Steady-state simulations using a revised regional groundwater-flow model based on MODFLOW were run to assess the potential long-term effects on the Upper Floridan aquifer (UFA) of pumping the Lower Floridan aquifer (LFA) at well 36Q398, located at Barbour Pointe in coastal Georgia near Savannah. Simulated pumping of well 36Q398 at a rate of 750 gallons per minute (gal/min; or 1.08 million gallons per day [Mgal/d]) indicated a maximum drawdown of about 2.19 feet (ft) in the UFA directly above the pumped well and at least 1 ft of drawdown within a nearly 190-square-mile area (scenario A). Induced vertical leakage from the UFA provided about 98 percent of the water to the pumped well. Simulated pumping of well 36Q398 caused increased downward leakage in all layers above the LFA, decreased upward leakage in all layers above the LFA, increased inflow to and decreased outflow from lateral specified-head boundaries in the UFA and LFA, and an increase in the volume of induced inflow from the general-head boundary representing outcrop units. Water budgets for scenario A indicated that changes in inflows and outflows through general-head boundaries would compose about 45 percent of the simulated pumpage from well 36Q398, with the remaining 55 percent of the pumped water derived from flow across lateral specified-head boundaries.

Additional steady-state simulations were run to evaluate a pumping rate in the UFA of 240 gal/min (0.346 Mgal/d), which would produce an equivalent maximum drawdown in the UFA as pumping from well 36Q398 in the LFA at a rate of 750 gal/min (called the “drawdown offset”; scenario B). Simulated pumping in the UFA for the drawdown offset produced about 2.18 ft of drawdown, comparable to 2.19 ft of drawdown in the UFA simulated in scenario A. Water budgets for scenario B also provided favorable comparisons with scenario A, indicating that 42 percent of the drawdown-offset pumpage (0.346 Mgal/d) in the UFA originates as increased inflow and decreased outflow across general-head boundaries from overlying units in the surficial and Brunswick aquifer systems and that the remaining simulated pumpage originates as flow across general- and specified-head boundaries within the UFA and LFA.

The revised model was evaluated for sensitivity by first altering horizontal and vertical hydraulic conductivity in the Lower Floridan semiconfining unit and then adjusting horizontal and vertical hydraulic conductivity in the LFA to match the 35.6 ft of drawdown at pumping well 36Q398. These adjustments also affected the maximum simulated drawdown in the UFA and the equivalent offset pumping in the UFA that would produce the same amount of drawdown. The maximum drawdown in the UFA ranged from 1.82 to 2.57 ft and the equivalent offset pumping in the UFA ranged from 199 to 278 gal/min.

The revised model reasonably depicts changes in groundwater levels resulting from pumping the LFA at Barbour Pointe at a rate of 750 gal/min. Results are limited, however, by the same model assumptions and design as the original model, and placement of boundaries and type of boundary used exert the greatest control on overall groundwater flow and interaquifer leakage in the system. Simulation results have improved regional characterization of the Floridan aquifer system, which could be used by State officials in evaluating requests for groundwater withdrawal from the LFA.