Multiscale Studies of U(VI) Desorption From Contaminated Sediments
Abstract
The migration of U(VI) in groundwater is controlled by both the adsorption and desorption from mineral surfaces as well as the heterogeneous distribution of reactive surfaces and hydraulic conductivity in an aquifer. These complex processes were investigated in both laboratory studies using sediments collected from a uranium contaminated aquifer located near Naturita, CO and in tracer test studies conducted at this field site. Batch experimental results described in a companion study indicate that at microscopic scales, U(VI) desorption is limited by slow mass transfer processes that occur on the time scale of approximately a month. These slow desorption rates decreased with increasing particle size suggesting a diffusive mechanism that was simulated using a multi-rate mass transfer (MRMT) model. Reactive transport simulations using the batch-derived MRMT model were applied to both laboratory column experiments and to small-scale tracer tests. The laboratory experiments used contaminated sediments that were collected at the field site and were leached with a U(VI)-free synthetic groundwater that had partial pressures of CO2 equal to 0.0003 or 0.02 atm. Both the observed tailing and the rebound of U(VI) concentrations during flow interruption events were reproduced by the model simulations. U(VI) desorption was also investigated in two small-scale tracer tests that were conducted at the field site. Desorption of U(VI) was studied by injecting uncontaminated groundwater with a NaBr tracer into the aquifer, using both forced- and natural-gradient experiments to vary rates of groundwater flow. The low concentration of U(VI) in the injected water caused both an initial decrease followed by a gradual rebound in U(VI) concentrations at several downgradient multilevel observation. The decrease and rebound in U(VI) concentrations were affected by numerous chemical and physical factors, including fluid mixing, desorption of U(VI) and experimentally-induced temporal changes in the chemical concentrations of U(VI), alkalinity, Ca and Na. Successful application of a one-dimensional MRMT model to the observed Br concentrations required that an additional immobile zone be added to the laboratory scale MRMT model to account for the heterogeneity in the field. In cases where the Br observations could be matched by the revised MRMT model, the U(VI) concentrations in the tracer tests could be predicted from the updated nonreactive transport model and batch geochemical model. However, Br breakthrough curves at several locations were highly nonideal and could not be reproduced by a one dimensional model. Two dimensional simulations are being conducted in these cases to evaluate the applicability of the MRMT model to describe U(VI) desorption under highly heterogeneous conditions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.H31B0786C
- Keywords:
-
- 1009 GEOCHEMISTRY / Geochemical modeling;
- 1832 HYDROLOGY / Groundwater transport;
- 1847 HYDROLOGY / Modeling