Summary
Complex pore geometry and composition, as well as anisotropic behavior and heterogeneity, can affect physical properties of rocks such as electrical resistivity and dielectric permittivity. The aforementioned physical properties are used to estimate in-situ petrophysical properties of the formation such as hydrocarbon saturation. In the application of conventional methods for interpretation of electrical-resistivity (e.g., Archie's equation and the dual-water model) and dielectric-permittivity measurements [e.g., complex refractive index model (CRIM)], the impacts of complex pore structure (e.g., kerogen porosity and intergranular pores), pyrite, and conductive mature kerogen have not been taken into account. These limitations cause significant uncertainty in estimates of water saturation. In this paper, we introduce a new method that combines interpretation of dielectric-permittivity and electrical-resistivity measurements to improve assessment of hydrocarbon saturation. The combined interpretation of dielectric-permittivity and electrical-resistivity measurements enables assimilating spatial distribution of rock components (e.g., pore, kerogen, and pyrite networks) in conventional models.
We start with pore-scale numerical simulations of electrical resistivity and dielectric permittivity of fluid-bearing porous media to investigate the structure of pore and matrix constituents in these measurements. The inputs to these simulators are 3D pore-scale images. We then introduce an analytical model that combines resistivity and permittivity measurements to assess water-filled porosity and hydrocarbon saturation. We apply the new method to actual digital sandstones and synthetic digital organic-rich mudrock samples. The relative errors (compared with actual values estimated from image processing) in the estimate of water-filled porosity through our new method are all within the 10% range. In the case of digital sandstone samples, CRIM provided reasonable estimates of water-filled porosity, with only four out of twenty-one estimates beyond 10% relative error, with the maximum error of 30%. However, in the case of synthetic digital organic-rich mudrocks, six out of ten estimates for water-filled porosity were beyond 10% with CRIM, with the maximum error of 40%. Therefore, the improvement was more significant in the case of organic-rich mudrocks with complex pore structure. In the case of synthetic digital organic-rich mudrock samples, our simulation results confirm that not only the pore structure but also spatial distribution and tortuosity of water, kerogen, and pyrite networks affect the measurements of dielectric permittivity and electrical resistivity. Taking into account these parameters through the joint interpretation of dielectric-permittivity and electrical-resistivity measurements significantly improves assessment of hydrocarbon saturation.
A New Method for Pore Structure Quantification and Pore Network Extraction from SEM Images