Formatted Title
Electrical Resistivity Imaging of a DNAPL Site: What Happens When Results Don’t Equal Success?
Background/Objectives
A former Naval Ordnance Plant, now Superfund site, in Georgia had a partially delineated trichloroethylene (TCE) plume in a sandy surficial aquifer which had absorbed into the underlying clay confining unit. From the 1940s to 1960s, ordnance manufacturing and metal plating were conducted at the site. Currently the site operates as an industrial park. Investigations by the United States Army Corps of Engineers (USACE) in the 1990s found TCE, both dissolved phase and a dense non-aqueous phase liquid (DNAPL), prevalent in the groundwater which led to vapor intrusion issues which have since been mitigated, leaving groundwater issues. Although some site characterization has been performed via monitoring wells and soil borings, the main source area near a historic storm sewer outfall at the western edge of the site found DNAPL near the base of the previous depth of investigation, as well as in the center of the site found groundwater concentrations indicative of NAPL. However, the vertical and horizontal extent into and across the clay confining unit was unknown. Site geology consists of stacked sand and clay beds gentling sloping to the south-southeast. The clay confining unit is approximately 15 m below ground surface (bgs), underlying the entire site. Depth to water is approximately 11 m bgs. The area of interest is an open field downgradient of the historic storm sewer outfall, located above the expected highest contaminant concentrations with a former (now demolished) explosives blending plant and storm drains to the north and east. The objective was to perform electrical resistivity imaging (ERI) surveys, providing thousands of data points in 2D space, and cross reference the results with previous high resolution lithological and chemical analysis of soil cores and groundwater samples which provide data points, few and far between, in only 1D space. The ERI surveys were verified with confirmatory drilling, lithologic logging, sample collection and analytical analysis. The goal of the effort was to expand the footprint of the previous investigations, and demonstrate that a relatively inexpensive, fast, high resolution, and non-invasive alternative geophysical field technique could delineate both the vertical and horizontal extent of the contaminant source area.
Approach/Activities
Eleven (11) ERI transects were performed with lengths of either 126 m or 157 m, allowing for vertical penetration of 25 to 31 m bgs, respectively. Some surveys were co-located with previous high resolution membrane interface probe/hydraulic profiling tool (MiHPT) data and where groundwater concentrations were indicative of NAPL – all previous data were provided by EPA Region 4 project lead and were not collected by this researcher. Surveys were planned to cover the known areas of high concentration to produce a model response for further comparison. Existing contaminant concentration data and lithology information were used where able for correlation and interpretation. Confirmatory drilling, core descriptions, photoionization detector (PID), and soil and groundwater sampling were performed to evaluate electrical anomalies.
Results/Lessons Learned
ERI results did not positively identify the expected DNAPL previously found. Three (3) confirmatory drilling targets were identified from the results, all in previously uninvestigated parts of the site to the east and northeast within the footprint of the explosives blending plant. Electrical resistivity values of the drilling targets were anomalous respective to background (whether sand or clay) and either >10,000 ohm-m or <10 ohm-m. The lithologic contact between the overlying sand and confining clay unit was identified by ERI. Depth to groundwater was not distinguished in the resistivity dataset, but depth was confirmed during confirmatory drilling, consistent with depths identified in previous data. No PID response was recorded in any of the locations drilled, but TCE was detected at depth in soil samples, of approximately 20 mg/kg, and in groundwater samples at about 5 µg/L. This presentation provides a true ‘lessons learned’ experience of when the results don’t confirm the known conceptual site model, advises data users on setting proper expectations, and data quality.