Summary Gravel is a common coarse-grained (2 mm) detrital component in Cretaceous Yageliemu Formation (K1y) sandstone reservoirs of Tarim Basin. Although its resistivity is similar to the sandstone, well logs often show abnormally high resistivity compared with conventional sandstone reservoirs with similar porosity, which affects saturation and fluid interpretation. Limited core samples make rock electrical test and computed tomography (CT)-based digital rock analysis insufficient to clarify how gravel increases the resistivity. The traditional 3D quartet structure generation set (QSGS) method can model gravelly sandstone structures, but it frequently exhibits undergrowth. Therefore, this study introduces an improved QSGS method by incorporating spatially weighted seed generation and tensor-based directional growth probability, with a dynamic volume fraction feedback mechanism to construct research models. Resistivity forward calculation using the finite element method is performed, and simulated resistivity is calibrated with laboratory data. The effects of gravel on microscopic pore structure parameters and rock electrical parameters are evaluated, and then a multiparameter coupled variable rock electrical model and porosity-segmented resistivity correction model are developed. An improved saturation model for gravelly sandstone reservoirs is further proposed. The results show that increasing gravel content and diameter raises resistivity, tortuosity, and pore-throat ratio while reducing coordination and connectivity, indicating that the abnormal resistivity is primarily controlled by gravel-induced complex pore structure. The cementation and saturation exponents are positively correlated with gravel content, diameter, tortuosity, and pore-throat ratio but negatively correlated with coordination and connectivity. At macroscale, the cementation exponent decreases with increasing average pore-throat radius, whereas the saturation exponent increases with the T2 cutoff. The improved saturation model enhances gas saturation inversion accuracy by 13.35% compared with the traditional Archie’s model, significantly improving saturation evaluation in deep gravelly sandstone reservoirs.
Lin et al. (Mon,) studied this question.