Bench-scale experiments to evaluate electrical resistivity tomography as a monitoring tool for geologic CO2 sequestration

Abstract

Field-scale studies have shown electrical resistivity tomography (ERT) to be an effective tool for imaging resistivity anomalies and monitoring infiltration events in the near subsurface. ERT also shows potential for monitoring supercritical-CO2 injections, despite deployment challenges in the deep subsurface. We present results from analog bench-scale experiments aimed at evaluating the ability of ERT to quantify the volume and spatial distribution of a resistive fluid injected into a brine-saturated porous medium. Our experiments represent a well-controlled analog for supercritical-CO2 injection into deep brine aquifers. We injected measured volumes of air into translucent chambers filled with quartz sand, lined with electrodes, and saturated with a low resistivity salt solution. Between injections, a CCD camera captured high-resolution images, and an ERT data acquisition system scanned the chamber. Processing of the CCD images using quantitative visualization techniques resulted in high-resolution measurements of the spatial distribution and saturation of the injected gas. Direct comparison to inverted resistivity fields then provided a quantitative measure of the ability of ERT to estimate the total volume of injected gas and its spatial distribution within the chamber. We present results from two experiments designed to represent different injection scenarios: (A) low injection rate and strong capillary barrier and (B) high injection rate and weaker capillary barrier. Results show that ERT provided good estimates of the shape, size and location of the primary plume, but overestimated brine saturation within the plume and did not detect thin pathways of gas from the injection port or within the overlying capillary barrier. ERT measurements also indicated a change in saturation within the primary plume that corresponded with observed leakage through the capillary barrier in (B), demonstrating the potential utility of ERT as a leakage-monitoring tool. Repeated ERT scans during our experiments led to degradation in data quality that corresponded with an increase in measured contact resistance. Decreased data quality over time is clearly a concern for ERT implementation as a long-term monitoring strategy and deserves further study to quantify the responsible mechanisms.