Development of a new coupled model for the characterization and monitoring of dense non-aqueous phase liquids with DC resistivity and induced polarization

ABSTRACT: The combined use of DC resistivity and induced polarization (DCIP) has long-standing potential as a valuable, non-invasive tool for improved characterization and monitoring at sites contaminated with dense non-aqueous phase liquids (DNAPLs). Knowledge on the performance of standalone DC resistivity is plentiful with an established understanding of the DC signatures (electrical resistivity) at DNAPL sites. However, the mechanisms that generate the IP signatures (chargeability) are not fully understood, resulting in the limited application of DCIP at DNAPL sites. Due to the prohibition of DNAPL experiments at field sites, numerical modeling is the primary method for studying DNAPLs at the field-scale. The objective of this study was to develop a new coupled DNAPL-DCIP model that simulates realistic DNAPL spill and remediation scenarios at the field-scale and their monitoring by DCIP surveys. The 3D multiphase flow model (DNAPL3D-MT), which simulates DNAPL migration and remediation in heterogeneous porous media, was integrated with a 3D geoelectrical forward/ inverse model (IP4DI), which predicts the corresponding DCIP response. Central to the model development was establishing a comprehensive linkage between the key hydrogeological parameters of DNAPL3D-MT (permeability, porosity, clay/sand content, water/DNAPL/air content) and the key geoelectrical parameters of IP4DI (electrical resistivity, chargeability). A sensitivity analysis was performed on a single model node to assess the behavior of the DCIP response to different hydrogeological and geoelectrical conditions and demonstrated the consistency of the DCIP model with a number of experimental and field observations from the literature. A field-scale simulation of a DNAPL release and subsequent remediation being monitored by time-lapse DCIP surveys demonstrated the unique platform that the DCIP model can provide to better understand the relationship between complex DNAPL distributions and associated DCIP signatures. The model will help improve interpretation of DCIP data at the field-scale and assist in DCIP becoming more commonplace at DNAPL sites.