ABSTRACT

Test drilling, field investigations, and digital modeling were completed at Fort Stewart, GA, during 2009–2010, to assess the geologic, hydraulic, and water-quality characteristics of the Floridan aquifer system and evaluate the effect of Lower Floridan aquifer (LFA) pumping on the Upper Floridan aquifer (UFA). This work was performed pursuant to the Georgia Environmental Protection Division interim permitting strategy for new wells completed in the LFA that requires simulation to (1) quantify pumping-induced aquifer leakage from the UFA to LFA, and (2) identify the equivalent rate of UFA pumping that would produce the same maximum drawdown in the UFA that anticipated pumping from LFA well would induce. Field investigation activities included (1) constructing a 1,300-foot (ft) test boring and well completed in the LFA (well 33P028), (2) constructing an observation well in the UFA (well 33P029), (3) collecting drill cuttings and borehole geophysical logs, (4) collecting core samples for analysis of vertical hydraulic conductivity and porosity, (5) conducting flowmeter and packer tests in the open borehole within the UFA and LFA, (6) collecting depth-integrated water samples to assess basic ionic chemistry of various water-bearing zones, and (7) conducting aquifer tests in new LFA and UFA wells to determine hydraulic properties and assess interaquifer leakage. Using data collected at the site and in nearby areas, model simulation was used to assess the effects of LFA pumping on the UFA.

Borehole-geophysical and flowmeter data indicate the LFA at Fort Stewart consists of limestone and dolomitic limestone between depths of 912 and 1,250 ft. Flowmeter data indicate the presence of three permeable zones at depth intervals of 912–947, 1,090–1,139, and 1,211–1,250 ft. LFA well 33P028 received 50 percent of the pumped volume from the uppermost permeable zone, and about 18 and 32 percent of the pumped volume from the middle and lowest permeable zones, respectively. Chemical constituent concentrations increased with depth, and water from all permeable zones contained sulfate at concentrations that exceeded the U.S. Environmental Protection Agency secondary maximum contaminant level of 250 milligrams per liter.

A 72-hour aquifer test pumped LFA well 33P028 at 740 gallons per minute (gal/min), producing about 39 ft of drawdown in the pumped well and about 0.4 foot in nearby UFA well 33P029. Simulation using the U.S. Geological Survey finite-difference code MODFLOW was used to determine long-term, steady-state flow in the Floridan aquifer system, assuming the LFA well was pumped continuously at a rate of 740 gal/min. Simulated steady-state drawdown in the LFA was identical to that observed in pumped LFA well 33P028 at the end of the 72-hour test, with values larger than 1 ft extending 4.4 square miles symmetrically around the pumped well. Simulated steady-state drawdown in the UFA resulting from pumping in LFA well 33P028 exceeded 1 ft within a 1.4-square-mile circular area, and maximum drawdown in the UFA was 1.1 ft. Leakage from the UFA through the Lower Floridan confining unit contributed about 98 percent of the water to the well; lateral flow from specified-head model boundaries contributed about 2 percent. About 80 percent of the water supplied to LFA well 33P028 originated from within 1 mile of the well, and 49 percent was derived from within 0.5 mile of the well. Vertical hydraulic gradients and vertical leakage are progressively higher near the LFA pumped well which results in a correspondingly higher contribution of water from the UFA to the pumped well at distances closer to the pumped well.

Simulated pumping-induced interaquifer leakage from the UFA to the LFA totaled 725 gal/min (1.04 million gallons per day), whereas simulated pumping at 205 gal/min (0.3 million gallons per day) from UFA well 33P029 produced the equivalent maximum drawdown as pumping LFA well 33P028 at 740 gal/min during the aquifer test. This equivalent pumping rate in the UFA underpredicts the area affected by vertical leakage resulting from pumping the new LFA well, and therefore underpredicts the pumping offset (reduction) required in the UFA to produce no net hydrologic effect on the UFA from pumping at the new LFA well. Two simulations that decreased pumping rates in existing UFA wells at Fort Stewart by a total of 205 and 370 gal/min while simultaneously pumping LFA well 33P028 at 740 gal/min reduced the magnitude and extent of drawdown in the UFA when compared with a simulation of pumping from well 33P028 without adjusting withdrawals at UFA wells.

First posted May 6, 2011