A chromate-contaminated site in southern Switzerland – Part 2: Reactive transport modeling to optimize remediation options

Abstract

► Development of a 2D reactive transport model of a Cr(VI) contaminated site. ► Model incorporation of site mineralogy, Cr isotopes, water chemistry and site hydrology. ► Simulation of measured present day Cr(VI) concentrations and δ 53 Cr values. ► Prediction of Cr(VI) plumes for alternate remediation strategies. ► Prediction of remaining time with ongoing subsoil Cr(VI) contamination. A 2D horizontal reactive transport model of a chromate-contaminated site near Rivera, Switzerland, was developed using the computer code CrunchFlow to evaluate site remediation strategies. Transport processes were defined according to the results of an existing hydrological model, and the definition of geochemical (reactive) processes is based on the results of a detailed mineralogical and geochemical site characterization leading to a comprehensive conceptual site model. Kinetics of naturally occurring Cr(VI) reduction by Fe(II) and natural solid organic matter is quantified by fitting measured Cr isotope ratios to a modeled 1D section along the best constrained flow line. The simulation of Cr isotope fractionation was also incorporated into the 2D model. Simulation of the measured present day Cr(VI) plume and δ 53 Cr value distribution was used for the 2D model calibration and corresponds to a situation where only monitored natural attenuation (MNA) is occurring. Other 2D model runs simulate alternate excavation scenarios. The simulations show that with an excavation of the top 2–4 m the groundwater Cr(VI) plume can be minimized, and that a deeper excavation depth only diminishes the plume if all the contaminants can be removed. A combination of an excavation of the top 2–4 m and monitoring of the ongoing natural Cr(VI) reduction is suggested as the most ecological and economical remediation strategy, even though a remaining time period with ongoing subsoil Cr(VI) contamination in the order of 1 ka is predicted.