The hydrologic and geochemical effects of aquifer storage recovery were evaluated to determine the potential for supplying the city of Charleston, South Carolina, with large quantities of potable water during emergencies, such as earthquakes, hurricanes, or hard freezes. An aquifer storage recovery system, including a production well and three observation wells, was installed at a site located on the Charleston peninsula. The focus of this study was the 23.2-meter thick Tertiary-age carbonate and sand aquifer of the Santee Limestone and the Black Mingo Group, the northernmost equivalent of the Floridan aquifer system.
Four cycles of injection, storage, and recovery were conducted between October 1999 and February 2002. Each cycle consisted of injecting between 6.90 and 7.19 million liters of water for storage periods of 1, 3, or 6 months. The volume of recovered water that did not exceed the U.S. Environmental Protection Agency secondary standard for chloride (250 milligrams per liter) varied from 1.48 to 2.46 million liters, which is equivalent to 21 and 34 percent of the total volume injected for the individual tests. Aquifer storage recovery testing occurred within two productive zones of the brackish Santee Limestone/Black Mingo aquifer. The individual productive zones were determined to be approximately 2 to 4 meters thick, based on borehole geophysical logs, electromagnetic flow-meter testing, and specific-conductance profiles collected within the observation wells. A transmissivity and storage coefficient of 37 meters squared per day and 3 x 10-5, respectively, were determined for the Santee Limestone/Black Mingo aquifer.
Water-quality and sediment samples collected during this investigation documented baseline aquifer and injected water quality, aquifer matrix composition, and changes in injected/aquifer water quality during injection, storage, and recovery. A total of 193 water-quality samples were collected and analyzed for physical properties, major and minor ions, and nutrients. The aquifer and treated surface water were sodiumchloride and calcium/sodium-bicarbonate water types, respectively. Forty-five samples were collected and analyzed for total trihalomethane. Total trihalomethane data collected during aquifer storage recovery cycle 4 indicated that this constituent would not restrict the use of recovered water for drinking-water purposes. Analysis of six sediment samples collected from a cored well located near the aquifer storage recovery site showed that quartz and calcite were the dominant minerals in the Santee Limestone/Black Mingo aquifer. Estimated cation exchange capacity ranged from 12 to 36 milliequivalents per 100 grams in the lower section of the aquifer.
A reactive transport model was developed that included two 2-meter thick layers to describe each of the production zones. The four layers composing the production zones were assigned porosities ranging from 0.1 to 0.3 and hydraulic conductivities ranging from 1 to 8.4 meters per day. Specific storage of the aquifer and confining units was estimated to be 1.5 x 10-5 meter-1. Longitudinal dispersivity of all layers was specified to be 0.5 meter. Leakage through the confining unit was estimated to be minimal and, therefore, not used in the reactive transport modeling.
Inverse geochemical modeling indicates that mixing, cation exchange, and calcite dissolution are the dominant reactions that occur during aquifer storage recovery testing in the Santee Limestone/Black Mingo aquifer. Potable water injected into the Santee Limestone/Black Mingo aquifer evolved chemically by mixing with brackish background water and reaction with calcite and cation exchangers in the sediment. Reactive-transport model simulations indicated that the calcite and exchange reactions could be treated as equilibrium processes.
Simulations with the calibrated reactive transport model indicated that approximately one-fourth of the total volume of water injected into