ABSTRACT

Ground water in the vicinity of several industrial facilities in Upper Gwynedd Township and Lansdale Borough, Montgomery County, Pa., is contaminated with several volatile organic compounds (VOCs). The 2-square-mile area was placed on the National Priorities List as the North Penn Area 7 Superfund Site by the U.S. Environmental Protection Agency (USEPA) in 1989. The U.S. Geological Survey (USGS) conducted geophysical logging, aquifer testing, water-level monitoring, and streamflow measurements in the vicinity of North Penn Area 7 from October 2002 through December 2006. This followed work that began in 2000 to assist the USEPA in developing an understanding of the hydrogeologic framework in the area as part of the USEPA Remedial Investigation.

The study area is underlain by Triassic- and Jurassic-age sandstones, siltstones, and shales of the Lockatong Formation and the Brunswick Group. Regionally, these rocks strike northeast and dip to the northwest. The sequence of rocks form fractured-rock aquifers that act as a set of confined to semi-confined layered aquifers of differing permeabilities. The aquifers are recharged by precipitation and discharge to streams and wells. The Wissahickon Creek headwaters are less than 1 mile northeast of the study area. This stream flows southwest approximately parallel to strike and bisects North Penn Area 7. Ground water is pumped in the vicinity of North Penn Area 7 for industrial use and public supply.

The USGS collected geophysical logs for 42 wells that ranged in depth from 40 to 477 ft. Aquifer-interval-isolation testing was done in 17 of the 42 wells, for a total of 122 zones tested. A multiple-well aquifer test was conducted by monitoring the response of 14 wells to pumping and shutdown of a 600-ft deep production well in November-December 2004. In addition, water levels were monitored continuously in four wells in the area from October 2002 through September 2006, and streamflow was measured quarterly at two sites on Wissahickon Creek from December 2002 through September 2005.

Geophysical logging identified water-bearing zones associated with high-angle fractures and bedding-plane openings throughout the depth of the boreholes. Heatpulse-flowmeter measurements under non-pumping, ambient conditions in 16 wells greater than 200 ft in depth indicated that borehole flow, where detected, was only upward in 2 wells and only downward in 5 wells. In nine wells, both upward and downward flow were measured. Geologic structure and pumping in the area affect the spatial distribution of vertical gradients. Heatpulse-flowmeter measurements under pumping conditions were used to identify the most productive intervals in wells. Correlation of natural-gamma-ray logs indicated bedding in the area probably strikes about 45 to 65 degrees northeast and dips about 9 degrees northwest.

Aquifer intervals isolated by inflatable packers in 17 wells were pumped to test productivity of water-bearing zones and to collect samples to determine chemical quality of water produced from the interval. Interval-isolation testing confirmed the vertical hydraulic gradients indicated by heatpulse-flowmeter measurements. The specific capacities of the 122 isolated intervals ranged over about three orders of magnitude, from 0.01 to 10.6 gallons per minute per foot, corresponding to calculated transmissivities of 1.2 to 2,290 feet squared per day. Intervals adjacent to isolated pumped intervals commonly showed little response to pumping of the isolated zone. The presence of vertical hydraulic gradients and lack of adjacent-interval response to pumping in isolated intervals indicate a limited degree of vertical hydraulic connection between the aquifer sections tested. Differences were apparent in inorganic water quality of water from isolated intervals, including pH, specific conductance, and dissolved oxygen. Concentrations of most VOC contaminants in most wells with predominantly upward vertical gradients were greatest in well-water samples from the relatively shallow isolated intervals. Trichloroethylene was the most frequently detected chlorinated VOC and was measured in samples from 14 of 17 wells tested; the maximum concentration was 640 micrograms per liter. Other halogenated VOCs detected in about half or less of the 17 wells tested included cis-1,2-dichloroethylene, tetrachloroethylene, freon-113, and freon-11 at maximum concentrations of 120, 37, 71, and 55 micrograms per liter, respectively.

Results of the aquifer test with multiple observation wells showed that water levels in 8 of 14 wells declined in response to pumping. Estimates of aquifer transmissivity determined using Theis-curve fitting ranged from 683 to 2,382 feet squared per day. The spatial distribution of the eight wells that responded to pumping are along strike and in the down-dip direction of producing zones of the pumped well. This spatial distribution suggests that hydraulic connections in the aquifer are structurally controlled by dipping beds to some extent.

Water-level monitoring in four wells from October 2002 through September 2006 shows the seasonal rise and decline of levels. Water levels in two wells near Wissahickon Creek were evaluated in relation to streamflow on dates of quarterly streamflow measurements. The Wissahickon Creek was a losing stream between the two measurement sites, and ground-water levels were lower than the stream-channel bottom for most dates. Water levels measured in the 2-square mile area around and including North Penn Area 7 in December 2004, June 2005, and September 2005 confirm previous findings that the potentiometric surface is relatively flat in the immediate vicinity of North Penn Area 7 and generally is similar to topography except in areas affected by large amounts of ground-water withdrawal. Comparison of water levels measured in boreholes completed at different depths in well clusters shows predominantly upward vertical gradients east of Wissahickon Creek, except where affected by nearby pumping, and predominantly downward vertical gradients west of Wissahickon Creek.

Results of geophysical logging, aquifer tests, and water-level measurements are consistent with the conceptual model of a layered leaky aquifer where the dip of the beds has a strong control on hydraulic connections in the ground-water system. Connections within and (or) parallel to bedding tend to be greater than across bedding.

Revised December 30, 2010

First posted October 8, 2008