Comparison of Chemical and Physical Non-Equilibrium Models to Simulate Multi-Scale Kinetic Mass-Transfer and Surface Complexation of U(VI) in Porous Media

Monday, April 12, 2010: 2:50 p.m.
Continental A (Westin Tabor Center, Denver)
Janek Greskowiak , Land and Water, CSIRO, Wembley, Western Australia 6913, Australia
Michael B. Hay , USGS
Henning Prommer, Ph.D. , Land and Water, CSIRO, Wembley, Western Australia 6913, Australia
Chongxuan Liu , Pacific Northwest National Laboratory, Richland, WA
Vincent Post, Ph.D. , Faculty of Earth and Life Sciences, VU University, Amsterdam, Netherlands
James A. Davis , USGS
The mobility of U(VI) in aerobic uranium contaminated groundwater is generally controlled by adsorption and is regulated by solution chemistry and  surface properties of the prevailing mineral phases. Thereby diffusional mass-transfer between water mobile and immobile domains has been found to play an important role for the overall mobility of U(VI) where groundwater velocities are high, e.g., as in the U(VI) contaminated aquifer of the Handford 300A site, Washington, USA. In the present study, two alternative approaches that both simulate coupled diffusional mass transfer and (non-linear) surface complexation processes were analysed and compared: (i) The chemical non-equilibrium approach, which approximates the bulk process as distributed rate surface complexation kinetics, and (ii) the physical non-equilibrium approach, in which the bulk kinetic processes result from a combination of diffusional mass-transfer and instantaneous surface complexation reactions in the immobile domains. The comparison was carried out for column and field scale scenario simulations that both considered varying solution chemistries that strongly impact the U(VI) desorption behaviour of the U(VI) contaminated model sediment. The field scale simulations reflect the highly dynamic character of groundwater flow and solution chemistries in the Handford 300A aquifer, driven by the rapidly changing water levels of the nearby Columbia River. The results of the column scale simulations illustrate that the two approaches differ depending on (i) the magnitude of variation in the apparent U(V) linear sorption coefficient Kd and (ii) the degree of sorption disequilibrium. In contrast, when the apparent Kd values were comparable both modelling approaches gave essentially identical results, unaffected by the degree of sorption disequilibrium. The field scale simulations indicated that the two approaches do not result in significantly differing U(VI) concentrations and U(VI) mass fluxes, despite high groundwater velocities and distinctly different water compositions of the ambient groundwater and the intruding and receding river water.
See more of: Reactive Transport Modeling
See more of: Topical Sessions