U.S. Department of Energy, Office of Science Environmental Remediation Sciences Division (ERSD) FY05 Third Quarter Performance Measure

In the third quarter of 2005, sampling and analysis of groundwater continued at the Old Rifle field site in Rifle, CO. Evaluation and publication of results is ongoing (see publication list on this website, http://www.pnl.gov/nabir-umtra/pubs.stm). ERSD FY05 Third Quarter Performance Measure entitled "Report results of Old Rifle field experiments and compare to previous laboratory studies" was successfully met by these activities.

1. U(VI) removal rates for field experiments and laboratory incubations
2. Observation of dissolved oxygen (DO) stratification in the Old Rifle alluvial sediments during spring runoff and its relationship to increases in U(VI) concentration
3. Progress in reactive transport modeling that accounts for bioreduction of U(VI), mineral equilibria, and groundwater flux and dispersion, including the concept of adapting the reactive transport model to modeling of bottle incubations.

U(VI) removal rates for bottle incubations of Rifle sediments are plotted by Ortiz-Bernad et al. (2004, see http://www.pnl.gov/nabir-umtra/pubs.stm). The highest rates observed over month-long time courses were ~0.24 uM/day for sediments without added U(VI) and ~1.39 uM/day for sediment with 20 uM U(VI) added. While direct comparison of the bottle incubation data and field data is not strictly appropriate, field observations of U(VI) removal during biostimulation under Fe(III) reducing conditions and are lower than during bottle incubation. In the 2002 experiment, the apparent U(VI) loss rate for well M-03 was ~0.02 uM/day. U(VI) removal rate for the 2004 experiment was similar overall to the 2002 experiment but in detail showed two different rates, ~0.05 and ~0.017 uM/day. It is not surprising that the U(VI) loss rate for field experiments is an order-of-magnitude lower than for bottle incubations. Bottle incubations are closed systems whereas field experiments are open systems with an on-going flux of U(VI) from up-gradient. This underscores the importance of reactive transport modeling which we have successfully used to track field geochemistry and Monod kinetics to estimate microbial growth and activity and other processes to match observed U(VI) concentrations during the field experiments (see discussion below). We are currently planning to use reactive transport modeling to directly link bottle incubations and field-scale experiments.

The 2005 spring runoff in the Colorado River at the Old Rifle site resulted in an increase in water table elevation of >1 m. As surmised from earlier observations, both DO and U(VI) increased in the shallowest part of the aquifer (see plots below).

The trends in DO and U(VI) with depth suggest a relationship between the two. A plot of DO vs. U(VI) for comparable depths in adjacent wells (M-07 and M-08) demonstrates the correlation (see plot below). U(VI) does not begin to increase significantly until DO reaches ~0.5 or 0.6 mg/l.

It is unclear if DO actually causes high U(VI) via oxidation of U(IV) or if the correlation is coincidental to the rising water table and the increased U(VI) is caused by dissolution and/or desorption of U(VI) "stranded" in vadose zone pores. Batch and column experiments currently being conducted by Peter Jaffe (Princeton) will directly address the reoxidation process using a sediment sample from Old Rifle (Rifle aquifer background sediment, or RABS). We also plan to provide Jaffe with vadose zone sediments to directly assess the origin of U(VI) in groundwater during periods of high water table.

Also completed were Biogeochemical reactive transport simulations of the 2002 acetate biostimulation field experiment at the Old Rifle UMTRA site in Colorado. These simulations include the response of the flowing groundwater system to the time-dependent release of acetate from a 16 m wide gallery of 20 injection wells aligned perpendicular to the predominant groundwater gradient. Terminal electon acceptor processes for Fe(III), U(VI), and S(VI) were modeled with energetics-based redox reactions that included the conversion of acetate to biomass and bicarbonate. A dual Monod rate law was employed with calibrated rates and utilization thresholds. In this case, the bioreduction of uranium is assumed to be performed by the iron reducers (i.e., Geobacter sp.). Upon depletion of the bioavailable Fe(III), there is a succession to acetate-oxidizing sulfate reducers which presumably do not reduce uranium. The simulation also includes abiotic sorption, mineral precipitation and dissolution reactions with the biological reduction byproducts. The plot below shows the comparison of simulated time-dependent concentrations of U(VI) at 5 monitoring wells 7.3 m downgradient from the injection gallery. Variability between the observations at the 5 monitoring wells is the result of spatial heterogeneity in material properties and temporal variations in the acetate release rate for each injection well. The dark blue line is the model result.

U.S. Department of Energy, Office of Science Environmental Remediation Sciences Division (ERSD) FY05 First Quarter Performance Measure

In the first quarter of 2005, sampling of groundwater was conducted at an experimental field site in Old Rifle, Colorado, and samples were analyzed for U concentrations. ERSD FY05 First Quarter Performance Measure entitled "Conduct monitoring at Old Rifle UMTRA experimental site and collect data on the bioreduction of uranium" was successfully met by this activity.

Monitoring (sampling) of groundwater wells was done as a follow up to an in situ acetate amendment experiment designed to stimulate growth of metal-reducing bacteria such as Geobacter. Uranium concentrations in groundwater down gradient decreased in a fashion similar to a 2002 field experiment except that the 2004 field experiment was terminated prior to development of extensive sulfate reduction. The decrease in U(VI) paralleled dominance of Geobacter in groundwater samples, strongly indicating that Geobacter is responsible for enzymatic reduction of U(VI) in situ (based on research of Derek Lovley et al. in progress). Monitoring of the minigallery in the first quarter of FY-2005 shows that U(VI) as measured by passive multi-level samplers rebounded by an average of 47% and a maximum of 78%. This result is interpreted to reflect the short duration of the 2004 acetate amendment (~1 month), producing less total biomass than longer experiments. Microbially mediated U(VI) reduction thus may not be sustained post-acetate injection as appears to be the case after longer experiments. The two figures below show U(VI) loss during the experiment followed by the rebound noted above for a specific well (M-18).

U.S. Department of Energy, Office of Science Environmental Remediation Sciences Division (ERSD) FY04 Fourth Quarter Performance Measure

Long-Term Monitoring of Uranium Remobilization

Injection of low concentrations of electron donor into the subsurface, shows that microbial mediated reduction of soluble U(VI) to relatively insoluble U(IV) can be readily achieved at the field scale (Anderson et al. 2003). This process is dominated by Geobacter (See www.geobacter.org) and results in removal of U(VI) from groundwater via precipitation. Given the effectiveness of biostimulation for U(VI) removal from groundwater, the next question is the rate at which U(IV) may be remobilized via oxidation, once electron donor addition is stopped. To address this question, monitoring of U(VI) concentration in groundwater down-gradient from the site of electron-donor amendment at the Old Rifle UMTRA Site has been performed after two different experiments, one in 2002 and the other in 2003.

Typical results are shown in the figure below. In general, the addition of electron donor, in this case, acetate (equivalent to dilute vinegar), results in variable rates of loss of U(VI) from groundwater. Following the 2002 experiment in the vicinity of well M-03 at 6 m depth (top panel), the amount of U(VI) removed from groundwater decreased over a nine month period but always remained positive ranging from 85 to 55%. Similar results were observed for this well after the 2003 experiment. Results for well M-08 (bottom panel) are even more striking as they actually increase with time, exceeding 85% several months after addition of acetate was stopped. The implication of these results is that addition of acetate to the subsurface at this site initiated bioreduction of U(VI) that was sustained at varying levels after acetate addition was stopped. So far, the question is not what the remobilization rate is, but rather the rate of ongoing U(VI) reduction. We are currently investigating the mechanism of U(VI) loss.

Contact and Collaborator Information:

Collaborators: Derek R. Lovley¹, Kelly Nevin¹, Regina O'Neil1, C. T. Resch², Aaron Peacock³, Helen Vrionis¹, Yun-Juan Chang³, Dick Dayvault⁴, Irene Ortiz Bernad¹, Ken Williams⁵, Susan Hubbard⁵, Steve Yabusaki², Yilin Fang², and D. C. White³

¹University of Massachusetts, Amherst, MA; ²Pacific Northwest National Laboratory, Richland, WA; ³ University of Tennessee, Knoxville, TN; ⁴ S. M. Stoller Corporation, U.S. Department of Energy, Grand Junction, CO; ⁵ Lawrence Berkeley National Laboratory, Berkeley, CA.