Three-dimensional reactive transport simulation of Uranium in situ recovery: Large-scale well field applications in Shu Saryssu Bassin, Tortkuduk deposit (Kazakhstan)

Uranium in situ recovery (ISR) is the most widely used uranium mining technique worldwide. It consists of the dissolution of the ore by a mining solution, directly within the deposit. By predicting fluid flow and geochemical reactions in reservoirs, reactive transport (RT) modelling is a powerful tool to better understand and pilot ISR production. However, very few industrial uses have been reported thus far. This paper fills the gaps by illustrating a large-scale well-field application of RT modelling at one of the largest ISR mines worldwide, which is operated by Katco. This study highlights the robustness of a complex workflow based on the coupled reactive transport software HYTEC and its added value for the operator in the context of uranium ISR. The robustness demonstration is performed on 2394 wells covering 39 different production areas (blocks). The model reproduces the observed uranium concentrations and pH of pumped solutions over time scales up to 12 years. Only three parameters are manually adjusted to calibrate the model: global initial grades in clays (beidellite), calcite, and iron hydroxide (goethite). The discrepancy between simulated and observed uranium production and acid consumption decreases as the observation scale widens, showing that local errors compensate for each other. These deviations are mainly explained by the uncertainties of the 3D geological models and not by the RT simulations. Furthermore, the robustness of the model is a key asset for decision-making as it enables accurate predictions. This accuracy is illustrated through a case study of four of the simulated blocks. In 2019, after 10 years of production, the well field was redesigned to target the remaining uranium using the RT-based workflow. Several scenarios were simulated and sequentially optimised using a geometallurgical approach. The final adopted design predicted a 28% increase in uranium production and 35% in economic gains over the first two years of simulation alone (2019–2021). These theoretical gains were validated in practice as the comparison between the 2019 predictions and observations over 16 months showed a deviation less than 10% in the total uranium production, which decreased to 1.9% using the observed operational conditions, which reinforces the predictability of the workflow and validates the forecasted gains. © 2022 Elsevier B.V.