During any future draw down and characterization activities DOE 
expects to stabilize exposed sediments and address the ``reasonable and 
prudent'' measures, discussed above.
    DOE considers continued operation of the RWS to be environmentally 
preferable in the short-term because L-Lake remain as a lake with its 
ecology unaffected.

Comments on Final RWEIS

    DOE received two letters commenting on the Final RWEIS. The first, 
a letter from EPA, Region IV, dated June 12, 1997, expressed concern 
that the RWEIS does not adequately consider injury or impacts to 
endangered species. To consider injury or impacts to these species DOE 
and USFWS entered into a formal consultation process regarding 
endangered species. The USFWS recommended specific ``reasonable and 
prudent'' measures to protect the bald eagle and wood stork during the 
L-Lake water level return to the original Steel Creek stream bed. DOE 
endorsed these reasonable and prudent measures. However, these measures 
will not be implemented at this time because the draw down of L-Lake 
will not occur as a result of this decision.
    The letter from EPA, Region IV also expressed concern that the 
RWEIS did not adequately consider the ecological risks associated with 
shutdown of the RWS. However, as explained in the Environmental Impacts 
section, the RWEIS ecological risk assessment (ERA) concluded that 
significant potential risks to ecological receptors from contaminants 
is not likely.
    Finally, as a general statement, the letter from EPA, Region IV 
stated that, ``This NEPA action should be coordinated to the fullest 
extent possible with FFA activities''. The selection of the No Action 
alternative has been made, in part, in response to this statement.
    In addition, the RWEIS documented several measures that were taken 
to coordinate the NEPA with the FFA process, which include the 
following: (1) use of FFA criteria as contamination level screening 
limits to estimate future potential remedial action decisions; (2) 
movement of the L-Lake unit from Appendix G to Appendix C of the FFA in 
order to avoid the unnecessary generation of a Site Evaluation Report 
and expedite the FFA process; and (3) preservation of the ability to 
refill L-Lake under the RWEIS shut down and maintain alternative in the 
event that such action is determined to be necessary under the FFA.
    The second letter, dated June 11, 1997, from the Office of 
Environmental Policy and Compliance within the Department of the 
Interior (DOI) addressed several concerns regarding the revised 
ecological risk assessment (ERA) provided in the RWEIS. Specifically, 
DOI commented that guidance DOE used to develop the RWEIS ERA ``is 
inadequate and inconsistent with the US EPA's guidance on ecological 
risk

[[Page 4243]]

assessment.'' To the contrary, DOE believes the RWEIS ERA was conducted 
in accordance with the most current EPA guidance.
    The objective of the RWEIS ERA was to determine, as accurately as 
possible under the existing L-Lake characteristics, the probable 
outcome of a CERCLA ERA.10 The RWEIS ERA is adequate because 
it used maximum contaminant concentrations in its risk assessments. In 
fact, all radiological risks were within dosimetry-based limits 
acceptable to the International Atomic Energy Agency. Where the maximum 
concentrations of some non-radiological contaminants showed a risk 
potential, either the average concentration of the contaminant was 
compared to background levels or, alternately, the contaminant 
concentration was compared with more contaminant-specific information 
available in accepted scientific literature. This procedure is part of 
the typical process of interpreting the results and uncertainties of an 
ERA and represents the general ERA approach recommended in the EPA 
guidance for Superfund. EPA, Ecological Risk Assessment for Superfund: 
Process for Designing and Conducting Ecological Risk Assessments, 
Review Draft, (1996).
---------------------------------------------------------------------------

    \10\ The CERCLA ERA will be used to aid the determination of a 
final remedial action at L-Lake. A final action is typically made 
only after the lake bed is characterized in detail using information 
such as groundwater hydrogeology, extent of groundwater 
contamination, and existing burial area contamination profiles. 
Presently, these areas are under as much as 50 feet of water and 
cannot be adequately characterized. As a result, a complete risk 
assessment cannot be performed and a final remedial alternative 
cannot be selected until L-Lake returns to the original Steel Creek 
stream bed.
---------------------------------------------------------------------------

    DOI also asserted that the ``[c]onclusions of no risk are 
inconsistent with actual research findings.'' DOI cited specific DOE 
studies sent to USFWS during the formal Endangered Species Act 
consultation. The studies DOI cited assessed DNA changes found in blood 
samples of various wildlife species present on the SRS. In response to 
this assertion, DOE notes that, in no case did any of the studies 
suggest that an observable change in a wildlife population would result 
from the exposure to low levels of radionuclides found in L-Lake. 
Accordingly, this supports the RWEIS ERA finding that significant 
potential risks to ecological receptors from contaminants are not 
likely.
    Finally, DOI commented on various DOE studies showing the presence 
of elevated mercury concentrations in the SRS environment. However, 
DOI's comment does not reflect the fact that the presence of mercury in 
the SRS environment is not the result of releases attendant to SRS 
operations. Indeed, mercury is elevated throughout areas of the 
southeastern United States due to atmospheric deposition, not due to 
SRS operations. Reflecting and restating this fact, SCDHEC issued a 
fish consumption advisory for numerous lakes and rivers in South 
Carolina based on mercury concentrations in fish. Again, the presence 
of this mercury was not and cannot be correlated to any SRS operations. 
Accordingly, DOE has no control over and is not responsible for the 
atmospheric deposition of mercury at SRS, or in other areas of the 
southeastern United States. Consequently, a returning L-Lake water 
level to the original Steel Creek stream bed would not exacerbate this 
regional phenomenon or increase ecological risk.

Conclusion

    After consideration of all relevant information and data, DOE 
selects the No Action alternative as the most appropriate action for 
the future of the River Water System at the Savannah River Site at this 
time. This operational decision is made in recognition of all 
beneficial and adverse environmental impacts, monetary costs, 
regulatory implications and commitments under the FFA, and dictates of 
relevant statutes.

    Signed this 23rd day of December, 1997, at Aiken, South 
Carolina.
Greg Rudy,
Acting Manager, Savannah River Operations Office.
[FR Doc. 98-1947 Filed 1-27-98; 8:45 am]
BILLING CODE 6450-01-P