Text

Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios at Vogtle Electric Generation Plant, Burke County, Ga, Open-File Report 2007-1363
	ML073310491
Person / Time
Site:	Vogtle
Issue date:	01/01/2007
From:	Cherry G, Clarke J; US Dept of Interior, Geological Survey (USGS)
To:	; Office of Nuclear Reactor Regulation
References
	+reviewedcja 2007-1363
	Download: ML073310491 (54)
	v • d • e

wwII Prepared in cooperation with the U.S. Nuclear Regulatory Commission Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios at Vogtle Electric Generation Plant, Burke County, Georgia Open-File Report 2007-1363 U.S. Department of the Interior U.S. Geological Survey Cover photograph:

Vogtle Electric Generation Plant cooling towers, Burke County, Georgia Photograph by Alan M. Cressler, U.S. Geological Survey Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios at Vogtle Electric Generation Plant, Burke County, Georgia By Gregory S. Cherry and John S. Clarke Prepared in cooperation with the U.S. Nuclear Regulatory Commission Open-File Report 2007-1363 U.S. Department of the Interior U.S. Geological Survey U.S. Department of the Interior DIRK KEMPTHORNE, Secretary U.S. Geological Survey Mark D. Myers, Director U.S. Geological Survey, Reston, Virginia:

2007 For product and ordering information:

World Wide Web: http://vwww.usgs.gov/pubprod Telephone:

1-888-ASK-USGS For more information on the USGS--the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment:

World Wide Web: http://www.usgs.gov Telephone:

1-888-ASK-USGS Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted materials contained within this report.Suggested citation: Cherry, G.S., and Clarke, J.S., 2007, Simulation and particle-tracking analysis of selected ground-water pumping scenarios at Vogtle Electric Generation Plant, Burke County, Georgia: U.S. Geological Survey Open-File Report 2007-1363, 51 p., Web-only publication at http://pubs.usgs.gov/usgspubs/of/2007/1363 Contents A bstract .....................................................................................................................................................

1 Introduction

................................................................................................................................................

1 Purpose and Scope .........................................................................................................................

1 Description of Study Area ..............................................................................................................

4 Savannah River Site and Ground-Water Contamination

..................

4 Clim ate and Runoff ................................................................................................................

4 Hydrogeologic Setting ........................................................................................................

5 Structural Features ...................................................................................................

6 Hydrogeologic Characteristics of the Savannah River Alluvial Valley ..............

6 Ground-W ater Flow ...............................................................................................................

6 Trans-River Flow ...........................................................................................................

13 Effect of Pen Branch Fault .......................................................................................

13 Ground-W ater Use .................................................................................................................

13 Sim ulation of Ground-W ater Flow ..........................................................................................................

14 Pen Branch Fault ..............................................................................................................................

16 Ground-W ater Pum ping Scenarios

..........................................................................................

17 2002 Base Case Condition

...............................................................................................

18 Scenario A ...............................................................................................................................

20 Scenario B ...............................................................................................................................

20 Scenario C ...............................................................................................................................

35 M odel Lim itations .............................................................................................................................

43'Sum m ary and Conclusions

..............................................................................................

........................

43 References Cited .......................................................................................................................................

44 iv Figures 1. Maps showing (A) study area, model grid, and model boundary, (B) location of Vogtle Electric Generation Plant production wells, Burke County, Georgia, and river and recharge cells in the Gordon aquifer (layer A2) and areal and local ground-water contamination at the Savannah River Site, South Carolina ...............

2 2. Schematic diagram showing hydrogeologic framework, model layers, and boundary conditions for the Vogtle Electric Generation Plant area, South Carolina, ground-water model .................................

5 3. Map showing subsurface extent of hydrogeologic units beneath the Savannah River alluvial valley, South Carolina and Georgia ..........................................................

7 4. Diagram of conceptualized hydrogeologic framework and related ground-water flow near Vogtle Electric Generation Plant, Georgia and South Carolina ................

8 5-8. Maps showing potentiometric surface, near Vogtle Electric Generation Plant, Georgia and South Carolina, September 2002, for the-5. U pper Three Runs aquifer .....................................................................................

..9 6. G o rd o n a q u ife r .........................

7 ....................................................................

...............

10 7. D ublin aq uife r system ..................................................................................................

11 8. M idville aquifer system ........................................................................................

..12 9. Map showing particle-tracking results for the year 2002, Base Case, near Vogtle Electric Generation Plant, Georgia and South Carolina ................................

19 10-15. Maps showing simulated water-level change for Scenario A, near Vogtle Electric Generation Plant, Georgia and South Carolina, in the-10. G o rd o n a q uife r ..............................................................................................................

2 1 11. M illers Pond aquifer ..............................................

...............................................

..22 12. U pper D ublin aquifer .............................................................................................

..23 13: Low er D ublin aquifer ...........................................................................................

..24 14. U pper M idville aquifer ..........................................................................................

..25 15. Low er M idville aquifer .........................................................................................

..26 16. Map s~howing particle-tracking results for the year 2002, Scenario A, near Vogtle Electric Generation Plant, Georgia and South Carolina ................................

27 17- 22. Maps showing simulated water-level change for Scenario B, near Vogtle Electric Generation Plant, Georgia and South Carolina, in the-17. G o rd o n a q uife r ..............................................................................................................

28 18. M illers Pond aquifer .............................................................................................

..29 o 19. U pper D ublin aquifer ............................................................................................

..30 20. Low er D ublin aquifer ...........................................................................................

..31 21. U pper M idville aquifer .........................................................................................

..32 22. Low er M idville aquife r ...................................................................

............................

33 23. Map showing particle-tracking results for the year 2002, Scenario B, study area, near Vogtle Electric Generation Plant, Georgia and South Carolina .............

34 V Figures-Continued 24-29. Maps showing simulated water-level change for Scenario C, near Vogtle Electric Generation Plant, Georgia and South Carolina, in the-24. G o rdo n aquifer .............................................................................................................

36 25. Millers Pond aquifer ....................................

37 26. U pper D ublin aquifer ...........................................................................................

..38 27. Low er D ublin aquifer ..............................................................................................

..39 28. U pper M idville aquifer ......................................................................................

... -40 29. Low er M idville aquifer ..........................................................................................

..41 30. Map showing particle-tracking results for the year 2002, Scenario C, near Vogtle Electric Generation Plant, Georgia and South Carolina ...............................

42 Tables 1. Ground-water use during 2000-2002 near Vogtle Electric Generation Plant, B urke C ounty, G eorgia ...................................................................................................

..13 2. Location and construction information for production wells at Vogtle Electric Generation Plant, Burke County, Georgia .....................................................................

14 3. Simulated and estimated values for transmissivity, Vogtle Electric Generation Plant m odel, Georgia and South Carolina .....................................................................

15 4. Simulated and estimated values for leakance, Vogtle Electric Generation Plant m odel, Georgia and South Carolina .....................................................................

15 5. Simulated pumpage by model layer for 2002 Base Case, Vogtle Electric Generation Plant model, Georgia and South Carolina .......................................................

16 6. Simulated pumpage at Vogtle Electric Generation Plant, Burke County, Georgia, for 2002 Base Case and pumping Scenarios A, B, and C ...........................

18 7. Summary of simulated time of travel for 2002 Base Case and for Scenarios A, B, and C, Vogtle Electric Generation Plant model, Georgia and South Carolina ........ 18 vi Conversion Factors and Datums Multiply By To obtain Length inch (in.) 2.54 centimeter (cm)foot (ft) 0.3048 meter (m)mile (mi) 1.609 kilometer (kin)Area square foot (fI2) 0.09290 square meter (M 2)square mile (mi 2) 2.590 square kilometer (kmi 2)Volume gallon (gal) 3.785 liter (L)million gallons (Mgal) 3,785 cubic meter (m 3)Flow rate gallon per minute (gal/min) 0.06309 liter per second (L/s)million gallons per day (Mgal/d) 0.04381 cubic meter per second (m 3/s)Vertical coordinate information is referenced to the North American Vertical Datum of 1988 (NAVD 88).Horizontal coordinate information is referenced to the North American Datum of 1983 (NAD 83)Altitude, as used in this report, refers to distance above the vertical datum.Concentrations of chemical constituents in water are given in micrograms per liter (pg/L).

Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios at Vogtle Electric Generation Plant, Burke County, Georgia By Gregory S. Cherry and John S. Clarke Abstract Introduction The source of ground water to production wells at Vogtle Electric Generation Plant (VEGP), a nuclear power plant in Burke County, Georgia, was simulated under existing (2002)and potential future pumping conditions using an existing U.S. Geological Survey (USGS) MODFLOW ground-water flow model of a 4,455-square-mile area in the Coastal Plain of Georgia and South Carolina.

Simulation results for three steady-state pumping scenarios were compared to each other and to a 2002 Base Case condition.

The pumping scenarios focused on pumping increases at VEGP resulting from pro-jected future demands and the addition of two electrical-generating reactor units. Scenarios simulated pumping increases at VEGP ranging from 1.09 to 3.42 million gallons per day (Mgal/d), with one of the scenarios simulating the elimination of 5.3 Mgal/d of pumping at the Savannah River Site (SRS), a U.S. Department of Energy facility located across the Savannah River from VEGP. The largest simulated water-level changes at VEGP were for the scenario whereby pumping at the facility was more than tripled, resulting in drawdown exceeding 4-8 feet (ft) in the aquifers screened in the production wells. For the scenario that eliminated pumping at SRS, water-level rises of as much as 4-8 ft were simulated in the same aquifers at SRS.Results of MODFLOW simulations were analyzed using the USGS particle-tracking code MODPATH to determine the source of water and associated time of travel to VEGP produc-tion wells. For each of the scenarios, most of the recharge to VEGP wells originated in an upland area near the county line between Burke and Jefferson Counties, Georgia, with none of the recharge originating on SRS or elsewhere in South Carolina.

An exception occurs for the scenario whereby pump-ing at VEGP was more than tripled. For this scenario, some of the recharge originates in an upland area in eastern Barn-well County, South Carolina.

Simulated mean time of travel from recharge areas to VEGP wells for the Base Case and the three other pumping scenarios was between about 2,700 and 3,800 years, with some variation related to changes in head gradients because of pumping changes.The Vogtle Electric Generation Plant (VEGP), near Waynesboro, Burke County, Georgia, is one of Southern Com-pany's two nuclear-generating facilities in Georgia (fig. 1 ).On August 15, 2006, Southern Nuclear Company applied to the U.S. Nuclear Regulatory Commission (NRC) for an early site permit (ESP) for an additional two reactors at the site. As part of the ESP permitting process, the NRC is charged with development of an environmental impact statement (EIS)to evaluate the effects of both construction and operation of these new reactors on the site and surrounding area. The EIS must describe the magnitude and nature of expected effects on ground water resulting from present and potential future ground-water withdrawal.

The assessment should include the area of VEGP and extend for distances great enough to cover potentially affected aquifers, including those located within the boundary of the U.S. Department of Energy, Savannah River Site (SRS), located in South Carolina across the Savan-nah River from VEGP (fig. IA, IB).The addition of two new reactors (Units 3 and 4) at VEGP will require an increase in pumping from the lower Dublin and upper and lower Midville aquifers, which cur-rently provide the water needed for reactor Units I and 2.NRC would like to evaluate the effects of additional pump-age on ground-water flow in the surrounding area. To help evaluate these effects, and improve understanding of regional ground-water flow in the area, the U.S; Geological Survey (USGS)-in cooperation with NRC-conducted a study using an existing ground-water flow model to simulate the source of ground water to VEGP production wells under current (2002)and potential future pumping conditions.

Purpose and Scope This report describes the effect of current (2002) and poten-tial future pumping on ground-water levels and flowpaths near VEGP for three pumping scenarios using an existing ground-water flow model (Clarke and West, 1998; Cherry, 2006) of a 4,455-square-mile (mi 2) area near Augusta, Ga. (fig. IA).

2 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios A.82'30'82°81030'33030'33'32'30'Base modified from U.S. Geological Survey 0 10 20 30 MILES State base maps I I I 0 10 20 30 KILOMETERS EXPLANATION Model boundary Fault--Approximately located;D, downthrown side; U, upthrown side Figure 1. (A) Study area, model grid, and model boundary, (B) location of Vogtle Electric Generation Plant production wells, Burke County, Georgia, and river and recharge cells in the Gordon aquifer (layer A2)(modified from Clarke and West, 1997) and areal and local ground-water contamination at the Savannah River Site, South Carolina (modified from Arnett and Mamatey, 1996; Cherry, 2006).

Introduction 3 B.81°50'81*40'81o30';0,00,640 A/M Are%- Sap , o~0 18. -k Id 0-chlorinated y M>d6o' volatile organics f 0 radionuclides, 60 metals 0 er~ 0Sanitary lalndfill

610) 0 oc)' nitrate !~ 0 6 0 chornae vo .0ile 000 organicstradio'

' Separations and Waste-00 ni mManagement Area, 64'%0 -N.: .chlorinated volatile orga ics, 0 6'0 100 tritium; other radionucli e n' d00o 00 0 metals, sulfate, nitrate-44 ,h r0d , a,1, B ranch(/ -i---jFault Pond 4A' ýw'1 0 R Area-Poi.t A 3/44. 4 0 a 0 radionuclides, 4'A- Pond cadmium 44,o 0 4 4.4 ZýP44" fO wel ,chlorinated volatile orais f -F " radionuclides, , pesticides, nitrate Vogtle Electric // "Z"'3 Generation Plant ,_ 'SArea-chlorinated nuclides, metals, sulfate -K'ndPrea "zz / -CX Are --o Q3 ra -- : chlorinated volatile organics, tritium, other radionuclides0

-" , metals /p siie, nirt 33010.33*00'Base modified from U.S. Geological Survey 1:24,000-scale digital data EXPLANATION Contaminated sites SAreal contamination 0 Localized contamination 0 31Z 0 Fault-Approximately located;u D, downthrown side; U, upthrown side 0 1 2 3 4 5 MILES I III .j 0 1 2 3 4 5KILOMETERS Simulated river cell V Simulated recharge cell Plant Vogtle production well 003 and identification number Figure 1. .(A) Study area, model grid, and model boundary, (B) location of Vogtle Electric Generation Plant production wells, Burke County, Georgia, and river and recharge cells in the Gordon aquifer (layer A2) (modified from Clarke and West, 1997) and areal and local ground-water contamination at the Savannah River Site, South Carolina (modified from Arnett and Mamatey, 1996; Cherry, 2006).-Continued 4 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Simulated water levels were compared to a Base Case repre-senting 2002 pumping rates throughout the model area.A particle-tracking analysis was conducted for each scenario to determine the source of water for VEGP production wells.For each scenario, the pumping distribution, simulated water-level changes, 'and ground-water flowpaths are described relative to the Base Case. Limitations of the model analysis also are provided.Description of Study Area The study area is in the northern part of the southeastern Coastal Plain physiographic province (Clark and Zisa, 1976)of Georgia and South Carolina (fig. IA). The Fall Line marks the boundary between Coastal Plain sediments and crystalline rocks of the Piedmont physiographic province and forms the approximate northern limit of the study area. Topographic relief generally is greatest near the Fall Line, becoming pro-gressively less toward the south and east. Altitudes range from as high as 650 feet (ft) near the Fall Line to less than 100 ft in the southern part of the study area and in the valleys of major streams, such as the Savannah River or Brier Creek. A steep bluff is present along the western bank of the Savannah River in southern Richmond County and most of Burke County, Ga.Relief along the Savannah River bluff is as much as 160 ft from the top of the bluff to the valley floor.The Coastal Plain province is well to moderately dis-sected by streams and has a well-developed dendritic drainage pattern. Streams that flow over the relatively soft Coastal Plain sediments develop wider floodplains and greater meander frequency than streams that flow over hard crystalline rocks of the Piedmont (Clark and Zisa, 1976). Most of the floodplain near the principal rivers, such as the Savannah River, has a wide expanse of swamp bordering both sides of the channel.Forestry and agriculture are the predominant land uses in the study area; major crops are pine timber, cotton, and soybeans.

Kaolin is mined in parts of the study area. The largest cities in the study area are Augusta, Ga., with a population of 194,950 during 2000; and Aiken, S.C., with a population of 25,460 during 2000 (U.S. Bureau of the Census, accessed February 3, 2003, at http://www.census.gov/).

Savannah River Site and Ground-Water Contamination The SRS encompasses a 310-mi 2 area across the Savannah River from VEGP in parts of Aiken, Barnwell, and Allendale Counties, S.C. (fig. I A, I B). The facility has manufactured nuclear materials for national defense since the early 1950s. A variety of hazardous materials--incltding radionuclides, vola-tile organic compounds, and heavy metals-are either disposed of or stored at several locations at SRS. Contamination of ground water has been detected at several locations on the site (fig. I B) with contamination mostly limited to the Upper Three Runs aquifer. The potential for movement of contaminated water into aquifers beneath the Upper Three Runs aquifer at SRSis dependent on the hydraulic conductivity of interven-ing confining units and the magnitude of downward hydraulic gradient (see Ground-Water Flow section).The only documented occurrence of ground-water contaminants into aquifers beneath the Upper Three Runs that the authors are aware of occurred in the vicinity of the A/M Area on SRS (see location, fig. IB). In this area, Chris-tensen and Gordon (1983) reported volatile organic com-pounds were detected at depths as great as 480 ft, affecting water-bearing zones in the "Congaree Formation" (Gordon aquifer) and "Tuscaloosa Formation" (lower Dublin aquifer).Contamination in the Congaree Formation was attributed to the discontinuity of the "green clay" that forms the confining unit beneath the Upper Three Runs aquifer. Contamination of the Tuscaloosa Formation was attributed to a poor grout seal in well 53A, which enabled downward migration of contami-nated ground water into deeper zones. During the first half of 2007, Washington Savannah River Company (2007) reported concentrations of trichloroethylene in the A/M Area of as high as 18,000 micrograms per liter (ýtg/L) in the composite "Lost Lake aquifer zone" (Gordon aquifer) and as high as 3,200 iAg/L in the "Crouch Branch aquifer" (upper and lower Dublin aquifers).

Contaminants in both aquifers Occur in southwest-trending plumes with concentrations exceeding 5 Vg/L across a distance of nearly 21,000 ft in the Lost Lake aquifer zone, and nearly 15,000 ft in the Crouch Branch.aquifer.

Contaminants in the source areas are being removed using recovery wells and above-ground air strippers (Washington Savannah River Company, 2007).Climate and Runoff A relatively mild' climate with warm, humid summers and mild winters characterizes the study area. Precipitation is highest during the winter months when continental storm fronts move through the region and during July and August when afternoon thunderstorms caused by daytime heating are common. Average annual precipitation in the study area for the period 1969-98, ranged from about 46 inches in Burke County, Ga., to greater than 52 inches in central Aiken County, S.C. (Southeast Regional Climate Center, accessed Febru-ary 11, 2004, at http://www.dnrsc.gov/climate/sercc/).

The Savannah River is the major surface-water feature in the study area and is the boundary between Georgia and South Carolina.

The river drains an area of about 10,580 mi 2 (1,140 mi 2 in the study area) and empties into the Atlantic Ocean near Savannah, Ga. During 1941-70, the average annual runoff in Georgia ranged from less than 0.9 cubic feet per second per square mile [(ft 3/s)/mi 2] of drainage area in southern Screven, Jenkins, Burke, and Jefferson Counties, and in northern Richmond and Jefferson Counties, to greater than 1.1 (ft 3/s)/mi 2 in eastern Richmond and Burke Counties (Faye and Mayer, 1990).

Introduction 5 North-west HYDROGEOLOGIC REPRESENTATION South-east EXPLANATION Hydrogeologic representation Stream alluviumAquifer Confining unit ICU)Basal confining unit and basement rock PE Post-Eocene Model representation

[ Source-sink specified head Active model cells Cells receiving recharge River cells Aquifer pinchout-Represented as zone of low transmissivity Lateral specified head boundary No-flow boundary Confining unit-Represented as vertical conductance term 1 Confining unit pinchout-Represented as zone of high vertical conductance Riverbed conductance At, C1 Modellayer Figure 2. Schematic diagram showing hydrogeologic framework, model layers, and boundary conditions for the Vogtle Electric Generation Plant area, South Carolina (ground-water model modified from Clarke and West, 1998; Cherry, 2006).Hydrogeologic Setting Coastal Plain sedimentary strata in the study area consist of layers of sand, clay, and limestone, which range in age from Upper Cretaceous through post-Eocene (fig. 2). The Fall Line (fig. I A) marks the approximate inner margin of Coastal Plain sediments.

The strata dip and progressively thicken fr'om the Fall Line to the southeast, with an estimated thickness of 2,700 ft in the southern part of the study area (Wait and Davis, 1986). The strata crop out in discontinuous belts that generally are parallel to the Fall Line. The sedimentary sequence uncon-formably overlies Paleozoic igneous and metamorphic rocks, and consolidated Mesozoic red beds (Chowns and Williams, 1983)Coastal Plain sediments comprise three principal aquifer systems near VEGP. In descending order, these aquifer systems are (1) the Floridan aquifer system, originally defined by Miller (1986) and later redefined by Aadland and others (1995)-comprised largely of Eocene calcareous sand and limestone; (2) the Dublin aquifer system (Clarke and others, 1985)-com-prised of Paleocene and Late Cretaceous sand; and (3) the Mid-ville aquifer systerm (Clarke and others, 1985)--comprised of Late Cretaceous sand. Although this subdivision was suitable for most regional-scale hydrogeologic studies, greater subdivi-sion of units was required to define vertical hydraulic heteroge-neity for detailed investigations of ground-water flow near the Savannah River. Accordingly, the three aquifer systems were divided into seven aquifers (fig. 2): " the Floridan aquifer system was subdivided into the Upper Three Runs aquifer and the Gordon aquifer (Aadland and others, 1995);* the Dublin aquifer system was subdivided into the Millers Pond aquifer, the upper Dublin aquifer and the lower Dublin aquifer (Falls and others, 1997); and" the Midville aquifer system was subdivided into the upper Midville aquifer and the lower Midville aquifer (Falls and others, 1997).The aquifers are separated and confined by layers of clay and silt, which become progressively sandy and discontinuous in updip areas. The aquifer systems coalesce where the confin-ing units become sandy. See Falls and others (1997) for a complete description of geologic and hydrogeologic units in the study area.

6 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Structural Features Major structural features in the study area (fig. IA) include the Belair Fault (Prowell and O'Connor, 1978) and the Pen Branch Fault (Price and others, 1991). The Belair Fault is a northeast-trending, high-angle reverse fault that dips to the southeast and has a maximum vertical displacement of 100 ft at the base of Coastal Plain strata (Prowell and O'Connor, 1978).The Pen Branch Fault is a northeast-trending, high-angle reverse fault that dips to the southeast.

On SRS, the fault consists of a 1.8-mile (mi)-wide zone of subparallel faults and some fault splays (Snipes and others, 1993). The fault is downthrown on the northwestern side, and maximum displacement ranges from 100 ft at the base of Coastal Plain strata to 30 ft at the top of the Eocene Dry Branch Formation (Price and others, 1991).Seismic data from Burke County, Ga., suggest the Pen Branch Fault zone is about 0.86-mi wide and includes "short fractures" or "stress-release faults" (Summerour and others, 1998). These features appear to cut confining units overlying the Millers Pond aquifer, upper and lower Dublin aquifers, and upper and lower Midville aquifers; however, it is unclear whether they cut into the confining unit overlying the Gordon aquifer (Summerour and others, 1998). The Pen Branch Fault zone includes the area of VEGP (fig. IA, 1B). The effects of the Pen Branch Fault on sediment deposition and hydraulic properties of hydrogeologic units is unknown; however, there may be some local effects on the hydrologic system (see Effect of Pen Branch Fault section).Hydrogeologic Characteristics of the Savannah River Alluvial Valley The hydrogeologic characteristics of the Savannah River alluvial valley (fig. 3) greatly influence the configuration of potentiometric surfaces, ground-water flow directions, and stream-aquifer relations.

To determine the effect of palen-river channel incision on hydrogeologic units, a map was constructed that shows the subsurface extent of hydrogeologic units beneath the mantle of alluvial'deposits in the Savannah River floodplain (fig. 3). The map indicates that each of the seven aquifers was incised by the paleo-Savannah River channel and covered with an infill of permeable alluvium, allowing direct hydraulic connection of the aquifers and river along parts of the river's reach. The lateral extent of the paleo-river channel incision corresponds to the width of the Savannah River alluvial valley and includes the modern-day alluvial bottom and ter-races as mapped by Prowell (1994). The width of the alluvial valley ranges from a minimum of about 0.5 mi near the Fall Line to about 7 mi near the Richmond-Burke County line.Summerour and others (1998) reported the possible pres-ence of several "channel features" along a seismic profile col-lected by Waddell and others (1995) in eastern Burke County, outside of the present Savannah River valley. These features are believed to cover an area about 3,000 ft wide, extending to about 500 ft deep; however, their presence was not confirmed by drilling.

Summerour and others (1998) suggested that "the channels, if real, could provide a potential pathway for the movement of groundwater (and pollutants) between aquifers." Because of the uncertainty of these features, they were not included in the ground-water model developed by Clarke and West (1998). If additional data become available to confirm the presence of these features, it may be desirable to incorpo-rate them into future ground-water models of the area.Ground-Water Flow The ground-water flow system near VEGP is gener-ally considered to be in a state of equilibrium (steady state), whereby rates of aquifer recharge and discharge are about equal, and there is an insignificant loss of water from aquifer storage (Clarke and West, 1998). Recharge enters the ground-water system in upland areas and moves downgradient toward points of discharge (fig. 4). Much of the recharge water is discharged from the shallow flow system into tributaries of the Savannah River. A smaller percentage of recharge infiltrates through clayey confining units and enters the deeper interme-diate and regional ground-water flow systems (fig. 4).Topography plays an important role in defining the posi-tion of areas of potential downward and upward flow. Clarke and West (1997) present maps showing head differences and the potential for flow between adjacent units in the study area.In interstream areas throughout most of the study area, the potential for flow between the Gordon aquifer and overlying Upper Three Runs aquifer is downward, indicating possible recharge by ground-water leakage from the Upper Three Runs aquifer to the Gordon aquifer (Clarke and West, 1997). Con-versely, in stream valleys and throughout much of the southern part of the study area, the potential for ground-water flow is upward, indicating possible discharge from the Gordon aquifer to the Upper Three Runs aquifer.The Savannah River serves as the major hydrologic drain in the VEGP area, with its floodplain considered to represent the same or nearly the same hydrologic condition as the river (Clarke and West, 1997). Each of the seven aquifers was incised by the paleo-Saannah River channel and covered with an infill of permeable alluvium (fig. 3), allowing direct hydraulic interconnection between the aquifers and the river (Clarke and West, 1997). This hydraulic connection allows water in confined aquifers to discharge into the river-by way of the alluvium-and may induce ground water to flow updip.Hydraulic connection between confined aquifers and the Savannah River can be inferred from potentiometric-surface maps (figs. 5-8) that show ground-water discharge areas along the Savannah River valley as lows or depressions in the poten-tiometric surface (Clarke and West, 1997). Ground water flows toward the depressions from all directions; however, down-stream from the depressions, the influence of the river on the aquifers becomes progressively diminished, and ground water resumes the regional gradient toward the southeast.

In these downstream areas, a ground-water divide or "saddle" (Siple, 1960, 1967) in the potentiometric surface is perpendicular to the river and separates upstream from downstream ground-water flow.

Introduction 7 82'81o30, 33o30'33'Base modified from U.S. Geological Survey State base maps Em EE Upper Three Runs aquifer Gordon confining unit Gordon aquifer Millers Pond confining unit Millers Pond aquifer 7 EXPLANATION Hydrogeologic units Upper Dublin confining unit Upper Dublin aquifer Lower Dublin confining unit Lower Dublin aquifer Upper Midville confining unit l Upper Midville aquifer Lower Midville confining unit Lower Midville aquifer Pre-Cretaceous basement rock Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side u Figure 3. Subsurface extent of hydrogeologic units beneath the Savannah River alluvial valley, South Carolina and Georgia (modified from Clarke and West, 1997).

8 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios South Carolina,iN .:ii Regional.. ... ..4 > .............. ....1/4. ....................Water table........................................................................

LOCAL FLOW SYSTEM.. ..... ... .. .. .. .......................I .......INTERMEDIATE FLOW SYSTEM.................

.......................................

..................... ..............................................................

REGIONAL FLOW SYSTEM'....

....-..--) ... ...........................L NOT TO SCALE EXPLANATION Aquifer Unsaturated zone Saturated zone M Alluvium Confining unit[ Pre-Cretaceous basement rock-*Direction of ground-water flow-Queried where unknown Figure 4. Conceptualized hydrogeologic framework and related ground-water flow near Vogtle Electric Generation Plant, Georgia and South Carolina (modified from Atkins and others, 1996).

Introduction 9 81050'81°40'81°30, 33020.33010'33o00'Base modified from U.S. Geological Survey 1:24,000-scale digital data 0 1 2 3 4 5 MILES 0 I I I2 I 4 I L I IO I 0 1 2 3 4 5 KILOMETERS EXPLANATION: Active area of source-sink layer Al-140 -Potentiometric contour- Shows altitude at which.water level would have stood in tightly cased wells in the Upper Three Runs aquifer during September 2002.Contour interval 40 feet. Datum is NAVD 88 0 U Direction of ground-water flow Pen Branch Fault-Approximately located;D, downthrown side; U, upthrown side Observation well Figure 5. Potentiometric surface for the Upper Three Runs aquifer, near Vogtle Electric Generation Plant, Georgia and South Car'olina, September 2002 (modified from Cherry, 2003).

10 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios 81050' 81040' 81 o30'240 33*2011 Par Pond 33o10'c4b 31Z074 31Z015 /"/31Z0730" 33o'0 31Z076/0S Base modified from U.S. Geological Survey 0 1 2 3 4 5 MILES 1:24,000-scale digital data [ I In I I II I I I EXPLANATION 0 1 2 3 4 5 KILOMETERS Active area of source-sink layer Al-140 -Potentiometric contour-Shows altitude at which Map area water level would have stood in tightly cased wells in the Gordon aquifer during September 2002.Contour interval 20 feet. Datum is NAVD 88 AledlDirection of ground-wtrfo

,', ,,.... 0 Pen Branch Fault-Approximately located;U, D, downthrown side; U, upthrown side crven Production well-Screened in the Gordon aquifer 31Z076 Observation well-Near Pen Branch Fault in which 0 simulated water level was 4-26 feet higher than observed.

Number is well identification Figure 6. Potentiometric surface for the Gordon aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina, September 2002 (modified from Cherry, 2003).

Introduction 11 81 050' 81 *40' 81 30'33°20'-10J j, Base modified from U.S. Geological Survey 0 1 2 3 4 5 MILES 1:24,000-scale digital data IlIII I I ll.I 0 1 2 3 4 5 KILOMETERS Aiken 7; Active area of source-sink layer Al\ ,, Ma area -140 -Potentiometric contour--Shows altitude at which; -water level would have stood in tightly cased wellsin the Dublin aquifer system during September 2002.Contour interval 20 feet. Datum is NAVD 88 Bure F-. Direction of ground-water flowU Pen Branch Fault--Approximately located;D, downthrown side; U, upthrown side* Production well--Screened in the upper and lower Dublin aquifer and Millers Pond aquifer Figure 1. Potentiometric surface for the Dublin aquifer system, near Vogtle Electric Generation PI'ant, Georgia and South Carolina, September 2002 (modified from Cherry, 2003).

12 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios 81W 50' 81'40' 81'30', :ý33*20" : 33°00' 0 Base modified from U.S. Geological Survey' 0 1 2 3 4 5 MILES 1:24,000-scale digital date I IiI 0 1 2 3 4 5 KILOMETERS Active area of source-sink layer Al, Map are -140 -- Potentiometric contour--Shows altitude at whichlevel would have stood in tightly cased wells-\Alenda Contour interval 20 feet. Datum is NAVD 88 Burke --Direction ot ground-water flow JekUs1 Pen Branch Fault--Approximately located;\,,..crve U 0, downthrown side; U, upthrown side@

well--Screened in the upper and lower Midville aquifer Figure 8. Potentiometric surface for the Midville aquifer system, near Vogtle Electric Generation Plant, Georgia and South Carolina, September 2002 (modified from Cherry, 2003).

Introduction 13 Because flow directions derived from potentiometric-surface maps do not account for the vertical component of flow, Clarke and West (1998) applied the USGS particle-tracking code MODPATH (Pollock, 1994) to characterize three-dimensional ground-water flow near the Savannah River.This technique is applied in this report to simulate ground-water flowpaths from the production wells at VEGP to their recharge areas.Trans-River Flow Trans-riverflow is a term that describes a condition whereby ground water originating on one side of a river migrates to the other side of the river through confined aquifers that underlie the river. Although some ground water could discharge into the river floodplain or alluvium on the opposite side of the river from its point of origin, this flow likely would return to the river. Return flow would occur because a slight hydraulic gradient exists toward the river along the floodplain.

Flow lines on potentiometric-surface maps of the confined Gordon aquifer and Dublin and Midville aquifer systems (figs. 6-8) suggest possible occurrences of trans-river flow for a short distance into Georgia prior to discharge into the Savannah River (Clarke and West, 1997).Flow lines on the map for the Upper Three Runs aquifer, however, do not indicate trans-river flow (fig. 5).Effect of Pen Branch Fault The Pen Branch Fault may have a local effect on ground-water flow. In the central part of the SRS, water levels in the Gordon aquifer near the P-19 well cluster site are anomalously high, producing a mound in the potentiometric surface (fig. 6).The high water level in this area may be the result of the offset of the Pen Branch Fault, whereby sediments of the Gordon aquifer are juxtaposed against sediments of the Upper Three Runs aquifer (Aadland and others, 1991). Because the units are in hydraulic connection near the fault, water levels and water chemistry of the Gordon aquifer are similar to those of the Upper Three Runs aquifer (Clarke and West, 1997).Ground-Water Use Ground-water use in the study area during 2000-2002 (W.J. Stringfield, U.S. Geological Survey, written commun., 2002; Fanning, 2003) was about 117 million gallons per day (Mgal/d) (table 1), most of which was for irrigation (54 per-cent) and public supply (26 percent).

In Georgia, most of the ground water used for irrigation is withdrawn from the Upper Three Runs aquifer in Jenkins County and southern Screven County, and from the Upper Three Runs and Gordon aquifers in Jefferson, Burke, and northern Screven Counties.

In South Carolina, most irrigation wells in Barnwell and Allendale Counties pump water from the Upper Three Runs and Gordon aquifers.

Ground-water use for public supply and industrial and mining purposes is mainly from the Dublin and Midville aquifer systems in both States.Table 1. Ground-water use during 2000-2002 near Vogtle Electric Generation Plant, Burke County, Georgia.[Modified from Cherry, 2006]Georgia Burke 3.87 21.23 0.15 Jefferson Jenkins Richmond Screven South Carolina Ak-en-t , Allendale

-Total-Georgia Total-South Carolina 1.84 6.92 3.82 0.54 3.94 0.01 14.88 5.22 2.87 1.15 15.62 1.82 4.82 .,07.98 .K6-06~1.20 5.62 <2.50.2.73 ~3i73 , : 0.41 22.28 52.93 8.67 8.75 10.33 8.97 0.90 0.03 0.64 0.03 0.33 0.02 0.22 0.02 0.74 0.03-.0.80 00o;:;, 0.27 0 :r .;000.: 2.83 0.13 1.70 0.00 0.78 26.96 0.00 13.25 0.00 4.84 0.00 23.21 0.00 19.36 0.00 12.66, 0.00 9.59 0.00 7.50ý0.78 87.62 0.00 29.75 Total-eight counties 31.03 63.26 17.64 4.53 0.13 0.78 117.37'See figure ]A for location Data sources: County totals for Georgia are from Fanning (1997, 2003) and Pierce and others (1982); total water-use data for South Carolina front Lonon and others (1983), and W.J. Stringfield (U.S. Geological Survey, written commun., 2002); site-specific data for irrigation wells located in Georgia from J.L. Fanning (U.S. Geological Survey, written commun., 2003) and V. P. Trent (Georgia Geologic Survey, written commun., 2003); and site-specific data for permitted wells located in South Carolina from Paul Bristol and Peter Stone (South Carolina Department of Health and Environmental Control, written commun., 2003).

14 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios At VEGP, three wells are each screened in the lower Dublin and upper and lower Midville aquifers, and are used to supply water to operating reactor Units I and 2 (wells 31Z002, 31Z003, and 31Z080; fig. IB; table 2). Screened intervals of wells at VEGP were obtained from drillers records and entered into the USGS National Water Information System (http://waterdata.usgs.gov/ga/nwis/inventory).

To determine the aquifer supplying water to each screened interval, altitudes of screened intervals were compared to maps showing the altitude of the tops of hydrogeologic units using a Geographic Information System (Harrelson and others, 1997).During 2002, the wells at VEGP supplied an average 724 gal/min (1.04 Mgal/d). The addition of two additional reactor units is projected to result in an increase in ground-water pumping of 1.09 Mgal/d (Mark Notich, U.S. Nuclear Regulatory Commission, written commun., April 10, 2007).If ground-water pumping during the startup of these reactors is similar to that during the startup of the original two reactors during 1988, then the initial increase in pumping could be as high as 3.42 Mgal/d.At SRS, estimated pumpage was 5.30 Mgal/d during 2002 (Cherry, 2006). A variety of multiaquifer wells com-pleted in the Gordon aquifer and Dublin and Midville aquifer systems provide water supply at SRS.Simulation of Ground-Water Flow The model used in this study is described in detail in Clarke and West (1998) and Cherry (2006); only a brief description is included herein. Clarke and West (1998)simulated predevelopment and 1987-92 conditions using the MODFLOW finite-difference simulator (McDonald and Harbaugh, 1988). Cherry (2006) updated the model to simu-late 2002 hydrologic conditions using the MODFLOW-2000 simulator (Harbaugh and others, 2000). Both studies simulated steady-state conditions for each time period. Steady-state simulations were deemed appropriate because of the minimal observed changes in hydraulic head or ground-water discharge to streams from predevelopment (pre- 1953) t6-1987-92 (Clarke and West, 1997). These minor fluctuations are an indi-cation that the ground-water system generally was in a state of equilibrium and any contributions from aquifer storage were minor. These assumptions are believed to remain valid for the study area for this investigation.

The model encompasses an area of about 4,455 mi 2 (fig. 1) and includes seven aquifers and seven confining units.These units crop out near the Fall Line and generally dip and thicken to the southeast.

Aquifer units are, in descending order (fig. 2): Table 2. Location and construction information for production wells at Vogtle Electric Generation Plant, Burke County, Georgia.[USGS, U.S. Geological Survey; 0, degree; ', minute; ", second; Aquifer: LD, lower Dublin; UM, upper Midville; LM, lower Midville]Land-surface Screened interval Pouto KUSGS well Well Latitde Lonitde lttue Year (feet below Aqie capaodcityo

'identification' number (ftd Lnnstructeiu dland surdace) Aqif allone Top Bottornm minute)3l1tZ002 TW-1 33'08'28

81'45'42" 219 1972 505 535 LD 1,200 555 585 LD 695 705 UM 730 750 UM 815 850 LM 31 Z080 IMU-2A 33'08'39" 81046'00" 235 1983 480 LD >2, 1 12 630< L650 690 ~79',/ 1M 31Z003 2 MU-1 33008'47" 81045'37" 197 1977 437 462 LD 3,334 468 483 LD 498 512 LD 536 546 LD 550 572 LD 676 696 UM 720 732 LM 788 820 LM Igpp JRt mnr In-r~ntinn 2 Now called MU-5 Simulation of Ground-Water Flow 15-the unconfined Upper Three Runs aquifer modeled as a source-sink layer (specified head layer A 1);-the confined Gordon aquifer (layer A2);-the Dublin aquifer system consisting of the Millers Pond aquifer (layer A3), upper Dublin aquifer (layer A4), and lower Dublin aquifer (layer A5); and-the Midville aquifer system consisting of the upper Midville aquifer (layer A6) and lower Midville aquifer (layer A7).The thickness, extent, and other hydraulic properties of these units, as well as the model development process are described in detail in Clarke and West (1998). A schematic dia-gram showing model layers and boundary conditions is shown in figure 2. As in the original model of Clarke and West (1998), confining units are not actively simulated, but instead use verti-cal conductance to simulate leakance between layers. Esti-mated and calibrated transmissivity values are listed in table 3 aid leakance values are listed in table 4 (Clarke and West, 1998). For the MODPATH particle-tracking analysis, a uniform porosity of 30 percent was assigned to aquifer layers and 50 per-cent was assigned to confining units (Clarke and West, 1998).The finite-difference grid for the model is aligned nearly parallel to the Savannah River and to the regional dip of the hydrogeologic units, and consists of 130 rows and 102 col-umns (13,260 grid cells) with a variable grid spacing ranging in size from 0.33 mi by 0.33 mi to 2 mi by 2.5 mi (fig. IA).The model grid area encompasses about 4,455 mi5, of which about 3,250 mi is actively simulated.

Grid density is higher near the Savannah River (including the Savannah River Site and VEGP) to enable simulation of steeper head gradients (fig. 1A). Each aquifer unit is represented with one layer of grid cells in the vertical dimension.

Table 3. Simulated and estimated values for transmissivity, Vogtle Electric Generation Plant model, Georgia and South Carolina.[#, number; from Clarke and West (1998)]" ~Transmnissivity, in square foot per day'Aquifr. , Estimated basedon field data'_ _ __ _ Simulated_...........

Aquir~ r;'- numiber2-#ofvle .Siutd',o vaue 'MinimumiP7Maximium

~Mean, Minimnum, Maximum Mean'.~Gordon aquifer A2 18 180 12,200 4,500 100 24,700 10,350 Millers Pond aquifer A3 10 195 2,000 1,000 10 3,900 1,310 Upper Dublin aquifer A4 17 555 25,200 5,830 10 20,000 7,220 Lower Dublin aquifer A5 21 40 8.900 3,940 10 25,500 10,030 Upper Midville aquifer A6 15 1,300 5,430 2,760 10 12,390 6,270 Lower Midville aquifer A7 37 800 25,500 8,900 515 34,395 19,020'Determined from aquifer tests and estimated from specific-capacity data and from borehole-resistivity logs.'Mean value weighted according to cell area.Table 4. Simulated and estimated values for leakance, Vogtle Electric Generation Plant model, Georgia and South Carolina.[#, number; -, not measured; from Clarke and West (1998)]e-ckance, in feet per day per foot of confining unit thickness' Hydrogeologic

' Layer Esimate'd eaIkance2' Simulated___________

~# of values "Minimunm Maximum ,Mean Minimum >Maximum Mea'iia...} ' 5 ? ....- 4, .44" Gordon confining unit Cl 6 4.7 x 10-1 1.2 x 10-1 2.1 x 10-1 9.0 x 10-8 1.3 x 10-1 1.7 x 10'Millers Pond confining unit C2 -3.9 x 10-1 8.7 x 10- 1.9 x 10-2 Upper Dublin confining unit C3 9 1.8x 10-6 1.6x 10-3 3.6x 10' 1.2 x 10-1 7.3 x 10-1 1.2 x 10-l Lower Dublin confining unit C4 1 2.4 x 10-5 2.4 x 10-5 2.4 x 10-1 3.0 x 10-7 6.5 x 10-3 6.6 x 10-3 Upper Midville confining unit C5 11 6.7 x 10-7 3.4 x 101 7.6 x 10-5 2.1 x 10-7 1.0 x 10-' 9.7 x 10-V Lower Midville confining unit C6 1 1.0 x 1075 1.0 x 10-1 1.0 x 10-5 7.7 x 10-1 3.6 x 10-' 9.0 x 10-3 Includes low permeability layers within aquifer layers.2 Estimated by dividing the vertical hydraulic conductivity unit by the thickness of the confining unit.'Mean value weighted according to cell area.

16 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Lateral model boundaries are a combination of no-flow and specified head for layers A2-A7. For all layers, the south-eastern boundary is simulated as a specified-head condition.

For layers A2-A3, the southwestern boundary is simulated as no-flow, corresponding to the position of a ground-water divide. Parts of the eastern boundary for layers A2 and A3 are simulated as specified head and no-flow. For layers A4-A7, most of the western boundary is simulated as no-flow, corre-sponding to the position of a ground-water divide. The eastern boundary for layers A4-A7 is simulated mostly as a specified-head condition.

Specified heads for each layer in the model are based on potentiometric surface maps for September 2002 (figs. 5-8) and generally are lower than in the original Clarke and West (1998) model to reflect effects of the 1998-2002 drought (Cherry, 2006).The bottom boundary of the model is no-flow, whereas the top boundary represented by layer A l is set as a source-sink specified-head condition with controlling specified heads based on water levels from the 2002 potentiometric-surface map of the Upper Three Runs aquifer (Cherry, 2003). Flow in the deeper active layers (A2-A7) of the model is simulated through a combination of active cells, specified head cells, recharge cells, and river cells.Most recharge to the simulated ground-water system was provided by leakage from layer A 1, with a comparatively smaller amount derived from recharge cells in layers A2-A7.Total simulated recharge is about 930 Mgal/d of which 777 Mgal/d were derived by leakage from layer A 1, and 153 Mgal/d were derived from recharge assigned to outcrop areas of hydrogeologic units (Cherry, 2006).Average annual pumpage for 2002 was assigned to model cells based on site-specific and county-aggregate data (table 1). Site-specific data are available for public supply, thermoelectric, industrial, and mining use and are assigned to known well locations.

County' aggregate agricultural pumping data were equally divided and assigned to known agricultural well locations Domestic and commercial and livestock use were not simulated by the model because these uses accounted for less than 4 percent of total study area pumpage during 2002 (table 1) with most of the withdrawal derived from shal-low wells completed in the Upper Three Runs aquifer.Where multi-aquifer wells are completed in several aquifer layers (such as at VEGP and SRS), pumpage was evenly proportioned to the various screened intervals in each well. Of the total study area ground-water use dur-.ling 2000-2002 of 117 Mgal/d, 67.2 Mgal/d were simulated from active layers (table 5) A2 (Gordon aquifer), A3 (Millers Pond aquifer), A4 (upper Dublin aquifer), A5 (lower Dublin aquifer), A6 (upper Midville aquifer), and A7 (lower Midville aquifer).

The remaining 49.8 Mgal/d were from the Upper Three Runs aquifer (layer AI), which is not actively simu-lated. The influence of pumpage from the Upper Three Runs aquifer on the overall flow system during 2002 is simulated by changing the head in that layer based on water-level measure-ments during September 2002 (fig. 5).Table 5. Simulated pumpage by model layer for 2002 Base Case, Vogtle Electric Generation Plant model, Georgia and South Carolina.[Modified from Cherry, 2006]Auir Model Year 20 upgi layer mi iion gallons perday, Gordon A2 10.7 Millers Pond A3 7.3 Upper Dublin A4 5.4 Lower Dublin A5 14.6 Upper Midville A6 9.8 Lower Midville A7 19.4 All layers 67.2* The calibrated model used for this study showed a reasonable fit to simulated water levels. Water-level residuals represent the difference between simulated and observed water levels, with positive values indicating that simulated values were greater than observed values. For the 1987-92 simula-tion, the model was calibrated using the average observed water levels at 313 model cells, with a mean of residuals of 0.8 ft and a root mean square (RMS) of the residuals of 10.6 ft (Clarke and West, 1998). For the 2002 simulation, model calibration was evaluated based on observations at 172 wells during September 2002, with a mean of residuals of 2.8 ft and a RMS of the residuals of 8.0 ft (Cherry, 2006).Pen Branch Fault The Pen Branch Fault may locally affect ground-water flow in the study area (see Effect of Pen Branch Fault section).Although hydraulic characteristics of the fault are unknown, the possible effects of the fault are incorporated into the model as follows: (1) by variations in depth, thickness, and hydraulic properties of model layers near the fault; and (2) by incor-poration of river cells where incision of the Gordon aquifer (layer A2) is believed to occur along the southern side of the fault in the Savannah River alluvial valley (fig. IA).Hydraulic properties of the upper and lower Dublin aquifers (model layers A4 and A5) were adjusted along the southern side of the fault during model calibration (Clarke and West, 1998). A zone of high transmissivity-greater than 15,000 feet squared per day (ft 2/d) and extending'as much as 6 mi south of the fault-was required in the two layers to achieve calibration of the model. Although there are no field data to confirm this zone of-higher transmissivity, it is possible that such a zone exists based on calibration results. Construc-tion of test wells and aquifer testing in this area would be required to confirm the presence of a high transmissivity zone.

Simulation of Ground-Water Flow 17 Maps showing the altitude of the top of hydrogeologic units (Falls and others, 1997) indicate that uplift along the southern side of the Pen Branch Fault resulted in shallower depths of units compared to equivalent units north of the fault.Near the Pen Branch Fault and the P-19 well cluster site on SRS, simulated head values for the Gordon aquifer (layer A2)during 1987-92 are considerably lower than observed values, with a residual of -81.1 ft (Clarke and West, 1998). Clarke and West (1997) reported that the anomalously high observed head in the Gordon aquifer in this area may be the result of (1) a high degree of aquifer interconnection between the Upper Three Runs and Gordon aquifers due to the Pen Branch Fault (Aadland and others, 1991), or (2) the possibility that the water-level measurement in the Gordon aquifer at the P-I19 well cluster site may not be representative of the head in layer A2 because of problems with well construction or mea-surement error. Because the reason for the high water level in the Gordon aquifer was not definitively established, the earlier study (Clarke and West- 1998) did not adjust model param-eters to attempt to match water levels in the Gordon aquifer at the P-19 well cluster. Simulation of higher head in the Gordon aquifer in this area would require increasing the leakance of the Gordon confining unit (layer C1) to enable greater connec-tion between the Upper Three Runs and Gordon aquifers.In the Savannah River valley, uplift along the southern*side of the Pen Branch Fault and erosion by the paleo-Savannah River appears to have resulted in exposure of the Gordon aquifer (layer A2) in a local area (fig. 3). This local exposure is simulated in the model as river cells in the Gordon aquifer that enables a higher degree of connection between the Gordon aquifer and the Savannah River (fig. IB).Despite this adjustment, simulated water levels in the Gordon aquifer (layer A2) near this feature during 2002 generally are high in the model, ranging from 4 to 26 ft higher than observed levels (wells 31Z015, 31Z073, 31 Z074, 31Z076, and 31Z077; fig. 6).Another area where uplift on the southern side of the fault is represented by the model occurs between Pen Branch and Four Mile Branch on SRS (fig. IB). In this upland area adja-cent to the Savannah River. Valley, units overlying the Gordon aquifer appear to have been eroded away and the aquifer is near land surface. Here, the Gordon aquifer is simulated using recharge cells, which enable direct infiltration of precipitation into the aquifer.Several hypothesized channel features (Summerour and others, 1998) along a seismic profile near the Pen Branch Fault in eastern Burke County, Ga., were ndt incorporated into the ground-water model because their existence was not con-firmed by test drilling.

According to Summerour and others (1998), these features potentially could affect a zone 3,000 ft wide and 500 ft deep, and provide a potential pathway for movement of ground water between aquifers.

If such a feature were simulated, high vertical hydraulic conductivity would be assigned to confining units overlying layers A2-A7, and would facilitate movement of water between the zones. This modification could result in reducing the simulated head in the Gordon aquifer by allowing water to discharge from the unit and would reduce the aforementioned difference between observed and simulated head. Because the presence, depth, and areal extent of these features, are unknown, they were not simulated by the current model.Ground-Water Pumping Scenarios The updated and calibrated model (Cherry, 2006) was used to simulate the effect of current and potential future pumping on ground-water levels and flowpaths near VEGP for a Base Case and three pumping-scenarios (table 6). The Base Case represents 2002 pumping rates throughout the model area (Cherry, 2006). The three scenarios were designed to simulate steady-state water levels resulting from (1) pumping increases at VEGP with pumping elsewhere in the study area held at 2002 rates (Scenarios A and C), and (2) the effects of increased pumping at VEGP combined with a shutdown of pumping at the SRS (Scenario B).Steady-state conditions in response to pumping changes are believed to be reached rapidly in the study area. Clarke and West (1998) indicated that for each of six stress periods during 1953-92, "heads showed an almost instantaneous stabiliza-tion, suggesting that the prevalence of steady-state conditions were achieved immediately following a change in pumpage." For each scenario, the pumping distribution, simulated water-level changes, and ground-water flowpaths are described rela-tive to the year 2002 Base Case.A particle-tracking analysis was conducted for the Base Case and for each scenario to determine the source of water withdrawn from the VEGP production wells. The USGS particle-tracking code MODPATH (Pollock, 1994) was used to generate advective water-particle pathlines and their associated time of travel based on the MODFLOW simula-tions. MODPATH was used to compute three-dimensional flow directions and time of travel using imaginary particles in a backtracking mode from the production wells at VEGP toward recharge areas in a map perspective.

Generally, the greater the number of particles applied vertically and horizon-tally in a model cell, the more accurate the definition of flow-paths for a given model layer. For this study, particles were placed at the center of each of the three grid cells containing a VEGP production well at increments representing 10 percent of the aquifer thickness (10 total particles per aquifer layer in each of the three grid cells). Particles were placed in the lower Dublin (layer A5), and upper (layer A6) and lower Midville (layer A7) aquifers, which provide water to the VEGP pro-duction wells. To avoid clutter and simplify display of particle flowpaths on maps, the number of particles displayed on the figure was reduced from 10 to 5. Simulated time of travel for all particles (10 per model cell) is summarized for the Base Case and for Scenarios A-C in table 7.

18 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Table 6. Simulated pumpage at Vogtle Electric Generation Plant, Burke County, Georgia, for 2002 Base Case and pumping Scenarios A, B, and C.[gal/min, gallons per minute; Mgal/d, million gallons per day],Pumping rate~Scenario Remarks Base Case 2002 724 1.04 Current conditions for existing reactor units A 1,482 2.13 Additional pumping capacity of new reactor units at average projected withdrawal rates B C 1,482 3,099 2.13 Additional pumping capacity of new reactor units at average projected withdrawal rates and elimination of 5.3 Mgal/d purnpage at Savannah River Site 4.46 Scenario represents a higher rate of withdrawal for the proposed new reactor units during their startup period (3.42 Mgal/d), and continuation of year 2002 pumping rates (1.04 Mgal/d) in the existing reactor units. The higher withdrawal in the new reactors is similar to that reported during 1988 for the startup of the existing reactors.

Southern Nuclear Company has noted that the high pumping rates during startup of Units I and 2 were related to achieving water-quality criteria and not to ground-water demand by the facilities.

Water treatment methods are now used to achieve the water-quality criteria and have greatly reduced ground-water pumping rates (Mark Nodich, U.S. Nuclear Regulatory Commission, written commun., September 10, 2007).Table 7. Summary of simulated time of travel for 2002 Base Case and for Scenarios A, B, and C, Vogtle Electric Generation Plant model, Georgia and South Carolina.[Ten particles were assigned to each aquifer layer in 3 model cells for a total of 30 particles per layer]bSim ulted tm mof travel iin~years

Allayer Statistic B~ase Case fScenario

Lower Dublin Mean 2,700 2,700 2,700 3,800-(A5) Median 2,700 2,600 2,700 3,000 Maximum 3,600 3,700 3,900 12,600 Minimum 2,100 2,100 2,100 1,800* ( 6 : :;,.,,,:-;-Median

2. ,2800 ,:::::2,700 Lower Midville Mean 3,100 3,100 3,200 2,800 (A7) Median 2,900 2,800 2,800 2,500 Maximum 3,800 4,200 4,600 4,000 Minimum 2,700 2,400 2,400 2,400 2002 Base Case Condition The year 2002 simulation represents the Base Case for comparison to each of the pumping scenarios.

The simulated hydrologic condition during 2002 represents effects of the 1998-2002 drought in which irrigation pumpage was above average and recharge and boundary-condition head were low due to decreased precipitation.

Results of model simulations for 2002 are presented in Cherry (2006); results of particle-tracking analyses at VEGP are presented herein.The simulated 2002 potentiometric-surface maps for the Dublin and Midville aquifer systems (figs. 7 and 8, respec-tively) indicate VEGP is within the Savannah River regional ground-water discharge zone in which the principal direction of ground-water flow is toward the Savannah River. None of the various scenarios resulted in large changes in the con-figuration of the simulated potentiometric surface and related ground-water flow directions.

The source of water to the VEGP production wells, as indicated by MODPATH analysis for year 2002, is recharge occurring in an upland area near the county line between Burke and Jefferson Counties, Ga. (fig. 9), with none of the water originating on SRS or elsewhere in South Carolina.Simulated mean time of travel from recharge areas to the VEGP production wells are about 2,700 years (yr) in the lower Dublin aquifer and about 3,100 yr in the upper and lower Midville aquifers (table 7). The fastest simulated time of travel, about 2,100 yr, was for a particle in the lower Dublin aquifer, and the longest was about 3,800 yr for a particle in the lower Midville aquifer.

Simulation of Ground-Water Flow 19 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 ' 5 10 KILOMETERS EXPLANATION S Vogtle Electric Generation Plant Model boundary D Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U Simulated ground-water flowpath-Particles placed at the center of each of the three grid cells containing a Plant Vogtle production well at increments representing 10 percent of the aquifer thickness (10 total particles per aquifer layer per cell). Five particle flowpaths for each layer are shown on map to avoid clutter Lower Dublin aquifer-.. .Upper Midville aquifer Lower Midville aquifer Figure.9.

Particle-tracking results for the year 2002, Base Case, near Vogtle Electric Generation Plant, Georgia and South Carolina.

20 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Scenario A Scenario B Scenario A simulates a 1.09-Mgal/d increase in average pumping rates at VEGP for the operation of existing reactors (Units I and 2) and an increase for the proposed new reactors (Units 3 and 4). The pumping increase was distributed evenly among the three production wells at VEGP. Simulated water-.level changes are shown in figures 10-15; particle-tracking results are shown in figure 16 and listed in table 7.For Scenario A, water-level changes were minimal, with maximum declines of greater than 0.25 ft in the Gordon aquifer (fig. 10), greater than 0.5 ft in the Millers Pond aquifer (fig. 11), greater than 1 ft in the upper and lower Dublin aqui-fers (figs. 12 and 13, respectively), and greater than 2 ft in the upper and lower Midville aquifers (figs. 14 and 15, respec-tively). Drawdown response in the shallow aquifers (Gordon, Millers Pond, and upper Dublin) is due to leakage through confining units in response to decreased head in the pumped zones (lower Dublin and upper and lower Midville aquifers).

In the upper and lower Dublin and upper and lower Midville aquifers, the zone of pumping influence (defined as greater than 0.5 ft of change) extends from about 3 to 4.5 mi onto SRS in South Carolina (figs. 12-15).For Scenario A, the source of water to VEGP production wells, as indicated by MODPATH analysis, is recharge in an upland area near the county line between Burke and Jefferson Counties, Ga. (fig. 16). Simulation results indicate that none of the recharge originated on SRS or elsewhere in South Caro-lina, despite the small amount of drawdown extending into that area in the lower and upper Dublin and lower and upper Midville aquifers (figs. 12-15). Because vertical-head gradi-ents are steep beneath the Savannah River alluvial valley, large changes in head are required to induce flow from the other side of the river. Mean simulated time of travel from recharge areas to the VEGP wells for Scenario A are about 2,700 yr in the lower Dublin aquifer and about 3,100 yr in the upper and lower Midville aquifers (table 7). The fastest simulated time of travel, about 2,100 yr, was for a particle in the lower Dublin aquifer and the slowest was about 4,700 yr for a particle in the upper Midville aquifer.Scenario B simulates a 1.09-Mgal/d increase in pumping at VEGP, as was simulated in Scenario A, and total elimina-tion of 5.3 Mgal/d pumping at the SRS facility (table 6). For this scenario, the 1.09 Mgal/d increase was distributed evenly among three production wells at VEGP completed in the lower Dublin and upper and lower Midville aquifers, and the 5.3 Mgal/d decrease was subtracted evenly among 12 produc-tion wells at the SRS completed in one or more of the follow-ing aquifers:

Gordon, Millers Pond, upper and lower Dublin, and upper and lower Midville.

Simulated water-level changes are shown in figures 17- 22; particle-tracking results are shown in figure 23 and listed in table 7.For Scenario B, the largest water-level changes were on SRS, with maximum increases of greater than 4 ft in the Gordon aquifer, greater than 1 ft in the Millers Pond aquifer, greater than 4 ft in the upper Dublin aquifer (fig. 19), greater than 8 ft in the lower Dublin aquifer (fig. 20), and greater than 4 ft in the upper and lower Midville aquifers (figs. 21 and 22, respectively).

At VEGP, the magnitude and extent of water-level decline resulting from increased pumping was less pronounced than that observed in Scenario A for an equiva-lent increase in pumping. The water-level rise resulting from elimination of SRS pumping reduced the effect of pumping at VEGP on ground-water levels. Maximum declines near.VEGP were greater than 2 ft in the upper and lower Midville aquifers (figs. 21 and 22, respectively), greater than 1 ft in the lower Dublin aquifer (fig. 20), and greater than 0.5 ft in the upper Dublin aquifer (fig. 19). There was no observed change at VEGP in the overlying Gordon and Millers Pond aquifers (figs. 17 and 18, respectively).

Despite the large water-level rise at SRS, the source of water to VEGP production wells, as indicated by MODPATH analysis (fig. 23), remained nearly identical to Scenario A (fig. 16). Simulation results indicate that ground-water recharge is provided in an upland area near the county line between Burke and Jefferson Counties, Ga. (fig. 23), with a mean simulated time of travel of about 2,700 yr in the lower Dublin aquifer; about 3,300 yr in the tipper Midville aquifer;and about 3,200 yr in the lower Midville aquifer (table 7). The fastest simulated time of travel was for a particle in the lower Dublin aquifer (about 2,100 yr), and slowest was for a particle in the upper Midville aquifer (about 5,200 yr). As was the case for Scenario A, none of the recharge originated on SRS or elsewhere in South Carolina.

Simulation of Ground-Water Flow 21 Base modified from U.S. Geological Survey 1:24,000-scale digital data 0 5 10 MILES i I I I 0 5 10 KILOMETERS EXPLANATION Up Vogtle Electric Generation Plant 0 Fault-Approximately located; D, downthrown side; U, upthrown side U--0,25 -Line of equal simulated water-level change-Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario A (see table 6 for description of scenario)Figure 10. Simulated water-level change for Scenario A in the Gordon aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

22 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data I I 0 5 10 KILOMETERS EXPLANATION Vogtle Electric Generation Plant 0 Pen Branch Fault-Approximately located; 0, downthrown side; U, upthrown side U--0.5 -Line of equal simulated water-level change-Interval, in feet, is 0.25.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario A (see table 6 for description of scenario)Figure 11. Simulated water-level change for Scenario A in the Millers Pond aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 23 Base modified from U.S. Geological Survey 0 5 .10 MILES 1:24,000-scale digital data I I I 0 5 10 KILOMETERS EXPLANATION t Vogtle Electric Generation Plant D Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U--0.5 -Line of equal simulated water-level change-Interval, in feet, is 0.5.Computed by subtracting the simulated-potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario A *(see table 6 for description of scenario)Figure 12. Simulated water-level change for Scenario A in the upper Dublin aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

24 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data I I I 0 5 10 KILOMETERS EXPLANATION tP Vogtle Electric Generation Plant Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-- Line of equal simulated water-level change-Interval, in feet, is 0.5.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario A (see table 6 for description of scenario)0 Production well-Completed in the lower Dublin aquifer in which pumping was adjusted for scenario Figure 13. Simulated water-level change for Scenario A in the lower Dublin aquifer, near.Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 25 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION Up Vogtle Electric Generation Plant 0--Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side-1 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from thesimulated potentiometric surface for Scenario A (see table 6 for description of scenario)0 Production well-Completed in the upper Midville aquifer in which pumping was adjusted for scenario Figure 14. Simulated water-level change for Scenario A in the upper Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

26 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION I Vogtle Electric Generation Plant Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U--1 -Line of equal simulated water-level change-Interval, in feet, is variable, Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario A (see table 6 for description of scenario)0 Production well-Completed in the lower Midville aquifer in which pumping was adjusted for scenario Figure 15. Simulated water-level change for Scenario A in the lower Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 27 Base modified from U.S, Geological Survey 0 5 10 MILES 1:24,000-scale digital data I I I 0 5 10 KILOMETERS EXPLANATION

£? Vogtle Electric Generation Plant Model boundary 0 Pen Branch Fault-Approximately located; 0, downthrown side; U, upthrown side U Simulated ground-waterflowpath--Particles placed at the center of each of the three grid cells containing a Plant Vogtle production well at increments representing 10 percent of the aquifer thickness (10 total particles per aquifer layer per cell). Five particle flowpaths for each layer are shown on map to avoid clutter Lower Dublin aquifer Upper Midville aquifer Lower Midville aquifer Figure 16. Particle-tracking results for the year 2002, Scenario A, near Vogtle Electric Generation Plant, Georgia and South Carolina.

28 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios

'Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data I I I I 0 5 10 KILOMETERS EXPLANATION U Vogtle Electric Generation Plant 0 Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-0.5 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenario)0 Production well-Completed in the Gordon aquifer in which pumping was adjusted for scenario Figure 17. Simulated water-level change for Scenario B inthe Gordon aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 29 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION

[P Vogtle Electric Generation Plant Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side-0.5 -Line of equal simulated water-level change-Interval, in feet, is 0.5.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenario)0 Production well-Completed in the Millers Pond aquifer in which pumping was adjusted for scenario Figure 18. Simulated water-level change for Scenario B in the Millers Pond aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

30 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION T Vogtle Electric Generation Plant o_ Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-0.5 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenario)0 Production well-Completed in the upper Dublin aquifer in which pumping was adjusted for scenario Figure 19. Simulated water-level change for Scenario B in the upper Dublin aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 31 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data I I I 0 5 10 KILOMETERS EXPLANATION E Vogtle Electric Generation Plant 0 Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U--1 -- Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenario)o Production well-Completed in the lower Dublin aquifer in which pumping was adjusted for scenario Figure 20. Simulated water-level change for Scenario B in the lower Dublin aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

32 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey u 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION U Vogtle Electric Generation Plant 0 Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U--1 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenariol, 0 Production well-Completed in the upper Midville aquifer in which pumping was adjusted for scenario Figure 21. Simulated water-level change for Scenario B in the upper Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 33 a 5 10MILES Base modified from U.S. Geological Survey 1:24,000-scale digital data I I 0 5 10 KILOMETERS EXPLANATION Ep Vogtle Electric Generation Plant O Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-0.5 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario B (see table 6 for description of scenario)0 Production well-Completed in the lower Midville aquifer in which pumping was adjusted for scenario Figure 22. Simulated water-level change for Scenario B in the lower Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

34 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey .0 5 10 MILES 1:24,000-scale digital data I I I I I 0 5 10 KILOMETERS EXPLANATION E Vogtle Electric Generation Plant Model boundary 0 Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U Simulated ground-water flowpath-Particles placed at the center of each of the three grid cells containing a Plant Vogtle production well at increments representing 10 percent of the aquifer thickness (10 total particles per aquifer layer per cell). Five particle flowpaths for each layer are shown on map to avoid clutter..... Lower Dublin aquifer Upper Midville aquifer-Lower Midville aquifer Figure 23. Particle-tracking results for the year 2002, Scenario B, study area, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 35 Scenario C Scenario C simulates a 3.42-Mgal/d increase in pumping at VEGP that represents a 2.33-Mgal/d higher rate of with-drawal than was simulated for Scenario A for the proposed new reactor units (table 6). The higher withdrawal for wells providing water to the new reactors is similar to that reported for the startup of the two existing reactors (Units I and 2) dur-ing 1988 (Mark Notich, U.S Nuclear Regulatory Commission, written commun., April 10, 2007). Southern Nuclear Com-pany has noted that the high pumping rates during startup of Units I and 2 were related to achieving water-quality criteria and not to ground-water demand by the facilities.

Water treatment methods are now used to achieve the water-quality criteria and have greatly reduced ground-water pumping rates (Mark Nodich, U.S. Nuclear Regulatory Commission, written commun., September 10, 2007).Although pumping rates simulated by Scenario C are viewed as implausible for long-term operation.of proposed Units 3 and 4, and the pumping rates are not proposed by Southern Nuclear Company, the scenario is designed to simulate pumping rates necessary to draw ground water from South Carolina to the VEGP production wells. The 3.42-Mgal/d increase was distributed evenly among three production wells at VEGP. Simulated water-level changes are shown in figures 24-29; particle-tracking results are shown on figure 30 and listed in table 7.The maximum simulated drawdown for Scenario C was greater than 1 ft in the Gordon aquifer (fig. 24), greater than 2 ft in the Millers Pond aquifer (fig. 25), greater than 4 ft in the upper Dublin aquifer (fig. 26), greater than 4 ft in the lower Dublin aquifer (fig. 27), and greater than 8 ft in the upper and lower Midville aquifers (figs. 28 and 29, respectively).

The extent of drawdown is largest for Scenario C when compared to Scenarios A and B, with the 0.5-ft drawdown contour in the upper and lower Midville aquifers extending about 29 mi to the southwestern model boundary in Jenkins and Screven Counties, Ga., and about 14 mi eastward into SRS (figs.'28 and 29, respectively).

In the overlying Gordon, Millers Pond, and upper Dublin aquifers (figs.. 24-26, respectively), drawdown response is due to leakage through confining units in response to decreased head in the production zones (lower Dublin and upper and lower Midville aquifers, figs. 27-29, respectively).

For Scenario C, the source of water to VEGP produc-tion wells, as indicated by MODPATH analysis (fig. 30), is recharge at a somewhat different location than that simulated for Scenario A (fig. 16). As was the case for Scenario A, simulation results indicate that much of the ground-water recharge for Scenario C occurs in an upland area near the county line between Burke and Jefferson Counties, Ga.;however, there is an additional source of water in an upland area in eastern Barnwell County, S.C. As was the case for *Scenarios A and B, none of the recharge originated on SRS.When compared to Scenarios A and B, simulated mean time of travel for Scenario C (table 7) was slower in the lower Dub-lin aquifer (about 3,800 yr), and faster in the upper and lower Midville aquifers (about 2,800 yr). For Scenario C, the fastest simulated time of travel of about 1,800 yr was for particles in the lower Dublin and upper Midville aquifers and slow-est (about 12,600 yr) was for a particle in the lower Dublin aquifer. The slower time of travel in the lower Dublin aquifer may be the result of the greater length and extent of flowlines shown on figure 30.

36 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I 0 5 10 KILOMETERS EXPLANATION tP Vogtle Electric Generation Plant 0 Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side u--0.5- Line of equal simulated water-level change-Interval, in feet, is 0.5.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)Figure 24. Simulated water-level change for Scenario C in the Gordon aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 37 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data I 0 5 10 KILOMETERS EXPLANATION 1:5: Vogtle Electric Generation Plant D Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side u--1 -- Line of equal simulated water-level change-Interval, in feet; is variable.Computed by subtracting the simulated potehtiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)Figure 25. Simulated water-level change for Scenario C in the Millers Pond aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

38 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I 0 5 10 KILOMETERS EXPLANATION EP Vogtle Electric Generation Plant D Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-- --- Line of equal simulated water-level change-[Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)Figure 26. Simulated water-level change for Scenario C in the upper Dublin aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 39 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I I 0 5 10 KILOMETERS EXPLANATION J:P Vogtle Electric Generation Plant.2. Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U-0.5 -Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)0 Production well-Completed in the lower Dublin aquifer in which pumping was adjusted for scenario Figure 27. Simulated water-level change for Scenario C in the lower Dublin aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

40 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I 0 5 10 KILOMETERS EXPLANATION t:? Vogtle Electric Generation Plant D Pen Branch Fault-Approximately located; 0, downthrown side; U, upthrown side U-0.5 -Line of equal simulated water-level change-[Interval, in feet, is variable.

-Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)Production well-Completed in the upper Midville aquifer in which pumping was adjusted for scenario Figure 28. Simulated water-level change for Scenario C in the upper Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Simulation of Ground-Water Flow 41 Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data 0 5 10 KILOMETERS EXPLANATION Up Vogtle Electric Generation Plant a) Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side u-1 Line of equal simulated water-level change-Interval, in feet, is variable.Computed by subtracting the simulated potentiometric surface for 2002 Base Case from the simulated potentiometric surface for Scenario C (see table 6 for description of scenario)0 Production well-Completed in the lower Midville aquifer in which pumping was adjusted for scenario Figure 29. Simulated water-level change for Scenario C in the lower Midville aquifer, near Vogtle Electric Generation Plant, Georgia and South Carolina.

42 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios Base modified from U.S. Geological Survey 0 5 10 MILES 1:24,000-scale digital data i I I I ]0 5 10 KILOMETERS EXPLANATION U Vogtle Electric Generation Plant Model boundary Pen Branch Fault-Approximately located; D, downthrown side; U, upthrown side U Simulated ground-water flowpath-Particles placed at the center of each of the three grid cells containing a Plant Vogtle production well at increments representing 10 percent of the aquifer thickness (10 total particles per aquifer layer per cell). Five particle flowpaths for each layer are shown on map to avoid clutter Lower Dublin aquifer Upper Midville aquifer Lower Midville aquifer Figure 30. Particle-tracking results for the year 2002, Scenario C, near Vogtle Electric Generation Plant, Georgia and South Carolina.

Summary and Conclusions 43 Model Limitations The steady-state simulations presented herein are believed to depict reasonably changes in ground-water levels that resulted from pumping increases of 1.09 to 3.42 Mgal/d at VEGP and from a pumping decrease of 5.3 Mgal/d at SRS, which together represent less than I percent of the total simu-lated ground-water flow of 1,035 Mgal/d for 2002 (Cherry, 2006). Because these pumping changes are of low magnitude and occur near the center of the simulated area, lateral bound-aries generally have little influence on simulation results.An exception occurs for Scenario C in which drawdown in the upper and lower Midville aquifers extends to the model's southwestern "no-flow" boundary.

Simulated drawdown that reaches a no-flow boundary results in higher Values than would occur if the boundary was not intercepted.

Steady-state simulations are believed to be representative of local hydro-logic conditions because response to changes in pumpage is short term and previous testing indicates the model is insensitive to changes in storage (Clarke and West, 1998).The revised ground-water flow model (Cherry, 2006)used for this investigation is based on a drought period (2002)in which boundary head was lowered to reflect decreased recharge to the ground-water system. It is likely that boundary conditions reflecting average or wet periods would result in somewhat different patterns of water-level change than were simulated for this study. Despite these possible variations, it is likely that ground-water flowpaths and recharge areas would be largely the same.The ground-water flow model used in this study is subject to the limitations described in Cherry (2006). These limita-tions include error and uncertainty in field measurements of water level and in estimates of pumping, limitations of the conceptual models, approximations made in representing the physical properties of the flow system and errors inherent in estimating the spatial distribution of these properties, approxi-mations made in the formulation and application of model boundary and initial conditions, errors associated with numeri-cal approximation and solution of the mathematical model of the flow system, and assumptions made in using the models to predict the future behavior Of the flow system.In some local areas near the Pen Branch Fault, simulated water leyels poorly, matched observed water levels in the Gordon aquifer (layer A2). Near the P- 19 well cluster site on SRS, and in the Savannah River alluvial valley near VEGP, simulated head was consistently lower than the observed head.The reasons for this mismatch are unknown; however, a local-ized hydraulic connection between layers A l (Upper Three Runs) and A2 (Gordon) near the Pen Branch Fault on SRS has been suggested by previous investigators (Aadland and others, 1991). Because the reason for the high water level in the Gordon aquifer was not substantiated by field investigations, it was decided by previous investigators (Clarke and West, 1998) not to account for this effect in the calibration of the ground-water flow model.Particle tracking using MODPATH is controlled largely by lateral and vertical head gradients, along with the hydraulic properties of the aquifers and confining units. In the VEGP area, data on the vertical hydraulic conductivity of aquifers, streambeds, and confining units are sparse. An additional limitation of particle tracking using MODPATHis the inability to determine whether a particle of water exits the flow system in a model cell containing a weak sink. A weak sink can be described as a discharge well that does not remove all the water entering a cell, so that some water continues to move through the system. Finally, the no-flow boundary condi-tion along the southwestern boundary of the model limits the available area for a simulated flowpath.

This limitation may have resulted in faster simulated time of travel in the lower Dublin aquifer for Scenario C than might have occurred if the no-flow boundary was located farther away from the pumping at VEGP.Summary and Conclusions An updated and calibrated MODFLOW ground-water flow model (Cherry, 2006) was used to simulate the effect of current and potential future pumping on ground-water levels and flowpaths near Vogtle Electric Generation Plant (VEGP), Ga., for a Base Case representing year 2002 conditions and three pumping scenarios:

Scenario A simulates a 1.09-million gallons per day (Mgal/d) increase in pumping at VEGP assuming average withdrawal rates with the operation of existing reactors (Units I and 2) and the proposed new reactors (Units 3 and 4).Scenario B simulates a 1.09-Mgal/d increase in pumping at VEGP, as was simulated in Scenario A, combined with a shutdown of the SRS facility (reduction of 5.3 Mgal/d).Scenario C simulates a 3.42-Mgal/d increase in pump-ing at VEGP that represents a higher rate of withdrawal for the proposed new reactor units during their startup period (3.42 Mgal/d), and continuation of year 2002 pumping rates (1.04 Mgal/d) in the existing-reactor units.Maximum water-level change resulting from increased pumping at VEGP (without changes at Savannah River Site (SRS) or elsewhere in the study area) were simulated in the pumped layers at VEGP-the lower Dublin and upper and lower Midville aquifers.

Simulated maximum declines in these units were from I to greater than 2 feet (ft) for Scenario A and from 4 to greater than 8 ft for Scenario C. Although none of the VEGP wells are completed in the upper Dublin aquifer, simulated water-level changes were similar to those observed in the pumped lower Dublin aquifer, suggesting a large degree of interconnection between the two aquifers.

A muted water-level decline from 0.25 to greater than 2 ft was simulated in the shallow Gordon and Millers Pond aquifers for Scenarios A and C as the result of leakage through confining units in response to decreased head in the production zones (lower Dublin and Lipper and lower Midville aquifers).

44 Simulation and Particle-Tracking Analysis of Selected Ground-Water Pumping Scenarios The largest simulated water-level changes at VEGP were for Scenario C, which represents a tripling of current pump-age at the facility.

Although such pumping rates are viewed as implausible for long-term operation of proposed Units 3 and 4, and are not proposed by Southern Nuclear Company, the sce-nario is designed to simulate pumping rates necessary to draw ground water from South Carolina to the VEGP production wells. For this scenario, drawdown was greater than 8 ft in the upper and lower Midville aquifers, and greater than 4 ft in the upper and lower Dublin aquifers.

Drawdown exceeding 0.5 ft in these aquifers extended about 29 miles (mi) to the south-western model boundary in Jenkins and Screven Counties, Ga., and about 14 mi eastward onto SRS in South Carolina.For Scenario B, elimination of pumping at SRS resulted in large water-level changes near SRS, with rises of greater than 8 ft in the lower Dublin aquifer and greater than 4 ft in the upper Dublin and upper and lower Midville aquifers.

At VEGP, the magnitude and extent of water-level decline result-ing from increased pumping was less than in Scenario A with maximum declines of greater than 2 ft in the upper and lower Midville aquifers, greater than 1 ft in the lower Dublin aquifer, and greater than 0.5 ft in the upper Dublin aquifer. The water-level rise resulting from elimination of SRS pumping reduced the effect of pumping at VEGP on ground-water levels.Results of MODFLOW simulations were analyzed using the USGS particle-tracking code MODPATH (Pollock, 1994)to determine the source of water and associated time of travel to VEGP production wells. For each of the scenarios, most of the recharge to VEGP wells originated in an upland area near the county line between Burke and Jefferson Counties, Ga., with none of the recharge originating on SRS or elsewhere in South Carolina.

An exception occurs for Scenario C, in which some of the recharge originates in an upland area in eastern Barnwell County, S.C. Simulated mean time of travel from recharge areas to the VEGP wells for theBase Case and the three scenarios was between about 2,700 and 3,800 years, with some variation related to changes in head gradients due to pumping changes.References Cited Aadland, R.K., Gellici, J.A., and Thayer, Paul, 1995, Hydrogeologic framework of west-central South Carolina: South Carolina Department of Natural Resources, Water Resources Division Report 5, 200 p.Aadland, R.K., Harris, M.K., Lewis, C.M., Gaughan, T.F., and Westbrook, T.M., 1991, Hydrostratigraphy of the general separations area, Savannah River Site (SRS), South Carolina: Aiken, S.C., U.S. Department of Energy Report WSRC-RP-91-13, Westinghouse Savannah River Company, 114 p.Atkins, J.B., Journey, C.A., and Clarke, J.S., 1996, Estimation of ground-water discharge to streams in the central Savannah River basin of Georgia and South Carolina:

U.S. Geological Survey Water Resources Investigations Report 96-4179, 46 p.Arnett, M.W., and Mamatey, A.R., eds.,1996, Savannah River Site environmental report for 1995: Prepared for U.S.Department of Energy by Westinghouse Savannah River Company, Aiken, S.C., WSRC-TR-96-0075, 271 p.Cherry, G.S., 2003, Precipitation, ground-water use, and ground-water levels in the vicinity of the Savannah River Site, Georgia and South Carolina, 1992-2002, in Hatcher, K.J., ed., Proceedings of the 2003 Georgia Water Resources Conference, April 23-24, 2003: Athens, Ga., The Univer-sity of Georgia, Institute of Ecology, CD-ROM, online at http://ga.

water, usgs.gov/pubs/other/gwrc2003/pdf Cherry, G.S., 2006, Simulation and particle-tracking analysis of ground-water flow near the Savannah River Site, Georgia and South Carolina, 2002, and for selected water-manage-ment scenarios, 2002 and 2020: U.S. Geological Survey Scientific Investigations Report 2006-5195, 156 p., Web-only publication at http://pubs.usgs.gov/sir/2006/51951.

Chowns, T.M., and Williams, C.T., 1983, Pre-Cretaceous rocks beneath the Georgia Coastal Plain-Regional implications, in Gohn, G.S., ed., Studies related to the Charleston, South Carolina earthquake of 1886-Tectonics and seismicity:

U.S. Geological Survey Professional Paper 1313-L, p. LI-L42, online at http://pubs.er.usgs.

gov/usgspubs/pp/ppl313 Christensen, E.J., and Gordon, D.E., 1983, Technical summary of groundwater quality protection program at Savannah River Plant-Volume 1, site geohydrology and solid and hazardous wastes: E.I. du Pond de Nemours and Company, Savannah River Laboratory, Aiken, S.C., pre-pared for the U.S. Department of Energy under contract DE-AC09-76SR00001, variously paged.Clark, W.Z., and Zisa,A.C., 1976, Physiographic map of Georgia: Georgia Geologic Survey, I sheet, scale 1:2,000,000.

Clarke, J.S., Brooks, Rebekah, and Faye, R.E., 1985, Hydro-.geology of the Dublin and Midville aquifer systems of east central Georgia: Georgia Geologic Survey Information Circular 74, 62 p.Clarke, J.S., and West, C.T., 1997, Ground-water levels, prede-Velopment ground-water flow, and stream-aquifer relations in the vicinity of the Savannah River Site, Georgia and South Carolina:

U.S. Geological Survey Water-Resources Investigations Report 97-4197, 120 p., 1 plate, online at http://pubs.

usgs.gov/wri/wri97-4197/.

Clarke, J.S., and West, C.T., 1998, Simulation of ground-water flow and stream-aquifer relations in the vicinity of the Savannah River Site, Georgia and South Carolina, prede-velopment through 1992: U.S. Geological Survey Water-Resources Investigations Report 98-4062, 134 p., online at http://pubs.

usgs.gov/wri/wri98-4062/.

References Cited. 45 Falls, W.F., Baum, J.S., Harrelson, L.G., Brown, L.H., and Jerden, J.L., Jr., 1997, Geology and hydrogeology of Cretaceous and Tertiary strata, and confinement in the vicinity of the U.S. Department of Energy Savannah River Site, South Carolina and Georgia: U.S. Geological Survey Water-Resources Investigations Report 97-4245, 125 p., onliný at http.//pubs.

er usgs.gov/usgspubs/wri/wri974245 Fanning, J.L., 1997, Water use in Georgia by county for 1995: Georgia Geologic Survey Information Circular 101, 95 p.Fanning, J.L., 2003, Water use in Georgia by county for 2000 and water-use trends for 1980-2000:

Georgia Geologic Survey Information Circular 106, 176 p., online at http://ga. water usgs.gov/pubs/other/ggs-icl06/.

Faye, R.E., and Mayer, G.C., 1990, Ground-water flow and stream-aquifer relations in the northern Coastal Plain of Georgia and adjacent parts of Alabama and South Carolina: U.S. Geological Water-Resources Investigations Report 88-4143, 83 p., online at http://pubs.er usgs.gov/usgspubs/wri/

wri884143 Harbaugh, AW., Banta, E.R., Hill, M.C., and McDonald, M.G., 2000, MODFLOW-2000, The U.S. Geological Survey modular ground-water model-User guide to modularization concepts and the ground-water flow process: U.S. Geologi-cal Survey Open-File Report 00-92, 121 p., online at http.//water usgs.gov/nrp/gwsoftware/modflow2000/ofrOO-92.pdf Harrelson, L.E., Falls, W.F., and Clarke, J.S., 1997, Selected well data in the vicinity of the Savannah River Site, South Carolina and Georgia: U.S. Geological Survey Open-File Report 97-657A, 215 p., 10 plates, online at http://pubs.

er usgs.gov/usgspubs/ofr/ofr96657A Lonon, G.E., Burnette, C.B., and Morris, H.J., 1983, Water use in South Carolina, 1980: South Carolina Water Resources Commission Report 138, 20 p.McDonald, M.G., and Harbaugh, A.W., 1988, A modular finite-difference ground-water flow model: U.S. Geologi-cal Survey Techniques of Water-Resources Investigations report, book 6, chap. Al, variously paged, online at http://pubs. er usgs.gov/usgspubs/twri/twriO6A 1 Miller, J.A., 1986, Hydrogeologic framework of the Floridan aquifer system in Florida and in parts of Georgia, Alabama, and South Carolina:

U.S. Geological Survey Professional Paper 1403-B, 91 p., online at http.//pubs.erusgs.gov/

usgspubs/pp/ppJ403B Pierce, R.R., Barber, N.L., and Stiles, H.R., 1982, Water use in Georgia by county forl980: Georgia Geologic Survey Information Circular 59, 180 p.Pollock, D.W., 1994, Users Guide for MODPATH/MOD-PATH-PLOT, Version 3-A particle tracking post-process-ing package for MODFLOW, the U.S. Geological Survey finite-difference ground-water flow model: U.S. Geological Survey Open-File Report 94-464, 248 p., online at http://pubs.

er usgs.gov/usgspubs/ofr/ofr94464 Price, Van, Fallaw, W.C., and McKinney, J.B., 1991, Geologic setting of the new production reactor within the Savannah River Site (U): Westinghouse Savannah River Company, report no. WSRC-RP-91-96, ESH-EMS-90171, 80 p.Prowell, D.C., 1994, Preliminary geologic map of the Barnwell 30 min x 60 min quadrangle, South Carolina and Georgia: U.S. Geological Survey Open-File Report 94-673, 38 p., 1 plate, scale 1:100,000, online at http://pubs.erusgs.

gov/usgspubs/ofr/ofr94673 Prowell, D.C., and O'Connor, B.J., 1978, Belair fault zone-Evidence of Tertiary fault displacement in eastern Georgia: Geology, v. 6, p. 681-684.Siple, G.E., 1960, Piezometric levels in the Cretaceous sand aquifer of the Savannah River Basin: Georgia Mineral Newsletter, v. 8, no. 4, winter 1960, p. 163-166.Siple, G;E., 1967, Geology and ground water of the Savannah River Plant and vicinity, South Carolina:

U.S. Geological Survey Water-Supply Paper 1841, 113 p., online at http ://pubs. er usgs.gov/usgspubs/wsp/wsp1841 Snipes, D.S., Fallaw, W.C., Price, Van, Jr., and Cumbest, R.J., 1993, The Pen Branch Fault-Documentation of Late Cretaceous-Tertiary faulting in the Coastal Plain of South Carolina:

Southeastern Geology, v. 33, no. 4, October 1993, p. 195-218.Summerour, J.H., Shapiro, E.A., and Huddlestun, RE, 1998, An investigation of Tritium in the Gordon and other aquifers in Burke County, Georgia, Phase II-Final report: Georgia Geologic Survey Information Circular 102, 72 p.Waddell, M.G., Keith,J.F., and Domoarcki, W.J., 1995, High resolution seismic characterization GGS-1, Burke County, Ga.: University of South Carolina Project Report to Georgia Geologic Survey, ESRI Technical Report 95-F129-1, 20 p.Wait, R.L., and Davis, M.E., 1986, Configuration and hydrol ogy of the pre-Cretaceous rocks underlying the southeast ern Coastal Plain aquifer system: U.S. Geological Survey Water-Resources Investigations Report 86-4010, 1 sheet, scale 1:2,000,000, online at http://pubs.er usgs.gov/usgspubs/wri/wri864010 Washington Savannah River Company, 2007, Semi-annual 2007 M-Area and metallurgical laboratory hazardous waste management facilities groundwater monitoring and corrective action report (U), volume 1: Aiken, S.C., U.S. Department of Energy Report WSRC-RP-2007-4039, prepared for the U.S. Department of Energy under contract no. DE-AC09-96-SR18500, variously paged.

Manuscript approved for publication, October 23, 2007 Prepared by USGS Georgia Water Science Center Edited and page layout by Patricia L. Nobles Graphics by Bonnie J. Turcott For more information concerning the research in this report, contact USGS Georgia Water Science Center, Atlanta, telephone:

770-903-9100

ML073310491

Text

Navigation menu

Search