Fluid flow visualization and resistivity variations during gas-water displacement in carbonate rock: Implications for reservoir evaluation

Carbonate reservoirs exhibit strong pore structure heterogeneity, leading to complex electrical properties and posing challenges for fluid saturation evaluation through well logging. To clarify the influence of fluid distribution heterogeneity on resistivity responses, a micro-etched glass model replicating real pore structures based on CT scans of core samples was developed. Gold-plated strips were integrated to enable simultaneous visualization of fluid distribution and resistivity measurement. Two sets of two-phase flow experiments were conducted: gas displacing water and water displacing gas. The results show that injected fluids preferentially form dominant flow paths through large pore-throats, while small pore-throats tend to trap residual fluids. During gas displacing water, the initially preserved water in small pores maintains a conductive network, resulting in a slow increase in resistivity. As gas invades the small throats, the conductive network breaks down, leading to a sharp rise in resistivity and a two-stage characteristic response, which reduces the applicability of Archie’s equation. Even at the same saturation level, different displacement methods (conductive phase displacing non-conductive phase vs. non-conductive phase displacing conductive phase) lead to significant differences in fluid distribution and resistivity behavior, indicating that pore structure and displacement history jointly influence electrical properties. Therefore, accurate evaluation of saturation in carbonate reservoirs requires comprehensive consideration of pore structure and displacement conditions, with interpretation models adjusted according to development stages (e.g., initial accumulation vs. water flooding). This study is limited by the use of a 2D model and ambient temperature and pressure conditions, which differ from the high-temperature and high-pressure conditions of actual reservoirs. Thus, the findings primarily offer qualitative insights and are not directly suitable for quantitative application.