SITE APPLICATIONS OF GENERALIZED COMPOSITE SURFACE COMPLEXATION MODELS FOR SIMULATING URANIUM TRANSPORT IN GROUNDWATER

Sorption of uranium in groundwater to solid-phase material is often simulated using single-component sorption, such as iron oxyhydroxides or organic carbon or clays, based on site knowledge. However, most geologic material has the potential for uranium sorption on a variety of materials. Batch laboratory studies were done on oxidized material from the proposed Dewey Burdock uranium in-situ recovery site in South Dakota. A synthetic groundwater with geochemistry similar to that of the local groundwater was spiked with uranium and brought into contact with solid-phase samples. Sorption on just iron oxyhydroxides does not fully explain the loss of uranium from the water to the solid phase. The batch test data were used to develop a generalized composite (GC) surface complexation model using PHREEQC (geochemical model) coupled with PEST (automated calibration program) following procedures that were first developed at a uranium mill tailings site in Naturita, Colorado. For the Dewey Burdock site, the GC surface complexation model provides calibration data for 1D reactive transport simulations of future uranium concentrations after site restoration.

GC surface complexation and reactive transport will be modeled for a uranium mill tailings site in Riverton, Wyoming, based on site conditions and batch studies with site core. Compared to a straight K_d sorption approach, a GC surface complexation model coupled with reactive transport allows for geochemical changes along the groundwater flow path. This more advanced technique will be compared to a K_d sorption and transport model developed for the site in 1998. The predictive capacities of the two modeling approaches will be compared to available groundwater data (1985 to 2013), which includes a flood event in 2010 that increased uranium concentration in several wells. The goal at the Riverton site is to use the most recent modeling techniques to better understand and simulate how flooding conditions changed the groundwater geochemistry. This modeling will be used to update the predictions of natural flushing rates and help guide site decision making.