Tests of Uranium (VI) Adsorption Models in a Field Setting (NUREG/CR-6911)
On this page:
Download complete document
Manuscript Completed: August 2006
Date Published: August 2006
G. P. Curtis, J. A. Davis
Water Resources Division
U.S. Geological Survey
Menlo Park, CA 94025
A. L. Schwartzman, NRC Project Manager
Division of Fuel, Engineering, and Radiological Research
Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington, DC 20555-0001
Job Code Y6462
Field-based techniques were tested for determining Kd values or for validating surface complexation model-derived Kd values. The field site used for the study was a uranium mill tailings site at Naturita, Colorado. The techniques tested included: 1) the use of downhole in-situ devices that held a sample of a single mineral or sediment that were deployed in the aquifer contaminated with uranium, 2) in-situ push-pull tracer tests conducted within the contaminated aquifer, and 3) small-scale U(VI) migration tracer tests conducted within the aquifer. The experiments were conducted on scales that ranged from centimeters for the downhole devices to several meters for the small-scale tracer tests.
The tracer test results showed that U(VI) that has been present in the Naturita aquifer for several decades could be readily desorbed by perturbing groundwater chemical conditions. The experimental observations clearly demonstrated that in aquifers where U(VI) concentrations are controlled by adsorption, dissolved U(VI) concentrations can be rapidly impacted by increases or decreases in alkalinity. The experimental results in U(VI) migration tests were reasonably well described by a reactive transport model that simulated adsorption reactions using a semimechanistic surface complexation modeling approach. The model predicted both increases and decreases of U(VI) concentration reasonably well when the transport parameters were calibrated to observed Br transport. The results demonstrate the adequacy of the model for simulation of transient geochemical effects under the influence of a natural gradient. In contrast, a constant-Kd model would have predicted neither the observed increases nor decreases in U(VI) concentration in tests in which only the alkalinity of the groundwater was perturbed. This indicates that the uncertainty of predictions using the SCM approach were much smaller than those that would have been obtained if a constant-Kd approach had been used to model U(VI) adsorption. A detailed understanding of U(VI) surface and aqueous speciation is required to predict changes in U(VI) concentration in the Naturita aquifer, and likely in other aquifers contaminated with U(VI).
It was concluded that push-pull test results were not useful in the Naturita aquifer for evaluating adsorption model parameters. The limitation results from the nearly reversible nature of the reactive transport simulations when applied to cases where sorption processes are probed by manipulating groundwater alkalinity. Velocity and/or dispersive processes in the aquifer were sufficiently large that geochemical effects were dampened by dilution in most of the experiments. Push-pull tests may be useful for studying rates of sorption and desorption in aquifers.
The results with downhole in-situ devices for measuring Kd values suggest that the method needs further improvement. Furthermore, the experiments suggest that surface complexation models for single minerals need to be calibrated in solutions with high dissolved carbonate concentrations in order to be applied to aquifers that contain groundwaters in equilibrium with partial pressures of carbon dioxide gas greater than 0.01 atmospheres, such as the Naturita aquifer.