The U.S. Geological Survey conducted borehole geophysical logging, collected and analyzed water-level data, and sampled sections of a rock core to determine the concentration of volatile organic compounds in the aquifer matrix of the Stockton Formation. Borehole geophysical logs were run in three monitor wells. At well 05MW04I, the vertical gradient was upward at depths above 42 feet below land surface (ft bls), downward between 42 and 82 ft bls, and upward below 82 ft bls. At well 05MW05I, a downward vertical gradient was present. At well 05MW12I, the vertical gradient was downward above 112 ft bls and upward below 112 ft bls.

Three water-bearing fractures in a 17-foot long rock core from 23.5 to 40.5 ft bls were identified and sampled. Three samples were analyzed from each water-bearing fracture—at the fracture face, 2 centimeters (cm) below the fracture, and 4 cm below the fracture. Fifteen compounds were detected; however, concentrations of seven compounds were less than 1 microgram per kilogram (mg/kg) when detected. Concentrations of benzene (from 0.39 to 3.3 mg/kg), 1,1-dichloroethene (1,1-DCE) (from 0.15 to 13 mg/kg), 1,1,1-trichloroethane (TCA) (from 0.17 to 22 mg/kg), and trichloroethylene (TCE) (from 0.092 to 9.6 mg/kg) were detected in all samples. The highest concentrations detected were for toluene, which was detected at a concentration of 32 and 86 mg/kg in the samples from unweathered sandstone at 2 and 4 cm below the fracture, respectively. Concentrations generally decreased with distance below the fracture in the mudstone samples. Concentrations of benzene and toluene increased with distance below the fractures in the unweathered sandstone samples. Concentrations of 1,1-DCE, TCA, and TCE were higher in the mudstone samples than in the samples from sandstone. Toluene concentrations were higher in unweathered sandstone than in weathered sandstone or mudstone.

The effect of the pumping of Horsham Water and Sewer Authority public supply well 26 (HWSA-26), 0.2 mile southwest of the base boundary, on groundwater levels on the base was determined by shutting the well down for 6 days to allow water levels to recover. Water levels in 22 nearby wells were measured. The only well (02MW01I) that showed an unambiguous response to the shutdown of well HWSA-26 is 1,350 feet directly along strike from well HWSA-26. The recovery of well 05MW11I in response to the shutdown of well HWSA-26 is masked by recharge from snowmelt but probably does not exceed about 0.2 feet on the basis of the water level in well 05MW11I, which showed a response to the pumping of well HWSA-26 that ranged from 0.5 to 0.15 foot.

Horizontal gradients differ with depth, and the rate and direction of ground-water flow and contaminant movement is depth dependent. The potentiometric-surface map for water levels measured in wells screened between 5 and 44 ft bls in the aquifer shows a ground-water mound that is the high point on a regional ground-water divide. From this divide, ground water flows both northwest toward Park Creek and southeast toward Pennypack Creek. The hydraulic gradient around this mound is relatively flat to the southeast and particularly flat to the northwest. The potentiometric-surface map for water levels measured in wells screened between 40 and 100 ft bls in the aquifer shows a very flat hydraulic gradient. Differences in the elevation of the potentiometric surface are less than 2 feet. The potentiometric-surface map for water levels measured in wells screened between 105 and 179 ft bls in the aquifer shows a steep hydraulic gradient between Sites 5 and 2 and a relatively flat hydraulic gradient between Sites 5 and 3. Water levels measured on October 7, 1999, showed downward vertical head gradients for all well clusters at Site 5. Vertical gradients ranged from 0.01 at well cluster 05MW10 to 0.2 at cluster 05MW11. Most gradients were between 0.01 and 0.026. Vertical head gradients vary with time. The variability is caused by a difference in the magnitude of water-level fluctuations between shallow and the deep fractures. The difference in the magnitude of water-level fluctuations is because of differences in lithology and aquifer storativity.