Time-lapse ERT monitoring of an injection/withdrawal experiment in a shallow unconfined aquifer

Abstract

To quantify performance of 3D time-lapse electrical resistivity tomography (ERT), a sequential injection/withdrawal experiment was designed for monitoring the pump-and-capture remediation of a conductive solute in an unconfined alluvial aquifer. Prior information is incorporated into the inversion procedure via regularization with respect to a reference model according to three protocols: (1) independent regularization involving a single reference model, (2) background regularization involving a reference model obtained via inversion of preinjection data, and (3) time-lapse regularization involving an evolving reference model obtained via inversion of data from previous experimental stages. Emplacement and sequential withdrawal of the solute is clearly imaged for all protocols. Time-lapse regularization results in greater amounts of model structure, while providing signifi-cant computational savings. ERT-estimated electrical conductivity is used to predict solute concentration and solute mass in the aquifer. At any experimental stage, we are able to estimate total solute mass in the aquifer with a maximum accuracy of 60%–85% depending on regularization protocol and survey geometry. We also estimate the withdrawn solute mass for every experimental stage (the change in mass between experimental stages). Withdrawn mass estimates are more reliable than total mass estimates and do not exhibit systematic underprediction or dependence on regularization protocol. Withdrawn mass estimates are accurate for changes in mass below [Formula: see text] of potassium bromide [Formula: see text] for horizontal and vertical dipole-dipole surveys, respectively. Estimating the withdrawn solute mass does not require background subtraction and, thus, does not require background data.