Real-time monitoring of in situ chemical oxidation (ISCO) of dissolved TCE by integrating electrical resistivity tomography and reactive transport modeling

Successful in situ chemical oxidation (ISCO) applications require real-time monitoring to assess the oxidant delivery and treatment effectiveness, and to support rapid and cost-effective decision making. Existing monitoring methods often suffer from poor spatial coverage given a limited number of boreholes in most field conditions. The ionic nature of oxidants (e.g., permanganate) makes time-lapse electrical resistivity tomography (ERT) a potential monitoring tool for ISCO. However, time-lapse ERT is usually limited to qualitative analysis because it cannot distinguish between the electrical responses of the ionic oxidant and the ionic products from contaminant oxidation. This study proposed a real-time quantitative monitoring approach for ISCO by integrating time-lapse ERT and physics-based reactive transport models (RTM). Moving past common practice, where an electrical-conductivity anomaly in an ERT survey would be roughly linked to concentrations of anything ionic, we used PHT3D as our RTM to distinguish the contributions from the ionic oxidant and the ionic products and to quantify the spatio-temporal evolution of all chemical components. The proposed approach was evaluated through laboratory column experiments for trichloroethene (TCE) remediation. This ISCO experiment was monitored by both time-lapse ERT and downstream sampling. We found that changes in inverted bulk electrical conductivity, unsurprisingly, did not correlate well with the observed permanganate concentrations due to the ionic products. By integrating time-lapse ERT and RTM, the distribution of all chemical components was satisfactorily characterized and quantified. Measured concentration data from limited locations and the non-intrusive ERT data were found to be complementary for ISCO monitoring. The inverted bulk conductivity data were effective in capturing the spatial distribution of ionic species, while the concentration data provided information regarding dissolved TCE. Through incorporating multi-source data, the error of quantifying ISCO efficiency was kept at most 5 %, compared to errors that can reach up to 68 % when relying solely on concentration data.