In the Albany area of southwestern Georgia, the Upper Floridan aquifer lies entirely within the Dougherty Plain district of the Coastal Plain physiographic province, and consists of the Ocala Limestone of late Eocene age. The aquifer is divided throughout most of the study area into an upper and a lower lithologic unit, which creates an upper and a lower water-bearing zone. The lower water-bearing zone consists of alternating layers of sandy limestone and medium-brown, recrystallized dolomitic limestone, and ranges in thickness from about 50 to 100 feet. It is highly fractured, and exhibits well-developed permeability by solution features that are responsible for transmitting most of the ground water in the aquifer. Transmissivity of the lower water-bearing zone ranges from about 90,000 to 178,000 feet squared per day. The upper water-bearing zone is a finely crystallized-to-oolitic, locally dolomitic limestone having an average thickness of about 60 feet. Transmissivities in the upper water-bearing zone are considerably less than those in the lower water-bearing zone. The Upper Floridan aquifer is overlain by about 20 to 120 feet of undifferentiated overburden consisting of fine-to-coarse quartz sand and noncalcareous clay. A clay zone about 10 to 30 feet thick may be continuous throughout the southwestern part of the Albany area, and where present, causes confinement of the Upper Floridan aquifer and creates perched ground water after periods of heavy rainfall. The Upper Floridan aquifer is confined below by the Lisbon Formation, a mostly dolomitic limestone that contains trace amounts of glauconite. The Lisbon Formation is at least 50 feet thick in the study area, and acts as an impermeable base to the Upper Floridan aquifer. The quality of ground-water in the Upper Floridan aquifer is suitable for most uses; wells generally yield water of the hard, calcium-bicarbonate type that generally meets the U.S. Environmental Protection Agency's Primary or Secondary Drinking Water Regulations.

The water-resource potential of the Upper Floridan aquifer was evaluated by compiling results of test drilling and aquifer testing in the study area, and by conducting computer simulations of the ground-water-flow system under the seasonal-low conditions of November 1985, and under conditions of pumping within a 12square-mile area located southwest of Albany. Results of test drilling, aquifer testing, and water-quality analyses indicate that, in the area southwest of Albany, geohydrologic conditions in the Upper Floridan aquifer, undifferentiated overburden, and Lisbon Formation were favorable for the aquifer to provide a large quantity of water without having adverse effects on the ground-water system. The confinement of the Upper Floridan aquifer by the undifferentiated overburden and the rural setting of the area of potential development decreases the likelihood that chemical constituents will enter the aquifer during development of the ground-water resources.

Computer simulations of ground-water flow in the Upper Floridan aquifer, incorporating conditions for regional flow across model boundaries, leakage from rivers and other surface-water features, and vertical leakage from the undifferentiated overburden, were conducted by using a finite-element model for groundwater flow in two dimensions. Comparison of computed and measured water levels in the Upper Floridan aquifer for November 1985 at 74 locations indicated that computed water levels generally were within 5 feet of the measured values, which is the accuracy to which measured water levels were known. Water-level altitudes ranged from about 260 feet to 130 feet above sea level in the study area during calibration. Aquifer discharge to the Flint River downstream from the Lake Worth dam was computed by the calibrated model to be about 1 billion gallons per day; about 300 million gallons per day greater than was measured for similar low-flow conditions. The excess computed discharge was attributed partially to stream withdrawals for industrial use, non-reported use, and channel evaporation, but mostly to increased gradients and increased flow from the aquifer to the river than existed during calibration.

Results from the calibrated finite-element model indicate that ground-water flow is dominated by inflow from regional-flow components to the west, north, and east of the study area, and by outflow to the Flint River downstream from the Lake Worth dam. Simulation results indicated that directions of ground-water flow were not changed appreciably by pumping at the November 1985 rates. However, vertical leakage from the undifferentiated overburden caused local deviations in the regional flow pattern.

A sensitivity analysis that was performed on 18 hydrologic factors affecting the flow system in the Upper Floridan aquifer showed that computed water levels changed the most (were the most sensitive) in response to changes in hydraulic conductivity of the aquifer, vertical leakage coefficient and water level in the undifferentiated overburden, and stage of the Flint River downstream from the Lake Worth dam. Computed water levels were least sensitive to changes in well pumpage, flow across the northern boundary and from Lake Worth, the boundary coefficient for the Flint River downstream from the Lake Worth dam, and flow from Cooleewahee Creek.

Simulations of six pumping scenarios in the area of potential development southwest of Albany showed that the Upper Floridan aquifer is capable of providing at least 72 million gallons per day from five locations (14.4 million gallons per day each) within this area without causing adverse affects on the flow system. The 72million-gallon-per-day scenario yielded a maximum drawdown of about 9.4 feet, which placed the water level in the Upper Floridan aquifer about 50 feet above the top of the lower water-bearing zone. Hence, the likelihood of aquifer dewatering, well interference, or sinkhole development from pumping as much as 72 million gallons per day from within the area of potential development is small. All pumping scenarios showed that about 81 percent of the ground-water pumpage was derived from regional flow that would have discharged to the Flint River downstream from the Lake Worth dam. The dominant ground-water-flow direction toward the Flint River was not changed and no induced recharge from the Flint River entered the potential-development area. Induced recharge from the undifferentiated overburden contributed to about 1.5 percent of the total volume pumped during the simulations.