Abstract Carbon dioxide‐rich brine formation during geological carbon sequestration induces calcite dissolution, governed by the physicochemical coupling of fluid flow, reactive transport, and pore structure evolution. Unveiling the mechanisms that control this dissolution, particularly under varying flow and structural conditions, is essential for predicting CO 2 plume migration and ensuring long‐term storage stability. While previous studies have explored these coupled processes, they often lack explicit resolution of fracture‐matrix interactions and are limited by computational scalability. In this study, we present a novel pore‐scale numerical framework that integrates the volumetric lattice Boltzmann method with a GPU‐CUDA parallel computing architecture, enabling efficient simulations of reactive flow in both fracture‐free system and fracture‐matrix system. Results reveal that injection velocity governs dissolution morphology and efficiency, with higher velocities reducing reactivity due to preferential flow, while temperature moderately enhances front heterogeneity but has limited impact on overall dissolution behavior. Based on the dissolution profiles observed in two types of 3D carbonate rock cores, three distinct calcite dissolution regimes (uniform, channel widening, and face dissolution) are identified. Moreover, the normalized permeability‐porosity relationship exhibits a negative correlation with temperature across all cases, except at higher injection velocities in the fracture‐matrix system, where a mixed correlation emerges under the influence of the fracture.
Xie et al. (Sun,) studied this question.