<p>Sandstone-type uranium deposits have complex pore structures that act as conduits for leaching solutions during in-situ leaching (ISL), and their geometry strongly influences fluid transport and leaching performance. Key parameters such as porosity, pore size distribution, and pore connectivity determine lixiviant migration paths and reaction interfaces. Accurate characterization of these parameters, followed by three-dimensional (3D) reconstruction and numerical simulation of fluid flow, is essential for improving ISL efficiency. In this study, micro-computed tomography (micro-CT) was combined with software platforms including Avizo, Dragonfly, and COMSOL to establish a workflow for quantitative pore structure analysis, multi-scale pore identification, and flow simulation. Quantitative characterization yielded a total porosity of 16.20%, an effective porosity of 14.90%, and only minor fluctuations in layer-by-layer interfacial porosity variation. Connected pores comprised 91.98% of total porosity, indicating pronounced heterogeneity yet good connectivity. Deep learning-based segmentation in Dragonfly successfully identified multi-scale pores, resulting in a total porosity of 16.66%, with intra-particle and inter-particle porosities of 0.25% and 16.41%, respectively, closely matching Avizo’s threshold segmentation. Flow simulations under three injection concentrations (1, 2, and 3&#xa0;mol·L^− 1) consistently revealed preferential flow channels and stagnant zones, with both maximum velocity and maximum pressure decreasing as injection concentration increased. The proposed workflow converts high-resolution 3D pore structures into structural parameters and permeability tensors for sandstone-type uranium deposit modeling. By linking pore-scale features with flow patterns, the workflow provides a framework to identify preferential flow paths and low-velocity zones, thereby enhancing understanding of seepage behavior and supporting more effective monitoring design and operational optimization.</p>