Hydrogeology of, and simulation of ground-water flow In, the Pohatcong Valley, Warren County, New Jersey
A numerical ground-water-flow model was constructed to simulate ground-water flow in the Pohatcong Valley, including the area within the U.S. Environmental Protection Agency Pohatcong Valley Ground Water Contamination Site. The area is underlain by glacial till, alluvial sediments, and weathered and competent carbonate bedrock. The northwestern and southeastern valley boundaries are regional-scale thrust faults and ridges underlain by crystalline rocks. The unconsolidated sediments and weathered bedrock form a minor surficial aquifer and the carbonate rocks form a highly transmissive fractured-rock aquifer. Ground-water flow in the carbonate rocks is primarily downvalley towards the Delaware River, but the water discharges through the surficial aquifer to Pohatcong Creek under typical conditions. The hydraulic characteristics of the carbonate-rock aquifer are highly heterogeneous. Horizontal hydraulic conductivities span nearly five orders of magnitude, from 0.5 feet per day (ft/d) to 1,800 ft/d. The maximum transmissivity calculated is 37,000 feet squared per day. The horizontal hydraulic conductivities calculated from aquifer tests using public supply wells open to the Leithsville Formation and Allentown Dolomite are 34 ft/d (effective hydraulic conductivity) and 85 to 190 ft/d (minimum and maximum hydraulic conductivity, respectively, yielding a horizontal anisotropy ratio of 0.46). Stream base-flow data were used to estimate the net gain (or loss) for selected reaches on Brass Castle Creek, Shabbecong Creek, three smaller tributaries to Pohatcong Creek, and for five reaches on Pohatcong Creek. Estimated mean annual base flows for Brass Castle Creek, Pohatcong Creek at New Village, and Pohatcong Creek at Carpentersville (from correlations of partial- and continuous-record stations) are 2.4, 25, and 45 cubic feet per second (ft3/s) (10, 10, and 11 inches per year (in/yr)), respectively. Ground-water ages estimated using sulfur hexafluoride (SF6), chlorofluorocarbon (CFC), and tritium-helium age-dating techniques range from 0 to 27 years, with a median age of 6 years. Land-surface and ground-water water budgets were calculated, yielding an estimated rate of direct recharge tothe surficial aquifer of about 23 in/yr, and an estimated net recharge to the ground-water system within the area underlain by carbonate rock (11.4 mi2) of 29 in/yr (10 in/yr over the entire 33.3 mi2 basin). A finite-difference, numerical model was developed to simulate ground-water flow in the Pohatcong Valley. The four-layer model encompasses the entire carbonate-rock part of the valley. The carbonate-rock aquifer was modeled as horizontally anisotropic, with the direction of maximum transmissivity aligned with the longitudinal axis of the valley. All lateral boundaries are no-flow boundaries. Recharge was applied uniformly to the topmost active layer with additional recharge added near the lateral boundaries to represent infiltration of runoff from adjacent crystalline-rock areas. The model was calibrated to June 2001 water levels in wells completed in the carbonate-rock aquifer, August 2000 stream base-flow measurements, and the approximate ground-water age. The ground-water-flow model was constructed in part to test possible site contamination remediation alternatives. Four previously determined ground-water remediation alternatives (GW1, GW2, GW3, and GW4) were simulated. For GW1, the no-action alternative, simulated pathlines originating in the tetrachloroethene (PCE) and trichloroethene (TCE) source areas within the Ground-Water Contamination Site end at Pohatcong Creek near the confluence with Shabbecong Creek, although some particles went deeper in the aquifer system and ultimately discharge to Pohatcong Creek about 10 miles downvalley in Pohatcong Township. Remediation alternatives GW2, GW3, and GW4 include ground-water withdrawal, treatment, and reinjection. The design for GW2 includes wells in the TCE and PCE source areas that wit