An electrical-resistivity survey was completed at the T tank farm at the Hanford nuclear site in Washington State, U.S.A. The purpose of the survey was to define the lateral extent of waste plumes in the vadose zone in and around the tank farm. The T tank farm consists of single-shell tanks that historically have leaked and many liquid-waste-disposal facilities that provide a good target for resistivity mapping. Given that the site is highly industrialized with near-surface metallic infrastructure that potentially could mask any interpretable waste plume, it was necessary to use the many wells around the site as long electrodes. To accommodate the long electrodes and to simulate the effects of a linear conductor, the resistivity inversion code was modified to assign low-resistivity values to the well’s location. The forward model within the resistivity code was benchmarked for accuracy against an analytic solution, and the inverse model was tested for its ability to recreate images of a hypothetical target. The results of the tank-farm field survey showed large, low-resistivity targets beneath the disposal areas that coincided with the conceptual hydrogeologic models developed regarding the releases. Additionally, in areas of minimal infrastructure, the long-electrode method matched the lateral footprint of a 3D surface-resistivity survey with reasonable fidelity. Based on these results, the long-electrode resistivity method may provide a new strategy for environmental characterization at highly industrialized sites, provided a sufficient number and density of wells exist.

No abstract

The distribution of subbottom geotechnical strength properties within the Panama Canal are needed to help with the Canal’s expansion. Core data already exist in the Canal, including lithological/stratigraphical descriptions and qualitative measurements of rock hardness. These data have been acquired within the Canal during previous expansion activities conducted over the past 60 years. Alone, the core data can be used to estimate rock hardness at unsampled locations using geostatistical methods. However, to help reduce uncertainty in the interpolation of rock hardness, a spatially continuous electrical resistivity survey was conducted to provide a better means of bridging information between cores. Although no direct causative link between rock hardness and resistivity exists, it was thought that the resistivity would be dependent upon jointly influencing parameters that comprise the geome‐chanical attributes of the rock, in this case porosity. For example, tuff generally had lower hardness and lower resistivity values compared to andesite and differences in porosity of these rock types would help explain the trend. When considering the resistivity in this geologic context, the spatial interpolation of rock hardness showed better agreement with measured data at sampled locations compared to methods that did not consider any geological context (including kriging of core data or a polynomial regression model between resistivity and rock hardness). Additionally, it is believed that full three‐dimensional inverse modelling of the resistivity data helped to correctly resolve the location of low‐resistivity features that could have been detected as off‐line effects in two‐dimensional processing algorithms. With these results, it is anticipated that the costs of dredging could be reduced by the simple fact that necessary resources can be anticipated for some of the harder rock types.

An electrical resistivity monitoring survey was conducted on a mine heap to track reagent movement during high-pressure injections. The injections were designed to increase the dissolution of metallic gold from low-grade ore and enhance recovery after surface leaching had ceased. The main objective of the geoelectrical monitoring was to observe the effectiveness of the injection technique and provide feedback to optimize injection parameters in real time. Real-time assessment was achieved by monitoring the raw output current and transfer resistance on a network of borehole electrodes installed around the injection well. It was demonstrated that the output current increased significantly on particular borehole electrodes after commencement of reagent injection, when the wetting front arrived at the electrodes. When injection ceased, the electrical current returned to the initial baseline current values. The timing and distribution of the electrodes demonstrating this behaviour varied with injection depth. The internal structure of the heap was likely a controlling factor in reagent movement. Resistance, converted to apparent resistivity, was also shown to change significantly in the region near the injection. Verification of the real-time assessment was conducted with post-injection time-lapse 3D tomographic inversion. While inverse modelling provides a truer 3D representation of reagent injection, the cost was shown to be a time-lag of 3.5 days to complete the modelling. The simplicity of monitoring the raw current output and voltage can make this a powerful tool for real-time tracking of fluid movement in the subsurface.

No abstract