Ground-water levels, predevelopment ground-water flow, and stream-aquifer relations in the vicinity of the Savannah River Site, Georgia and South Carolina
Ground-water levels, predevelopment ground-water flow, and stream-aquifer relations in the vicinity of the U.S. Department of Energy Savannah River Site, Georgia and South Carolina, were evaluated as part of a cooperative study between the U.S. Geological Survey, U.S. Department of Energy, and Georgia Department of Natural Resources. As part of this evaluation: (1) ground-water-level fluctuations and trends in three aquifer systems in sediment of Cretaceous and Tertiary age were described and related to patterns of ground-water use and precipitations; (2) a conceptual model ofthe stream-aquifer flow system was developed; (3) the predevelopment ground-water flow system, configuration of potentiometric surfaces, trans-river flow, and recharge-discharge relations were described; and (4) stream-aquifer relations and the influence of river incision on ground-water flow and stream-aquifer relations were described. The 5,147-square mile study area is located in the northern part of the Coastal Plain physiographic province of Georgia and South Carolina. Coastal Plain sediments comprise three aquifer systems consisting of seven aquifers that are separated hydraulically by confining units. The aquifer systems are, in descending order: (1) the Floridan aquifer system—consisting of the Upper Three Runs and Gordon aquifers in sediments of Eocene age; (2) the Dublin aquifer system—consisting of the Millers Pond, upper Dublin, and lower Dublin aquifers in sediments of Paleocene-Late Cretaceous age; and (3) the Midville aquifer system—consisting of the upper Midville and lower Midville aquifers in sediments of Late Cretaceous age. The Upper Three Runs aquifer is the shallowest aquifer and is unconfined to semi-confined throughout most of the study area. Ground-water levels in the Upper Three Runs aquifer respond to a local flow system and are affected mostly by topography and climate. Ground-water flow in the deeper, Gordon aquifer and Dublin and Midville aquifer systems is characterized by local flow near outcrop areas to the north, changing to intermediate flow and then regional flow downdip (southeastward) as the aquifers become more deeply buried. Water levels in these deeper aquifers show a pronounced response to topography and climate in the vicinity of outcrops, and diminish southeastward where the aquifer is more deeply buried. Stream stage and pumpage affect ground-water levels in these deeper aquifers to varying degrees throughout the study area. The geologic characteristics of the Savannah River alluvial valley substantially control the configuration of potentiometric surfaces, ground-water-flow directions, and stream-aquifer relations. Data from 18 shallow borings indicate incision into each aquifer by the paleo Savannah River channel and subsequent infill of permeable alluvium, allowing for direct hydraulic connection between aquifers and the Savannah River along parts of its reach. This hydraulic connection may be the cause of large ground-water discharge to the river near Jackson, S.C., where the Gordon aquifer is in contact with Savannah River alluvium, and also the cause of lows or depressions formed in the potentiometric surfaces of confined aquifers that are in contact with the alluvium. Ground water in these aquifers flows toward the depressions. The influence of the river is diminished downstream where the aquifers are deeply buried, and upstream and downstream ground-water flow is possibly separated by a water divide or "saddle". Water-level data indicate that saddle features probably exist in the Gordon aquifer and Dublin aquifer system, and also might be present in the Midville aquifer system. Ground-water levels respond seasonally or in long term to changes in precipitation, evapotranspiration, pumpage, and river stage. Continuous water-level data and water-levels measured in a network of 271 wells during the Spring (May) and Fall (October) in 1992, indicate that seasonal water-level changes generally are less than 4 feet, and that larger changes occur near supply wells. Long-term data (more than 10 years of record) from 283 wells indicate that water levels in most aquifers declined during some period prior to 1993. Although most declines were less than 15 feet and were limited to scattered areas influenced by supply wells, widespread declines of as much as 59 feet occurred in the lower Midville aquifer in Richmond County, Georgia. Water-level data at well-cluster sites indicate that in the vicinity of major ground-water divides, head decreases with depth; and in the vicinity of a regional drain, such as the Savannah River, head increases with depth. Although this vertical head distribution holds true over much of the study area, water-level data indicate that the Gordon, Millers Pond, and lower Dublin aquifers are apparent hydrologic "sinks" in parts of the study area, indicating a potential for vertical leakage from both above and below. Reasons for anomalously low heads in these three aquifers are unclear, but probably are related to: (1) subsurface pinchout of the aquifer, that influences flow patterns in the ground-water flow system, (2) hydraulic connection of the aquifer to river alluvium and associated large ground-water discharge, or (3) water-level declines as a result of pumpage. Estimated average ground-water discharge to the Savannah River based on data from 1941, 1942, and 1949, was 1,220 cubic feet per second; of which 46 percent was from the local flow system, 41 percent was from the intermediate flow system, and 13 percent was from the regional flow system. Estimated ground-water discharge during the 1954 and 1986 droughts indicates that the contribution from the intermediate flow system decreased in a downstream direction, and that most ground-water discharge was from the local flow system. The decreased contribution from the intermediate flow system in downstream reaches may be related to a downdip (downstream) thickening of confining units underlying the stream, that reduces upward leakage of water. Flow lines based on contours of the estimated predevelopment potentiometric surface of the confined Dublin and Midville aquifer systems, suggest ground-water flow beneath the floodplain of the Savannah River from one side of the river to the other side. This phenomenon, termed trans-river flow, is assumed to occur for a short distance into Georgia prior to discharge into the Savannah River alluvial valley. Trans-river flow is influenced by changes in hydraulic gradient (ground-water levels) near the river. Trans-river flow cannot be inferred from contours of the potentiometric surface of the Gordon or Upper Three Runs aquifers.