Uranium Natural Attenuation Downgradient of an in Situ Recovery

Article pubs.acs.org/est

Cite This: Environ. Sci. Technol. 2019, 53, 7483−7493

Uranium Natural Attenuation Downgradient of an in Situ Recovery Mine Inferred from a Cross-Hole Field Test Paul W. Reimus,*,† Martin A. Dangelmayr,† James T. Clay,‡ and Kevin R. Chamberlain§ †

Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, New Mexico 87545, United States Cameco Resources, Inc., 762 Ross Road, Douglas, Wyoming 82633, United States § University of Wyoming, Department of Geology and Geophysics, 1000 East University Avenue, Dept. 3006, Laramie, Wyoming 82071, United States Downloaded via NOTTINGHAM TRENT UNIV on August 13, 2019 at 12:00:20 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

S Supporting Information *

ABSTRACT: A field test was conducted at a uranium in situ recovery (solution mining) site to evaluate postmining uranium natural attenuation downgradient of an ore zone. Approximately 1 million liters of water from a previously mined ore zone was injected into an unmined ore zone that served as a proxy for a downgradient aquifer, while a well located approximately 23 m away was pumped. After 1 year of pumping, only about 39% of the injected U(VI) was recovered, whereas essentially 100% of coinjected chloride was recovered. A geochemical/transport model was used to simultaneously match the chloride and uranium concentrations at the pumping well while also qualitatively matching aqueous 238U/235U ratios, which reflect uranium removal from solution by reduction. It was concluded that ∼50% of the injected U(VI) was reduced to U(IV), although the reduction capacity in the flow pathways between the injection and production wells was estimated to be nearly exhausted by the end of the test. Estimating the reduction capacity of the downgradient aquifer can inform restoration strategy and offer a useful metric for regulatory decisions concerning the adequacy of restoration. U(VI) reduction should be effectively irreversible in these anoxic environments, which differ greatly from shallow oxic environments where U(IV) is readily reoxidized.

1. INTRODUCTION In situ recovery (ISR) of uranium is an economically favorable mining technique for extracting low-grade uranium ore ( 0.025 mmol/kg-sediment, with the values of kr and Sred being perfectly inversely correlated for these larger Sred values (identical fits as long as the product krSred remained constant). Fourth, the best overall match to both the uranium concentration data and the δ238U data was obtained with Sred

3. RESULTS AND DISCUSSION 3.1. Chloride and Uranium Concentration Data. Uranium was both delayed in its arrival and had lower dilution-adjusted normalized concentrations than chloride at the pumping well in the cross-hole test (Figure 2, which also shows selected model fits to the uranium data, discussed in section 3.2). The calculated recovery of Cl− at 7P-124 was approximately 120% of the Cl− mass injected into 7I-233, which is reflected in Figure 2 in that the Cl− model curve falls below the observed Cl− concentrations during the latter half of the test because the modeled recovery was limited to 100%. The reasons for this result are discussed in detail in section S4 of the Supporting Information. Uranium recovery was ∼39% over the time of the test. However, because the uranium concentrations were still elevated above background at the end of the test, the actual recoveries would have been higher had pumping continued. A log−linear extrapolation of the last 94 days of the tail of the breakthrough curve yielded an estimated ultimate recovery of ∼49% if pumping had continued indefinitely (the extrapolation was carried out until uranium concentrations dropped below 7487

DOI: 10.1021/acs.est.9b01572 Environ. Sci. Technol. 2019, 53, 7483−7493

Article

Environmental Science & Technology

Figure 3. δ238U values at the production well 7P-124 (red triangles) and in the MP-423 injection water (orange circles) during the cross-hole test plotted with model-predicted fractions of uranium reduced for the four uranium model curves shown in Figure 2. All δ238U values are relative to the mass-weighted average 238U/235U value in the injected MP-423 water. Note that the fractions reduced are referenced to the right-hand y axis, and the scale of this axis is reversed so that larger values are plotted lower. Also, the scale of this axis goes beyond a range of 0−1 so that zero corresponds to a δ238U value of zero on the left-hand axis, and the final values for the two curves with Sred = 0.0063 mmol/kg-sediment correspond visually to the final values of δ238U in well 7P-124. Error bars for δ238U reflect 2 standard deviations of measured values for numerous measurements of isotopic standards.

= 0.0063 mmol/kg-sediment. Also, the match to the δ238U data for Sred = 0.0063 mmol/kg-sediment could be improved by making minor adjustments to the parameter values associated with the best fit to the uranium concentration data without significantly compromising the fit to the concentration data. In contrast, parameters providing good fits to the concentration data for other Sred values had to be adjusted more to obtain reasonable matches to the δ238U data, with the result being significantly compromised fits to the concentration data. In the case of Sred > 0.025 mmol/kg-sediment, the qualitative matches to the δ238U data were consistently poor regardless of how much the model parameters were adjusted. The four uranium model fits shown in Figure 2 illustrate some of the points mentioned in the preceding paragraph. The fit corresponding to Sred = 0 (no reduction) represents the best fit that could be obtained without a reduction reaction, and it clearly results in significant overprediction of the uranium tailing behavior. The fit for Sred > 0.025 mmol/kg-sediment and the two fits for Sred = 0.0063 mmol/kg-sediment are arguably comparable, illustrating the inability of the uranium concentration data alone to place an upper bound constraint on the value of Sred. Of the two fits for Sred = 0.0063 mmol/kgsediment, one is the best overall fit to the concentration data and the other has slightly adjusted parameters to provide a better match to the δ238U data (see Figure 3) at the expense of a slightly poorer fit to the concentration data. The model parameters and SOWSR values associated with each of the uranium model curves of Figure 2, along with those of selected simulations not shown in Figure 2, are given in Table 1. The results in Table 1 for simulations not shown in Figure 2 serve

Table 1. Model Parameters and SOWSR Values for Selected Model Simulationsa Sred (mmol/kg-sediment)

log β

kr (kg/(mmol s))

SOWSR

0 0.0038 0.0050 0.0050* 0.0063 0.0063* 0.0075 0.0075* 0.0101 0.0126 >0.025

2.33 2.1 2.1 1.9 1.96 1.8 2.1 1.7 2.2 2.2 2.21

N/A 6.0e-05 4.9e-05 2.2e-04 1.1e-04 2.6e-04 2.6e-05 3.1e-04 7.9e-06 5.6e-06 5e-8/Sred

154 38 15.4 23.7 9.8 15.3 16.9 28 14.1 14.9 15.7

a Rows in italics correspond to simulations plotted in Figures 2 and 3. Rows with asterisks next to Sred values correspond to simulations in which the parameters that best fit the uranium concentration data were adjusted to achieve a better qualitative match to the δ238U data. log β values are all based on the assumption that Ssorp = 60.2 mmol/ kg-sediment.

to further illustrate some of the points mentioned in the preceding paragraph. Because of the perfect inverse correlation of log β and Ssorp, the latter parameter was fixed in all simulations to be 60.2 mmol/kg-sediment, a value estimated from the product of sediment-specific surface area measurements and literature values of uranium sorption densities per unit area. Prior estimates of SRH sediment specific surface area ranged from 3.9 to 13.1 m2/g,10 and uranium sorption site densities for quartz have been reported at 4.81 and 5.1 sites/ 7488


Article


Figure 4. δ238U fractionation curve (inset plot) obtained by plotting observed δ238U values vs model-predicted fractions of uranium not reduced during the cross-hole test. The fractions of uranium not reduced were taken from the adjusted-parameter model fit for Sred = 0.0063 mmol/kgsediment shown in Figures 2 and 3. The larger plot shows δ238U values predicted for the test using the deduced fractionation factor of 1.00035.

which reflect goodness of fit to the concentration data, increased for Sred values of 0.005 and 0.0075 mmol/kgsediment when model parameters were adjusted to better match the δ238U data). Thus, whereas a reasonable fit to the concentration data could be obtained for any Sred value greater than about 0.0038 mmol/kg-sediment, the δ238U data constrained Sred values to the narrow range of 0.005−0.0075 mmol/kg-sediment, which constitutes a restrictive upper bound on downgradient aquifer reduction capacity. Section S7 of the Supporting Information provides additional discussion and plots illustrating the inability to match both the uranium concentration data and δ238U data outside the range of 0.005 < Sred < 0.0075 mmol/kg-sediment and why 0.0063 mmol/kg-sediment is selected as an optimal Sred value. Dilution-adjusted Ca2+ and SO42− concentrations and alkalinity (as mg/L CaCO3) together with model predictions are provided in Figure S3 in the Supporting Information. The ability of the model to approximate the observed Ca2+ concentrations and alkalinity is important because these two variables (along with pH) have the greatest effect on uranium aqueous speciation and modeled sorption behavior. 3.3. Inferred δ238U/235U Isotope Fractionation in Field Test. In cases where the PEST/PHREEQC model produced reasonable matches to the observed shifts in δ238U values, the δ238U shifts were plotted against the model-predicted fractions of U not reduced to produce an isotope fractionation curve deduced from the field test. An estimate of the 238U/235U Rayleigh fractionation factor was then obtained by fitting the

nm2,33,34 producing Ssorp estimates that range from 23.7 mmol/ kg-sediment to 96.7 mmol/kg-sediment, with a mean value of 60.2 mmol/kg-sediment. We recognize that many sorption sites on aquifer sediments (and almost all of those at which reduction likely occurred) would likely not be associated with quartz. However, our model interpretation can accommodate any sorption site density that might be considered more appropriate as long as it is greater than about 0.1 mmol/kgsediment, which is the point at which there simply were not enough sorption sites to achieve good matches to the uranium concentration data even if log β was set to very large values. Discussions of how the estimated sorption and reduction parameters given in Table 1 compare with laboratory-derived estimates of these parameters are provided in sections S5 and S6, respectively, of the Supporting Information. Figure 3 shows model-predicted fractions of uranium reduced vs time for the four uranium model curves of Figure 2, plotted along with measured δ238U values vs time. The criteria for a good qualitative match to the δ238U data was to reproduce the early maximum in fraction reduced (minimum in δ238U values), as well as the lower (but nonzero) relatively steady fraction reduced in the latter part of the test. Clearly, only the curves with Sred = 0.0063 mmol/kg-sediment in Figure 3 meet these qualitative matching criteria. It was found that Sred could be varied only from about 0.005 to 0.0075 mmol/kgsediment before either the fit to the concentration data or the qualitative match to the δ238U data was significantly compromised (see Table 1 for how much SOWSR values, 7489


Article

Environmental Science & Technology resulting trend, following the procedure of Brown et al.28 Figure 4 (inset) shows the results of this exercise for the crosshole test using the predicted fractions of uranium reduced vs time from the adjusted model fit with Sred = 0.0063 mmol/kgsediment in Figure 3. Figure 4 also shows the predicted δ238U values in the field test if the resulting best-fitting fractionation factor of 1.00035 is applied to the predicted fractions of uranium reduced (the match in Figure 4 looks slightly better than the qualitative match between fractions reduced and δ238U shown in Figure 3 because of the curvature of the fractionation curve). The best-fitting fractionation factor of 1.00035 is somewhat smaller than would be predicted from the trends reported by Brown et al.,28 but this is not surprising given the differences between the aquifer solids (a heterogeneous assemblage of mineral phases and possibly organic phases) and those used in the laboratory batch reactors (synthetic FeS). Furthermore, calcium concentrations and alkalinities were higher in the MP-423 water than in any of the laboratory experiments.28 Even so, the field-deduced fractionation factor of 1.00035 falls within the range of values reported by Brown et al.26 for all their experiments (1.00023 to 1.00083), and it is also in good agreement with a fractionation factor of 1.00048 ± 0.00008 reported by Basu et al.35 for an ISR mine in Texas. 3.4. Possible Underestimation of Downgradient Uranium Reduction Capacity. In the preceding sections, it was shown that honoring both the δ238U data and the uranium concentration data in interpreting the cross-hole test provided a rather restrictive upper-bound estimate of downgradient reduction capacity. There are several reasons why the field test and its interpretation may have underestimated the true reduction capacity of the downgradient aquifer (in addition to the aforementioned inference from well logs that the injection well was located on the upgradient, oxidized side of the roll-front deposit). First, any oxygen that managed to get into the MP-423 water during its 2 km transit in the surface pipeline and subsequent free-fall of over 60 m to the water table in well 7I-233 would have consumed reduction capacity that otherwise might have gone toward reducing U(VI). Dissolved oxygen was not routinely measured during the test, but even if concentrations were only 0.1 mg/L (vs a saturation level of ∼7 mg/L), oxygen would have constituted nearly 10% of the electron-acceptor equivalents of U(VI) injected during the test. Second, the short residence time of the test relative to downgradient residence times under natural flow conditions (as much as 30 years to reach the legal mine-unit boundary) may have precluded the observation of some reduction reactions that have time scales longer than that of the test. We assumed a single rate for all reduction reactions, which was constrained mainly by qualitatively matching the δ238U data, but in reality there are likely many reductants and many rates, some of which may have been too slow to observe. Third, the untreated nature of the MP-423 water used in the cross-hole test (no measures taken to reduce total dissolved solids concentrations, especially alkalinity and Ca2+ concentrations, both of which greatly stabilize U(VI) against reduction36,37) should have resulted in less U(VI) reduction in the test than would be expected in an actual postmining downgradient transport scenario. Fourth, drilling, completion, and development of the injection and production wells could have resulted in “halos” of oxidized sediments in the immediate vicinity of these wells that would have reduced the reduction capacity between the wells. Fifth, it is possible that some of the uranium

recovered at the pumping well could have been mobilized from the U-rich ore zone rather than originating from well MP-423, in which case the reduction capacity would have been underestimated (and the observed decreases in δ238U values would have likely been artificially suppressed). Finally, the estimated reduction capacity of 0.0063 mmol U(VI)/kgsediment translates to a U loading of about 0.00015 wt/% in the downgradient sediments, but roll-front deposits at SRH are typically on the order of 0.1 wt % U, which implies that there is the potential for significantly more downgradient reduction capacity in the sediments than was deduced from the test. Even if only a small fraction of this apparent gap in reduction capacity were available to reduce U(VI) over the much longer time scales associated with natural flow rates, downgradient U(VI) migration distances would be predicted to be trivial. For all these reasons, we conclude that the cross-hole test provided a conservatively low estimate for an upper bound of downgradient reduction capacity. 3.5. Uncertainties and Limitations. In this section, we discuss additional uncertainties and limitations of the study that go beyond the potential underestimation of reduction capacity. Perhaps most importantly, the study results should not be extrapolated to other ISR sites, or even to other locations at SRH, without carefully considering differences in aqueous and sediment geochemistry at the different locations. Our intent here was to demonstrate a methodology that could be applied at different sites, with the expectation that the results will vary from site to site. At a minimum, different parametrizations of the model can be expected at different sites, and it might also prove necessary to alter the model to properly describe observed transport behavior (for instance, including an additional uranium sorption reaction in the GC SCM). A significant uncertainty associated with the model interpretation was distinguishing between uranium reduction and strong sorption, particularly in the latter part of the test. We felt we could not rely on the observation of reduced aqueous species or model-calculated redox equilibria to infer uranium reduction because the primary reduced species, U(IV), is too insoluble to be measured in solution and using redox equilibria to infer reduction can be misleading for many reasons (e.g., inconsistencies in the oxidation−reduction potentials deduced from different redox couples, incomplete knowledge of solid phases forming and the degree to which solids catalyze reduction reactions, and gaps/uncertainties in thermodynamic data, particularly for reactions involving the highly stable ternary calcium−uranyl−carbonate complexes that dominated aqueous uranium speciation36,37). Therefore, we used shifts in δ238U values as an objective indicator of reduction. However, the δ238U data did not extend all the way to the end of the test, and it was not possible to continue the test until U concentrations approached background levels, both of which would have helped distinguish between reduction and strong sorption late in the test. If a longer test would have revealed a higher tail in uranium concentrations in comparison to that predicted by the best-fitting model (or obtained from a log−linear extrapolation), a better model fit could have been obtained by including a second, stronger sorption reaction in the model, which would have slightly decreased the amount of reduction deduced. However, it is important to keep in mind that the inclusion of a second (or third) sorption reaction without reduction offered little improvement over the fit of the observed uranium concen7490


Article


water is ∼0.08. Conventional restoration practices typically reduce U(VI) aqueous concentrations to a few milligrams per liter or less, which implies that, even with the relatively small downgradient reduction capacity inferred from the field test, the U(VI) left behind would migrate downgradient only a short distance from the ore zone before being reduced. However, the inventory of adsorbed U(VI) left behind in the ore zone (which could later desorb and become mobile) could potentially amount to dozens of milligrams per liter, which would require the equivalent of several times as much downgradient volume as ore-zone volume to reduce all the residual sorbed U(VI) (a sorption inventory of 50 mg/L U(VI) would require 4 times as much downgradient volume). Even so, such calculations using the reducing capacity estimated from the cross-hole test do not suggest extremely large downgradient migration distances and, given the considerations discussed in section 3.4, the above calculations are more likely to overestimate than underestimate migration distances. Ultimately, we would advocate using a reactive transport model in conjunction with transport parameter estimates derived from a field test (as in this study) to predict downgradient transport distances for a given amount of U(VI) inventory remaining behind in an ore zone. Alternatively, it may be more practical to use a conservative estimate of reduction capacity obtained from such a combined field testing and modeling exercise and work backward to determine how much ore-zone restoration needs to be done to lower U(VI) inventories enough so there is sufficient reduction capacity between the ore zone and the aquifer exemption boundary to reduce all the remaining U(VI). Either way, the results of this study should provide useful information and perspective to support decisions related to restoration and regulatory compliance for uranium ISR mining sites.

trations with only one sorption reaction and no reduction; thus, reduction was clearly necessary to achieve a reasonable fit. The only way that a second sorption reaction without reduction could have offered a fit to the uranium concentration data similar to that of the best-fitting model with reduction was if at least one sorption reaction was specified to have slow kinetics. That is, at least one sorption reaction would have to be described mathematically in a manner similar to that of the reduction reaction, with slow enough desorption kinetics to make desorption negligible over the time scale of the test. However, this approach would have been inconsistent with the implementation of the GC SCM in PHREEQC and, more importantly, it could not explain the observed shifts in δ238U because sorption is a nonfractionating process (or, if anything, it causes δ238U to shift slightly in the opposite direction of what was observed). We also rejected the possibility that there could have been a significant influence of uranium diffusion into and out of secondary (nonflowing or slow-flowing) porosity, which can have an effect similar to that of slow sorption and desorption. Besides being inconsistent with the δ238U data, significant diffusion into secondary porosity is inconsistent with the deduced ∼100% recovery of Cl− in the test. Given all these considerations, any realistic elevated uranium tailing beyond the end of the test could have lowered the estimate of injected uranium reduced by only a few percent (from about 50% of the injected uranium to no less than 45−46%). 3.6. Implications. The simultaneous matching of the PEST/PHREEQC model to both the uranium concentrations and the δ238U data in the field test provides convincing evidence that U(VI) reduction occurred during the test. From a regulatory perspective, this result is significant in that uranium reduction should be effectively irreversible in downgradient reducing environments at SRH (mimicking processes that caused ore zone formation in the first place), whereas sorption is generally considered a fully reversible process that only delays uranium transport, but does not immobilize uranium. However, the qualitative matches to the δ238U data also suggest that the reduction capacity within the test flow pathways was limited to ∼0.0063 mmol/kg-sediment, which for the assumed aquifer porosity of 0.25 translates to ∼0.00005 mol/L of solution or 12.5 mmol/m3 of total aquifer volume. An estimate of the volume of aquifer pore space interrogated during the test is given by the product of the volumetric injection flow rate (11 L/min) and the mean residence time of a conservative tracer (49 days for chloride), which in this case is ∼800000 L of water. Thus, ∼40 mol of reductant (800,000 L × 0.00005 mol/L) is estimated to have been present in the test flow pathways, which is only a little over half of the total moles of U(VI) injected during the test (∼71 mol). Furthermore, the model predicted that ∼50% of the injected U(VI), or ∼35 mol, was reduced during the test (in good agreement with the ∼50% recovery estimated from the log−linear extrapolation of uranium concentrations mentioned in section 3.1). This implies that ∼88% of the available model-deduced reduction capacity in the flow pathways (∼35 mol of uranium reduced divided by ∼40 mol of available reductant) was consumed in the test. If the downgradient reduction capacity is 0.0063 mmol/kgsediment and the aquifer porosity is 0.25, it can be shown that the ratio of downgradient aquifer volume to ore zone volume needed to reduce every 1 mg/L of U(VI) present in ore zone

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.9b01572. Comparison of Ore Zone and Downgradient Geochemistry 238 U/234U Activity Ratio Data from Cross-Hole Test and Additional Detail on 238U/235U and 238U/234U Isotope Ratio Measurements Additional Details of the PHREEQC Model, including Predictions of Concentrations of Aqueous Species that Affect Uranium Speciation Explanation for 120% Recovery of Cl- in Cross-Hole Test, and Implications for Uranium Results Sorption Parameter Comparison with Previous Laboratory Experiments Reduction Reaction Parameter Comparison with Previous Laboratory Experiments Supporting Information for Best Model Matches to both Uranium Concentration Data and δ238U Data (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*P.W.R.: tel, 505-670-5297; e-mail, [email protected]. ORCID

Paul W. Reimus: 0000-0002-7422-9540 Notes

The authors declare no competing financial interest. 7491


Article


■

parameters of sediments from a uranium in situ recovery site. J. Contam. Hydrol. 2018, 213, 28−39. (14) Liger, E.; Charlet, L.; Van Cappellen, P. Surface catalysis of uranium (VI) reduction by iron (II). Geochim. Cosmochim. Acta 1999, 63 (19−20), 2939−2955. (15) Fox, P. M.; Davis, J. A.; Kukkadapu, R.; Singer, D. M.; Bargar, J.; Williams, H. Abiotic U(VI) reduction by sorbed Fe(II) on natural sediments. Geochim. Cosmochim. Acta 2013, 117, 266−282. (16) Scott, T. B.; Allen, G. C.; Heard, P. J.; Randell, M. G. Reduction of U(VI) to U(IV) on the surface of magnetite. Geochim. Cosmochim. Acta 2005, 69 (24), 5639−5646. (17) WoldeGabriel, G.; Boukhalfa, H.; Ware, D.; Cheshire, M.; Reimus, P.; Heikoop, J.; Conradson, S. D.; Batuk, O.; Havrilla, G.; House, B.; Simmons, A.; Clay, J.; Basu, A.; Christensen, J. N.; Brown, S. T.; DePaolo, D. J. Characterization of cores from an in-situ recovery mined uranium deposit in Wyoming: Implications for postmining restoration. Chem. Geol. 2014, 390 (18), 32−45. (18) Gallegos, T. J.; Campbell, K. M.; Zielinski, R. A.; Reimus, P. W.; Clay, J. T.; Janot, N.; Bargar, J. R.; Benzel, W. M. Persistent U(IV) and U(VI) following in-situ recovery (ISR) mining of a sandstone uranium deposit, Wyoming, USA. Appl. Geochem. 2015, 63, 222−234. (19) Senko, J. M.; Istok, J. D.; Krumholz, L. R. In-Situ Evidence for Uranium Immobilization and Remobilization. Environ. Sci. Technol. 2002, 36 (7), 1491−1496. (20) Istok, J. D.; Senko, J. M.; Krumholz, L. R.; Watson, D.; Bogle, M. A.; Peacock, A.; Chang, Y. J.; White, D. In situ bioreduction of technetium and uranium in a nitrate-contaminated aquifer. Environ. Sci. Technol. 2004, 38 (2), 468−475. (21) Wu, W.-M.; Carley, J.; Luo, J.; Ginder-Vogel, M. A.; Cardenas, E.; Leigh, M. B.; Hwang, C.; Kelly, S. D.; Ruan, C.; Wu, L.; Van Nostrand, J.; Gentry, T.; Lowe, K.; Carroll, S.; Luo, W.; Fields, M. W.; Gu, B.; Watson, D.; Kemner, K. M.; Marsh, T.; Tiedje, J.; Zhou, J.; Fendorf, S.; Kitanidis, P. K.; Jardine, P. M.; Criddle, C. S. In Situ Bioreduction of Uranium (VI) to Submicromolar Levels and Reoxidation by Dissolved Oxygen. Environ. Sci. Technol. 2007, 41 (16), 5716−5723. (22) Moon, H. S.; Komlos, J.; Jaffé, P. R. Uranium Reoxidation in Previously Bioreduced Sediment by Dissolved Oxygen and Nitrate. Environ. Sci. Technol. 2007, 41 (13), 4587−4592. (23) Komlos, J.; Peacock, A. D.; Kukkadapu, R. K.; Jaffé, P. R. Longterm dynamics of uranium reduction/reoxidation under low sulfate conditions. Geochim. Cosmochim. Acta 2008, 72 (15), 3603−3615. (24) Stylo, M.; Wang, Y.; Monga, N.; Romaniello, S.J.; Weyer, S.; Bernier-Latmani, R. Uranium isotopes fingerprint biotic reduction. Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (18), 5619−5624. (25) Bopp, C. J.; Lundstrom, C. C.; Johnson, T. M.; Sanford, R. A.; Long, P. E.; Williams, K. H. Uranium 238U/235U Isotope Ratios as Indicators of Reduction: Results from an in situ Biostimulation Experiment at Rifle, Colorado, U.S.A. Environ. Sci. Technol. 2010, 44 (15), 5927−5933. (26) Basu, A.; Sanford, R. A.; Johnson, T. M.; Lundstrom, C. C.; Löffler, F. E. Uranium isotopic fractionation factors during U(VI) reduction by bacterial isolates. Geochim. Cosmochim. Acta 2014, 136, 100−113. (27) Shiel, A. E.; Johnson, T. M.; Lundstrom, C. C.; Laubach, P. G.; Long, P. E.; Williams, K. H. Reactive transport of uranium in a groundwater bioreduction study: Insights from high-temporal resolution 238U/235U data. Geochim. Cosmochim. Acta 2016, 187, 218−236. (28) Brown, S. T.; Basu, A.; Ding, X.; Christensen, J. N.; DePaolo, D. J. Uranium isotope fractionation by abiotic reductive precipitation. Proc. Natl. Acad. Sci. U. S. A. 2018, 115 (35), 8688−869320. (29) Dang, D. H.; Novotnik, B.; Wang, W.; Georg, R. B.; Evans, R. D. Uranium isotope fractionation during adsorption,(co) precipitation, and biotic reduction. Environ. Sci. Technol. 2016, 50 (23), 12695−12704.

ACKNOWLEDGMENTS This work was supported by funds from the State of Wyoming Legislature In-Situ Recovery Technologies Research Program (administered by the University of Wyoming School of Energy Resources under Grant WYDEQ46764 via Los Alamos Workfor-Others Agreement NFE-110063), by the U.S. Environmental Protection Agency (under Interagency Agreement DW89-92306201-0), and by the U.S. Department of Energy Nuclear Energy Office, through the DOE Office of Technology Transitions (Project TCF-16-12154). Cameco Resources, Inc., provided in-kind support for the field testing efforts under Cooperative Research and Development Agreement 10760.0 DOE-TCF. This project could not have been done without the support of Brent Berg, former president of Cameco Resources. Los Alamos National Laboratory is operated by Triad National Security, LLC, for the National Nuclear Security Administration of the U.S. Department of Energy (Contract No. 89233218CNA000001).

■

REFERENCES

(1) Catchpole, G.; Kuchelka, R. Groundwater restoration of uranium ISL mines in the United States, 1993. (2) Hall, S. Groundwater restoration at uranium in-situ recovery mines, South Texas coastal plain; US Geological Survey: 2009. (3) Reimus, P. W.; Pohll, G.; Mihevc, T.; Chapman, J.; Haga, M.; Lyles, B.; Kosinski, S.; Niswonger, R.; Sanders, P. Testing and parameterizing aconceptual model for radionuclide transport in a fractured granite using multiple tracers in a forced-gradient test. Water Resoures Research 2003, 39 (12), 1356−1370. (4) Hunter, J., Fluvial architecture and paleo-ground water infiltration of the Fort Union Formation near the Highland Uranium Mine, southern Powder River Basin, Wyoming, in Coalbed methane and Tertiary geology of the Powder River Basin: Wyoming Geological Association Guidebook; Miller, W. R., Ed.; Wyoming Geological Association: 1999; pp 119−139. (5) Bernhard, G.; Geipel, G.; Reich, T.; Brendler, V.; Amayri, S.; Nitsche, H. Uranyl (VI) carbonate complex formation: Validation of the Ca2UO2 (CO3) 3 (aq.) species. Radiochim. Acta 2001, 89 (8), 511−518. (6) Dong, W.; Brooks, S. C. Determination of the formation constants of ternary complexes of uranyl and carbonate with alkaline earth metals (Mg2+, Ca2+, Sr2+, and Ba2+) using anion exchange method. Environ. Sci. Technol. 2006, 40 (15), 4689−4695. (7) Stewart, B. D.; Mayes, M. A.; Fendorf, S. Impact of uranylcalcium- carbonato complexes on uranium (VI) adsorption to synthetic and natural sediments. Environ. Sci. Technol. 2010, 44 (3), 928−934. (8) Parkhurst, D. L.; Appelo, C. A. J. User’s guide to PHREEQC (V2); US Geological Survey: 1999; p 312. (9) Doherty, J. E.; Hunt, R. J. Approaches to highly parameterized inversion-A guide to using PEST for groundwater-model calibration; US Geological Survey: 2010. (10) Dangelmayr, M. A.; Reimus, P. W.; Wasserman, N. L.; Punsal, J. J.; Johnson, R. H.; Clay, J. T.; Stone, J. J. Laboratory column experiments and transport modeling to evaluate retardation of uranium in an aquifer downgradient of a uranium in-situ recovery site. Appl. Geochem. 2017, 80, 1−13. (11) Davis, J. A.; Kent, D. B. Surface complexation modeling in aqueous geochemistry. Reviews in Mineralogy and Geochemistry 1990, 23 (1), 177−260. (12) Davis, J. A.; Meece, D. E.; Kohler, M.; Curtis, G. P. Approaches to surface complexation modeling of Uranium (VI) adsorption on aquifer sediments. Geochim. Cosmochim. Acta 2004, 68 (18), 3621− 3641. (13) Dangelmayr, M. A.; Reimus, P. W.; Johnson, R. H.; Clay, J. T.; Stone, J. J. Uncertainty and variability in laboratory derived sorption 7492


Article

Environmental Science & Technology (30) Brennecka, G. A.; Wasylenki, L. E.; Bargar, J. R.; Weyer, S.; Anbar, A. D. Uranium Isotope Fractionation during Adsorption to Mn-Oxyhydroxides. Environ. Sci. Technol. 2011, 45 (4), 1370−1375. (31) Jemison, N.; Johnson, T. M.; Shiel, A. E.; Lundstrom, C. C. Uranium Isotopic Fractionation Induced by U(VI) Adsorption ontoCommon Aquifer Minerals. Environ. Sci. Technol. 2016, 50 (22), 12232−12240. (32) Shiel, A. E.; Laubach, P. G.; Johnson, T. M.; Lundstrom, C. C.; Long, P. E.; Williams, K. H. No Measurable Changes in 238U/235U due to Desorption−Adsorption of U(VI) from Groundwater at the Rifle, Colorado, Integrated Field Research Challenge Site. Environ. Sci. Technol. 2013, 47 (6), 2535−2541. (33) Arnold, T.; Zorn, T.; Zänker, H.; Bernhard, G.; Nitsche, H. Sorption behavior of U (VI) on phyllite: experiments and modeling. J. Contam. Hydrol. 2001, 47 (2−4), 219−231. (34) Kitamura, A.; Fujiwara, K.; Yamamoto, T.; Nishikawa, S.; Moriyama, H. Analysis of adsorption behavior of cations onto quartz surface by electrical double-layer model. J. Nucl. Sci. Technol. 1999, 36 (12), 1167−1175. (35) Basu, A.; Brown, S. T.; Christensen, J. N.; DePaolo, D. J.; Reimus, P. W.; Heikoop, J. M.; Woldegabriel, G.; Simmons, A. M.; House, B. M.; Hartmann, M.; Maher, K. Isotopic and Geochemical Tracers for U(VI) Reduction and U Mobility at an in Situ Recovery U Mine. Environ. Sci. Technol. 2015, 49 (10), 5939−5947. (36) Ginder-Vogel, M.; Criddle, C. S.; Fendorf, S. Thermodynamic constraints on the oxidation of biogenic UO2 by Fe (III)(hydr) oxides. Environ. Sci. Technol. 2006, 40 (11), 3544−3550. (37) Maher, K.; Bargar, J. R.; Brown, G. E. Environmental Speciation of Actinides. Inorg. Chem. 2013, 52 (7), 3510−3532.

7493


Uranium Natural Attenuation Downgradient of an in Situ Recovery

Recommend Documents