From Microstructure to Macroscopic Flow: FEM-Based Multiscale Permeability Prediction via Digital Rock Models

Abstract

Accurate prediction of fluid flow in subsurface reservoirs is critical for optimizing hydrocarbon recovery, particularly in heterogeneous formations where microscopic rock structure dictates macroscopic behavior. Traditional laboratory methods often struggle to capture the complex spatial variations inherent in these rocks. To address this problem, this study establishes an integrated digital rock physics framework that bridges the gap between microscale pore geometry and macroscale permeability. The research uses high-resolution X-ray computed tomography (CT) to image sandstone samples, followed by advanced image processing to reconstruct a high-fidelity three-dimensional digital rock model. From this model, multiple representative elementary volumes (REVs) are extracted to rigorously capture the internal structural heterogeneity. The methodology employs the FEM to solve the incompressible Stokes equations directly on the complex pore geometry, simulating fluid dynamics at the pore scale without relying on empirical simplifications. The results reveal significant variations in pore connectivity across the sample. Flow simulations demonstrate that permeability is not solely a function of porosity but is critically controlled by the topology of the pore network. Specifically, the study identifies distinct flow behaviors where well-connected high-permeability zones act as preferential flow channels, whereas adjacent tight zones serve as flow barriers. The numerical results are rigorously validated against experimental data, achieving a close agreement with an average deviation of approximately 11.3%, confirming the reliability of the computational workflow. These findings highlight that strong microscale heterogeneity is the root cause of macroscopic channeling issues, such as early water breakthrough during water flooding operations. Consequently, this work provides a robust theoretical basis for designing optimized reservoir management strategies, such as profile control and targeted stimulation, to improve sweep efficiency and maximize economic recovery in complex sandstone reservoirs.