As discussed previously, the Rifle UMTRA site in Colorado is an old uranium processing facility which suffers from various contamination issues, including uranium. Uranium is predominantly found in the mobile U(VI) form in the subsurface due to an insufficient supply of electron donors to stimulate anaerobic respiration and/or consume dissolved oxygen. Laboratory studies have demonstrated the potential of microbes to reduce U(VI) to immobile U(IV) in an aquifer system,110 and the in situ treatment of U(vi) using the same method was tested at the Old Rifle site. Contaminated soil has been removed from the site, leaving only groundwater contamination within the local aquifer. Concentrations of uranium in this area range from 0.4 to 1.4 mM, above the maximum UMTRA contamination limit of 0.18 mM.28

The method used in this field-scale test was describe in detail by Anderson et al.,28 and is summarised briefly here. Injection wells were installed in two rows of ten, perpendicular to groundwater flow (which is typically towards the Colorado River). Each well contained three injection points positioned at different depths in the subsurface. A storage tank was filled periodically with native groundwater and was amended with sodium acetate as an electron donor to stimulate uranium-reducing bacteria and potassium bromide as a conservative tracer at concentrations of 100 and 10 mM, respectively. Oxygen was removed from the groundwater through nitrogen sparging. During operations, the injections were set to provide 1 to 3 ml of the solution from the storage tank per minute corresponding to 1 to 3 mM acetate and 100 to 300 pM bromide per day. Monitoring wells were installed at intervals downgradient corresponding to groundwater travel of approximately 4, 9, and 18 days with a further three wells placed upgradient to serve as controls. Acetate was injected continuously over a three month period from June to October 2002, with groundwater samples collected at regular time intervals from all monitoring wells. Groundwater conditions were monitored, including pH, conductivity, redox potential and dissolved oxygen, with further samples taken for U(VI), anion (bromide, nitrate, and sulfate), Fe(n), sulfide and acetate analysis. A second round of acetate injections were made over the same months in 2003, after which no further amendments were made.111

Bromide, added as a groundwater tracer, was not detected in any of the upgradient wells but was detected after 4, 9, and 18 days at each of the corresponding downgradient wells confirming the injection solution had reached the targeted area. After the first set of injections, U(VI) concentra­tions were observed to decrease 9 days after the injections began with concentrations dropping to or below 0.18 pM within 50 days at some wells.28 The decrease in U(VI) was concurrent with the accumulation of Fe(II) and prior to any sulfate reduction. After 50 days, the U(vi) concentration began to increase, coincident with a decrease of Fe(II) and acetate falling to non-detectable levels. Bromide levels were still detected at wells where acetate levels had fallen suggesting that an increase in consumption of acetate was occurring near the point of injection. This correlated with observations following the second injection stage of a depletion of reducible iron oxide near the injection point and an accumulation of sulfide111 suggesting all the available Fe(III) had been consumed and that sulfate reducers were now actively consuming acetate at the injection point.

A substantial shift in the microbial community was observed throughout the injection trials. Organisms in the family Geobacteraceae (which includes the known U(VI)-reducing genus Geobacter) became dominant early on,28 with the greatest enrichment of Geobacteraceae correlated to the greatest proportion of U(iv) detected.111 As reducible Fe(m) became depleted and sulfide accumulation occurred, the dominance of the Geobacteraceae decreased as they were replaced by species related to known sulfate-reducers.111 After the second round of injections in 2003, U(vi) continued to be removed from the groundwater for over a year after the cessation of acetate injections.112 This casts doubt on the suggestion following on from the first round of injections that U(VI) removal is acetate dependent. Flow-through column experiments suggested that the continued decrease in groundwater U(VI) levels could be linked to increased sorption to soils in a reduced environment.112

This series of field studies suggest that the stimulation of metal-reducing bacteria is an effective method for the removal of U(vi) from groundwater. However, when the supply of reducible Fe(III) oxides runs out, sulfate-reducers become dominant and do not appear to be as effective at reducing U(vi) to U(iv). Promising data from the second round of injections indicates that in sufficiently reduced soils, U(vi) removal may continue, without the continued need for acetate injections, via sorption to soils.