A Precise Pore-Structure Matching Method for Microfluidic Chips and Core Porous Media and Its Oil-Displacement Mechanisms

Abstract

Microfluidic chips are widely used to investigate pore-scale oil-displacement mechanisms, but the quantitative matching between chip pore structures and real reservoir cores remains insufficient. In this study, a quantitative pore-structure matching workflow was developed by integrating image processing, vector conversion, a self-developed MATLAB-based two-dimensional (2D) image editor, pore-structure characterization, microfluidic visualization experiments, and numerical simulation. Key parameters, including porosity, permeability, pore-throat radius, pore-throat ratio, and coordination number, were used to evaluate the matching between microfluidic chips and reservoir cores. The results show that the proposed method enabled pixel-level editing and synchronous regulation of pore-throat structures. Three chip models, including homogeneous high-permeability, homogeneous low-permeability, and heterogeneous models, were constructed, with a controllable permeability range of 200–1000 mD. By combining image analysis, core fluorescence analysis, constant-rate mercury intrusion, and COMSOL simulation, the matching degree between the chip models and real core pore-structure parameters exceeded 90%, indicating that the chips effectively reproduced the main pore-throat characteristics of the target reservoir. Visualization oil-displacement experiments further demonstrated that 15–30 μm preformed particle gel (PPG) particles achieved effective pore-throat matching through selective plugging, flow diversion, and deep profile control. The adaptive plugging, profile-control, and displacement system significantly improved sweep efficiency in the heterogeneous model, with recovery factors of 95.03% and 85.7% in the high- and low-permeability layers, respectively, higher than those obtained by weak-alkali ASP flooding. Numerical simulations identified the optimal displacement parameters as a viscosity of 70 mPa·s and an interfacial tension of 0.03 mN/m. This study provides a quantitative method for designing representative microfluidic chips and offers experimental and numerical support for pore-scale EOR mechanism analysis.