An investigation of Tritium in the Gordon and other aquifers in Burke county, Georgia [1994]

An Investigation of Tritium in the Gordon and Other Aquifers
in Burke County, Georgia
Joseph H. Summerour, Earl A. Shapiro, Jerry A. Lineback, Paul F. Huddlestun, and Allan C. Hughes
Work Performed as Part of a Cooperative Agreement with U.S. Department of Energy
Agreement Number DE-FG-09-92SR12868
GEORGIA DEPARTMENT OF NATURAL RESOURCES ENVIRONMENTAL PROTECTION DIVISION GEORGIA GEOLOGIC SURVEY
Atlanta 1994

An Investigation of Tritium in the Gordon and Other Aquifers in Burke County, Georgia
Joseph H. Summerour, Earl A. Shapiro, Jerrry A. Lineback, Paul F. Huddlestun, and Allan C. Hughes
GEORGIA DEPARTMENT OF NATIJRAL RESOURCES Joe D. Tanner, Commissioner
ENVIRONMENTAL PROTECTION DNISION Harold F. Reheis, Director
GEORGIA GEOLOGIC SURVEY William H. McLernore, State Geologist
Atlanta 1994
INFORMATION CIRCULAR 95

EXECUTIVE SUMMARY
Tritium was initially detected in the ground water of eastern Burke County, Georgia, in April, 1988 by the Environmental Radiation Program ofthe Georgia Environmental Protection Division (EPD) as part ofa long-term monitoring program. Tritium was not detected again until 1991, when repeated sampling showed elevated tritium concentrations consistently present in groundwater samples from what was reported to bea deep public watersupply well. At the direction ofGovernorZell Miller, the Georgia Geologic Survey (GGS) Branch of the EPD initiated an investigation, in late 1991, to evaluate if Georgia aquifers had become pollutedby tritium, the extentofthat pollution, thepossible threat to public health, and the pathways for tritium to enter the groundwater regime in Georgia. In March, 1992, the EPD received authorization for funding for this study from the U.S. Department of Energy (DOE).
The GGS conducted or sponsored seven sub-investigations in eastern Burke County as part of this project. These studies include: 1. tritium analysis of water samples from 109 water wells; 2. two studies of tritium abundance in the base flow of streams; 3. installation of fifteen ground-water monitoring wells at six sites, and tritium analyses of ground-water samples from
those wells; 4. analysis of the subsurface geology; 5. analysis of the hydrogeology of the Upper Three Runs (water table) and Gordon aquifers; 6. geochemical evaluation of the Upper Three Run and Gordon aquifers; and 7. a seismic survey of the Savannah River channel in the study area.
The analysis of water samples from 109 water wells in eastern Burke County indicates that fifteen wells contain average tritium concentrations of 500 picoCuries or more. Nine of these fifteen wells are definitely drawing their water from the Upper Three Runs aquifer, and the remaining six wells are probably drawing their water from this source (based upon the evidence of this study). At leasteightofthe fifteen water wells are either ungrouted, lack casing, lack a surface pad, or have experiencedfailure of the casing.
The base flow studies indicate tritium pollution of the Upper Three Runs aquifer throughout all ofeastern Burke County. All samples contain less than 18 percent of the of the Maximum Contaminant Level (MCL) for tritium set by the Environmental Protection Agency (EPA). The highest concentrations of tritium (up to 3,500 picoCuries per liter) occur north of Hancock Landing. Tritium levels decrease to the northwest, west, and southeast of this area.
The ground-water monitoring wells show no detectable concentrations oftritium in the uppermost confined aquifer (Gordon aquifer) in eastern Burke County in five out of six wells. The sixth Gordon monitoring well may be detecting a very local point source of tritium pollution. Although tritium is present in all of the monitoring wells sampling the Upper Three Runs aquifer, to this date, all water samples have been well below the MCL for tritium set by EPA.
The analysis of the subsurface geology indicates that there are complex facies changes within many of the sedimentary units. These facies changes complicate the flow of ground water in the study area. The Savannah River has cut through the sediments that make up the Upper Three Runs aquifer in at least the northern half of the study area. McBean Creek and several other small streams in the study area have also cut through the aquifer.
Analysis of the hydrogeology of the shallow aquifers indicates that the Upper Three Runs aquifer is compartmentalized by streams in the study area resulting in numerous small, local, ground-water flow regimes. Recharge and discharge of the Upper ThreeRunsaquiferisprimarilyrestrictedtotheselocalregimes. TherechargeareafortheGordonaquiferismostlikelyinsouthern Richmond and northwestern Burke Counties, outside of the study area. Within the study area, the Gordon aquifer discharges to the Savannah River. The potential direction of vertical groundwater flow is generally downward from the Upper Three Runs aquifer into the Gordon aquifer. However, the confining bed between these two aquifers retards such movement ofground water.
The Upper Three Runs aquifer consists of two hydrogeochemical layers and the Gordon aquifer consists of a single hydrogeochemical layer. Lateral geochemical variation occurs in the direction of ground-water flow. The primary conclusions of this study are as follows: 1. there is no evidence of a threat to public health due to tritium pollution of aquifers in Burke County; 2. there is widespread tritium pollution of the Upper Three Runs (water table) aquifer in eastern Burke County, but this
pollution is below the concentration of tritium allowed for drinking water by EPA; 3. there is no evidence of regional tritium pollution of the Gordon aquifer in eastern Burke County; and 4. existingdata are not adequate to fully resolve the issue ofthe pathway for tritium into the Upper Three Runs aquifer. However, the investigation indicates that some pathways are more likely than others and suggests specific pathway models for future investigation.
A series of thirteen recommendations are made including one recommendation for increasing public awareness, two recommendations concerning public health issues, seven recommendations for further technical studies of the aquifers, and three recommendations for long-term monitoring.

TABLE OF CONTENTS INTRODUCTION ..................................................................................................................................................................... 1
Statement of Problem ......................................................................................................................................................... 1 Location of Study Area ...................................................................................................................................................... 1 Geology/flydrology ........................................................................................................................................................... 2 Weather ............................................................................................................................................................................... 4 Cultural Features ................................................................................................................................................................ 7 History of Tritium Releases from Savannah River Site .................................................................................................... 8 Tritium in Rainfall in Burke County .................................................................................................................................. 12 Tritium in Ground Water ................................................................................................................................................... 14 Acknowledgements ............................................................................................................................................................ 50 PROCEDURES ......................................................................................................................................................................... 22 Study of Water Wells ......................................................................................................................................................... 22 Base Flow Studies .............................................................................................................................................................. 22
1991 Base Flow Study ............................................................................................................................................... . 24 1992 Base Flow Study ............................................................................................................................................... . 24 Installation of Monitoring Wells ........................................................................................................................................ 24 Description of Cores ........................................................................................................................................................... 24 Seismic Survey of Savannah River Channel ..................................................................................................................... 24 Ground-Water Geochemistry ............................................................................................................................................. 30 Aquifer Testing ................................................................................................................................................................... 30 RESULTS .................................................................................................................................................................................. 30 Tritium in Water Wells ...................................................................................................................................................... 30 Tritium in Base Flow .......................................................................................................................................................... 34 1991 Base Flow Study ................................................................................................................................................ 34 1992 Base Flow Study ................................................................................................................................................ 37 Comparison of 1991 and 1992 Base Flow Results .................................................................................................... 38 Comparison of 1992 Base Flow Results with Water Well Data ............................................................................... 40 Tritium in Monitoring Wells .............................................................................................................................................. 42 Geology ............................................................................................................................................................................... 42 Seismic Survey of Savannah River Channel ..................................................................................................................... 50 Hydrogeology ..................................................................................................................................................................... 50 Upper Three Runs Aquifer ......................................................................................................................................... 50 Gordon Aquifer .......................................................................................................................................................... . 52 Vertical Hydraulic Gradient ....................................................................................................................................... 54 Results of Aquifer Tests ............................................................................................................................................. 54 Ground-Water Geochemistry ............................................................................................................................................ . 54 Vertical Geochemical Variation ................................................................................................................................. 54 Lateral Geochemical Variation .................................................................................................................................. 58 CONCLUSIONS ...................................................................................................................................................................... . 58 Extent of Pollution .............................................................................................................................................................. 58 Stratigraphic Extent of Pollution ................................................................................................................................ 58 Geographic Extent of Pollution .................................................................................................................................. 58 Public Health ..................................................................................................................................................................... . 60 Pathway into the Aquifer ................................................................................................................................................. .. 60 Direct Lateral Transport ............................................................................................................................................. 62 Indirect Lateral Transport .......................................................................................................................................... . 62 Upwards Transport ................................................................................................................................................... .. 62 Downwards Transport ................................................................................................................................................ 62
Point Source Pollution ........................................................................................................................................ 64 Other Pathways ........................................................................................................................................................... 64 Trans-River Flow (Underflow) ......................................................................................................................................... . 64 RECOMMENDATIONS .......................................................................................................................................................... 64 Public Awareness ............................................................................................................................................................... 64 Public Health ..................................................................................................................................................................... . 64 Technical Studies ................................................................................................................................................................ 64 Long-Tenn Monitoring ...................................................................................................................................................... 66
ii

REFERENCES CITED ............................................................................................................................................................. 67 GLOSSARY ............................................................................................................................................................................... 69
FIGURES 1. Index map of eastern Burke County .................................................................................................................................. 3 2. Stratigraphic and hydrostratigraphic units ......................................................................................................................... 4 3. Geologic map of the study area ......................................................................................................................................... 5 4. Average monthly precipitation at Bush Airport ................................................................................................................ 6 5. Windrose diagram for SRS/Burke County ........................................................................................................................ 6 6. Bar graph showing atmospheric releases-1954-1992 ....................................................................................................... 9 7. Locations of EPD rainfall collection sites ......................................................................................................................... 11 8. Yearly rainfall tritium-EPD Station 11 .............................................................................................................................. 12 9. Yearly rainfall tritium-EPD Station 35 .............................................................................................................................. 13 10. Distribution of tritium in rainfall in the SRS-Burke County area ..................................................................................... 14 11. Map of surface water and monitoring well collection sites-Plant Vogtle property ......................................................... 17 12. Tritium concentrations-Plant Vogtle spring-River Mile 150.9 ......................................................................................... 19 13. Tritium concentrations-Mallards Pond .............................................................................................................................. 21 14. Wells sampled by EPD, prior to Tritium Project .............................................................................................................. 23 15. Wells sampled by EPD, during Tritium Project ................................................................................................................ 25 16. Brier Creek hydrograph ...................................................................................................................................................... 26 17. Locations of collection sites for the 1991 base flow study ............................................................................................... 27 18. Locations of collection sites for the 1992 base flow study ............................................................................................... 29 19. Locations of Tritium Project well cluster sites .................................................................................................................. 31 20. Construction details for a typical ground-water monitoring well ..................................................................................... 32 21. Locations of cores used to develop the geologic framework ......................................,..................................................... 33 22. Map of sampled wells with Tritium Project well numbers ............................................................................................... 35 23. Inset map of sampled wells-with Tritium Project well numbers ...................................................................................... 37 24. Photograph of the shattered portion of Delaigle Trailer Park well #3 .............................................................................. 38 25. Isopleth map based on surface water tritium concentrations of the 1991 base flow study .............................................. 39 26. Isopleth map based on surface water tritium concentrations of the 1992 base flow study .............................................. 41 27. 1992 isopleths with polluted residential wells ................................................................................................................... 43 28. 1992 isopleths with cluster sites ......................................................................................................................................... 45 29. North-south stratigraphic cross-section in central-east Burke County ............................................................................. 47 30. Congaree-equivalent sand isopach ..................................................................................................................................... 49 31. Snapp Formation isopach ................................................................................................................................................... 51 32. Correlation chart ................................................................................................................................................................. 53 33. Structure contour map of top of Lisbon Formation ........................................................................................................... 55 34. Structure contour map of top of Steel Creek Formation ................................................................................................... 57 35. Upper Three Runs aquifer water table surface map .......................................................................................................... 59 36. Comparison of well elevations/water table elevations ...................................................................................................... 60 37. Local potentiometric surface map of the Gordon Aquifer, based on Tritium Project data ............................................. 61 38. Potentiometric surface map of the Gordon Aquifer, based on a regional study (Brooks and others, 1985) ................... 63 39. Lateral variability of total dissolved solids in hydrogeochemical unit 2 .......................................................................... 65
TABLES 1. Daily water use for Burke County, Georgia ...................................................................................................................... 7 2. Summary of unscheduled tritium releases from the Savannah River Site (SRS) ............................................................ 8 3. Yearly SRS atmospheric releases ...................................................................................................................................... 10 4. Tritium analyses from surface waters collected by Georgia Power Co. (1982-1991) ..................................................... 15 5. Tritium analyses from Upper Three Runs aquifer monitoring wells on Plant Vogtle property (1985-1991) ................ 16 6. Construction data for Upper Three Runs aquifer wells on Plant Vogtle property ........................................................... 18 7. Results of initial sampling of wells within the Savannah River corridor by EPD Environmental Radiation Program .. 20 8. Summary of Tritium Project ground-water monitoring wells .......................................................................................... 28 9. Results of initial sampling of residential showing average tritium values >500 picoCuries per liter ............................. 34 10. Results of geophysical logging of polluted water wells .................................................................................................... 36 11. Comparison of tritium values from public and private wells and values estimated from 1992 base flow study ........... 40 12. Results of tritium analyses of water samples from Tritium Project ground-water monitoring wells ............................. 44
iii

13. Comparison of tritium values from monitoring wells and values estimated from 1992 base flow study ....................... 46 14. USGS Trans-River Row Project monitoring wells at Miller's Pond ............................................................................... 48 15. Water levels in the Upper Three Runs and Gordon aquifers ............................................................................................ 52 16. Transmissivity values of the aquifers at cluster siteTR92-1 ............................................................................................. 54 17. Ground-water geochemistry of the Upper Three Runs and Gordon aquifers .................................................................. 56 APPENDICES 1. Identification of the wells sampled before and during the Tritium Project by EPD personnel ......................................... 69 2. Radioactive decay of tritium ................................................................................................................................................. 76 3. Tritium in rainfall samples collected in Burke and Screven Counties ................................................................................ 77 4. Tritium in rainfall categorized by tritium concentration ..................................................................................................... 78 5. Results of tritium analyses from fifteen tritium polluted residential and public water wells ............................................. 79 6. Results of tritium analyses from ground-water monitoring wells ....................................................................................... 82 7. "As built" construction diagrams and data for Tritium Project monitoring wells .............................................................. 86 8. Gordon aquifer wells used in construction ofpotentiometric map shown in Brooks and others, 1985 and Figure 38 .... 93
lV

An Investigation of Tritium in the Gordon and Other Aquifers in Burke County, Georgia
INTRODUCTION
Statement of Problem
The Environmental Radiation Program of the Georgia Environmental Protection Division (EPD) has conducted routine sampling of water wells for tritium in Burke County, Georgia since 1979 (see Appendix 1, p. 70, for sampling dates and analytical results). This sampling program was initiated because of concerns over environmental releases of tritium from the nearby Savannah River Site (SRS) in South Carolina. Public water supply wells (i.e., those used by communities and trailer parks) have been of special interest in this program. As part of this regular sampling program, a water sample was collected on April 4, 1988 from one of three public water supply wells at the Delaigle Trailer Park in eastern Burke County (Figure 1, p. 3). The chemical analysis of the sample from this well (designated by EPD as Delaigle Trailer Park well #3) indicated a tritium concentration of 600 ( 100) picoCuries per liter, which is equivalent to three percent of the U.S. Environmental Protection Agency (EPA) Maximum Contaminant Level (MCL) for tritium. [Note: The EPA MCL is based on the requirement that a person not receive a radiation dose of more than 4 millirem per year from drinking water. For comparison, the natural background dose at sea level from cosmic rays is approximately 26 millirem, and the total radiation dose from all natural sources of radiation is approximately 300 millirem (National Council on Radiation Protection and Measurements, 1987). Drinking two liters of water, containing 20,000 picoCuries per liter, every day for a year results in a radiation dose of 4 millirem]. A follow-up sampling on April 10, 1990 did not detect any tritium in the water from this well (detection limits were only 300 picoCuries per liter at that time). However, samples, collected on January 8 and July 18, 1991 measured 1200 (200) picoCuries per liter, which is six percent of the EPA MCL.
The well data report for public water systems, submitted to the EPD by the driller of the Delaigle well, stated that this well was drilled into the Gordon aquifer to a depth of 300 feet. As reported, this well was constructed using 200 feet of six inch polyvinyl chloride (PVC) casing, which was pressure grouted from 200 feet to the surface. Below this interval, from 200 to 300 feet, the well was constructed of four inch PVC screen. This information and the results of the tritium analyses raised the following concerns: 1. a public water supply well, used by dozens of local residents, was polluted by tritium, in amounts significantly above normal background concentrations (background is approximately 39 picoCuries per liter); 2. at least one of the deep, confined, drinking water aquifers (the Gordon aquifer) in Burke County appeared to

be polluted with tritium; and 3. if the Gordon aquifer were polluted with tritium, that pollution might represent the dilute leading edge ofa tritium plume entering the aquifer through an undetermined pathway from the Savannah River Site in South Carolina.
To address these concerns, Georgia Governor Zell Miller directed EPD to conduct an investigation to: 1. map the extent of tritium pollution in the ground water of eastern Burke County; 2. to evaluate if there was any current or future threat to public health; and 3. to assess how the tritium entered the ground-water regime.
The Georgia Geologic Survey (GGS) Branch of EPD was assigned the responsibility to conduct the investigation. Drilling was initiated at Tritium Project cluster site TR921 (seep. 24 for a description of the monitoring well program) in December, 1991. In March of 1992, the U. S. Department ofEnergy (DOE) provided EPD with $800,000 to perform the study.
Previously, in late July, 1991, the DOE entered into an agreement with the U.S. Geological Survey (USGS) for an investigation of the conditions under which ground water from SRS in South Carolina can migrate beneath the Savannah River, into Georgia aquifers. Originally referred to as the Underflow Project, this study is now known as the Trans-River Flow Project. The USGS Trans-River Flow Project and the GGS Tritium Project have different but complimentary goals and objectives. The focus of the Trans-River Flow Project is the modeling of the deeper aquifers in Georgia and South Carolina. The focus of the tritium investigation is on a documented pollution occurrence in the shallow aquifers. However, each project generates information that contributes to the success of the other project.
Location of Study Area
Burke County is located on the eastern margin of Georgia, along the Georgia-South Carolina border (Figure 1). The northern boundary of the County is approximately 15 miles southeast of Augusta, Georgia, while the southern boundary is approximately 70 miles northwest ofSavannah, Georgia. Burke County is bounded on the east by the Savannah River and to the north by McBean Creek. The DOE's Savannah River Site is located directly across the Savannah River from Burke County.
The northern boundary of Burke County lies 12 to 19 miles south of the Fall Line, which separates the Piedmont from the Coastal Plain physiographic province. Burke County is located in the Vidalia Upland District of the Coastal Plain physiographic province (Clark and Zisa, 1976). The Vidalia Upland District is moderately dissected by streams, and relief varies from 100 to 150 feet. Stream valleys are narrow except for major rivers. Major rivers usually have wide flood plains occupied by wetlands.
The study area is located in the eastern third of Burke County. The study area extends from the Savannah Riveron

the east to Brier Creek on the west, and from the RichmondBurke County line (McBean Creek) on the north to the Burke-Screven County line on the south.
Geology/Hydrogeology
The Cretaceous and Tertiary sedimentary rocks that underlie Burke County were deposited in the form ofseveral alternating layers of permeable sands and limestones separated by less permeable layers of calcareous or kaolinitic clay. Ground water flows more easily through the permeable layers, which, if they yield significant quantities of water to wells and springs, are termed "aquifers" (a glossary of technical terms is provided on p. 70). The less permeable clay layers, termed "aquitards" (or confining beds), form barriers to the upward or downward movement of ground water and yield little water to wells. Aquitards have a strong influence on local and regional ground-water flow and may serve to protect high quality ground water in one aquifer from natural contamination or human induced pollution that may exist in an overlying or underlying aquifer.
In the Tritium Project study area, the entire sedimentary section from the ground surface downward to the unweathered Paleozoic igneous and metamorphic basement is included in the Southeastern Coastal Plain hydrogeologic province (Aadland and others, 1992). The hydrostratigraphic terminology used in this report (Figure 2, p. 4) was developed following the USGS guidelines.
The sandy Tobacco Road Formation and the Irwinton Sand Member of the Dry Branch Formation (Barnwell Group) make up the top 100 to 200 feet of the stratigraphic section in eastern Burke County (Figures 2 and 3). Some beds, lenses, and laminae of clay occur within the Irwinton Sand (Huddlestun and Hetrick, 1986). The calcareous sands of the Griffins Landing Member of the Dry Branch Formation underlie the Irwinton Sand Member and crop out along the bluffs of the Savannah River and in the deeper tributary streams along with the underlying Utley Limestone Member of the Clinchfield Formation. The Barnwell Group forms the local unconfined or water table aquifer and was initially referred to as the "Jacksonian" aquifer. The chrono-stratigraphic term "Jacksonian" was used by Vincent (1982) and Brooks and others (1985) to identify the water table aquifer (on the basis of its Jacksonian (Upper Eocene) age). At the USGS Trans-River Flow Miller's Pond site (Figure 1), the water table aquifer is referred to Upper Floridan (Clarke and others, in review), however Miller (1986) defines the Floridan aquifer as a carbonate aquifer; thus the term "Floridan" is inappropriate. The SRS equivalent of the "Jacksonian" aquifer is referred to as the Upper Three Runs (Aadland and others, 1992). As this is a more proper hydrostratigraphic term, hereinafter, the water table aquifer in eastern Burke County will be referred to as the Upper Three Runs aquifer.
The Lisbon Formation underlies the Barnwell and consists of two members, the McBean Member in the northern part of the study area and the Blue Bluff member in the southern part of the study area (Figure 2). The Blue

Bluff member is dense, clayey, calcareous, and has a low permeability. In central and southern Burke County, the Blue Bluff member lies between the Barnwell Group sands and limestones that make up the Upper Three Runs aquifer and the underlying Gordon aquifer. In this position, the Blue Bluff member of the Lisbon Formation serves as a local aquitard (confining layer). The McBean Member, on the other hand, is a permeable calcareous sand to sandy limestone and has the potential to permit ground-water recharge (by downward leakage) from the Upper Three Runs aquifer into the Gordon aquifer.
The Gordon aquifer in Burke County is made up of two sandy units: the Bennock Millpond sand/Still Branch sand, and unnamed Congaree-equivalent sand (Huddlestun, in preparation). [Note: for this report, the Bennock Millpond sand and the Still Branch sand are informal descriptive names, with Bennock Millpond and Still Branch being the general locations where these lithologic units crop out]. The Gordon aquifer is very porous, transmissive, and yields large amounts of potable water to wells in eastern Burke County. In the central and southern part of the study area, the Gordon has a slight hydrogen sulfide (rotten egg) odor and its use as water supply is less viable.
The Gordon aquifer is underlain by an aquitard formed by the regionally extensive kaolinitic clay at the top of the Oconee Group (Figure 2). This kaolin is characteristically mottled, strongly pigmented, dense, impermeable, and varies in thickness from 10 feet to 50 feet and is stratigraphically equivalent to the Snapp Formation ofSouth Carolina. Hereinafter, in this report, this kaolinitic clay unit will be referred to as the Snapp Formation. In South Carolina, the Snapp Formation is part of the upper portion of the Meyers Branch Confining System (Aadland and others, 1992).
The Snapp Formation is underlain by sand and kaolinitic sand which contains occasional streaks or thin beds of lignite. This unit is the "Ellenton" Formation, which is lowest Tertiary formation in the section (Figure 2). In Tritium Project monitoring well TR92-1C (described on p. 24), this sand aquifer is referred to as "sand within the Meyers Branch Confining System". For the sake ofbrevity, hereinafter, this aquifer will be referred to as the Meyers Branch "aquifer". Beneath the "Ellenton" Formation, in the Cretaceous part of the Oconee Group in eastern Burke County, the stratigraphic section consists of sand beds interbedded with the lenticular kaolinitic clay beds of the Steel Creek Formation, which forms the lower part of the Meyers Branch Confining System, in South Carolina (Aadland and others, 1992). Underlying the Steel Creek Formation is the Gaillard Formation, within which lies the Dublin aquifer. Generally, the Oconee Group consists of a series of fining upward sand sequences with the most permeable sands near the base of each sequence and poorly permeable to impermeable, fine-grained kaolinitic sands or kaolins in the upper part of each sequence. As these clays or kaolins are not areally extensive, the stack of Oconee Group sands is probably hydrologically interconnected and forms a single aquifer in the study area, especially in the updip areas.

2

Formation/

Burke Co.

Lithology

Member

Hydrologic Units

. :C:::,.

-:..:-.:.:-:-..:::.:..::.:..:.:..:..-:::.:.:::-.::..::...::.:..:.:....:......:.::..=:.-::..::.=::..:.:..:::.:.:.:....:.::.







Tobacco
1----

Road Sand
-------

-

1

2 :;::_............. :.-:::::=:=:::::, Irwinton Sand Mbr.

=c., .......... ....-:_.:.-.:::-:-:...:.....

Dry Branch Fm.

;E ~.. ~ . .-.....~. .~.-.~ . .~ . .:_.;_-:.T

Griffins Landing Mbr.

Ill .-;-, ,-,- . - , - - r
IJl : -':- -;'L-.' ... ---:- ~ :

Dry Branch Fm.

<II

C <II

Utley Ls. Mbr.

(.)

w 0

Clinchfield Fm.

.....L

.....L

.....L

.....L

c.

_L_-,

- r - ......L.- r - _j_

_...L_ -,-

McBean Mbr.

J Blue Bluff t mbr.

ii~J;;;g'.;Ift;i:EfiiiI M;:=~ ...6 -,- -,- -,-- -,-

Lisbon Fm.

...,...,.,.,.,..,...,.,..,,.,.,...,..,....,,....,,......,,,.,...,...,.,....,..,..;.------,------,

- .............. -...........

SfiD ::ooh

um...:.:::.;:.:.:.:.....:-.:::.:.-::::.!.:::-:.:::.::::~:=.~......

Unnamed Congaree-equivalent sd.

Upper Three Runs Aquifer
Aquitard Gordon Aquifer

SAS Hydrologic Units
Upper Three Runs Aquifer
(updip)
Floridan Aquifer system
(downdip)
Gordon Confining unit
Floridan Aquifer system

<II C

Snapp Fm.

Aquitard

<II
(.)

Meyers Branch

0

5 <II Q. : :.-......: : : : : : : : : : : : :~..;_~~

iii
11.

.:::.:::..r---=----:;:___--;::,.

I,. : - - -

"Ellenton" Fm.

Meyers Branch "aquifer"

Confining system

. ~ I:{IJ:/)\~:~t~iI~~ ~ C, ............................:::.~.-,-:-----:---:: ..=;=+----------+---------;

Steel Creek Fm.

Aquitard

.................. ... . ... ...=:......:. [ ~ 0 ......:.. ; ":;:...: :=-'=-.-=-.1-----------1-----------------

::Q:,. ca
f

.: : . . . ; .:::

Gaillard Fm.

Dublin Aquifer

Dublin Aquifer

. ..... 0

~: :: : : :.-:-:.::;: :.-.-..::::::

~ Coarse sand
b/{J Medium sand

~ Limestone Li=..:-:7d Marl

b-=-:J Clay
Marl and clay are dotted where sandy.

Figure 2. Stratigraphic and hydrostratigraphic units for Burke County and SAS area. SAS terminology from Aadland and others, 1992.

Regional dip on the top of the Blue Bluff member in eastern Burke County is approximately 10 feet per mile to the southeast. However, there may be local or formational deviations from this dip. No faults have been firmly documented in eastern Burke County. Faye and Prowell (1982) postulated a northeast-southwest trending fault (the Millett fault) extending from near Barnwell, South Carolina to near the town ofPerkins in Jenkins County, Georgia. The town of Girard, in southern Burke County lies along this postulated fault. An investigation that included core drilling, seismic surveys, and other methods, conducted by the Bechtel Corporation in 1982 for the Georgia Power Company, found no evidence for the existence of the Millett fault. Snipes and others (1993) identified a fault (the Pen Branch fault) trending northeast-southwest across SRS in South Carolina and projected this fault into Georgia, citing

elevation differences in the Utley Limestone exposed along the bluffs of the Savannah River immediately south ofPlant Vogtle. Henry (unpublished data, 1994) has identified the possible extension of the Pen Branch fault in the sediments underlying the Savannah River in the vicinity of Hancock Landing. The possibility of the Pen Branch fault occurring in Burke County is suggested by variances in subsurface elevations shown in structural contour and formational thicknesses shown in isopach maps. No surface expression of the fault plane has been observed.
Weather
Burke County has a climate that is characterized by warm, humid summers and mild winters. Data collected at the National Weather Service office at Bush Airport (south

4

14
Average 12 Precipitation (Inches) 10
8
6 4
2 0
Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec

maximum mean minimum

Figure 4. Mean and maximum monthly precipitation values at Bush Airport, Augusta, Georgia (from Gorday, 1985).

N

w

E

s
Figure 5. Wind rose plot (1987 - 1991) for the SAS-Burke County area, showing direction from which the wind blows and frequency of occurrence. (Modified from Aadland and others, 1993. )
6

Table 1 Daily water use for Burke County, Georgia, in 1987 and 1990.
la. Withdrawals in million gallons per day.

Source

1987

%of 1990

%of

volume 1987 total volume 1990 total

(1)

(1)

(2)

(2)

Ground water

8.53

13.82

6.99

9.88

Surface water

53.19

86.18

63.77

90.12

Total

61.72 100.00 70.76 100.00

lb. Withdrawals by water-use categories in million gallons per day.

Category Thermo-electric
generation Irrigation
Public supply Domestic/ Commercial Livestock

1987 (1) 52.09 6.91 1.48 1.05 0.19

1990 (2) 64.54 3.02 2.11 0.91 0.18

(1) Bachtel and Boatright, 1992. (2) Fanning and others, 1992.

of Augusta) indicate that monthly mean high temperatures range from 91 Fin July to 58F in December and January,
0
while monthly mean low temperatures range from 39 F in 0
December to 72 Fm July (Baker, 1979). Mean annual precipitation at Bush Airport measures
approximately 44.6 inches per year (Baker, 1979). The highest monthly precipitation rates usually occur in July and August, which coincides with peak thunderstorm activity. The lowest precipitation occurs in October and November (Figure 4) (Baker, 1979). Information furnished by Plant Vogtle indicate yearly rainfall totals similar to those measured at Bush Airport. The primary wind direction, mea-

sured at the SRS in South Carolina, is from the northeast with secondary wind directions from the west and south (Figure 5).
Cultural Features
Burke County is the second largest county in Georgia with an area of 834.1 square miles (Bachtel and Boatright, 1992). As of the 1990 Census, the population of Burke County is 20,579, of which 72.3 percent is rural and 27.7 percent urban. As of 1989, the per capita income for the county was estimated at $10,380, and ranked 147th of the

7

Table2 Yearly totals of SRS planned tritium atmospheric releases (1954 - 1992).

Year
1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966

Released Tritium-Ci
(Curies) 216 (1)
36,100 (1) 469,000 (1) 1,200,000 (1) 2,340,000 (1) 1,050,000 (1) 951,000 (1) 886,000 (1) 1,110,000 (1) 1,130,000 (1) 1,520,000 (1) 744,000 (1) 675,000 (1)

(1) Murphy and others, 1991. (2) Arnett and others, 1992. (3) Aadland and others, 1993.

Year
1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

Released Tritium-Ci
(Curies) 689,000 (1) 762,000 (1) 469,000 (1) 513,000 (1) 621,000 (1) 822,000 (1) 601,000 (1) 937,000 (1) 518,000 (1) 304,000 (1) 381,000 (1) 360,000 (1) 333,000 (1)

Year
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992

Released Tritium-Ci
(Curies) 317,000 (1) 395,000 (1) 434,000 (1) 618,000 (1) 786,000 (1) 667,000 (2) 425,000 (3) 590,000 (3) 462,000 (3) 310,000 (3) 250,000 (3) 200,000 (3) 156,000 (3)

159 counties in Georgia. The largest contributor to the economy ofBurke County is the Vogtle Electric Generating Plant, operated by Georgia Power Co. As of 1991, there were 20 manufacturing plants in the county, producing apparel, furniture, lumber, and textiles. As defined by 1987 data, 38.1 percent (317.79 square miles) of Burke County land is utilized as farmland, ranking 45th of 159 counties. Of the farmland, 38 percent (120.76 square miles) is identified as harvested cropland. As of 1990, irrigated cropland accounts for 21.92 percent (26.48 square miles) of the harvested cropland (Fanning and others, 1992). The major crops of Burke county are peanuts, soybeans, com, cotton,

wheat, oats, and rye. The production ofelectricity by Georgia Power's Plant
Vogtle represents the county's largest user of water, with 1990 estimates of 64.54 million gallons per day, while irrigation (3.03 millon gallons per day) and public supply (2.11 millon gallons per day) rank second and third (Fanning and others, 1992). 1987 estimates (Bachtel and Boatright, 1992) indicate that 37 percent of the Burke County population is served by public water supplies. Water use data for Burke County, for 1987 and 1990, are shown in Table 1 (p. 7).

8

Table3 Summary of significant unplanned releases from the Savannah River Site.

Date of Release 05102n4 (1) 12131n5 (1) 03/37/81 (1) 07/16/83 (1) 03/23/84 (1) .09/02/84 09/07/84 (1) 01/31/85 (1) 03/27/85 (1) 07/31/87 (1) 03/01/88 (1) 06/07/88 (1) 10/06/88 (1) 12/07/88 (1) 12/22/91 12/25/91 (2) 05/25/92 (3) 07/12/92 (3)

Release in Curies 479,000 182,000 33,000 56,000 7,500 57,900 9,285 19,422 172,000 20,000 3,650 7,000 3,500 5,700 80 12,000

Pathway Atmospheric

Percent Tritiated
Water
<1

Atmospheric

99.4

Atmospheric

99.7

Atmospheric

~1

Atmospheric

~70

Atmospheric

99

Atmospheric

54

Atmospheric

99.9

Atmospheric

2.7

Atmospheric

14

Atmospheric

4

Atmospheric

-11

Atmospheric

99.5

Percent Elemental
Tritium >99 0.6 0.3 ~99 ~30 1
46 0.1 97.8 86 96 ~89 0.5

Release Area
Separations area
Separations
area
Separations
area
Separations
area
Separations
area
Separations
area
Separations
area
Separations
area
Separations
area
Separations
area
Separations area
Separations
area
Separations area

Surface water

100

0

Pen Branch

Atmospheric Atmospheric

Not specified (4)
Not specified (4)

Not specified (4)
Not specified (4)

Separations
area
Separations
area

(1) Murphy and others, 1991. (2) Arnett and others, 1992. (3) Aadland and others, 1993. (4) Primarily elemental tritium.
10

5000---.-----------------------~

4000 - - - - - - - - - - - - - - - - - - - - - - - -
3000 - Tritium values picoCuries per liter
2000 - -

1000 - -

o-~~~-1----..

1982

1984

1986

1988

1990

1992

Figure 8. Bar graph showing average yearly rainfall tritium values at EPD station # 11. Location is shown in Figure 7.

History of Tritium Releases from Savannah River Site
Tritium is a radioactive isotope of hydrogen with a halflife of 12.35 years (Fritz and Fontes, 1980). The unit of measurement for tritium used throughout this report is the picoCurie per liter (1 trillionth of one Curie per liter), which represents 0.037 electron releases per minute.
From the initiation of SRS operations in 1954 through 1988, approximately 25.6 million Curies (Ci) of tritium were released to the atmosphere and surface waters of the site, and approximately seven million Ci of tritium were placed into seepage basins and burial grounds (Murphy and others, 1991 ), with the largest year!y release (2.4 million Ci) in 1958 (Aadland and others, 1993). The yearly totals of planned atmospheric releases are listed in Table 2 (p. 10) and graphically indicated on Figure 6 (p. 9). Because radioactive decay should have removed over three fifths of this material, there are currently about 9.9 million Ci of tritium from the atmospheric and stream releases remaining in the environment and approximately 3.2 million Ci remaining in the seepage basins and burial grounds (Murphy and others, 1991). A brief table showing the radioactive decay of tritium over a 100 year period is provided in Appendix 2 (p. 77).
Tritium is released to the atmosphere as water vapor and as a gas from routine reactor operations, recovery of transuranic elements, recovery of tritium, heavy water rework, and laboratory research. Of the original 24 million Ci in atmospheric releases, 28.0 percent were from reactor

areas, 71.6 percent were released from separations areas, and 0.4 percent were from other facilities (Murphy and others, 1991). Of the 1.5 million Ci released to Savannah River Site streams, 75.0 percent were derived from reactor areas, 15.5 percent were from separations facilities, and 9.5 percent were from other facilities.
In addition to releases during normal operations, there have been inadvertent (unplanned) releases of over 1 million Ci of tritium due to mechanical or human process errors. A brief summary of the most significant unplanned releases is provided in Table 3 (p. 10).
As of 1993, none of the reactors at the Savannah River Site are operating. Total (planned) atmospheric releases of tritium have steadily declined in recent years from 595,000 Ci in 1987 to approximately 156,000 Ci in 1992 (Aadland and others, 1993). This represents a decline of approximately 20 percent per year (Aadland and others, 1993). In 1992, 69 .75 percent of the atmospheric releases were from the separations facilities, 29.95 percent from the reactor areas, 0.31 percent from heavy water rework, and 0.005 percent from diffuse sources including ponds and polluted land areas. Unplanned atmospheric releases during 1992 included 80 Ci from the K reactor (May 25) and 12,000 Ci from the H separations area (July 12).
Tritium in Rainfall in Burke County
The natural abundance of tritium in the terrestrial environment is extremely low. Tritium is produced natu-

12

9000-------------------------------,
~0--------- - --------------

7000 - - - - - - - -

6000 - - - - - - - -

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

-- - 5000------
Tritium values picoCuries per liter
4000- -----

3000- -

2000 - -

- 1000 -

- - -------

I
0

1983

1985

1987

1989

1991

1993

Figure 9. Bar graph showing average yearly rainfall tritium values at EPD station #35. Location is shown in Figure 7.

rally, in small quantities, in the upper atmosphere through the interaction of cosmic rays with atmospheric nitrogen. This natural production results in tritium concentrations in rainfallof13 to80picoCuriesperliter(Gat, 1980). Robertson and Cherry (1989) calculated from field data that natural concentrations of tritium in rainfall, in Ohio, between 1920 and 1952, were approximately 10 picoCuries per liter. However, there is significant latitudinal, seasonal, and geographical (continental versus oceanic) variation in natural levels of tritium in rainfall (Gat, 1980; Fontes, 1980). Starting in 1952, the low background concentrations of tritium in rainfall were overwhelmed by tritium produced

during the atmospheric testing of nuclear weapons. This "bomb" tritium has declined since the cessation of atmospheric testing in 1963, and, as of 1978, "bomb" tritium accounts for over 16picoCuries per literoftritium in rainfall (Fontes, 1980) above the natural background concentrations.
Current concentrations of tritium in rainfall near Atlanta, Georgia, including both natural and "bomb" tritium are 39 picoCuries per liter (Rose, 1993). Because Atlanta is approximately 150 miles from the Tritium Project study area, SRS activities should have only minor effects on these measurements. Therefore, the concentration of tritium

13

Figure 10.

Directional distribution of tritium in rainfall for SAS-Burke County area, based on analyzed rainfall samples (1982 - 1986). Values are stated in picoCuries per liter. (Modified from Murphy and others, 1991.)

found in rainfall near Atlanta can be treated as background levels for comparison with measurements taken in Burke County.
Tritium concentrations in rainfall have been measured by the Environmental Radiation Program of EPD from selected sites in eastern Burke County since 1981 (Figure 7, p. 11). In addition, DOE maintains rainfall stations in Burke, Richmond, and Screven counties but also has stations in Augusta, Savannah, and Macon, Georgia. Elevated tritium concentrations have been observed at all EPD and DOE stations. Tritium values from samples collected from EPD sites on Hancock Landing Road (Station 11) and at the Plant Vogtle Simulator (Station 35) (Figure 7), both near the Savannah River, are of particular interest to the present study because these sites lie closest to the area of maximum concentration of tritium seen in Burke County ground water. Average yearly rainfall tritium values for Stations 11 and 35 are shown on Figures 8 and 9, respectively.
The highest tritium concentration measured in rainfall at the Hancock Landing Road site between 1981 and 1994 was 11,500 picoCuries per liter, from a sample collected December 24, 1984. As late as January 7, 1993, a rainfall sample collected from this site showed a tritium content of 7,000 picoCuries per liter, however, this was the first sample to measure above 5,000 picoCuries per liter since December 17, 1987 (Appendix 3, p. 78). A compilation of tritium values above 5000 picoCuries perliter (25 percent of

EPA MCL) from EPD rainfall monitoring stations in eastern Burke and Screven Counties (Figure 7), including the Hancock Landing Road site, is included in Appendix 3. However, most rainfall samples collected by the EPD in Burke County contain only moderately elevated concentrations of tritium. Eighty-seven percent of the EPD samples
(n =726) from seven sites in Richmond, Burke and Screven
Counties contain less than 2,500 picoCuries per liter of tritium (Appendix 4, p. 79).
Murphy and others (1991) produced a map showing the distribution of average tritium in rainwater around the Savannah River Site between 1982 and 1986 (Figure 10). This map was based on rainwater samples collected from 33 stations within a 25 mile radius around the Savannah River Site, including four stations in Georgia. The map shows that rainfall over eastern Burke County, during the period from 1982 to 1986, had average tritium concentrations exceeding 1,500 picoCuries per literand that, near the Savannah River, the rainfall had average tritium concentrations exceeding 6,000 picoCuries per liter.
Tritium in Ground Water
The potential sources of tritium in ground water are recharge from rainfall, recharge from rivers or lakes, and transport from other aquifers. Recharge from rainfall includes both tritium in current rainfall and tritium that has

14

Table4
Tritium concentrations ofsurface water samples collectedby GeorgiaPowerCo/Southern Nuclear Operating Co., August, 1982 - October, 1991. Locations shown in Figures 7 & 11. Analyses courtesy of Southern Nuclear Operating Co. Samples collected on 19 November 1991, 27 October 1992, and 20 October 1993 by EPD personnel. Tritium concentrations in picoCuries/liter.

Date
08ft)3/82 08/10/82 07ft)5/83 10/04/83 0lft)3/84 04ft)3/84 07/10/84 10ft)8/84 03ft)5/85 05/13/85 07/22/85 10/15/85 01/16/86 04ft)3/86 07/14/86 10ft)7/86 01/12/87 04/09/87 10/22/91 11/19/91 lOfl.7/92 (1) 07/27/93 lOfl.0/93 (3)

Blue Bluff SpringRiver Mile 150.1
3810 1610 1480 1850
1400200 (2) 1300200 (2)

Plant Vogtle SpringRiver Mile 150.9
3333 2460 2490 3930 2820 2000 2300 2710 2580 2650 2600 3310 3100 2430
1800200

Mallards Pond
1280 2540 2320 2290 2120 2120 1820
2030
2080 1810 1670 1850 1850 1750 1700 1660
1400200
Dry

Beaverdam Creek at River Road 1910
1390 1300200 1100200 800100

(1) Collected during 1992 base flow study. (2) Collected from nearby springs on Blue Bluff at River Mile 149.5 (approximate). (3) Collected during 1993 base flow study.
15

Tables
Tritium concentrations in ground water samples from Upper Three Runs (water table) aquifer monitoring wells on Plant Vogtle property (see Figure 11). Analyses courtesy of Southern Nuclear Operating Company. Tritium values in picoCuries per liter.

Date

129 142 800 801 802 803A SOSA 806B 807A

03/05/85

1950 940 1270

1270 1940

05/13/85

1970 1740 1190 1880

1340 1950 3510

07/22/85

2000

920 2090

1320 2080

3770

10/15/85

2530

01/15/86-01/ 18/86

2080

04/03/86-04/ 04/86

2110

10/17/86-07/ 18/86

2160

10/07/86-10/ 10/86

01/06/87-01/ 12/87

04/09/87-04/ 13/87

2050

08/26/87-08/ 27/87

2090

12/28/88

580 2020 2220 1750

637

2430 922

1220

934

1730

1160

1090 1990 969 1370

673

1910 996 1150

766

1770

1200

886

1730

1060

827

1640

775

1480

1290 1300

3940 3360 2980 1340 3200 2680 2960 3020
2360

10/22/91

1530

1750

2170

accumulated in the vadose zone from previous rainfall. Current background tritium concentrations (as measured in the shallow ground water of the north Georgia Piedmont) average 62 picoCuries per liter and range up to 109 picoCuries per liter in soil and weathered rock material (Rose, 1992). Tritium behaves as a conservative tracer of ground-

water flow (Fontes, 1980; Davis and others, 1985; Allison, 1988; Rose, 1992; Rose, 1993) and has been used extensively as a tracer in ground-water studies since the 1950's. Tritium has many properties that make it an ideal tracer. It primarily occurs as tritiated water (Gat, 1980), replacing one (or both) ofthe hydrogens in the water molecule. It does

16

Table6
Construction data for Upper Three Runs aquifer monitoring wells on Plant Vogtle property. Wells in use from 1985 - 1991. Locations shown in Figure 11. Infonnation from Southern Nuclear Operating Company and USGS.

Plant Vogtle well#
129 142 800 801 802 803A 805A 806B 807A

Ground elevation (feet above M.SL.)
215.8 231.2 213.7 214.8 235.0 218.3 232.7 210.0 202.0

Casing depth (feet below
surface) 92 85 69 62.5
76.75 57 95 55 65

Screened interval (feet)
5
10 20 20 9 20 20 10 10

not adsorb onto the surface of rock particles, nor is it absorbed into the minerals' structure (Davis and others, 1985). Isotopic fractionation of tritium in ground water is of relatively minor importance (Fontes, 1980). Tritium is not significantly affected by reactions other than radioactive decay (Freeze and Cherry, 1979; Dincer and Davis, 1984).
The potential mechanisms of movement of tritium in ground water include advection, convection, mechanical dispersion, and molecular diffusion (Freeze and Cherry, 1979; Davis and others, 1985; Domenico and Schwartz, 1990). Advection is the bulk movement of ground water in response to local or regional hydraulic gradient. In most aquifers, advection is the dominant process of water movement. Convection (free or natural convection) is the movement ofground water due to difference in density. Although tritiated water molecules are slightly heavier than normal water molecules, the difference in density is minor. Mechanical dispersion is the mixing ofground water during the process of advection caused by local variations in velocity and path followed by water molecules. Mechanical dispersion is a major factor in the dilution and spread ofa pollutant away from the advective path. Molecular diffusion is the mixing of ground water due to random molecular motions.

Molecular diffusion in rapidly moving ground water is generally considered to be of minor importance compared to mechanical dispersion (Davis and others, 1985). However, Robertson and Cherry (1989) found that, when dispersion and linear ground-water velocity are low, molecular diffusion controls the mixing of tritium in ground water.
The earliest surface water samples indicating the presence of measurable tritium in eastern Burke County, were collected by Georgia Power Company and/or Southern Nuclear Operating Company (both Southern Company subsidiaries), prior to the start-up of Plant Vogtle. From August, 1982 through October, 1991, surface water samples (Table 4, p. 15) were collected from spring sites along the Savannah River bluffs at River Mile 150.1 and River Mile 150.9, from a site at the head of Mallards Pond (Figure 11, p. 17) and the River Road crossing of Beaverdam Creek (Figure 17, p. 27). Bar graphs show comparative tritium concentrations from the Plant Vogtle spring at River Mile 150.9 (Figure 12, p. 19) and Mallards Pond (Figure 13, p. 21), which were sampled on a sporadic basis from April, 1984 through July, 1993.
The initial collection of water samples from Upper Three Runs aquifer monitoring wells on Plant Vogtle property ( Table 5, p. 16; Figure 11) began on March 5, 1985.

18

I I
I I
I

Table7
Results of initial sampling of wells within the Savannah river corridor by the EPD Environmental Radiation Program, prior to the initiation of the Tritium Project Tritium Project numbers and locations are shown in Figure 14 and listed in Appendix 1.

Tritium Project Number

EPD Station Number

Name

Date of First Sample

Results

34

1

1-20 Welcome Center, Augusta

04/13/87

Below detection

7

6

Augusta Lock & Dam

04/13/87

Below detection

39

8

Residence - Bennock Mill Rd.

10!04/88

Below detection

100

12

Plant Vogtle Visitor Center

04/13/87

Below detection

37

14

Girard Main Well

04/13/87

Below detection

Below

87

17

Stoney Bluff Park

10/05/88

detection

/,

35

20

U.S. 301 Welcome Center, Screven Co.

04/13/87

Below detection

105

25

Waynesboro Main Well

04/13/87

Below detection

107

26

Old Shell Station - Hancock Landing Rd./ River Rd.

04/13/87

Below detection

3

29

Delaigle Well #3 (aka A&A Store#3)

09/22/87

Below detection

98

35

Plant Vogtle Simulator

09/22/87

Below detection

94

37

Viola Brigham Residence

04/13/87

Below detection

From this date through October 22, 1991, samples were collected on a sporadic basis from a total of nine monitoring wells. Basic construction details for the Plant Vogtle shallow monitoring wells (depth and screened intervals) are listed in Table 6.
On April 13, 1987, the Georgia EPD Environmental Radiation Program began the periodic sampling of eight

wells (Table 7; Figure 14, p. 23) along the Savannah River corridor from the Georgia Welcome Center, Interstate Highway 20, in Augusta, southward to the Georgia Welcome Center, U.S. Highway 301, near Sylvania, in northeastern Screven County. Four more wells were added later, including the Delaigle Trailer Park well #3.

20

Acknowledgments
We wish to extend our appreciation to the Burke County property owners and residents for their invaluable assistance and cooperation with our drilling and sampling activities during the Tritium Project. For permission to install cluster sites on their properties, we wish to thank the following: Georgia Power Company, Mr. Avner Delaigle, Mr. R. W. Mobley, Mr. Earl Nally, and Thomson Oak Flooring Company. For allowing the geophysical logging of residential wells, we wish to thank the following: Mr. Avner Delaigle, Mr. Mark Jackson, Mr. Lamar Paul, Ms. Rose Johnson, Mr. Frank Wimberly, Mr. Larry Sconyer (Hug-A-Hog Plantation), Mr. Ralph Greer, Mr. George Wilson, and Ms. Marie Vann (Walnut Run Ostrich Farm). For allowing the storage of materials, we wish to thank the following: Mr. Marshall A. Miller and Mr. Avner Delaigle. For technical assistance and information, we wish to thank the following: Mr. Gerald Grainger (Southern Company), Mr. Bill Ollinger (Southern Nuclear Operating Company), Mr. Carl Carswell (Georgia Power Company, Plant Vogtle), Mr. Carlton Chambers (Georgia Power Company, Plant Vogtle Land Office), Dr. Chet Nichols (Westinghouse Savannah River Corporation); Mr. John Clarke, Mr. Fred Falls, Ms. Joan Baum, and Mr. Chris Leeth (USGS); Dr. Seth Rose, Mr. Philip James (Georgia State University); Mr. Cliff Blackman, Mr. James Gary, and Mr. Robert Rosson (Environmental Radiation Laboratory, Georgia Institute of Technology); and Dr. Vernon J. Henry (Georgia Southern University).
The Tritium Project Technical Advisory Committee includes the following persons: Mr. John Clarke (USGS), Mr. Tom Temples (DOE); Dr. Van Price, Jr. (Westinghouse Savannah River Corporation); Dr. Leland T. Long and Dr. William Chameides (Georgia Institute of Technology); Dr. Wade Nutter and Dr. James Spaulding (University of Georgia).
This project was supported, in part, through a Cooperative Agreement with DOE (Cooperative Agreement Number DE-FC09-92SR18268). This report does not constitute an endorsement by the Department of Energy of the views expressed in this report.
PROCEDURES
Study of Water Wells
Prior to the initiation of the Tritium Project, the EPD Environmental Radiation Program periodically tested twelve water wells in eastern Burke County for tritium (Figure 14, Table 7, p. 20;
Following the discovery of tritium in the Delaigle Trailer Park well #3, EPD expanded the sampling of private and public supply wells in eastern Burke County. Private water supply wells include domestic (residential), supply, commercial (business and institutional), and agricultural water wells; public water wells include both publicly owned

and privately owned wells that provide water to at least 25 people or have at least 15 hookups. The wells were selected to provide a relatively uniform geographic coverage throughout the study area. Other wells were sampled at the request of property owners. By February, 1994, a total of 109 wells had been sampled (see Figures 15, 22 and 23 for locations and Appendix 1 for analytical results).
During the current study, water samples were taken at each residence from the outdoor faucet closest to the well. The water was allowed to run for approximately ten minutes prior to sampling in order to clear the waterpipes ofstanding water. The Environmental Radiation Laboratory at the Georgia Institute ofTechnology analyzed the water samples for tritium. The detection limit for tritium at this laboratory is 100 picoCuries per liter. As part of the EPD's quality assurance procedures, blank samples of deionized water (with no tritium) and duplicate samples were submitted along with the well samples. Within the sample of 109 water wells, all wells that yielded tritium concentrations at or above 500 picoCuries per liter were considered radiologically anomalous or polluted. The 500 picoCuries per liter "threshold" is 2.5 percent of the EPA MCL for tritium of 20,000 picoCuries per liter, and is slightly higher than the analytical detection limit of 100 picoCuries per liter. The polluted water wells were periodically resampled during the course of the investigation.
The USGS conducted geophysical logging of nine of the anomalous water wells. The specific types ofgeophysical logs used depended on the availability of equipment. Geophysical logging provided information on the total depth of the well, the screened or open interval within the well (indicating the aquifer being sampled), whether the well had been properly grouted, and any damage to the well casing. Water level measurements were also taken during the geophysical Jogging.
Base Flow Studies
Water moving through a stream channel is derived from three sources: overland flow, interflow, and base flow (Freeze and Cherry, 1979; Domenico and Schwartz, 1990). Overland flow is rain water that moves over the land surface into stream channels. Interflow is the water derived from soils and sediments above the water table. Base flow is water derived primarily from the water table aquifer. During the Fall season in Burke County, precipitation is at its lowest level and soil moisture has been depleted by crops and other vegetation. During this time period, streamflow is typically very low and is primarily derived from base flow. Water samples collected from streams during such periods provide a reasonable sample of water in the Upper Three Runs (water table) aquifer. "Ideal" base flow conditions are identified from recorded stream flow discharge rates on streams with USGS gaging stations (Figure 16, p. 26), one of which lies within the study area. Because the true depths of most wells in eastern Burke County are uncertain and the number ofsprings and streams far exceeds

22

the number of wells in eastern Burke County, a base flow study is a good method to evaluate the areal distribution of tritium in the Upper Three Runs aquifer.
The preferred sampling locations for base flow studies are springs and small first-order streams (close to the origin of the streams). Second and third-order streams are less preferred sampling sites as these larger streams represent mixing of ground water from several sources, producing a composite sample. Areas immediately downstream from swamps, ponds, and lakes are the least preferred collection sites because tritium concentrations in these surface water bodies may be biased by recent rainfall and surface runoff. Second and third-order streams may be utilized, where more favorable sites are not available due to time, personnel, and accessibility constraints.
1991 Base Flow Study On November 19 and 20, 1991, EPD personnel conducted a base flow study over an area covering parts of three USGS 7.5 minute quadrangles (see Figure 17, p. 27, for sampling locations). A total of 53 samples were collected, primarily from road crossings of area streams.
1992 Base Flow Study From October 26 through 28, 1992, GGS personnel conducted the second base flow study over an area covering parts of seven USGS 7.5 minute quadrangles (Figure 18, p. 29). A total of 126 samples were collected, primarily from springs and small first-order streams, and road crossings when more preferable sites were not available. In order to achieve the proper spacing, a grid system was established for the area to be sampled. The grid spacing was not uniform over the sampling area. Because the Shell BluffLanding 7.5 minute quadrangle map contained thirteen of the fifteen polluted domestic water wells, this area was designated as a "dense sampling zone." Within this "dense sampling zone," four samples were collected per square mile wherever possible. This "dense sampling zone" extended two to three miles into the margins of the adjacent quadrangles. A buffer, approximately four miles wide, surrounding the "dense sampling zone" was defined as a "moderate sampling zone" in which one sample per square mile was collected. The area beyond the "moderate sampling zone" buffer was defined as the "light sampling zone." Within this "light sampling zone" one water sample was collected within every four square miles. The general boundaries of the 1992 study were the Burke-Screven County line to the southeast, the Savannah River to the east, McBean Creek to the north, the Norfolk Southern System railroad tracks west of Georgia Highway 56 to the west, and U.S. Highway 25 to the southwest.
Installation of Monitoring Wells
During the Tritium Project, GGS drilling crews installed fifteen ground-water monitoring wells in clusters at six sites in eastern Burke County (see Figure 19, p. 31, for

well cluster locations) in order to obtain in situ samples of ground water for tritium analyses. Continuous cores were taken at each monitoring well site prior to the installation of the monitoring wells. Following the completion of drilling, the USGS geophysically logged each core hole. Correlation of the cores and geophysical logs allowed the selection of the most appropriate interval to place the screen within each monitoring well. Screened intervals were selected on the basis of high permeability ofthe sediment, low clay content, thickness of the high permeability zones (as interpreted from core samples and/or geophysical logs), and the projected elevation of the aquifer to be sampled. In most wells, ground-water sampling was restricted to a ten foot screened interval. See Table 8, p. 28, for aquifers, depths, and screened intervals.
The construction details for a typical monitoring well are shown in Figure 20 and "as built" construction diagrams are illustrated in Appendix 7. After drilling and installation, each well was developed using a surge block and an air compressor. The development process removes the remnants of drilling mud and formational clays, silts, and fine sands from the vicinity of the screen and otherwise corrects damage to the aquifer caused by the drilling process. After development, a concrete pad was poured around the top of each well and a locked, protective steel housing was placed over the well. A permanent submersible pump was installed in each well.
The first four cluster sites (TR92-l through TR92-4) were placed between the locations of the polluted water wells. The locations of the two subsequent cluster sites (TR92-5 and TR92-6) were based on the results of tritium analyses from the first four cluster sites, from the polluted water wells, from base flow sampling, as well as site availability and suitability.
Seven monitoring wells at the USGS Miller's Pond cluster site (Figure 19, p. 31) were also available for periodic sampling. These wells are part of the ongoing USGS Trans-River Flow Project.
Using the calculated water volume of each well, a minimum of three well volumes were purged prior to sampling. Periodic water level measurements were also taken in each well.
Description of Cores
The rock units of eastern Burke County act as the "plumbing" through which the ground water flows. Careful examination and description of cores allow the reconstruction of that "plumbing". Staff of the GGS and the USGS examined 9,244 feet of core including material from all six cluster sites, a core from the USGS Miller's Pond site, one core drilled by Georgia Power Company at Plant Vogtle, one core drilled by the GGS in 1991 near McBean Creek (as part of another project), and eight cores drilled by the Bechtel Corporation in southern Burke County in 1982 (Figure 21, p. 33).

24

1,000

-w

C,
a:

"O C:

<( 0

I

CJ
Cl)

(/)

u ... ~

~ -0 aCl).
~ Cl) $

-z CJ

w<(

:0
:::,

CJ

~

-

"'"' , ....I.I I/""" ,....
I / I/

100 31
10,000

5

10

15

1920

25

30

NOVEMBER 1991
Collection dates: November 19 and 20, 1991.

-w

C,
a:

"O C:

<( 0

u ... I

CJ
Cl)

(/)

~

0 aCl). 1,000

jt\

~ Q)

~ $

..... '-

-

-z -~

<(
w

.:0::, CJ

~

Ii,.

I,"

~I"'

....
'L-
-- "'i...........

100

30

5

10

15

20

25262728 30

OCTOBER 1992
Collection dates: October 26, 27 and 28, 1992.

02197830 Brier Creek near Waynesboro, Georgia. Note difference in vertical scales.

Figure 16. Hydrographs showing stream discharge rates for Brier Creek (near Waynesboro) during the 1991 and 1992 base flow studies. (From USGS data.)

Seismic Survey of Savannah River Channel

surveys were conducted by Dr. Vernon J. Henry, of Georgia

Southern University, and others, along the Savannah River

During the course of the Tritium Project, two seismic channel, within the study area. The seismic surveys were

26

Table8 Summary of Tritium Project ground water monitoring wells.

Well Number 1R92-1A 1R92-1B 1R92-1C 1R92-1D 1R92-2A 1R92-2B 1R92-3A 1R92-3B 1R92-4A 1R92-4A2 1R92-4B 1R92-5A 1R92-5B 1R92-5C 1R92-6A 1R92-6B

Aquifer Upper Three Runs

Elevation - feet above M.S.L.
235

Gordon

235

Meyers Branch

235

Dublin

235

Upper Three Runs

285

Gordon

285

Upper Three Runs

195

Gordon

195

Upper Three Runs

192

Upper Three Runs

192

Gordon

192

Upper Three Runs

235

Gordon

235

Gordon

235

Upper Three Runs

240

Gordon

240

Depth below Surface (feet)
105 225 310 370 120 330 75 205 60 90 190 165 305 315 141 250

Screened Interval (feet)
90- 100 210-220 290-300 345 -355 105 - 115 310- 320
55 -65 185 - 195 55-65 75 - 85 (1) 175 - 185 145 - 155 275 - 285 200- 300 (2) dry hole (3) 180- 200 (4)

(1) Redrilled to greater depth to produce more water. (2) Designed to duplicate driller's records for Delaigle Trailer Park well #3. (3) Screened in impermeable part of Utley Limestone. (4) Designed to also meet the needs of the USGS Trans-River Flow Project.

28

conducted to address the following objectives: 1) evaluation of possible breaching of the shallow aquifers (Upper Three Runs and Gordon) by the Savannah River; 2) evaluation of the thickness of the Savannah River alluvium; 3) correlation ofseismic sequences with the upper Cretaceous/ lower Tertiary stratigraphic units identified in regional water wells and Tritium Project cores; and 4) possible identification of the Pen Branch fault.
A high resolution seismic survey was conducted during October 10-13, 1992 between theRichmond-BurkeCounty line and the Burke-Screven County line, a distance of 70 river miles. This study was performed using a boat-drawn EG&G Model 225 UNIBOOM system, with data recorded on both a magnetic tape and a graphic (analog) recorder. The data were used to estimate the thickness ofriver channel alluvium, to assess breaching of shallow aquifers, and to correlate the stratigraphic units, where possible. The information from the seismic study were compared with stratigraphic data derived from eighteen auger holes drilled along the adjacent river floodplain in Georgia and South Carolina by the USGS (Leeth and Nagle, in preparation).
A medium resolution seismic survey was conducted during August 23-27, 1993 between Hancock Landing and the Georgia Power Boat Ramp, approximately one mile down-river from Plant Vogtle. This survey was carried out using a boat-drawn Bolt 600B air gun with one cubic inch and ten cubic inch chambers.
Ground-Water Geochemistry

(Ni), chromium (Cr), and molybdenum (Mo) using inductively coupled plasma spectrometry, and for chloride (Cl), sulfate (SO4), nitrate (NO3), fluoride (F), and phosphate (PO4) using ion chromatography. Calcite saturation indices were calculated for all samples. The University ofGeorgia CooperativeExtension Serviceperformedall chemicalanalyses, except for tritium analyses, which were performed by the Environmental Radiation Laboratory at the Georgia Institute of Technology.
Aquifer Testing
The four monitoring wells at site TR92-1 (Figure 19, p. 51) were subjected to pumping tests by a group ofresearchers from Clemson University, Department of Earth Sciences (Moore and others, 1992). This aquifer testing was carried out as part of the previously mentioned USGS Trans-River Flow Project. The purpose of the testing was to obtain transmissivity data on each aquifer and to estimate the degree of interconnection (leakage) between the aquifers. The aquifer testing was accomplished by pumping each well in turn, beginning with the shallowest (TR92-1 A), while using the other wells as observation wells. Pumping was conducted using temporarily installed multi-stage submersible pump, to allow for aquifer testing at varying flow rates. While the pumping was underway, pressure transducers installed in the observation wells were used to record any drawdown of the water level, which would indicate aquifer interconnection.

As rainwater falls and seeps into the soil, its chemistry is initially affected by the gases and dust-sized particulate matter in the atmosphere and then by the particulate matter and gases within the soil. As the water percolates downward, its chemistry is further altered by chemical reactions with the sediments. Because the chemical composition of the sediments of each aquifer is often different, the ground water contained within each aquifer may also be chemically different.
Dr. Seth Rose and Philip James, of Georgia State University, performed a study of ground-water geochemistry to define and interpret geochemical variation within the Upper Three Runs and Gordon aquifers in eastern Burke County (Rose and James, 1993). The goal of the study was to characterize the geochemistry of each aquifer and to suggest the pathways for water movement within the aquifers. The Rose and James (1993) study included seventyfive water samples from a variety of sources including private and public water supply wells (56 samples), groundwater monitoring wells (10 samples), springs (4 samples), and first-order streams during periods dominated by base flow (5 samples). Each sample was measured for alkalinity, pH, and specific conductance within 48 hours ofcollection. Samples were analyzed for ionic concentrations of calcium (Ca), magnesium (Mg), sodium (Na), potassium (K), silicon (Si), manganese (Mn), iron (Fe), aluminum (Al), boron (B), copper (Cu), zinc (Zn), lead (Pb), cadmium (Cd), nickel

RESULTS
Tritium in Water Wells
As of early 1994, 109 private and public water supply wells had been sampled during the Tritium Project (Figures 22, p. 35, and 23, p. 37). Tritium Project numbers (#1 through #109) were assigned to every sampled well and the subsequent tritium analysis results are shown in Appendix 1. Using the previously identified "threshold" value of 500 picoCuries per liter, as of early 1994, a total of fifteen wells had been identified as polluted (Table 9, p. 34). The first ten polluted wells, identified between July, 1991 and October, 1992, were selected for USGS geophysical logging, to establish the accurate depths of the wells. Property owners granted permission for geophysical logging of nine of these first ten polluted wells. One property owner, however, declined permission for geophysical logging because of the age and condition of his well. The results of the geophysical logging, which took place in September, 1992 and February, 1993, are shown in Table 10. The geophysical logging of the nine wells demonstrate that eight of the nine wells are drawing their water from the Upper Three Runs aquifer. The logging results of the ninth well (Delaigle Trailer Park well #3) were inconclusive, but a downhole video camera study of the well in April, 1993, showed that the 6 inch PVC casing was shattered from 154 feet below the top of the

30

Locked protective 1------1 steel housing

Casing

Concrete pad

4" PVC casing

5' bentonite clay seal

Figure 20. Details of a typical ground water monitoring well constructed for the Tritium Project.

casing downward for at least three feet (Figure 24, p. 38). The total shattered interval is unknown, because below 157 feet, the well is filled with sand, limestone fragments, and PVC fragments. The ground water entering the well from the 154 to 157 foot interval is probably from the Utley Limestone Memberof the Clinchfield Formation. The Utley Limestone is included in the Upper Three Runs aquifer.

Six of the nine geophysically logged wells lack grouting or casing (in the case of the two cistern type wells). Only three of the wells have a concrete pad at the surface protecting the wells from leakage around the well casing.
Prior to geophysical logging, the GGS obtained "reported" depths for six of the first ten polluted wells from either the well driller or the home owner (Table 10). In all

32

Table9
Water wells with elevatedconcentrations oftritium ~00picoCuries per liter) in eastern BurkeCounty. 109 wells were sampled. Well locations are shown in Figures 22 and 23.

Name Arthur Jackson Residence

Tritium Project Well Number
6

Date of Initial Sample
10/08/91

TritiumpicoCuries/liter
1300200

Delaigle Trailer Park Well #3

3

07/07/91

1200200

Ralph Greer Residence

79

11/20/91

1000 100

Hug-a-Hog Plantation

46

10/17/91

1000 100

George Wilson Residence

36

02/10/92

600 100

Lamar Paul Residence

59

08/21/92

900 100

Walnut Run Ostrich Farm

77

08/14/92

700 100

Ricky Greer Residence

80

02/10/92

500 100

Rose Johnson Residence

84

10/08/91

500 100

Frank Wimberly Residence

33

10/01/92

1000 100

Julian Morris Residence

56

07/07/92

500 100

Bill Sturgeon Residence

9

02/02/93

1200200

Mary Johnson Residence

65

Alma Crook Residence

108

05/18/93 02/03/94

1600200 2000200

Earl Mills Residence

109

02/03/94

800 100

cases the measured depth of the well is significantly shallower than the reported depth. The discrepancy may be due to gradual filling of the wells with sediment, failure of the well (as in the Delaigle well), or improper "reporting" of depths. For whatever reason, the discrepancy indicates that "reported" depths of wells in Burke County are not usable for hydrogeologic analysis.

The five additional polluted wells were located (after the USGS logging) by re-examination ofEPD records (#56) and additional sampling (#9, #65, #108, and #109) (Figure 23, p. 37). Four of these five polluted wells are "reported" to have been drilled into the Upper Three Runs aquifer by the owners (#56 and #108) or by the driller of the other two wells (#9 and #65) (who orally stated that the wells were

34

Table 10
Results of geophysical logging of anomalous water wells, and surface inspection of well sites. Well locations shown in Figures 22 and 23. All wells are in Three Runs Aquifer.

Name
Arthur Jackson Residence (1)
Delaigle Trailer Park Well #3 (1)
Ralph Greer Residence (2)

Well Number
6
3
79

Well Construction
no grout, no surface pad
grouted, surface pad.damaged
casing
no grout, no surface pad

Screened/Open Interval (feet below surface)
100-138.9
154 - 157 + (5)
87 - 118.1

Original Reported Depth
(feet) 220 (6)
300 (6)
180 (6)

Hug-a-Hog Plantation

36" cistern, no

46

casing

0-27

not reported

George Wilson Residence (2)

grouted, surface

36

149- 168.9

pad

~

220 (6)

Lamar Paul Residence (4)

grouted(?), no

59

surface pad

90-114

140 (6)

Walnut Run Ostrich Farm (3)

77

no grout, no surface pad

99- 114

220 (6)

Ricky Greer

Residence (7)

80

unknown

unknown

not reported

Rose Johnson Residence

84

Frank Wimberly

Residence

33

no grout, no surface pad
36" cistern, no casing, surface
pad

105 - 112 0- 80.8

not reported not reported

(1) Geophysically logged on 04 September 1992. (2) Geophysically logged on 11 September 1992. (3) Geophysically logged on 02 February 1993. (4) Geophysically logged on 03 February 1993.

(5) Casing shattered from 154' to 157'+ below surface. Damage revealed through downhole video-camera examination.
(6) Oral communication from well driller or property owner. (7) Owner declined offer to perform geophysical logging of well
because of concern about condition of old pump. The pistontype pump used can lift water <100'.
36

----......;._Ri:v.e.:r:R::o:a;d::~- 068

..71 70

N

88
..s 1

0

2 miles

Polluted Non-polluted
Figure 23. Detailed map of the central portion of the Tritium Project study area. Numbers identify wells sampled by EPD personnel before and during the Tritium Project (1987 - present). Polluted wells average average >500 picoCuries per liter.

"drilled above the (Utley) limestone"). The other polluted well (#109) was "reported" by the owner to be a "deeper" well. As with the nine logged wells, these five orally reported depths are considered unreliable. Sampling dates and results for subsequent testing of all fifteen polluted water wells are listed in Appendix 5 (p. 80).
Beyond the nine wells measured by geophysical logging, the well depths of the remaining 100 (unlogged) wells are unknown (or unreliable). Other than variances in well depth (Upper Three Runs aquifer vs. Gordon aquifer), the

significance of the interspersal of polluted wells within the non-polluted wells is uncertain, as 100 percent of the local wells have not yet been sampled.
Tritium in Base Flow
1991 Base Flow Study The tritium concentrations of the 53 samples collected during the 1991 base flow study ranged from 400 ( 100) to 1900 ( 200) picoCuries per liter. The areal distribution of

37

PVC casing

'PVC
' 1"

Figure 24. Photograph of the shattered portion of the Delaigle Trailer Park Well #3.

tritium concentrations based on the 1991 base flow study is shown in Figure 25 (p. 39). As the water present in streams under base flow conditions is derived almost entirely from Upper Three Runs aquifer discharge, tritium concentrations in the streams should approximate the tritium concentrations in the aquifer. As shown on the isopleth map (Figure 25, p. 39), the area with the highest concentrations of tritium

lies near the Savannah River,just north ofHancock Landing Road and east of River Road. From this area, tritium concentrations decrease to the northwest, west, and southeast.
1992 Base Flow Study The tritium values of 126 samples collected during the

38

Table 11
Comparison of tritium concentrations estimated from 1992 base flow study with measurements from public and private supply wells. Well locations are shown in Figures 22 and 23.

Tritium Project Well Numbers

Expected Tritium Concentrations (based on 1992 Base Flow Study)

3

1500

6

2100

9

1200

33

1000

36

1100

46

700

56

1500

59

1600

65

1600

77

1200

79

1200

80

1300

84

1600

108

1700

109

1700

Measured Tritium Concentrations (picoCuries/liter) 1200 1000 1200 1000 700 900 500 800 1600 700 800 600 500 2000 800

Date Measured
08/06/';)2 12/30/92 02/02/';)3 10/01/92 09/03/';)2 08/14/';)2 07/07/92 09/04/';)2 05/18/93 12/30/';)2 09/03/92 03/04/92 01/15/';)3 02/03/94 02/03/';)4

1992 base flow study ranged from <100 (below detection limits) to 2200 (200) picoCuries per liter. The areal distribution of tritium concentrations based on the 1992 base flow study is shown in Figure 26 (p. 41). As with the 1991 base flow study, the area with the highest tritium values lies just north of Hancock Landing Road and just east of River Road. Subsequent sampling within this area ofpeak tritium concentration showed a tritium concentration of 3,500 picoCuries per liter( 200) from a spring along the bluffs ofthe Savannah River near drilling site TR92-6 (see Figure 19 for location of this cluster site).

Comparison of 1991 and 1992 Results In general, the distribution of tritium in 1992 was similar to the 1991 distribution. Detailed comparisons of the 1991 and 1992 studies are not appropriate because ofthe differences in the numberofsamples used in the two studies. However, the data suggest that in 1992 there was a slight contraction ofthe 500 and 1000 picoCurie perliter isopleths on the southeastern side of the area of pollution. Whether this represents a long-term trend towards reduction in tritium pollution or is the result of year to year fluctuations is unknown.

40

Comparison of1992 Base Flow Results with Water Well Data
The results from the 1992 base flow study are generally corroborated by tritium concentrations observed .in water supply wells (Table 11; Figure 27, p. 43), based on the geographic location of each polluted water supply well. Thus, a water supply well located half way between the 500 and 1,000 picoCurie per liter isopleths would have a predicted tritium concentration of 750 picoCuries per liter. One water well sample has a tritium concentration higher than the predicted tritium concentrations from the 1992 base flow samples. Three of the fifteen water wells have tritium concentrations that match the predicted base flow concentrations, five wells have tritium concentrations between 100 and 500 picoCuries per liter (one contour interval) of the predicted base flow concentrations, and six wells have tritium concentrations between 600 and 1,100 picoCuries per liter (two contour intervals) of predicted base flow concentrations. The fact that these public and private water supply wells generally have lower concentrations of tritium than the base flow samples may be due to these water wells drawing their water from various depths in the aquifer, whereas the base flow samples generally are derived from the uppermost part of the aquifer. The five water supply wells that have tritium concentrations 600 picoCuries per liter or more below the predicted concentrations lie along a line between Hancock Landing and the intersection of River Road and Georgia Highway 80. This zone of significantly lower than expected tritium concentrations may be due to the presence of discontinuous clay layers within the Upper Three Runs aquifer, which differentially retard the downward movement of tritium-polluted ground water, or to effects from faulting which has been postulated for that area.
Tritium in Monitoring Wells
The average tritium concentrations for samples from each of the fifteen ground-water monitoring wells at the six cluster sites are shown in Table 12 (p. 44). The results of all monitoring well tritium analyses are presented in Appendix 6 (p. 84).
After development and purging of the seven Gordon aquifer monitoring wells, only one Gordon aquifer monitoring well (TR92-5C) showed consistent detectable concentrations of tritium (above 100 picoCuries per liter). This result has been confirmed by repeated sampling at all six cluster sites.
Gordon aquifer monitoring well TR92-5B showed low concentrations of tritium (500 to 600 picoCuries per liter) when initially sampled, on June 15, 1993. The well was later resampled after extensive purging of the well prior to sampling. These later samples had tritium concentrations below the detection limits. The initial positive results for this well were probably due to tritium introduced into the Gordon aquifer from the overlying polluted Upper Three Runs aquifer, during drilling. This tritium was flushed out after thorough purging of the well. Monitoring well TR92-

SC, screened between 200 and 300 feet, consistently shows low concentrations of tritium (300 to 400 picoCuries per liter) even after extensive purging. This well is located 50 feet from the radiologically anomalous Delaigle Trailer Park well #3. The Delaigle well is screened between the depths of200 and 300 feet (according to the driller's report), but the casing has been shattered at a depth of 154 feet (Figure 24, p. 38). It is likely that the Delaigle Trailer Park well #3 is acting as a conduit allowing tritium polluted water from the Upper Three Runs aquifer to enter the Gordon aquifer, and that monitoring well TR92-5C is detecting this pollution. Monitoring well TR92-5B, which is also screened in the Gordon aquifer, is located 94 feet from the Delaigle Trailer Park well #3 and does not show detectable concentrations of tritium. This suggests that the area of pollution within the Gordon aquifer is relatively small.
After development and purging of the six Upper Three Runs aquifer monitoring wells (excluding TR92-6A, which was dry), all of the samples from the Upper Three Runs aquifer have detectable concentrations of tritium, with values ranging from 200 to 1700 picoCuries per liter (Appendix 6), well below the EPA MCL of 20,000 picoCuries per liter. Tritium concentrations in the Upper Three Runs aquifer monitoring wells TR92-2A, TR92-3A, TR92-4A, TR92-4A2, TR92-5A, and TW-4 (USGS Miller's Pond site) are consistent with the tritium concentrations measured in the 1992 base flow studies (Figure 28, p. 45; Table 13, p. 46). Tritium concentrations detected in the Upper Three Runs well TR92-1A, however,aresignificantly lower than concentrations expected based on the results of the base flow studies. Results from the 1991 and 1992 base flow studies suggest that tritium concentrations in the Upper Three Runs aquifer in the vicinity of cluster site TR92-l should be within the range of 1200 to 1800 picoCuries per liter. Repeated sampling ofTR92-IA, however, resulted in a maximum tritium concentration of 300 picoCuries per liter (Appendix 6). Examination of the core samples from this cluster site indicates that monitoring well TR92-IA is screened in a zone that is separated from the upper part ofthe Upper Three Runs aquifer by a thin clay layer. This clay layer may serve as a partial confining unit that retards the tritium polluted water in the upper part of the aquifer from freely entering the strata sampled by the well. In addition to the information supplied by the GGS Tritium Project monitoring wells, seven USGS Trans-River Flow Project monitoring wells at Miller's Pond were sampled for tritium. The depths and aquifers sampled for each well at the Miller's Pond Cluster Site are shown in Table 14 (p. 48). Of the samples tested, only the Upper Three Runs aquifer well (TW-4) had detectable tritium concentrations (700 picoCuries per liter 100).
Geology
Seventeen cores (see Figure 21, p. 33, for core locations) were examined and used to reconstruct a three dimensional geologic framework for eastern Burke County. Figure 29 (p. 47) shows an approximate north-south slice

42

Table 12 Results of tritium analyses of water samples from Tritium Project ground-water monitoring wells.

Well Number TR92-1A TR92-IB
TR92-1C (1) TR92-1D TR92-2A TR92-2B TR92-3A TR92-3B TR92-4A
TR92-4A2 (2) TR92-4B TR92-5A TR92-5B
TR92-5C (3) TR92-6A(4)
TR92-6B

Aquifer Upper Three Runs

Tritium (picoCuries/Iiter)
200

Gordon

below detection

Meyers Branch

below detection

Dublin

below detection

Upper Three Runs

1400

Gordon

below detection

Upper Three Runs

1200

Gordon

below dectection

Upper Three Runs

1600

Upper Three Runs

1600

Gordon

below detection

Upper Three Runs

1000

Gordon

below detection

Gordon

400

Upper Three Runs

dry hole

Gordon

below detection

Sampling Date ll/17f.)2 11/171')2 12/14f.)2 12/141')2 03/02f.)3 02/181')3 03/021')3 01/051')3 03/021')3 02/031')4 01/061')3 09/011')3 06(241')3 09/0lf.)3
09/0lf.)3

(1) Sand within Meyers Branch Confining System. (2) Redrilled slightly deeper for better water production. (3) 100' screened interval, to duplicate reported construction of Delaigle Trailer Park well #3. (4) Well is screened in an impermeable portion of the Utley Limestone.
44

Table 13
Comparison oftritium concentrationspredictedfrom 1992base flow study withmeasurements from Tritium Projectground-water monitoring wells.

Well

Estimated Tritium Concentrations Based on
1992 Base Flow Study (picoCuries/Iiter)

Measured Tritium Concentrations (picoCuries/Iiter)

Date Measured

1R92-1A

1500

200

11/17/92

1R92-2A

2000

1700

02/18/93

1R92-3A 1R92-4A

1100 1700

1200 1600

03/02/93 03/02/93

1R92-5A
1R92-6A
TW-4 (Miller's
Pond)

1500 2000 900

1000

08/26/93

dry

-------

700

01/15/93

through the study area that includes five of the six cluster sites.
The sediments of the Upper Eocene Barnwell Group (consisting of the Tobacco Road Formation, the Irwinton Sand Member and the Griffin's Landing Memberofthe Dry Branch Formation, and the Utley Limestone Member of the Clinchfield Formation) make up the upper part of the stratigraphic section (Figure 29).
The Tobacco Road Sand is a deeply weathered, highly pigmented, massively-bedded, moderately to poorly sorted, burrowed and bioturbated, medium to coarse-grained sand, containing minor amounts of clays, chert, calcite, glauconite, and other minor constituents (Huddleston, in preparation). The Irwinton Sand Member of the Dry Branch Formation is generally a well sorted, variously bedded, fine to medium grained sand. Smectitic clay beds, lenses, and laminaeofvarying thickness and lateral extentoccurthroughout the Irwinton Sand, resulting in locally perched water tables. Down dip from and underlying the Irwinton Sand is the Griffin's Landing Member of the Dry Branch Formation. The Griffin's Landing Member is a moderately sorted, massive to vaguely bedded calcareous sand, with local

occurrences of thin limestone beds, calcareous beds or lenses, oyster shell beds and bioherms, and chert and silicacemented sandstones (Huddleston, in preparation). The basal unit of the Barnwell Group is the Utley Limestone Member ofthe Clinchfield Formation. Occurring primarily in eastern Burke County, the Utley Limestone is typically a moldic, fossiliferous sandy limestone that is variably glauconitic, with occasional beds of calcareous sands. The Utley Limestone is highly variable in thickness, perhaps due to dissolution, ranging in thickness from 0 (where missing) to 13 feet, at the type locality at Mallards Pond, near Plant Vogtle (Figure 11, p. 17). In summary, approximately the upper 95 percent of the Barnwell Group is
composed ofsandy units; the lower 5 percent is a limestone.
There appears to be little change offacies within the Barnwell Group across the study area except that the top of the Griffin's Landing Member grades both upwardand laterally into the Irwinton Sand.
Recent geological mapping by Hetrick (1992) shows that, within the study area, the Savannah River has cut through the sediments of the Barnwell Group at least as far south as Beaverdam Creek, downstream from Plant Vogtle.

46

Table 14 USGS Trans-River How Project monitoring wells at Miller's Pond Site in Burke County, Georgia. All data provided by USGS.

Well Number

Screened Interval (feet below surface)

Tritium (picoCuries/
liter)

Aquifer

TW-4 TW-5A TW-6 TW-7 TW-3 TW-2 TW-1

80- 199 211-251 299 - 325 450 - 475 518-548 595 -625 705 - 735

700 100
below detection
below detection
below detection
below detection
below detection
below detection

Upper Three Runs
Upper Dublin
upper Lower Dublin
lower Lower Dublin
Allendale confining unit
upper Midville
lower Lower Midville

This incision by the Savannah River probably extends further to the south, but the contact between the Barnwell Group and the Lisbon Formation is covered by Quaternary alluvium. Geologic mapping also shows that McBean Creek has cut through the Barnwell Group sediments in the study area as have several smaller tributaries to the Savannah River including Boggy Gut Creek, Newberry Creek, Beaverdam Creek, and several small unnamed creeks.
The Lisbon Formation occurs throughout the study area, with the McBean Member in the northern part and the Blue Bluff member underlying the remainder of the area. The McBean Member is typically a fine grained sandy limestone that is variably calcarenitic, argillaceous, micaceous, and carbonaceous (Huddlestun, in preparation), with some glauconite in the lower part. In the McBean core, the McBean Member of the Lisbon Formation is a hard to mildly indurated, friable and brittle limestone that is massively-bedded and occasionally fossiliferous, with occasional beds of soft, sandy limestone.
In the northern and central parts of the study area, the McBean grades laterally through an "undifferentiated Lisbon" consisting of a soft, sandy limestone, overlying an unconsolidated calcareous, argillaceous, glauconitic, bioclastic, phosphatic sand, as seen in core from the TR924 site, and into the Blue Bluff member (Huddleston, in preparation). The Blue BluffMemberofthe Lisbon Formation ranges in composition from a bioturbated to thinly-

bedded, silty to finely sandy, calcareous clay to a very argillaceous limestone, with minor lenses ofsand. The Blue Bluff member ranges in thickness from 39 feet in areas where it intertongues with the updip McBean Member to 64 feet in core from the TR92-3 site and 84.5 feet in the Georgia Powercore VG- I, near Girard (Huddleston, in preparation). Because of its thickness and argillaceous nature, the Blue Bluff member (where present) appears to serve as a local aquitard.
Underlying the Lisbon Formation in the northern part of the study area are the sediments of the Claiborne Group, with the uppermost portion consisting of the Bennock Millpond sand (informal name) and, in the southern part of the study area, its downdip equivalent, the Still Branch sand (informal name) (Figure 29, p. 47; The Bennock Millpond sand broadly consists of three distinctive sand lithofacies: 1) a fossiliferous, calcareous, fine sand; 2) a massive to thinly bedded, noncalcareous, fine to very fine sand; and 3) a noncalcareous, bioturbated sand. The Bennock Millpond sand appears to be correlative with the Warley Hill Formation ofSouth Carolina. Downdip, in the vicinity ofHancock Landing, the Bennock Millpond sand intertongues with and grades laterally into the Still Branch sand. The Still Branch sand consists of a calcareous, massively bedded, fine to medium grained sand, with an upper part which is a moldic, fossiliferous sandy limestone/calcareous sandstone. Where permeability permits, the Bennock Millpond and Still Branch

48

sands compose the upper portion of the Gordon aquifer. Underlying both the Bennock Millpond and Still Branch
sands is an unnamed Congaree-equivalent sand. This Congaree-equivalent sand is soft, barely cohesive to loose, noncalcareous sand, with scattered thin beds of clay, with no appreciable facies changes throughout the study area. These characteristics permit this unit to serve as the main portion of the Gordon aquifer in the study area. The thickness of this sand unit varies from absent in the McBean core to 62 feet in Georgia Power core VG- I (Figure 30, p. 49). This sand appears to grade laterally (eastward) into the Congaree Formation in South Carolina (Huddlestun, in preparation).
Beneath the Claiborne Group lies the Oconee Group, consisting of the Lower Paleocene Snapp Formation (Figure 31, p. 51) and the "Ellenton" Formation and the Upper Cretaceous Steel Creek and Gaillard Formations (Figure 32, p. 53). The Snapp Formation marks the uppermost occurrence of kaolinitic clays in the study area. The kaolinitic beds of the Snapp Formation mark the top ofa fining upward sequence, with a basal sand unit (Huddlestun, in preparation). The Snapp Formation kaolin is massively bedded and structureless, variably silty, finely micaceous, occasionally pyritic, and generally hackly with an irregular fracture. The Snapp Formation is absent in the core at site TR92-6, possibly permitting some vertical leakage between the Gordon aquifer and the underlying aquifer.
Gradationally underlying the Snapp Formation is the "Ellenton" Formation. Within the study area, the sands of the "Ellenton" Formation are fine to medium-grained and moderately to well-sorted, with the bedding ranging from massively structureless to rudely bedded and faintly bioturbated, with some interlaminated thin clay beds. Most commonly, the sands of the "Ellenton" are loose to barely cohesive and soft (Huddlestun, in preparation), characteristics which define this unit as the lower portions of the Gordon aquifer in Georgia and the sand within the Meyers Branch Confining System in South Carolina (Aadland and others, 1992). Disconformably underlying the "Ellenton" is the Upper Cretaceous Steel Creek Formation, which consists of interbedded, mottled, varicolored, fining upward kaolin and sand units. The quartz sands of the Steel Creek are variably cohesive to loose (depending on the kaolin content), massively bedded and structureless (Huddlestun, in preparation). The kaolin beds are generally silty to sandy, micaceous, massively bedded, structureless, and contain varying amounts of heavy minerals. As with the overlying Snapp Formation, this kaolin is dense, with a hackly, irregular fracture.
The deepest unit encountered by a Tritium Project monitoring well (TR92- l D) is the Upper Cretaceous Gaillard Formation, which conformably underlies the Steel Creek Formation. The sands of the Gaillard Formation are similar to those of the Stee! Creek, except there are fewer clay beds present. The Gaillard Formation is defined as the Dublin aquifer within the study area, as well as in South Carolina (Aadland and others, 1992).

Structurally, the study area can be characterized as a region of gently dipping sediments, with a regional dip (between the Richmond-Burke County line and the BurkeScreven County line) on the top of the Blue Bluff member of 10 feet per mile to the southeast (Figure 33, p. 55). The Pen Branch fault, as defined at SRS, cannot be detected within the spacing of the current series of cores. There are, however, two minor structural features of uncertain significance near cluster site TR92-3 south of Beaverdam Creek. The Hancock Landing structural high appears to influence mainly the deeper rock units (Paleocene and Cretaceous formations) (Figure 34, p. 57) whereas the Beaverdam Creek structural high influences only the shallower rock units (Eocene formations) (Figure 33).
Seismic Survey of the Savannah River Channel
The information provided by the high resolution seismic survey and the auger boring suggests the breachment of the Lisbon Formation (aquitard) by the paleo-river channel and the incision of the underlying Gordon aquifer between the Point Comfort area and Hancock Landing (Figure 33) (Henry, unpublished data, 1994 and Leeth and Nagle, in preparation). North of McBean Creek, the Dublin aquifer has also been breached by the paleo-river channel.
The medium resolution seismic survey suggests the presence of a high angle reverse fault or related drape fold approximately 1000 feet down-river from Hancock Landing. This structural feature extends upward to within approximately 200 feet of the paleo-river channel base and terminates in Paleocene or early Eocene deposits (Henry, unpublished data, 1994).
Within SRS, the location ofthe Pen Branch Fault is well defined by drill core and seismic reflection data (Snipes and others, 1993). If projected to the Savannah River from its known location (Snipes and others, 1993), the Pen Branch Fault possibly could intersect the river in the vicinity of the structural feature described by Henry (unpublished data, 1994).
Hydrogeology
Upper Three Runs Aquifer The Upper Three Runs aquifer in the study area is formed by the permeable sands and limestones of the Barnwell Group. A general map of the approximate water table surface has been constructed (Figure 35) using the elevation of the springheads (point of origin) of perennial streams (as interpreted from USGS 1:24,000 topographic maps) supplemented by water level information from monitoring wells. As perennial steams mark the intersection between the land surface and the Upper Three Runs aquifer, the springhead of a perennial stream approximates the elevation of the water table. Because there are many more springheads than monitoring wells in the study area, the use ofspringheads improves the resolution ofa water table map.

50

Table 15 Ground-water levels in the Upper Three Runs and Gordon aquifers. Water levels are in feet above mean sea level.

Cluster Site
Miller's Pond Site (1)
1R92-1

Land Surface Elevation
230
235

Upper Three Runs Aquifer
165.19 (6)

Gordon Aquifer 122.70 (2, 6)

Potential Vertical Flow
Downwards

186.16 (4)

150.58 (3)

Downwards

1R92-2

285

208.95 (4)

108.50 (4)

Downwards

1R92-3

195

136.64 (4)

146.66 (4)

Upwards (3)

1R92-4

192

128.00 (5)

116.76 (4)

Downwards

1R92-5

235

156.03 (4)

140.74 (4)

Downwards

1R92-6

240

dry hole

91.00 (4)

unknown

(1) USGS Trans-River Project site. (2) The Gordon aquifer is missing at Miller's Pond. Instead, the upper Dublin aquifer
occurs immediatedly below the Blue Bluff member aquitard. (3) See text (Section 3.6.3) for discussion. (4) Water levels measured (1:)/10/93. (5) Water levels measured (1:)/14/93. (6) Water levels measured (1:)/13/93. Data from USGS.

As the major streams in eastern Burke County are incised completely through the Barnwell sediments that constitute theUpperThree Runs aquifer, ground-water flow in the Upper Three Runs aquifer is compartmentalized and local (Figure 35). A major, northwest-southeast trending compartment is formed between the Savannah River, Brier Creek, and McBean Creek. Sub-compartments are formed by the major tributaries to the principal streams. Dissection of the land surface by streams tributary to the Savannah River has divided the eastern side of the study area into several small, local, ground-water flow regimes. Dissection of the land surface by streams tributary to Brier Creek is much less marked, resulting in less compartmentalization of ground-water flow. Consequently, some down gradient flow to the southeast is possible, parallel to Brier Creek. However, most water in the Upper Three Runs aquifer in eastern Burke County appears to discharge directly to local streams.
As expected, there is a close relationship between the

elevation of the land surface and the elevation of the water table (Figure 36, p. 60). With the possible exception of monitoring wells in very close proximity to the Savannah River (1R92-6A), land surface elevation appears to be the dominant factor in determining the depth to the water table and proximity to a perennial stream a secondary factor. Each upland area in eastern Burke County serves as a recharge area for local ground-water flow in the Upper Three Runs aquifer. Since flow regimes in the Upper Three Runs aquifer are local, the chemistry of the ground water discharging from each individual recharge area should closely reflect local conditions. Recharge to the downdip portion of the Upper Three Runs aquifer probably takes place primarily in the area near Girard.
The Upper Three Runs aquiferin eastern Burke County is an anisotropic multi-layered unconfined aquifer. The upper layer of the aquifer is composed of the sands of the Tobacco Road Formation and the Irwinton Sand Member of the Dry Branch Formation.

52

System/ European

Series

Stage

Q)
ni

Priabonian

...J

Provincial

Stage

Western Georgia

'

Jacksonian

Ocala Limestone

Eastern Georgia Prowell & others, 1985
Barnwell Group

Eastern Georgia this report
Barnwell Group

W South Carolina E
-Barnwell
Group_ Cooper Group

Q)

C

Q)

CJ
0 UJ

Q)
i5
"t:l
~

Bartonian Lutetian

"~jffiiir.ffmiw:mr.~mwmnm 111111111111111111111111111111111111111 -------------------------------------IIIIIIIIIIIIIIIIIIIIIII

II llllllllll 11111111111111111111111111

Lisbon Formation
I

l !Uil:N Lisbon/McBean Formation 1

Still Branch sd.

=~ ij

Ypresian

UJ

Tallahatta Formation
Hatcheti bee/Bashi Fm.

Q)

C
Q)

Q)
ni

CJ ...J

0

Q)

a<.t.i

=> ~

UJ

Selandian Danian

Sabinian

Midwayan

Clayton Formation
111111111111111111111111111111111111111 :

Vw I

Maastrichtian
I

Navarroan

Providence Sand Ripley Formation

Unnamed Congareeequivalent sand

lllllllllllllllllllllllllllllllllll 1111111111111

I

. . . . . Snapp Formation

Williamsburg Fm.

lllll 11111111111111111111111111111111

11 Ellenton II Fm.

:

1111111111111111111111111111111111111

llllllllllllllllllllllllllllllllllll11 Peedee Formation

Campanian

Tayloran

Cusseta Sand

Steel Creek Fm.
I

Black Creek Group

(/)
:::,
Santonian Coniacian

Blufftown Fm.

11111111111111111111111111111111111111111 Middendorf Fm. Cape Fear Fm.

Turonian

Cenomanian

Table 16
Transmissivity values of the aquifers at Tritium Project cluster site 1R92-l. Data based on Clemson University aquifer tests. Location is shown in Figure 19.

Aquifer
Upper Three Runs (water table)

Transmissivity 479ft2/d

Lisbon Formation) ofthe Gordon aquifer extend throughout most of the study area, the major recharge area for the Gordon aquifer in eastern Burke County is probably in southernRichmondCountyand northwestern Burke County (inferred from the potentiometric surface map of the Gordon aquifer in Brooks and others, 1985). Thebeds that make up the Gordon aquifer only crop out along stream channels in Richmond County and do not crop out at all in northwestern Burke County (Hetrick, 1992). Recharge ofthe Gordon aquifer in southern Richmond and northern Burke Counties is probably by downward vertical leakage through the relatively permeable McBean Member of the Lisbon Formation which overlies the Gordon aquifer in those areas.

Gordon Meyers Branch "aquifer"
Dublin

112ft2/d 637 ft2/d 92ft2/d

Discontinuous clay beds and lenses in the Irwinton Sand Member create vertical and possibly lateral differences in the hydraulic conductivity of this layer. The lower layer of the Upper Three Runs aquifer is composed of the limestones of the Utley Member of the Clinchfield Formation. During drilling operations for the installation of ground-water monitoring wells, voids were encountered within the Utley. This suggests that at least some of the ground-water flow in the Utley may be through solution channels.
Gordon Aquifer A potentiometric surface map of the Gordon aquifer within the Tritium Project study area has been constructed (Figure 37, p. 61) based on water level information from the six monitoring wells (Table 15). In contrast to the localized recharge and flow patterns ofthe Upper Three Runs aquifer, recharge and flow through the Gordon aquifer is regional in nature. As indicated by published potentiometric maps (Brooks and others, 1985), the regional flow of the Gordon aquifer is to the southeast, except where locally influenced by the incisement of local streams. In the case of northern Burke County, the incisementby the upper waters ofMcBean Creek and Brier Creek (Brooks and others, 1985; Hetrick, 1992) locally influences the potentiometric contour lines. East of Brier Creek, as indicated by the potentiometric map shown in Brooks and others (1985), as well as Figures 37 and 38 (pages 61 and 63) of this report, and from unpublished USGS data, the direction of local flow for the Gordon aquifer (in the study area) is to the northeast, towards the Savannah River, indicating that the river forms a regional discharge zone for the Gordon. Because the calcareous clays of the upper aquitard (Blue Bluff member of the

Vertical Hydraulic Gradient At all but one of the monitoring well cluster sites in Burke County, the head in the Upper Three Runs aquifer is higher than the head in the Gordon Aquifer (Table 15, p. 52). As a result of this difference in potential energy, there is the potential for ground water to move downwards from the tritium polluted Upper Three Runs aquifer into the tritium-free (i.e., below the limits of detection) Gordon aquifer. Because the Blue Bluff member aquitard, which lies between theUpperThree Runs and the Gordon aquifers, underlies most ofthe study area, such downward movement appears to be retarded. As no confining bed provides a perfect barrier against vertical transport, it is likely that some vertical transport may eventually occur. However, if the rate of flow is slower than the decay rate for tritium, tritium may never reach the upper Gordon aquifer in any significant quantity. As previously shown in Table 15, the cluster site water levels in the Upper Three Runs aquifer are higher than the water levels in the Gordon aquifer, except for 1R92-3. This reversal in water levels (or hydraulic heads) may be due to: topographic effects and/or structural effects. The low water level in well 1R92-3A may be due to the previously discussed relationship between topography and water levels in the Upper Three Runs aquifer (p. 50), as site 1R92-3 is at a relatively low elevation ( 195 feet above mean sea level). In this case, the confined Gordon aquifer would be unaffected by surface topography. The relatively high water level in well 1R92-3B may be related to a minor structural high in the Barnwell, Lisbon (Figure 33, p. 55), and Congaree-equivalent sediments, which may locally affect the hydraulic head of the Gordon aquifer. The possibility exists that the head reversal may due to both topographic and structural effects. In the northern part of the study area, where the Blue Bluff member of the Lisbon Formation is replaced by the more permeable McBean Member, there is the potential for vertical transport of both water and tritium from the Upper Three Runs aquifer into the Gordon. However, tritium has not been detected in the Gordon aquifer at the cluster sites at which the McBean Member has replaced the Blue Bluff member. This suggests that thereare sufficient clay and fine grained sands in the Barnwell Group sediments and the

54

Table 17
Ground-water geochemistry of the Upper Three runs and Gordon aquifers. The values given are medians for that parameter from all samples within the gydrogeochemical unit Data from Rose and James (1993, umpublished).

Chemical Parameter
pH TDS (1)
Na Ca Mg HCO3 SO4 K Si Cl

Unit 1: upper Upper Unit 2: lower Upper Three Runs Aquifer Three Runs Aquifer

Unit 3: Gordon Aquifer

7.3

7.7

7.8

100mg/L

226mg/L

242mg/L

1.2mg/L

1.7 mg/L

2.7mg/L

19.4 mg/L

44.7mg/L

44.9mg/L

0.4 mg/L

1.2 mg/L

1.2mg/L

55.4mg/L

132.9 mg/L

136.3 mg.L

1.3 mg/L

4.4 mg/L

10.3 mg/L

0.2mg/L

0.3 mg/L

0.4 mg/L

15.6mg/L

32.4 mg/L

34.0mg/L

4.0mg/L

4.1 mg/L

2.8mg/L

(1) TDS is Total Dissolved Solids.

McBean Member of the Lisbon to minimize vertical transport.
Results ofAquifer Tests The aquifer pumping tests carried out on the four monitoring wells at site TR92-l (see Figure 19, p. 31) by Cl.emson University (Moore and others, l 992J showed transmissivity values between 92 ft2/d and 637 ft /d (Table 16). These data show that, at site TR92-l, the Upper Three Runs aquifer and the Meyers Branch "aquifer" have much higher transmissivity values than the Gordon and Dublin aquifers. These variations are most likely due to the grain size and clay content of the sediments forming the aquifers. No drawdown was observed in any of the observation wells during the pumping tests, indicating that the three lower aquifers are confined with respect to the Upper Three Runs

aquifer and to each other.
Ground-Water Geochemistry
Vertical Geochemical Variation
Based primarily on ground-water geochemistry, but also considering differences in hydrostatic head and lithologic information from cores, the upper three hundred feet of the ground-water system in eastern Burke County can be divided into three hydrogeochemical units (Rose and James, 1993) (Table 16, p. 54). This results in a partition of the Upper Three Runs aquifer into two hydrogeochemically different units. Hydrogeochemical Unit 1consists ofground water within the sands and clays of the Tobacco Road and Dry Branch Formations. These two formations make up 90 to 95 percent of the Upper Three Runs aquifer. The ground

56

water within Unit 1 has approximately neutral pH and is low of eastern Burke County;

in dissolved solids (Table 17). Hydrogeochemical Unit 2 2. evaluate if there was any current or future threat to

consists of ground water within the Utley Limestone Mem- public health;

ber of the Clinchfield Formation and carbonate rich portions 3. assess how the tritium entered the ground-water re-

of the Griffins Landing Member of the Dry Branch Forma- gime.

tion. The portions of the Griffins Landing Member that are

dominated by sand and clay are part of Hydrogeochemical

Extent of Pollution

Unit I. Hydrogeochemical Unit 2 is equivalent to the

lowermost part of the Upper Three Runs aquifer. The

Stratigraphic Extent of Pollution

ground water within Unit 2 is characterized by significantly

At the present time, there is no evidence that the Gordon

higher concentrations of total dissolved solids, Na, Ca, Mg, aquifer in eastern Burke County is regionally polluted by

HCO3, and SO4 than Unit 1.

tritium. None of the polluted water wells are withdrawing

Hydrogeochemical Unit 3 consists of the sandy Bennock water from the Gordon aquifer, and the few water wells in

Millpond and Still Branch sands and the unnamed Congaree- eastern Burke County that are in the Gordon have no

equivalent sands that lie beneath the Lisbon Formation (the detectable concentrations of tritium. Five of the six ground-

Gordon aquifer). Except for a higher concentration of water monitoring wells installed into the Gordon aquifer as

sodium and possibly sulfate, the ground water of Unit 3 is part of this project contain no detectable concentrations of

chemically similar to the water of Unit 2.

tritium. The sixth monitoring well may be detecting tritium

All three hydrogeochemical units are chemically char- that has entered the Gordon aquifer through a nearby water

acterized as a calcium-bicarbonate water type. Analysis of well in which the casing has shattered.

the data shows that 80 to 90 percent of the ionic content of

There is definite, low concentration, tritium pollution

these waters has evolved from the dissolution of calcite of the Upper Three Runs aquifer in eastern Burke County.

(Rose and James, 1993).

Nine of the ten geophysically logged water wells that have

elevated concentrations of tritium are drawing their water

Lateral Geochemical Variation

from the Upper Three Runs aquifer, and there is good

Rose and James (1993) document lateral chemical circumstantial evidence that the tenth well is also in the

gradients for total dissolved solids, calcium, bicarbonate, Upper Three Runs aquifer. Of the five tritium polluted

magnesium, sodium, sulfate, dissolved silica, and calcite water wells that were identified after geophysical logging

saturation indices in Hydrogeochemical Unit 2 within the was completed (#9, #56, #65, #108, and #109, see Figure

study area. Such lateral chemical gradients confirm the 23), there is good circumstantial evidence that all five wells

general direction of flow within the aquifer. Figure 39 (p. are in the Upper Three Runs aquifer. All of the ground-

65) is modified from their figure and shows the lateral water monitoring wells screened in the Upper Three Runs

variability in total dissolved solids for Unit 2 (the total aquifer contain detectable concentrations of tritium. Analy-

dissolved solids are dominated by calcium and bicarbon- sis of 1992 base flow studies detected tritium in 177 out of

ate). The contour line pattern shown in Figure 39 has a 179 samples (99 percent).

general resemblance to the water table surface map for the

An unresolved question regarding the sampling of

Upper Three Runs aquifer (Figure 35, p. 59). The upland water wells in Burke County concerns the 94 water wells

recharge areas east of Brier Creek have high concentrations that did not contain anomalous concentrations of tritium.

of total dissolved solids. Dissolved solids decrease towards Many of these wells are interspersed among the anomalous

Brier Creek and, particularly, towards the Savannah River. wells. The majority of these low-tritium (less than 500

This trend, however, is inconsistent with chemical trends picoCuries per liter) wells are also probably drawing water

normally expected within an aquifer. Undernormal ground- from the Upper Three Runs aquifer, however, the lack of

water conditions, total dissolved solids should increase in records concerning the depths of the wells prevents a clear

the direction of ground-water flow (Freeze and Cherry, identification of the aquifer. A possible explanation for the

1979). For example Brooks and others (1985) show in- lack of tritium in these wells is the known heterogeneity of

creasing down-gradient concentrations of total dissolved the sediments that constitute the Upper Three Runs aquifer.

solids in the Gordon aquifer within the Georgia Coastal Variable clay content, discontinuous clay beds, and clay

Plain. The observed decrease in total dissolved solids along lenses cause local perched water tables, variable rates of

the projected flow path of the Upper Three Runs aquifer . vertical and lateral ground-water movement, and complex

may be due to the precipitation of calcite or the vertical flow paths. An alternative explanation is that the high-

mixing of ground waters (Rose and James, 1993).

tritium wells are the result of faulty well construction such

as lack of grout, failure of the casing, or lack of a cement

CONCLUSIONS

pad. This second explanation, however, is not supported by

the distribution of tritium in the Upper Three Runs aquifer

As presented in the introduction of this report there as observed in the base flow study (Figures 25 and 26). The

were three goals for the Tritium Project. These goals were widespread nature of the tritium pollution, its concentric

to:

distribution pattern, and the occurrence of tritium up-gradi-

1. map the extent of tritium pollution in the ground water ent from the anomalous wells, all argue against the anoma-

58

300"T"""---------,,.------------,,.-----------r--,---------,

------, I

I

250 - - - - - - - - 1

I

C: 0
i
~ ai
:s~ 200 - - - - - - - -
I
150 - - - - - - - - -I-

TR92-1A Miller's Pond
., - - - - --1TR92 -3A I
TR92-4Ae I

TR92-2A e

100 - - - - - - - - - - , , . - - - - - - - - - - - - , , . - - - - - - - - - - . - - , - - - - - - - - -

100

150

200

250

300

Land surface elevation

Figure 36. Relationship of water table and land surface elevations for the Upper Three Runs aquifer in eastern Burke County. Elevations are in feet above mean sea level.

lous wells acting as point sources for regional pollution of the Upper Three Runs aquifer. The question of low-tritium wells dispersed among the high-tritium wells can only be resolved by additional studies of the construction of the low-tritium wells, and detailed studies of the Upper Three Runs aquifer.
Geographic Extent of Pollution Tritium pollution of the Upper Three Runs aquifer is widespread throughout eastern Burke County. The base flow studies show that the pollution extends from the Savannah River on the northeast to beyond Brier Creek on the southwest, and from at least McBean Creek on the northwest to at least the Burke-Screven county line on the southeast. Surface water samples taken after the 1992 Base Flow Study from sites in Louisville (23 miles WSW from Waynesboro)and Wrens(23 miles WNW from Waynesboro) (both in Jefferson County) and Butler Creek (Richmond County, south of Augusta) (21 miles north of Waynesboro) measured below detection limits for tritium (<100 picoCu-

ries per liter). Other surface water samples collected in the vicinity of the McBean community, in southern Richmond County, measured from 400 to 500 picoCuries per liter. Samples taken from southeast Burke County, near the Screven County line also measure near the detection limits. From this information, it appears that the tritium pollution of the Upper Three Runs aquifer does not extend far beyond the boundaries of Burke County.
Public Health
There is no evidence of a threat to public health at the present time. All analyses of water samples from the Upper Three Runs aquifer show that the concentrations of tritium are significantly lower than the EPA MCL for tritium (which is 20,000 picoCuries per liter). The highest observed tritium concentration was 3,500 picoCuries per liter, which is 17.5 percent of the EPA MCL. More than half of the samples from the 1992 base flow study have tritium concentrations less than 4.5 percent of the EPA MCL.
Future threats to public health are a function of the

60

pathway for tritium into the Upper Three Runs aquifer and future production of tritium at the Savannah River Site. No reactors have been operating at the Savannah River Site since 1988. Asa result, the amount of tritium released to the environment declined by approximately two thirds between 1987 and 1991 (Arnett and others, 1992).
Pathway into the Aquifer
Existing data are not adequate to resolve the issue of the pathway for tritium into the Upper Three Runs aquifer. However, the data generated by the investigation place major constraints on reasonable transport models, greatly reduce the likelihood of some models. The method of multiple working hypotheses was used in evaluating the pathway for tritium into the Upper Three Runs aquifer. This scientific method requires the development of a large number of falsifiable alternative hypotheses (a falsifiable hypothesis is one that can be tested and has the potential to be proven wrong). These hypotheses are tested against the data, and hypotheses that are inconsistent with the data are eliminated. Repeated testing of an hypothesis provides greater and greater confidence in a specific model. There are several models for the transport of tritium into the Upper Three Runs aquifer. The tritium might enter the aquifer laterally, it might enter from below, or it might enter the aquifer from above.
Direct Lateral Transport The model of direct lateral transport postulates that tritium has polluted the Upper Three Runs aquifer in South Carolina, and that tritiated water from the South Carolina aquifer has moved as a plume directly into the Upper Three Runs aquifer of Georgia. We know that the first part of this model is correct; that the Upper Three Runs aquifer at the Savannah River Site has been polluted with tritium (Murphy and others, 1991; Arnett and others, 1992). However, the geometry of the Upper Three Runs aquifer deposits in Georgia and the direction of ground-water flow within that aquifer indicate that direct lateral transport of tritium from the Savannah River site is unlikely. Because the Savannah River has cut completely through the sediments that make up the Upper Three Runs aquifer, there is no subsurface path that would allow a plume oftritiated water within the Upper Three Runs aquifer at the Savannah River Site to move directly into Georgia. In addition, the direction of ground-water flow in the Upper Three Runs aquifer in Burke County is to the east (towards the Savannah River) rather than to the west.
Indirect Lateral Transport The model of indirect lateral transport postulates that tritium from the Savannah River Site in South Carolina has polluted the water of the Savannah River and the ground water in the alluvium of the river. The model then postulates that the tritium-laden water from the Savannah River and its alluvial fill recharge the unconfined aquifer in Georgia.

We know that the water of the Savannah River has received tritium from the Savannah River Site. However, existing data from Burke County do not support the other components of the model. The sediments that make up the Upper Three Runs aquifer in Burke County are topographically higher than the floodplain of the Savannah River in the area of pollution. Therefore, the waters of the Savannah River have no direct connection with that aquifer. This model is also inconsistent with the observation that the hydrostatic head in the Upper Three Runs aquifer is higher than the level of the Savannah Riverin the area of pollution. This suggests that ground water in the Upper Three Runs aquifer is discharging to the river rather than being recharged by the river. The conclusion that the Upper Three Runs aquifer is discharging to the Savannah River is supported by the map of the water table surface (Figure 35) and the occurrence of a line of springs along the bluffs of the Savannah River, near the base of the Upper Three Runs aquifer and above the flood plain of the river.
Upwards Transport The model of upwards transport postulates that tritium from the Savannah River Site in South Carolina has polluted deeper, confined aquifers in South Carolina, and that tritium-laden water has moved under the Savannah River into Georgia. The model then postulates that the tritium polluted water in the confined aquifer moves upwards, through breaches in the confining bed, into the Upper Three Runs aquifer. Upwards transport of tritium into the Upper Three Runs aquifer from underlying aquifers may be possible but is not supported by the existing data. First, no tritium has been detected in the aquifers below the Upper Three Runs aquifer in eastern Burke County. Second, the hydrostatic head in the Upper Three Runs aquifer in eastern Burke County is generally higher than the head in the shallowest of the confined aquifers (the Gordon aquifer) directly below it. This means that there is the tendency for water to move downwards from the Upper Three Runs aquifer into the confined aquifer rather than upwards into the Upper Three Runs aquifer. Third, current data for the Gordon aquifer in eastern Burke County indicate that the direction of groundwater flow within the aquifer is to the east rather than to the west. Therefore, any tritium leaking upwards into the Upper Three Runs aquifer should be carried towards the Savannah River rather than into the interior of Burke County. However, the model can not be totally rejected. Tritium has been detected in the Congaree Formation at the Savannah River Site (Cummins and others, 1991; Murphy and others, 1991; Arnett and others, 1992). The number of sites in Burke County for sampling the Gordon aquifer for tritium is small. There is at least one area in Burke County (Site TR92-3) in which the hydrostatic head is reversed so that, at that site, it may be possible for ground water to move from the Gordon aquifer into the Upper Three Runs aquifer. There are areas in Burke County in which the confining bed that separates the Upper Three Runs and Gordon aquifers is

62

m1ssmg. Moreover, the data for the direction of groundwater flow in the Gordon aquifer are sparse, thus the eastward direction of ground-water flow may have some exceptions.

Upper Three Runs aquifer. Because of the overall inconsistency of the data with the predictions of the model, the point source model for regional pollution of the Upper Three Runs aquifer is highly unlikely.

Downwards Transport The model of downwards transport postulates that tritium from the Savannah River Site in South Carolina has polluted rainwater in Georgia. The model postulates that this tritium-polluted rainwater percolates downwards and recharges the Upper Three Runs aquifer in eastern Burke County. Downwards transport includes recharge of the aquifer from both tritium in current rainfall and tritium that has accumulated in the vadose zone from previous rainfall. Tritium pollution of the Upper Three Runs aquifer through recharge by tritiated rainwater is the most viable hypothesis, although not fully tested. We have good records showing that rainfall in Burke County is polluted with tritium. The map published by the Westinghouse Savannah River Company (Murphy and others, 1991) (Figure 10), illustrating the distribution of tritium in rainwater in Georgia shows a pattern which is in reasonable agreement with the geographic distribution of tritium observed in the base flow studies. However, although the existing data arc consistent with the rainfall model, they do not prove that this is the actual pathway for tritium into the Upper Three Runs aquifer in eastern Burke County. Even though rainfall data demonstrate that rainwater in Georgia contains elevated concentrations of tritium, the data are too sparse to map the distribution oftritiated rainfall in Georgia. The Wcstinghouse map (Murphy and others, 1991) was based on only four data points in Georgia. There is also no information on tritium in the unsaturated zone in Burke County, and no information on the vertical distribution of tritium in the unconfined aquifer.
Point Source Pollution A variant of the downwards transport model postulates that tritium-polluted rainwater has entered the Upper Three Runs aquifer through poorly constructed wells. This is a point source model for the pollution. This model would predict plumes of tritiatcd water moving in the direction of ground-water flow from the point source wells. We know that at least eight of the fifteen water wells showing anomalous concentrations of tritium arc either ungrouted, lack casing, lack a surface pad, or have experienced failure of the casing (Table 9). However, the pattern of distribution of the tritium in the Upper Three Runs aquifer (Figures 25 and 26) docs not support a point source origin for the tritium. The tritium is widespread throughout eastern Burke County. The distribution pattern is approximately concentric around the Hancock Landing area. There are significant tritium occurrences up-gradient from the anomalous wells. The distribution of tritium shows no relationship to the water table surface (Figure 35) and, therefore, is inconsistent with the directions of flow in the

Other Path ways The method of multiple working hypotheses can only test models that have been proposed. The possibility that the tritium observed in the Upper Three Runs aquifer has entered the aquifer by some pathway other than those that have been considered can not be dismissed.
Trans-River Flow (Underflow)
The discussion of potential pathways for tritium into the Upper Three Runs aquifer in Burke County (Section 4.1.4) only considers the observed pollution of that one aquifer. The potential for the pollution of deeper confined aquifers through direct transport of tritium underneath the Savannah River remains a separate unresolved issue. This issue is being addressed by another investigation, the TransRiver Flow Project (initially known as the Underflow Project), presently being carried out by the USGS. The Trans-River Flow Project will include the construction of several ground-water monitoring well clusters in Burke and Screven Counties, performance of aquifer tests at selected well clusters, study of ground-water quality, development ofpoten tiometric surface maps, investigation ofstreamflow conditions in the Savannah River, development ofa conceptual model of the hydrologic flow system, and testing of the conceptual model using a digital ground-water flow model.
RECOMMENDATIONS
Based on the results of the investigations conducted by EPD on the occurrence of tritium in the ground water of Burke County, thirteen recommendations are made. The recommendations fall into four broad categories: public awareness, public health, technical studies, and long-term monitoring.
Public Awareness
Recommendation 1. We recommend that the Department of Natural Resources (DNR) issue a press release in Burke County to explain the situation to the public in the affected area. Such a press release should: l) provide the public with technically correct information about the occurrence of tritium in their ground water; and 2) provide the public with information that their health is not being threatened.
Public Health
Recommendation 2. We recommend that EPD, through the Water Well Standards Advisory Council, issue an advisory notice to all well drillers working in Burke

64

County, that, until further notice, DNR recommends that all new domestic water supply wells within the area of tritium pollution be drilled down to the Gordon aquifer and that all wells be grouted through the entire interval of the Upper Three Runs aquifer. This recommendation will result in some additional cost to well owners in that wells would be drilled approximately 70 to 100 feet deeper than current practice in eastern Burke County. Slight additional costs would also be incurred by grouting the 100 to 200 foot thickness of the Upper Three Runs aquifer. Because the concentrations of tritium in the Upper Three Runs aquifer are well below the EPA Maximum Contaminant Level (MCL), the advisory would only be a recommendation, and would leave the matter up to the landowner contracting for the construction of the well. The EPD should issue a similar advisory to the Public Health Division of the Georgia Department of Human Resources.
Recommendation 3. We recommend that EPD issue no new permits, within the area of Burke and surrounding counties affected by tritium pollution of ground water, for public water supply wells that draw their water from the Upper Three Runs aquifer.
Technical Studies
Recommendation 4. We recommend that the DOE conduct or fund additional studies of the Upper Three Runs aquifer in eastern Burke County. Specifically we recommend that studies be conducted to evaluate the following characteristics of the Upper Three Runs aquifer:
(1) The vertical distribution of tritium should be evaluated within the water table aquifer. This will provide information on which parts of the aquifer arc polluted as well as on the pathway of the tritium into the aquifer.
(2) The lateral or geographic extent to the tritium distribution in the Upper Three Runs aquifer should be defined. Current studies have not firmly established the lateral limits of the pollution. At least one additional base flow study be conducted covering a larger area than the two previous base flow studies.
(3) The hydraulic properties of the Upper Three Runs aquifer should be established through aquifer tests. The hydraulic properties will allow the Department of Natural Resources to predict the horizontal and vertical flow rates within the aquifer and the effects of pumping on the aquifer.

eastern Burke County. Specifically we recommend that all of the domestic water wells, public drinking water supply and municipal water supply wells that are used for monitoring tritium in eastern Burke County be logged geophysically to determine well depth, aquifer, and well construction. Our current study has demonstrated that drillers' records are inadequate for this purpose.
Recommendation 7. We recommend that DOE conduct or fund additional studies of tritium transport in the Upper Three Runs aquifer in eastern Burke County. Specifically we recommend that studies be conducted to test several models of tritium transport into the Upper Three Runs aquifer including pollution through recharge by tritiated rainwater and pollution through upwards transport. Specific investigations include: analysis of the vertical distribution of tritium in the Upper Three Runs aquifer (covered in recommendation 5); age dating of the water in the Upper Three Runs aquifer; evaluation of tritium accumulation in the vadose zone; and evaluation of the effect of faults on tritium transport.
Recommendation 8. We recommend that DOE provide funding to accelerate the drilling program for the USGS Trans-River Flow Project and bring that study to conclusion. Recommendation 9. We recommend that DOE provide funding to fully evaluate the tritium pollution of the Delaigle Trailer Park well #3. This public water supply well was the original well that alerted EPD to the general problem of tritium pollution of ground water in Burke County. Serious questions remain about the construction of this well.
Recommendation 10. We recommend that the DOE conduct or fund high resolution tritium analyses of the ground-water monitoring wells in eastern Burke County. The tritium analyses in the present study had a detection limit of 100 picoCuries per liter. It is possible that tritium has begun to leak into the Gordon aquifer but that present analytical methods can not detect the pollution. There are analytical methods that have detection limits for tritium of 1 picoCurie per liter or less. Such high resolution analysis would provide a relatively inexpensive early warning of possible contamination of the confined aquifer.

Recommendation 5. We recommend that the DOE conductor fund additional studies of the confining bed at the base of the Upper Three Runs aquifer in eastern Burke County. Specifically we recommend that measurements be made to assess the vertical hydraulic conductivity of the confining bed as well as the potential for leakage of tritiated ground water from the Upper Three Runs aquifer into the upper Gordon aquifer.
Recommendation 6. We recommend that DOE conduct or fund additional studies of water supply wells in

Long-Term Monitoring
Recommendation 11. We recommend that DOE establish or fund a long-term monitoring network for periodic sampling of ground water from both the Upper Three Runs aquifer and the upper Gordon aquifer in eastern Burke County. This network should be made up of wells constructed specifically as ground-water monitoring wells. Monitoring wells already constructed as part of the current study should form the base for the network. Four to twelve additional monitoring wells may be required.

66

Recommendation 12. We recommend that annual base flow studies be incorporated as part ofa long-term monitoring program. The current investigation has clearly demonstrated that base flow studies arc the best available method for evaluating the extent of pollution in the Upper Three Runs aquifer in Burke County. Annual base flow studies will show whether the area of pollution is expanding or contracting.
Recommendation 13. We recommend that the DOE establish or fund along-term monitoring network for monthly sampling of tritium in rainwater in Georgia. The hypothesis that the pathway for tritium into the Upper Three Runs aquifer is through recharge by tritiatcd rainwater can not be adequately evaluated with the current number of monitoring stations in Georgia. The proposed monitoring network must contain a sufficient number ofstations to allow reasonable mapping of tritium distribution.
REFERENCES
Aadland, R. K., Thayer, P. A., and Smits, A. D., 1992, Hydrostratigraphy ofthe Savannah River Site Region, South Carolina and Georgia: in Fallaw, Wallace and Price, Van, editors; Geological Investigations of the Central Savannah River Area, South Carolina and Georgia; Carolina Geological Society, p. B-X-1-6.
Aadland, R. K., and others, 1993, Savannah River Site Environmental Report for 1992, 396 p.
Allison, G. B., 1988, A review of the physical, chemical and isotopic techniques available for estimating groundwater recharge: in Simmers, I., editor, Estimation of Natural Groundwater Recharge: Dordrccht, D. Reidel Publishing Company, p. 49-72.
Arnett, M. W., Karapatakis, L. K., Mamatcy, A. R., and Todd, J. L., 1992, Savannah River Site Environmental Report for 1991 (U): U.S. Department of Energy, WSRCTR-92-186, 562 p. plus Appendices.
Bachtel, D. C. and Boatright, S. R., 1992, The Georgia County Guide (11th Edition): Athens, Georgia, The University of Georgia, College of Agricultural and Environmental Sciences and Department of Housing and Consumer Sciences, Cooperative Extension Service, 203 p.
Baker, Michael, Jr., Inc., 1979, Fort Gordon, Georgia terrain analysis: Unpublished report for the U.S. Army, Nov. 1979, 58 p.

rainfall and runoff in Georgia, 1941-70: Georgia Geologic Survey Hydrologic Atlas 9, I sheet.
Clark, W. Z., Jr., and Zisa, A. C., I 976, Physiographic map of Georgia: Georgia Geologic Survey, scale 1:2,000,000, 1 sheet.
Cummins, C. L., Martin, D. K., Todd, J. L., and Exploration Resources, Inc., 1991, Savannah River Site Environmental Report (U), Volume II Groundwater Monitoring: Westinghouse Savannah River Company report WSRCIM-91-28, 64 p., plus appendices.
Davis, S. N., Campbell, D. J., Bentley, H. W., and Flynn, T. J., 1985, Ground-water Tracers: Ada, Oklahoma, Robert S. Kerr Environmental Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, 200 p. Dincer, T., and Davis, G. H., 1984, Application of environmental isotope tracers to modeling in hydrology: Journal of Hydrology, v. 68, p 95-113.
Domenico, P.A., and Schwartz, F. W., 1990, Physical and Chemical Hydrogeology: New York, John Wiley & Sons, 824 p.
Fanning, J. L., Doonan, G. A., and Montgomery, L. T., I 992, Water use in Georgia by county for I 990: Georgia Geologic Survey Information Circular 90, 98 p.
Faure, Gunter, 1986, Principles of isotope geology: New York, John Wiley and Sons.
Faye, R. E. and Prowell, D. C., 1982, Effects of Late Cretaceous and Cenozoic faulting on the geology and hydrogeology of the Coastal Plain near the Savannah River, Georgia and South Carolina: U.S. Geological Survey OpenFile Report 82-156, 73 p.
Fontes, J. Ch., 1980, Environmental isotopes in groundwater hydrology: in Fritz, P. and Fontes, J. Ch., editors, Handbook oflsotope Geochemistry; Volume 1, The Terrestrial Environment, A: Amsterdam, Elsevier Scientific Publishing Company, p. 75- 140.
Freeze, R. A., and Cherry, J. A., 1979, Groundwater: Englewood Cliffs, New Jersey, Prentice-Hall, Inc., 604 p.
Fritz and Fontes, 1980, Introduction: in Fritz, P. and Fontes, J. Ch., editors, Handbook of Isotope Geochemistry; Volume 1, The Terrestrial Environment, A: Amsterdam,Elsevier Scientific Publishing Company, p. 1-19.

Brooks, R., Clarke, J. S., and Faye, R. E., 1985, Hydrogeology of the Gordon aquifer system of east-central Georgia: Georgia Geologic Survey Information Circular 75, 4 I p.
Carter. R. F., and Stiles, H. R., I983, Avcrage annual

Oat, J. R., 1980, The isotopes of hydrogen and oxygen in precipitation: in Fritz, P. and Fontes, J. Ch., editors, Handbook of Isotope Geochemistry; Volume 1, The Terrestrial Environment, A: Amsterdam, Elsevier Scientific Publishing Company, p. 21-47.

67

Gorday, L. L., 1985, The hydrogeology of the Coastal Plain strata of Richmond and northern Burke Counties, Georgia: Georgia Geologic Survey Information Circular 61, 43 p.
Henry, Vernon J., 1994 Summary of results of a seismic survey of the Savannah River adjacent to the Savannah River Plant Site, Burke County, Ga.; unpublished progress report to the Georgia Geologic Survey Tritium Project, February, 1994, 3 pages.
Hetrick,J. H., 1992, A geologic atlas of the Wrens-Augusta area: Georgia Geologic Survey Geologic Atlas 8.

Rose, S. and James, P., 1993, Geochemical description of ground water in the Gordon and other aquifers in Burke County, Georgia; Preliminary draft report: Unpublished report submitted to the Georgia Geologic Survey, 62 p.
Snipes,D. S.,Fallaw, W.C.,andPrice, V.,Jr.,andCumbest, R. J., 1993, The Pen Branch fault: Documentation of Late Cretaceous and Tertiary faulting in the Coastal Plain of South Carolina: Southeastern Geology, vol. 33, #4, pg. 195218.
GLOSSARY

Huddlestun, Paul F., in preparation, Stratigraphy of eastern Burke County (manuscript in progress), Georgia Geologic Survey Open File Report.
Huddlestun, P. F. and Hetrick, J. H., 1979, The stratigraphy of the Barnwell Group of Georgia: Georgia Geologic Survey Open File Report 80-1, 89 p.
Huddlestun, P. F. and Hetrick, J. H., 1986, Upper Eocene stratigraphy of central and eastern Georgia: Georgia Geologic Survey Bulletin 95, 78 p.
Leeth, D. C. and Nagle, D. N., 1994 Geomorphology and geologic characterization of the Savannah River Floodplain, Georgia and South Carolina; Geologic Society of America Abstracts with Programs, vol. 26, #4.
Moore, J., James, A., Daggett, J., and Price, S., 1992, Aquifer Test Report, Tritium Site TR92-1, Georgia, August 11-12, 1992: Clemson University Department of Earth Sciences Draft Report, 93 p.
Murphy, C. E., Jr., Bauer, L. R., Hayes, D. W., Marter, W. L., Ziegler, C. C., Stephenson, D. E., Hoel, D. D., and Hamby, D. M., 1991, Tritium in the Savannah River Site environment (U): Westinghouse Savannah River Company report WSRC-RP-90-424-1, 133 p., plus appendices.
National Council on Radiation Protection an Measurements, 1987, Ionizing radiation exposure of the population of the United States: NCRP Report 93, 87 pp.
Robertson, W. D., and Cherry, J. A., 1989, Tritium as an indicator of recharge and dispersion in a groundwater system in central Ohio: Water Resources Research, v. 25, no. 6, p 1097-1109.
Rose, S., 1992, Tritium in groundwater of the Georgia Piedmont: implications for recharge and flow paths: Hydrological Processes, v. 6, p. 67-78.
Rose, S., 1993, Environmental tnt1um systematics of baseflow in Piedmont Province watersheds, Georgia (USA): Journal of Hydrology, v. 143, p. 191-216.

"Aquifer"-a permeable body of rock or sediment with the ability to store and transmit ground water. The permeability is due to the presence of fractures and minute passageways between grains. Examples in the study area are sand and/or gravel bodies and fractured limestones.
"Confined aquifer"-an aquifer which is bounded above and below by confining beds. Recharge is by rainfall onto distant exposed portions of the aquifer, or downward leakage through the overlying confining bed.
"Confining bed"-an impermeable body of rock or sediment which prevents the flow of ground water. The impermeability is due to the density of the rock. Examples in the study area are clays and dense limestones.
"Discharge"-methods by which water leaves the aquifer.
"picoCurie"- the quantity of a radioactive element which releases 0.037 electrons per second.
"Potentiometric surface"-water in confined aquifers is under hydraulic pressure. When a well penetrates the upper confining bed, the potentiometric head is the distance above the aquifer that the water level reaches (because of the pressure). Contour maps constructed from water level measurements give indications of the direction of groundwater flow.
"Recharge"-methods by which water enters the aquifer.
"Unconfined aquifer"-an aquifer which is exposed to the ground surface. Recharge is by rainfall and sometimes by surface stream leakage. Other terms for this type of aquifer include: "Water Table aquifer," "Surficial aquifer," "Shallow aquifer," and "Upper Three Runs aquifer" (in this report).
"Water table"-the level at which subsurface rock or sediment becomes saturated with water. Measurements are taken in wells or at spring sites, where the water table intersects the ground surface. The water table generally follows the surface topography.

68

Appendix 1 Identification of wells sampled before and during the Tritium Project by EPD personnel

Owner/Resident 1. A & A Store#l

Date 10/30/91 12/06/91

Lab# S-5397 S-5466

Tritium-pC/liter 300 100 100 100

2. A & A Store#2

09/04/92 01/05/93

S-5937 S-6298

200 100 <100

3. A & A Store #3 (aka Delaigle Trailer Park well #3)

09/22/87 04/04/88 04/10/90 01/08/91 07/18/91 09/04/91 10/08/91 10/30/91 11/06/91 12/06/91 01/14/92 02/10/92 03/04/92 07/07/92 08/06/92 01/28/93 04/07/93 05/18/93 08/26/93

SR-3272 SR-3577 SR-4852 SR-5382 SR-5688 SR-5799 SR-5838 SR-5943 SR-5954 SR-6048 S-5482 S-5548 SR-6213 SR-6419 S-5905 S-6326 S-6356 S-6389 S-6455

<200 600 100 <300 1200200 1200 200 1500200 1400200 1200200 800 100 1400200 1200200 1300200 1100200 1300200 1200200 1200200 1000 100 1200200 1100200

4. Albert Patrick Residence

11/05/91

S-5509

<100

5. Allen Chapel Church

10/17/91

S-5504

<100

6. Arthur Jackson Residence

10/08/91 11/05/91 12/06/91 02/10/92 03/04/92 07/07/92 12/30/92 07/15/93

S-5493 S-5507 S-5464 S-5547 S-5673 SR-6417 S-6289 S-6420

1300200 1100200 1100200 1000 100 1100 200 1200200 1000 100 1100200

7. Augusta Lock and Dam

04/13/87 09/22/87 04/04/88 10/04/88 04/11/89 10/10/89 04/10/90 01/08/91 07/18/91 10/08/91 07/07/92

SR-3008 SR-3262 SR-3568 SR-3845 SR-4203 SR-4474 SR-4844 SR-5377 SR-5683 SR-5857 SR-6387

<200 <300 <300 <300 <300 <300 <300 <100 100 100 100 100 <100

8. Avner Delaigle Residence

10/30/91 08/06/92

S-5398 S-5906
69

100 100 <100

Owner/Resident 9. Bill Sturgeon Residence
10. Botsford Springs Church
11. Burke County EMA #2
12. Burke County EMA #11-4"
13. Burke County EMA #11-6" 14. Charles Beazley Residence 15. Christine Williams Residence 16. Clyde Dixon Residence 17. Clarence Overton Residence
18. Dale Heath Residence 19. Daniel Grove Church
20. Dave Rogers Residence 21. Derossie Delaigle Residence
Residence 22. E. J. Weis, Sr. Residence 23. E. J. Weis, Jr. Residence 24. Earl Bush Residence 25. Ebenezer Church 26. Emma Moton Residence 27. Ethel Wesby Residence 28. Eugene Carter Residence 29. Eula Hopkins Residence 30. Fairfield Church 31. Flakes Auto

Date
02/02/93 05/18/93 10/17/91 09/04/92 10/17/91 09/04/92 10/17/91 08/06/92 08/06/92
01/14/92
10/01/92
10/01/92 01/14/92 07/07/92 10/20/93
08/06/92 10/17/92 11/10/92
08/21/92
08/06/92
09/11/92
09/11/92
08/06/92
10/17/91 07/07/92
01/28/93
09/03/92
10/01/92
10/17/91 10/08/91 02/10/92 08/21/92

Lab#
S-6334 S-6384
S-5501 S-5911
S-5500 S-5933
S-5496 S-5909
S-5901
S-5475
S-5976
S-5972
S-5514 SR-6418 S-6683
S-5908
S-5497 S-6240
S-5916
S-5907
S-5944
S-5945
S-5904
S-5495
SR-6422
S-6328
S-5939
S-5971
S-5499 SR-5846 S-5545 S-5915 70

Tritium-pC/Iiter
1200 200 1300200
<100 <100
<100 <100
<100 100 100 <100
300 100
<100
<100
<100 200 100 <100
100 100 <100 <100
<100
<100
200 100
200 100
100 100 <100
200 100
<100
100 100
100 100
<100 <100 100 100 <100

Owner/Resident 32. Flakes Residence 33. Frank Wimberly Residence 34. Ga Welcome Ctr.
1-20 Augusta
35. Ga Welcome Ctr. 1-20 Augusta
36. George Wilson Residence
37. Girard Mall
38. Glen Saxon Residence 39. Glenn Stroud Residence 40. Gobbie Grove Church
(aka Eva Grubbs Residence) 41. Greg Hawkins 42. Hancock Landing
Lodge

Date 10/30/91
10/01/92 01/06/93
04/13/87 09/22/87 04/04/88 10/04/88 04/12/89 10/10/89 04/10/90 01/08/91 07/18/91 10/08/91 07/07/92
04/13/87 09/22/87 10/05/88 04/12/89 10/10/89 04/10/90 01/08/91 0 7/18/91 10/08/91 07/07/92
02/10/92 03/04/92 07/07/92 09/03/92 01/06/93
04/13/87 09/22/87 04/04/88 04/11/89 10/10/89 04/10/90 01/08/91 07/18/91 10/08/91 07/07/92
10/01/92
07/07/92
09/11/92 01/05/93
09/04/92
04/10/90 10/08/91

Lab# S-5399
S-5974 S-6296
SR-3007 SR-3261 SR-3567 SR-3844 SR-4202 SR-4475 SR-4843 SR-5376 SR-5682 SR-5856 SR-6386
SR-3011 SR-3266 SR-3851 SR-4207 SR-4470 SR-4850 SR-5380 SR-5687 SR-5859 SR-6390
S-5544 S-5670 S-5880 S-5940 S-6293
SR-3010 SR-3263 SR-3571 SR-4205 SR-4471 SR-4848 SR-5378 SR-5685 SR-5860 SR-6388
SR-5969
SR-6405
S-5948 S-6297
S-5935
SR-4846 SR-5848
71

Tritium-pC/liter <100
1000 100 800 100
<200 <300 <300 <300 <300 <300 <300 <100 100 100 100 100 200 100
<200 <300 <300 <300 <300 <300 <100 <100 100 100 <100
600 100 700 100 1100200 700 100 500 100
<200 <300 <300 <300 <300 <300 <100 100 100 100 100 <100
100 100
<100
<100 <100
400 100
<300 100 100

Owner/Resident 43. Harry Williams Residence 44. Hatcher Fann Well 45. Hatcher Fann Irrigation weII 46. Hug-a-Hog Plantation
47. Jeanette Powell Residence 48. Jeff Chandler Residence 49. Jerry Collins Residence 50. Jim Garrison Residence 51. Job Springs Church 52. John Sturdivant Residence 53. Jones Residence on River Rd. 54. J.C. Mallard Residence 55. Chris Mallard Residence 56. Julian Morris Residence
57. Kennedy Store 58. Kwik Way Store 59. Lamar Paul Residence
60. Leonard Williams Residence 61. Lewis Sullivan Trailer 62. Lonnie Griffin Residence 63. Lucy Rouse Residence

Date
01/14/92
11/20/91 07/07/92
02/10/92 07/07/92
10/17/91 11/05/91 08/14/92 01/06/93
10/01/92
08/21/92
09/04/92
10/27/92
10/17/91
02/10/92
01/14/92
08/14/92
08/14/92
02/12/92 07/07/92 02/03/94
10/17/91 08/21/92
02/10/92 08/21/92
08/21/92 09/04/92 01/06/93 02/03/93 07/15/93
08/06/92
08/06/92
10/01/92
11/17/92

Lab#
S-5519
S-5407 SR-6412
S-5541 SR-6412
S-5494 S-5508 S-5897 S-6294
S-5970
S-5919
S-5936
S-6236
S-5503
S-5550
S-5517
S-5898
S-5899
S-5617 SR-6414 S-6709
S-5498 S-5913
S-5549 S-5912
S-5918 S-5934 S-6295 S-6332 S-6421
S-5902
S-5903
S-5978
S-6260 72

Tritium-pC/Iiter
200 100
300 100 100 100
<100 100 100
1000 100 900 100 900 100 800 100
<100
<100
100 100
<100
100 100
<100
<100
100 100
200 100
400 100 500 100 400 100
<100 <100
<100 <100
900 100 800 100 900 100 800, 100 800 100
200 100
<100
<100
<100

Owner/Resident 64. Margaret Williams Residence 65. Mary Johnson Residence
66. Mattie Stevens Residence 67. M. A. Miller Residence 68. M. A. Miller Trailer 69. M. A. Miller flowing well
east of River Rd. 70. M. A. Miller flowing well,
downstream from dam 71. M.A. Miller flowing well,
upstream from dam 72. Mary Fincher Residence 73. Mayonnaise Residence 74. McBean Fire Station 75. McCoy Residence 76. Bill Rucker Residence 77. Walnut Run Ostrich Farm
78. Percy Dixon Residence 79. Ralph Greer Residence
80. Ricky Greer (trailer)out of service mid-'92
80A. Ricky Greer Hog Pen faucet
81. Rickey Johnson Residence 82. Robert Lynch

Date
11/17/92
05/18/93 02/03/94
02/20/93
07/02/92
07/02/92
09/03/92
07/02/92
10/22/91 07/02/92
08/21/92
02/10/92
01/17/92 01/15/92
02/08/94
08/14/92 12/30/92 02/02/93 05/18/93
10/01/92
02/10/92 03/04/92 07/07/92 09/03/92 01/06/93 07/15/93
11/20/91 02/10/92 03/04/92
02/03/93
09/11/92
09/03/92

Lab#
S-6261
S-6386 S-6708
S-5941
S-5847
S-5848
S-5941
S-5844
S-5392 S-5845
S-5917
S-5551
S-5477
SR-6110
S-6716
S-5900 S-6290 S-6333 S-6385
S-5973
S-5543 S-5669 SR-6407 S-5938 S-6292 S-6422
S-5408 S-5542 S-5671
S-6336
S-5947
S-5930 73

Tritium-pC/Iiter
<100
1600200 1600200
<100
<100
100 100
<100
100 100
100 100 <100
<100
100 100
<100
200 100
<100
700 100 700 100 600 100 900 100
100 100
600 100 600 100 700 100 800 100 500 100 600 100
1000 100 500 100 600 100
<100
100100
200 100

Owner/Resident 83. Roman Powell Residence 84. Rose Johnson Residence
(aka Willie Brown Residence Gobbie Grove Church)
85. Sharon Jackson Residence 86. Shell Bluff Store
87. Stoney Bluff Park
88. Thomas J. Rouse Residence 89. Thompson Bridge Church 90. Thomson Oak Flooring
(Big T Hunting Lodge) 91. Thomson Oak Flooring
(flowing well - Cox Point) 92. Tommy Greer Residence
93. Unknown Yellow House 94. Viola Brigham Residence

Date
10/28/92
10/08/91 11/05/91 02/10/92 03/04/92 07/07/92 08/21/92 01/28/93 02/03/93
10/01/92
10/08/91 01/14/92 07/07/92 09/03/92
10/05/88 04/12/89 04/10/90 01/08/91 07/18/91 10/08/91 07/07/92
09/16/92
10/17/91
10/26/92 01/28/93
05/18/93
01/14/92 02/10/92 07/07/92
07/07/92 07/07/92
04/13/87 09/22/87 10/04/88 04/11/89 10/10/89 04/10/90 01/08/91 07 /18/91 10/08/91 01/14/92 07/07/92 10/01/92

Lab#
S-6239
S-5492 S-5506 S-5546 S-5672 SR-6415 S-5914 S-6327 S-6335
S-5975
S-5490 S-5520 SR-6406 S-5942
SR-3850 SR-4206 SR-4849 SR-5379 SR-5686 SR-5858 SR-6389
S-5949
S-5502
S-6237 S-6325
S-6387
S-5476 S-5540 SR-6411
SR-6413 SR-6413
SR-3014 SR-3274 SR-3858 SR-4211 SR-4489 SR-4854 SR-5384 SR-5690 S-5485 S-5518 SR-6423 S-5977
74

Tritium-pC/liter
<100
500 100 400 100 400 100 500 100 600 100 500 100 400 100 400 100
100 100
<100 <100 100 100 100 100
<300 <300 <300 <100 <100 <100 <100
<100
<100
100 100 100 100
<100
<100 <100 100100
400 100 400 100
<200 <300 <300 <300 <300 <300 <100 <100 <100 <100 100 100 <100

Owner/Resident 95. Vogtle Construction well 96. Vogtle Make Up well 97. Vogtle Rec. Ctr. well 98. Vogtle Simulator
99. Vogtle Training Center 100. Vogtle Visitor Center 101. Walter Grubbs Residence 102. William Cox Residence 103. Wanda Siller Residence 104. Waynesboro Filter Plant 105. Waynesboro Main Well
106. William Edwards Residence 107. Shell Station-River Road/
Hancock Landing Road (closed in 1989)
108. Alma Crook Residence 109. Earl Mills

Date
10/08/91
10/08/91
10/08/91
09/22/87 04/08/88 10/04/88 04/11/89 10/10/89 04/10/90 01/08/91 07/18/91 10/08/91 01/14/92 07/07/92
10/08/91
07/07/92
09/03/92
08/06/92
07/07/92
02/12/92
04/13/87 09/22/87 04/04/88 10/05/88 04/11/89 10/11/89 04/10/90 01/09/91 07/18/91 10/08/91 02/12/92 05/01/92
07/07/92 09/04/92
04/13/87 09/22/87 04/04/88 10/04/88 04/11/89 10/10/89
02/03/94
02/03/94

Lab#
S-5489
S-5486
S-5488
SR-3273 SR-3579 SR-3856 SR-4210 SR-4488 SR-4853 SR-5383 SR-5689 S-5484 S-5521 SR-6421
S-5487
SR-6420
S-5931
S-5910
SR-6408
S-5616
SR-3012 SR-3269 SR-3575 SR-3853 SR-4208 SR-4469 SR-4851 SR-5381 SR-5680 SR-5861 S-5515 SR-6289
SR-6416 S-5977
SR-3013 SR-3270 SR-3576 SR-3855 SR-4209 SR-4487
S-6706
S-6707
75

Tritium-pC/Iiter
<100
<100
<100
<200 <300 <300 <300 <300 <300 <100 100 100 <100 <100 200 100
<100
100 100
200 100
<100
<100
<100
<200 300 100 <300 <300 <300 <300 <300 <100 <100 <100 <100 <100
200 100 100 100
<200 <200 <300 <300 <300 <300
2000200
800 100

Appendix 2 Radioactive decay of tritium

Tritium isa radioactive isotope ofhydrogen with a half-life of 12.35 years (i.e. after 12.35 years, 50 percent of the original amount of tritium remains). Below is a short table to illustrate the breakdown of tritium over an approximate 100 year time period.

Time Elapsed: Amount of Tritium Remaining

Release date:

100.00%

12.35 years:

50.00%

24.70 years:

25.99%

37.05 years:

12.50%

49.40 years:

6.25%

61.75 years:

3.12%

74.10 years:

1.56%

86.45 years:

0.78%

98.80 years:

0.39%

76

Appendix 3 Tritium in rainfall samples collected in Burke and Screven counties

EPD Station Number
11 11 11 11 11 11 11 11 11 11 11 11 14 14 14 20

Sample Number
1115

Collection Date
07/12/83

Tritium (picoCuries/
liter)
8,900

EPD Station Number
20

1348

12/28/83

7,600

20

1810

12/24/84

11,500

25

1863

01/24/86

10,400

25

1918

02/27/85

6,800

25

1972

03/28/85

13,7000

25

2387

01/23/86

9,000

25

2728

09/25/86

10,900

35

2884

12/12/86

7,700

35

2959

02/26/87

10,700

35

3419

12/17/87

5,200

35

6682

01/07/93

7,000

35

1969

03/28/85

5,200

35

2386

01/23/86

13,000

35

2822

12/12/86

12,300

35

858

12/02/82

6,000

Sample Number
2519 4692 857 923 1102 1813 1970 1353 1811 1859 1967 2683 2961 4803 5389

Collection Date

Tritium (picoCuries/
liter)

04/17/86

7,500

01/23/90

5,700

12/02/82

31,700

01/27/83

6,300

06/16/86

6,700

12/18/84

9,500

03/28/85

5,300

12/28/83

8,000

12/18/84

5,400

01/24/85

9,900

03/28/85

8,900

08/27/86

11,000

02/26/87

9,400

03/15/90

5,300

01/08/91

5,700

The table shows rainfall tritium concentrations >5000 picoCuries per liter, from a total of25 EOD rainfall collection stations in the study area. Locations are shown in Figure 7. Station 11: Hancock Landing Road, near the Savannah River Station 14: Georgia Highway 23, I mile north of Girard Station 20: Welcome Center, U.S. Highway 301 at the Savannah River Station 25: Georgia Power company Maintenance Office in Waynesboro Station 35: Georgia Power Plant Vogtlc Simulator

77

Appendix4 Tritium in rainfall categorized by tritium concentration
This table categorizes rainfall tritium samples into three concentration classes. All samples are from collection sites in Broke and Screven counties in Georgia. Of the 25 EPD rainfall collection sites in the area, the seven presented have the longest and most consistent collection history. Values are in picoCuries per liter. Rainfall collection sites are shown in Figure 7.

Collection Site
Number

Total Number
of Samples

Number(%) of Samples
<2500 picoCuries per
liter

Number(%) of Samples 2500 - 4999 picoCuries per
liter

Number(%) of Samples
>5000 picoCuries per
liter

11

110

80 (72.2)

17(15.4)

13 (11.8)

14

59

52 (88.1)

4 (6.8)

3 (5.1)

20

116

110 (94.8)

3 (2.6)

3 (2.6)

23

38

34 (89.5)

4 (10.5)

0

25

144

132 (88.5)

12 (8.2)

5 (3.3)

35

1114

84 (75.7)

19 (17.1)

8 (7.2)

48

148

142 (95.9)

6 (4.1)

0

Site 11: Hancock Landing Road, near the Savannah River Site 14: Georgia Highway 23, 1 mile north of Girard Site 20: Welcome Center on U.S. Highway 301 at the Savannah River Site 23: Intersection of Georgia Highways 23 and 24 in Sardis Site 25: Georgia Power Company Maintenance Office in Waynesboro Site 35: Georgia Power Plant Vogtle Simulator Site 48: Augusta Youth Center on Georgia Highway 56

78

Appendix 5 Results of analyses from fifteen tritium polluted residential and public water wells The locations of the anomalous water wells are shown in Figures 22 and 23. All tritium values are in picoCuries per liter.

Site 6: Arthur Jackson Residence Latitude: 3309'40" Longitude: 81 46'32"

Site 3: Delaigle Trailer Park well #3 Latitude: 3308'09" Longitude: 81 47'03"

Collection Date
10/08/91 11/05/91 12/06/91 02/10/92 03/04/92 07/07/92 12/30/92 07/15/93

Lab ID Number S-5493 S-5507 S-5464 S-5547 S-5673 SR-6417 S-5905 S-6420

Tritium Content 1300 200 1100200 1100200 1100 200 1100 200 1200 200 1200 200 1100 200

Site 36: George Wilson Residence Latitude: 3312'32" Longitude: 81 50'56"

Collection Date

Lab ID Number

Tritium Content

02/10/92

S-5544

600 100

03/04/92

S-5670

700 100

07/07/92

S-5880

1100 200

09/03/92

S-5490

700 100

01/06/93

S-6293

500 100

79

Collection Date
09(22/87 04/04/88 04/10/90 01/08/91 07/18/91 09/04/91 10/08/91 10/30/91 11/05/91 12/06/91 01/14/92 02/10/92 03/04/92 07/07/92 08/06/92 01(28/93 04/07/93 05/18/93 08(26/93

Lab ID Number SR-3272 SR-3577 SR-4852 SR-5382 SR-5688 SR-5799 S-5483 S-5396 S-5505 S-5465 S-5516 S-5548 SR-6213 SR-6419 S-5905 S-6326 S-6356 S-6389 S-6455

Tritium Content
<200 600 100
<300 1200200 1200200 1500200 1400200 1200200 800 100 1400200 1200 200 1300200 1100200 1300200 1200200 1200200 1000 100 1200200 1100200

Site 79: Ralph Greer Residence Latitude: 3312'37" Longitude: 8112'37"

Site 46: Hug-a-Hog Plantation Latitude: 3303'10" Longitude: 8147'39"

Collection Date
12/20/91 12/06/91 02/10/92 03/04/93 07/07/92 09/03/92 01/06/93 07/15/93

Lab ID Number S-5408 S-5463 S-5543 S-5669 SR-6407 S-5938 S-6292 S-6422

Tritium Content 100 100 <100(?) 600 100 600 100 700 100 800 100 500 100 600 100

Site 84: Rose Johnson Residence Latitude: 3310'23" Longitude: 8153'00" (first 2 samples recorded as Gobbic Grove Church)

Collection Date

Lab ID Number

Tritium Content

10/08/91

S-5492

500 100

11/05/91

S-5506

500 100

02/10/92

S-5546

400 100

03/04/92

S-5672

500 100

07/07/92

SR-6415

600 100

08/21/92

S-5914

500 100

01/28/93

S-6327

400 100

02/03/93

S-6335

400 100

80

Collection Date
10/17/91 11/05/91 08/14/92 01/06/93

Lab ID Number S-5494 S-5508 S-5897 S-6294

Site 77: Walnut Run Ostrich Farm Latitude: 3307'57" Longitude: 81 50'48"

Collection Date
08/14/92 12/30/92 02/04/93 05/18/9.

Lab ID Number S-5900 S-6290 S-6333 S-6385

Site 59: Lamar Paul Residence Latitude: 3311 '09" Longitude: 8149'20"

Collection Date
08/21/92 09/04/92 01/06/93 02/04/93 07/15/93

Lab ID Number S-5918 S-5934 S-6295 S-6332 S-6241

Tritium Content 1000 100 900 100 900 100 800 100
Tritium Content 700 100 700 100 600 100 600 100
Tritium Content 900 100 800 100 900 100 800 100 800 100

Site 56: Julian Morris Residence Latitude: 3311 '05" Longitude: 8149'15"

Collection Date
02/12/92
07/07/92 02/03/94

Lab ID Number S-5617
SR-6414 S-6709

Tritium Content
100 100
500 100 400 100

Site 65: Mary Johnson Residence Latitude: 3307'57" Longitude: 8150'22"

Collection Date
05/18/93
02/03/94

Lab ID Number
S-6386
S-6708

Tritium Content
1600 200
1600 200

Site 80: Ricky Greer Residence Latitutde: 3312'24" Longitude: 81 50'48" (Well out of service as of mid-1992)

Collection Date
02/10/92
03/04/92

Lab ID Number
S-5542
S-5671

Tritium Content
500 100
600 100

Site 109:Earl Mills Residence Latitude: 3309'26" Longitude: 8147'38"

Site 9: Bill Sturgeon Residence Latitude: 3307'57" Longitude: 81 "50'22"

Collection Date
02/02/93
05/18/93

Lab ID Number
S-6334
S-6384

Tritium Content
1200200
1300200

Site 108: Alma Crook Residence Latitude: 3309'24" Longitude: 81 47'38"

Collection Date
02/03/94

Lab ID Number
S-6707

Tritium Content
2000200

Site 33: Frank Wimberly Residence Latitude: 3310'34" Longitude: 8153'00"

Collection Date
10/01/92
01/06/93

Lab ID Number
S-5974
S-6296

Tritium Content
1000 100
800 100

Collection Date

Lab ID Number

Tritium Content

02/03/94

S-6706

800 100

81

Appendix6 Results of Tritium Analyses from Ground-Water Monitoring Wells The locations of the ground-water monitoring wells arc shown in Figure 19. All tritium values are in picoCuries per liter.
SiteTR92-l

Well Designation and Aquifer Collection Date Lab ID Number Tritium Content

TR92-1A Upper Three Runs

06/25/92

S-5389

<100

TR92-1A Upper Three Runs

07/16/92

S-5853

200 100

TR92-1A Upper Three Runs

08/14/92

S-5895

300 100

TR92-1A Upper Three Runs

08/14/92

S-5896

200 100

TR92-1A Upper Three Runs

11/17/92

S-6258

200 100

TR92-1A Upper Three Runs TR92-1B Gordon

02/08/94 06/25/92

S-6714 S-5840

300 100 <100

TR92-1B Gordon

07/08/92

S-5851

<100

TR92-1B Gordon

11/17/92

S-6259

<100

TR92-1B Gordon
TR92-1C Meyers Branch "aquifer" (1)
TR92-1C Meyers Branch "aquifer"
TR92-1C Meyers Branch "aquifer"
TR92-1D Dublin

02/08/94 06/25/92 07/16/92 12/14/92 06/25/92

S-6715 S-5841 S-5854 S-6284 S-5842

<100 <100 <100 <100 <100

TR92-1D Dublin

07/07/92

S-5852

<100

TR92-1D Dublin

12/14/92

S-6285

<100

(1) Same as "Sand within Meyers Branch Confining System". 82

SiteTR92-2

Well Designation and
aquifer

Collection Date

TR92-2A Upper Three Runs

02/18/93

TR92-2A Upper Three Runs

03/02/93

TR92-2A Upper Three Runs

03/02/93

TR92-2A Upper Three Runs

07/15/93

TR92-2A Upper Three Runs
TR92-2B Gordon TR92-2B Gordon TR92-2B Gordon

02/03/94
01/06/93 02/18/93 07/15/93

Lab ID Number
S-6345
S-6348
S-6349
S-6418
S-6712 S-6286 S-6346 S-6419

Tritium Content
1700200
1400200
1400 200
1400 200
1400200 <100 <100
100 100

Site TR92-3

Well Designation and
Aquifer
TR92-3A Upper Three Runs
TR92-3B Gordon
TR92-3B Gordon

Collection Date
03/02/93 01/05/93 09/01/93
83

Lab ID Number
S-6350
S-6287 S-6462

Tritium Content
1200 200
<100 <100

Site1R92-4

Well Designation and Aquifer
1R92-4A Upper Three Runs
1R92-4A-2 Upper Three Runs (1)

Collection Date
03/02/93
02/03/64

1R92-4B Gordon

01/06/93

Lab ID Number
S-6351

Tritium Content
1600200

S-6713

1600200

S-6288

<100

1R92-4B Gordon

08/26/93

S-6454

(1) Second well drilled for better water production.

<100

84

SiteTR92-5

Well Designation and
Aquifer TR92-5A Upper
Three Runs
TR92-5A Upper Three Runs
TR92-5A Upper Three Runs
TR92-5A Upper Three Runs
TR92-5B Gordon TR92-5B Gordon
TR92-5B Gordon TR92-5B Gordon TR92-5B Gordon
TR92-5B Gordon TR92-5B Gordon
TR92-5C (2)
TR92-5C TR92-5C
TR92-5C TR92-5C
TR92-5C TR92-5C

Collection Date
08/26/93
09/01/93
09/01/93
02/03/94 06/15/93 06/15/93 06/24/93 06/24/93 06/24/93 07/15/93 07/15/93 08/26/93 09/01/93 09/01/93 09/01/93 09/07/93 09/07/93 02/03/94

Lab ID Number

Tritium Content

S-6450

1000 100

S-6456

800 100

S-6457

1000 100

S-6710
S-6395 S-6396 S-6399 S-6400 S-6401 S-6415 S-6416 S-6452 S-6458 S-6459 S-6460 S-6478 S-6479 S-6711

1000 100
500 100 (1) 600 100 (1)
<100 100 100
<100 <100 <100 300 100 400 100 300 100 400 100 300 100 300 100 300 100

(1) Positive results due to tritium introduced during drilling. Samples taken after adequate purging had no detectable tritium. (2) Constructed to duplicate Dclaiglc Trailer Park well #3.
85

Site1R92-6

Well Designation and
Aquifer
1R92-6A Upper Three Runs
1R92-6B Gordon
1R92-6B Gordon
1R92-6B

Collection Date
09/01/93 08/26/93 09/01/93 09/03/93

Lab ID Number
S-6463
S-6453 S-6461 S-6473

Tritium Content
dry hole
<100 <100 <100

Miller's Pond Site (USGS Trans-River Flow Project)

Well Designation and
Aquifer
TW-1 Lower Midville
1R-2 Upper Midville
TW-3 Allendale Confining Bed
TW-4 Upper Three Runs
TW-5A Upper Dublin
TW-6 Lower Dublin
TW-7Lower Dublin

Collection Date 12/14/92 11/03/92 01/08/93 01/15/93 02/11/93 01/23/93 03/03/93

Lab ID Number S-6282 S-6283 S-6329 S-6330 S-6344 S-6331 S-6353

Tritium Content
<100 <100 <100 700 100 <100 <100 <100

Appendix 7 Tritium Project cluster site data
The following six pages include the pertinent data for each Tritium Project cluster site: 1) Site elevation 2) Well construction data and diagrams 3) Geophysical data 4) Lithologic data 5) Water levels
86

Well Construction Diagrams

TR92

Depth below

ground surface

2a

'

2b *

Gamma

Geophysical Logs
Spontaneous Potential

Single-Point Resistivity

Lithologic Symbols

Lithology
Formations/Members

...........:..........:......:...-:....:........:..........:...........-..:.................

Spontaneous Potential

Long-Normal Resistivity

Lithologic Symbols

---------:---------:.--------..----------_------------------_-------------------:----.---_--------------------:---------..-------..-----------_------------------_--------_------------------_---------------_--------------:-------_---------:-.-------..--.-------.-------------------:.--------..-

..........:..:...:............ . ' :

: :

I:::... .

. .-...... '. .---::--:--.--::--:-.--~--I:-:-,'-_---:-.-~-I---~------Ie---.

e

e I

II

. ..-..-.-...--....-.-..-.-..-...-.-...-:.-..-.-..-.-....--..--.:.-.:-:-:.-.:

Appendix 8
Identification of wells used in construction of Gordon aquifer potentiometric map shown in Brooks and others, 1985. Well locations are shown in Figure 38.

Well Number
29Z3 29Z4 30X4 30Z6
31Z9
32Y4
32X23

Owner
F.P.Saxon
W.T. Stone
Perkins
Miller's Pond
Plant Vogtlc Construction
well #8
William Cox #2
Mill haven Company

Well depth (feet) 170 225 446 92
251
415
375

Depth of casing (feet)
-----
127 400 42
220
360
-----

93

For convenience in selecting our reports from your bookshelves, they are color-keyed across the spine by subject as follows:

Red Dk.Purple Maroon Lt Green Lt Blue Dk.Green Dk. Blue Olive
Yellow
Dk.Orange Brown Black Dk.Brown

Valley and Ridge mapping and structural geology Piedmont and Blue Ridge mapping and structural geology Coastal Plain mapping and stratigraphy Paleontology Coastal Zone studies Geochemical and geophysical studies Hydrology Economic geology Mining directory Environmental studies Engineering studies Bibliographies and lists of publications Petroleum and natural gas Field trip guidebooks Collections of papers

Colors have been selected at random, and will be augmented as new subjects are published.

Editor: Melynda Lewis

The Department of Natural Resources is an equal opportunity employer and offers all persons the opportunity to compete and participate in each area of DNR employment regardless of race, color, religion, national origin, age, handicap, or other non-merit factors.