Modeling uranium and ²²⁶Ra mobility during and after an acidic in situ recovery test (Dulaan Uul, Mongolia)

Highlights

• A single reactive transport model was used for production and environmental modeling

• Six years after the end of the ISR test, natural attenuation was effective for U and ²²⁶Ra

• Key geochemical mechanisms are clay sorption and ²²⁶Ra co-precipitation into BaSO₄

• Gravity flow has a strong impact on acidic plume dilution

Abstract

This article presents the results of groundwater monitoring over a period of six years and the interpretation of these results by a reactive transport model, following an In Situ Recovery (ISR) test on the Dulaan Uul uranium deposit in Mongolia. An environmental monitoring survey was set up using 17 piezometers, from which it has been possible to describe the changes in the water composition before, during and after the ISR test. The water quality before the start of mining activities rendered it unfit for human consumption. During and after the test, a descent of the saline plume was observed, resulting in a dilution of the injection solutions. After a rapid decrease to pH = 1.13 during the production phase of the ISR test, the pH stabilized at around 4 in the production area and 5.5 below the production cell one year after the end of the test. Uranium and radium were being naturally attenuated. Uranium returned to background concentrations (0.3 mg/L) after two years and the measured ²²⁶Ra concentrations represent no more than 10% of the expected concentrations during production (75 Bq/L). The modeling of the contaminants of concern mobility, namely pH and concentrations of sulfate, uranium and ²²⁶Ra, is based on several key complementary mechanisms: density flow, cation exchange with clay minerals and co-precipitation of ²²⁶Ra in the barite. The modeling results show that the observed plume descent and sulfate dilution can only be predicted if consideration of a high-density flow is included. Similarly, the changes in pH and ²²⁶Ra concentration are only correctly predicted when the cationic exchanges with the clays and the co-precipitation reaction within the barite using the solid solution theory are integrated into the models. Finally, the proper representation of the changes in water composition at the scale of the test requires the use of a sufficiently fine mesh (1 m × 1 m cell) to take into account the spatial variability of hydrogeological (permeability distribution in particular) and geological (reduced, oxidized and mineralized facies distributions) parameters.

Introduction

World uranium production is now mainly accomplished by In Situ Recovery (ISR), with 29,492 tons produced in 2017, or nearly 50% of world production (WNA, 2018). This mining technique, the first evidence of which dates back to Roman antiquity, is also used to extract evaporites (potash, natron, and halite). In addition to uranium, ISR is also used for other metals such as copper and gold (Kuhar et al., 2018; Martens et al., 2012; Seredkin et al., 2016; Sinclair and Thompson, 2015). This mining method consists of dissolving the uranium contained in permeable, mainly sandstone, geological formations “in situ” using leaching solutions. The solubilized ore is then pumped to the surface and the uranium extracted (IAEA, 2001). The leaching solution is then re-injected into the well system after its properties have been renewed. The leaching solutions used for uranium production by ISR are of two types: alkaline and acidic. The other possibility is acid leaching (most often sulphuric) as practiced since the 1960s in countries of Eastern Europe, Central Europe and Central Asia (Kazakhstan, Uzbekistan, Russia, Ukraine, Bulgaria, Czech Republic) but also in China and Australia (IAEA, 2016; Märten, 2006). Compared to more traditional mining methods, ISR has attractive investment costs due to the absence of ore extraction and mechanical processing. ISR also has low operating costs that make it possible to bring low- or very low-grade deposits, in the order of 100 ppm of uranium, into production (Seredkin et al., 2016).

The deposits targeted for ISR extraction are mainly roll-front (Fig. S1, Supplementary information). These are epigenetic concentrations of uranium within the terrigenous sedimentary series in intermountain and epicontinental basins (Dahlkamp, 2010). Mineralized orebodies are found within permeable sandy layers, with several superimposed mineralized layers generally present in the same basin, sometimes several hundred meters thick. Mineralization corresponds to the presence of a redox front separating a reduced downstream area rich in organic matter and often sulphides, and an upstream area where the sediments appear leached and/or oxidized. Since uranium is insoluble under reducing conditions, it precipitates in contact with the reducing zone in the form of uraninite and coffinite (WoldeGabriel et al., 2014; Zammit et al., 2014). Unlike many other types of uranium deposits, roll-front deposits are low grade whilst also often having considerable tonnage and extension, at depths that can reach several hundred meters. Another necessary condition for ISR extraction is the confinement of the mineralized reservoir, generally by low-permeability clay screens, thus isolating the exploited reservoir from other adjacent aquifers.

It is the mine operator's responsibility to monitor the hydraulic balance between the injection and the production wells in order to limit the risks of a potential spread of leaching solutions outside the mine's boundary (Jeuken et al., 2007). These processes had not been verified in the case of the Stráž mine in the Czechoslovak Republic, resulting in significant pollution of an aquifer dedicated to drinking water supply (Smetana et al., 2002). Reactive transport models are increasingly developed to optimize the operational (recovery) parameters, such as injection flow rates and the pH of the injection solution, by taking into account the heterogeneity of the geological formation being mined (see for example Bonnaud et al., 2015; Johnson and Tutu, 2016; Lagneau et al., 2019; Regnault et al., 2015). However, these studies are often not published for confidentiality reasons (see, for example, Martens et al., 2012; Simon et al., 2014; Staub et al., 1986; Woods and Jeuken, 2015).

ISR mining is considered to result in a limited environmental impact on the surface compared to other mining techniques. In addition, no mill tailings is generated (Ballini et al., 2020; Chautard et al., 2020; Déjeant et al., 2016). The main environmental impacts associated to ISR mining are related to water quality in the mineralized aquifer and the risk of contaminating adjacent aquifers (Mudd, 2001a, Mudd, 2001b; Saunders et al., 2016). This last point is often questioned by stakeholders acting for the societal and environmental acceptance of mines operated by ISR. This is partly due to the limited data available to date in the scientific literature, particularly regarding the environmental impact of acid ISR exploitation. Knowledge on this topic had just started to emerge (Clay, 2015).

The initial water quality of uranium-mineralized aquifers is generally poor with high natural levels of metals and radioelements (²³⁸U and its descendants) but also salinity levels of up to several g/L. They are unsuitable for human consumption (Ariunbileg et al., 2016) and belong to a class for industrial or technical use.

ISR extraction generates a saline plume in the mining zones following the injection of the leaching solutions and the chemical reactions with the mineralized host (dissolution, precipitation, sorption…). The criteria for the remediation of mineralized aquifers exploited using ISR depend on the legislation of the countries where the mines are located. However, it is possible to identify a general consensus, that is to return to a physicochemical water quality comparable to that observed before mining. The definition of water quality may vary from one country to another. For example, it may be required to return it to a strictly identical state in terms of chemical composition, or instead to verify a range of concentrations which will maintain the initial class of the aquifer being mined. This illustrates the importance of establishing the initial hydrogeochemical background levels, by taking into account the spatial variability in the chemical composition of the water inherent in the geochemistry of the mineralized aquifer. For example, we can distinguish between the hydraulic upstream part of the deposit characterized by oxidizing conditions, and the mineralized and the hydraulic downstream zones, both characterized by reducing conditions and highly variable dissolved uranium contents. In the case of acid ISR mining, it is possible to distinguish several contaminants and physicochemical parameters of interest: pH, linked to the acidification of the mineralized zone; SO₄ ions, also linked to the acidification of the extraction zone (use of sulphuric acid); dissolved residual uranium; ²²⁶Ra from the dissolution of the ore (as well as other radioactive descendants of uranium); and finally, other metals and cations solubilized by the acidification and oxidation of the mined mineralized zone.

Three main approaches to the remediation of aquifers mined by ISR currently stand out: (1) the implementation of regular environmental monitoring and modeling to assess the natural attenuation potential of the mined reservoir, (2) enhanced natural attenuation (notably via biostimulation) and (3) “pump and treat”.

Natural attenuation is based partly on the geochemical ability of the geological host to buffer the chemical disturbance associated with the injection of the oxidizing solution, and partly on the natural dilution of the salt plume generated during mining. This remediation strategy has been used beyond the single context of acid ISR (see Christensen et al., 2004; Jørgensen et al., 2010). Natural attenuation has a clear economic advantage since this solution is accompanied simply by environmental monitoring with, if required, predictive geochemical modeling (Davis and Curtis, 2007; Jeuken et al., 2009). The durations involved are longer than for “pump and treat” and bioremediation solutions, but can be reduced in the case of assisted natural attenuation as proposed by (Park et al., 2007). As a result, the remediation of ISR sites by natural attenuation remains poorly documented, although some examples are available in the scientific literature (Bakarzhiyev et al., 2004; Dong et al., 2016; Fyodorov, 2002; IAEA, 2005; Yazikov and Zabaznov, 2002).

Enhanced natural attenuation is one of the alternative aquifer remediation strategies that have been developed, initially and mainly outside the ISR context. The solutions developed are mainly based on immobilizing the contaminants within the aquifer by optimizing the physicochemical conditions of the host environment. It is interesting to note the development in recent years of so-called biostimulation solutions, resulting in particular from work carried out on the treatment of polluted soils. Biostimulation aims to bio-catalyse the chemical reactions necessary to achieve the remediation objectives (Jemison et al., 2020; Long et al., 2008). These solutions have sometimes gone beyond the simple experimental stage and have been applied on an industrial scale, notably for certain organic compounds. In the case of uranium, several industrial biostimulation tests have recently been carried out (see for example Groudev et al., 2008, Groudev et al., 2010; Long et al., 2015). In the more specific context of ISR operations, biostimulation work is currently underway both at alkaline sites (Hall, 2009; Wu et al., 2006a, Wu et al., 2006b; Yabusaki et al., 2007, Yabusaki et al., 2014) and acid sites (Coral et al., 2018; Descostes et al., 2014). Finally, “pump and treat” has been widely deployed in the United States in the context of alkaline ISR operations. It focuses mainly on water treatment, where appropriate, in several stages (IAEA, 2001, IAEA, 2005): (i) Pumping out the contaminated water after the end of mining operations. The leached area is thus in contact with the uncontaminated aquifer; ii) Treating the contaminated water on the surface (usually by reverse osmosis) and then re-injecting it into the leached area; iii) if necessary, adding a chemical reducer (e. g. H₂S or Na₂S) (in-situ or ex-situ) to promote the precipitation of insoluble metals under reducing conditions; iv) Pumping out the injected water again, and then re-injecting it to enhance the mixing phase. In view of the feedback from experience gained mainly under alkaline ISR operating conditions, it is now considered that this expensive solution does not necessarily meet the initial objectives of remediation (Hall, 2009; Yabusaki et al., 2014). Indeed, the sensu stricto treatment of water does not take into consideration the stock of contaminants of interest in their entirety (Plaue and Czerwinski, 2002). Thus, the contaminants of interest may be immobilized within the aquifer either as neoformed minerals or sorbed onto the surfaces of the aquifer's host minerals, mainly of clay (Dangelmayr et al., 2018; Johnson and Tutu, 2016; Robin et al., 2015b, Robin et al., 2017, Robin et al., 2020). The gradual return of the natural hydraulic gradient leads to chemical reactions between the uncontaminated groundwater and the solid fraction of the leachate zone, which will tend to remobilize the adsorbed contaminants (Anastasi and Williams, 1984). Also, even if the water from the leached area is treated in the short term, a remobilization of contaminants can be observed in the longer term when the equilibrium conditions of the leached area are modified by the return of uncontaminated water (see examples given in Hall, 2009).

In order to assess the best solution for the remediation of mineralized aquifers mined by acid ISR, several complementary approaches are available, in particular: (1) observations at mine sites operated by acid ISR; (2) laboratory tests; (3) the establishment of pilot sites; (4) geochemical modeling. Recent exploration work carried out by COGEGOBI LLC (a subsidiary of the Orano group) has led to the discovery of two major roll-front uranium deposits in the Tsagaan Els basin of Mongolia: Dulaan Uul and Zuuvch Ovoo (Cardon et al., 2015; Grizard et al., 2019). These deposits are currently the subject of feasibility studies for in situ acid leaching, and it is in this context that an ISR test has been carried out at the Dulaan Uul deposit. The pilot site of Dulaan Uul can be considered for modelers as a relevant work tool. This site, equipped with numerous environmental monitoring piezometers (17), made it possible to monitor the post-operations natural attenuation from December 2010 to May 2017 with the benefit of a comprehensive set of data (broad sampling, both spatially and temporally). The Dulaan Uul site is therefore a high-quality calibration point for environmental monitoring models which aim to propose an environmental remediation scenario (solely by natural attenuation, or coupled with reprocessing/bioremediation) adapted to the local geochemistry. This article presents the results of the environmental monitoring of the ISR test carried out at the Dulaan Uul deposit over a six-year period, as well as the geochemical modeling of the behaviour of the various contaminants of interest, which are pH, SO₄ ions, and also uranium and ²²⁶Ra. This work makes it possible to understand the nature of the impact caused by acid ISR exploitation and to develop reactive transport models focused on the chemical changes during and after the ISR test, based on water quality monitoring and available mineralogical and hydrogeological data. As a result, this work helps to identify the key geochemical mechanisms and the degree of precision of their consideration, to correctly represent these changes both during the test phase and during environmental monitoring after the test has been completed. Ultimately, the operator will have access to a tool and to some key information that will enable him to choose the appropriate remediation method based on the legislative and environmental constraints of the site being operated.