Assessing the effect of chlorinated hydrocarbon degradation in aquitards on plume persistence due to back-diffusion.

This modeling study aims to investigate how reactive processes in aquitards impact plume persistence in adjacent aquifers. For that purpose the migration of a trichloroethene (TCE) plume in an aquifer originating from dense non-aqueous phase liquid (DNAPL) source dissolution and back-diffusion from an underlying reactive aquitard was simulated in a 2D-numerical model. Two aquitard degradation scenarios were modeled considering one-step degradation from TCE to cis-dichloroethene (cDCE): a uniform (constant degradation with aquitard depth) and a non-uniform scenario (decreasing degradation with aquitard depth) and were compared with a no-degradation scenario. In the no-degradation scenario, a long-term TCE tailing above the Maximum Contaminant Level (MCL) caused by back-diffusion after source removal was observed. In contrast, in the aquitard degradation scenarios, TCE back-diffusion periods were shorter, whereby the extent of back-diffusion reduction depended on the aquitard degradation depth and the rate. For high degradation rates (half-life: 30-80days), an aquitard degradation depth greater than 65cm prevented TCE plume persistence after source removal but generated a long-term tailing above the MCL for the produced cDCE. For slow degradation rates (half-life: <200days), TCE was only partially degraded after source removal, independent of the aquitard degradation depth, leading to a long-term dual contamination of the aquifer by cDCE and TCE. A sudden enrichment of C in TCE and cDCE was observed after source removal in the uniform and non-uniform degradation scenarios that was distinct from δC patterns observed when aquifer degradation occurs (continuous enrichment of C along the plume axis) and for when there is absence of degradation (no change of isotope ratios). This demonstrates that δC measurements in the aquifer can be used as a diagnostic tool to demonstrate aquitard degradation, which simplifies the identification of reactive processes in aquitards, as aquifers are usually easier to monitor than aquitards.