Resolution Effect on Image-Based Conventional and Tight Sandstone Pore Space Reconstructions: Origins and Strategies

Abstract A suitable choice of spatial resolution is essential to the three-dimensional (3D) image-based reconstruction of rock pore space: a too low spatial resolution may result in the omission of sub-resolution porosity while too high spatial resolution may lead to a field of view smaller than the representative volume. In this work, the 3D pore spaces of conventional and tight sandstone specimens are reconstructed by X-ray computed tomography with spatial resolutions of 2 μm, 5 μm, 10 μm, and 20 μm and equal image voxel quantities (600×600×600) to study the influence of spatial resolution on the reconstruction quality. Petrophysical properties such as porosities, permeabilities and pore-throat structure parameters are calculated from the 3D images and measured by helium porosity, Klinkenberg-corrected permeability and mercury intrusion capillary pressure porosimetry for comparison. High-resolution images are sampled from different positions in the specimens to evaluate the core-scale heterogeneities. Rock textures and pore-throat structures of the specimens are analyzed through casting thin sections and MICP to find out the origin of the resolution effect. The results show that 3D images with any single spatial resolution have limited detectivities to recover the full pore and throat distribution of sandstone specimens, which is the root cause of the resolution effect. The contributions of different pore genetic types with various pore sizes to petrophysical properties vary with simulation targets and methods. For conventional sandstone specimens with only one dominate pore genetic type, the appropriate spatial resolution can be chosen by calibrating the simulated petrophysical properties to actual measurements. For tight sandstone specimens with several pore genetic types, the contents and contributions to the petrophysical properties should be investigated before pore space reconstructions.