Steady State DNAPL Dissolution in Three-Dimensional Fractured Sandstone Network Experiments

Abstract

The distribution of residual dense nonaqueous phase liquid (DNAPL) in the subsurface plays a critical role in the DNAPL dissolution kinetics. However, measuring residual DNAPL at the field scale in fractured bedrock settings is generally impractical. This research uses a three-dimensional (3D), bench-scale, fractured-rock network comprised of low-porosity sandstone to evaluate the dissolution kinetics of tetrachloroethylene (PCE) DNAPL at residual saturation during ambient groundwater conditions. To our knowledge, this work presents the first experiments to investigate DNAPL dissolution in 3D bench-scale fractured systems. DNAPL dissolution in the relatively uniform fracture network was evaluated and described using an effective parameter, the bulk mass transfer coefficient (KL). Results from dissolution experiments revealed a positive, statistically significant correlation between KL and DNAPL-water interfacial area, and between KL and DNAPL saturation, analogous to porous media experiments. While aperture size and uniformity influenced DNAPL trapping and interfacial area in the fracture network experiments, DNAPL-altered aperture size did not have a statistically significant influence on dissolution rates. The mass transfer behavior for our 3D fracture network was more similar to one-dimensional (1D) porous media experiments than to two-dimensional (2D) and 3D porous media experiments, which primarily used nonaqueous phase liquid (NAPL) pools. This result is likely due to a combination of the coexistence of the nonwetting NAPL phase and the flowing-aqueous phase in the larger fractures or at fracture intersections, as well as better mixing due to dynamic flow pathways in the network. Finally, residual NAPL saturations could not be replicated in the networks, possibly due to the unpredictable flow-switching behavior of multiphase fluids in fractured rock.