Quantification of mineral accessible surface area and flow-dependent fluid-mineral reactivity at the pore scale

Abstract

Accessible surface areas (ASAs) of individual rock-forming minerals exert a fundamental control on the maximum mineral reactivity with formation fluids. Notably, ASA efficiency during fluid-rock reactions can vary by orders of magnitude, depending on the inflow fluid chemistry and the velocity field. Due to the lack of adequate quantification methods, determining the mineral-specific ASAs and their reaction efficiency still remain extremely difficult. Here, we first present a novel joint method that appropriately calculates ASAs of individual minerals in a multi-mineral sandstone. This joint method combines SEM-image processing results and Brunauer-Emmett-Teller (BET) surface area measurements by a Monte-Carlo algorithm to derive scaling factors and ASAs for individual minerals at the resolution of BET measurements. Using these atomic-scale ASAs, we then investigate the impact of flow rate on the ASA efficiency in mineral dissolution reactions during the injection of CO2-enriched brine. This is done by conducting a series of pore-scale reactive transport simulations, using a two-dimensional (2D) scanning electron microscopy (SEM) image of this sandstone. The ASA efficiency is determined employing a domain-averaged dissolution rate and the effective surface area of the most reactive phase in the sandstone (dolomite). As expected, the dolomite reactivity is found to increase with the flow rate, due to the on average high fluid reactivity. The surface efficiency increases slightly with the fluid flow rate, and reaches a relatively stable value of about 1%. The domain averaged method is then compared with the in-out averaged method (i.e the “Black-box” approach), which is often used to analyzed the experimental observations. The in-out averaged method yields a considerable overestimation of the fluid reactivity, a small underestimation of the dolomite reactivity, and a considerable underestimation of the ASA efficiency. The discrepancy between the two methods is becoming smaller when the injection rate increases. Our comparison suggests that the result interpretation of the in-out averaged method should be contemplated, in particular, when the flow rate is small. Nonetheless, our proposed ASA determination method should facilitate accurate calculations of fluid-mineral reactivity in large-scale reactive transport simulations, and we advise that an upscaling of the ASA efficiency needs to be carefully considered, due to the low surface efficiency.