Objectives - Image-Based Rock Physics (IBRP) simulation of petrophysical properties based on sub-micron to micron-scale images of very fine-grain rocks is constrained by the resolution and range of various imaging techniques used. Unlike some conventional sandstone and carbonate reservoir rock where a single Micro-Computed X-ray Tomography (μCT) volume images nearly all of the significant pores and pore throats, many low-permeability rock types contain phase regions with micro-pores and pore throats, including intergranular microcrack pores, that are not accurately resolved at the required μCT scale needed for a representative elementary volume (REV) for the whole rock. Properties for these regions are obtained at a finer-scale or using a different measurement method and these properties then assigned to the phase regions at the larger REV scale. This study explores the methodology involved in obtaining and assigning microcrack properties in μCT rock images and demonstrates a workflow to handle uncertainty in the location and properties of microcracks using two representative low-permeability sandstones.
Methods/Procedures/Process - The workflow combines μCT images of a mini-plug sample (~50mm3), which represents the rock REV, with Focused Ion Beam - Scanning Electron Microscopy (FIB-SEM) images (~200μm3) of regions of various types of observed microporosity (including intergranular microcrack pores) which occur within the REV sample. Different representative types of microporosity regions were imaged and properties calculated from the higher-resolution FIB-SEM image volumes. For some fraction of μCT microporosity regions, such as micro-fractures, their locations in the REV μCT sample was known but the micro-fracture properties were not known. A sub-resolution micro-fracture model was numerically constructed, honoring the mineral facies morphology and microporosity types assigned based on their respective distributions as observed in high resolution SEM images. Resultant porosity, capillary pressure and flow properties on the larger REV volume were cross-validated with independent core analysis measurements.
Results/Observations/Conclusions - This study illustrates a workflow for assigning properties, obtained at finer scales or using other measurement methods, to regions in the REV at larger scale but lower resolution. The resulting rock model produces the same porosity, permeability, and capillary pressure as core analysis measurements, and has the potential to predict relative permeability.
Applications/Significance/Novelty - It is expected that the majority of low-permeability rocks require an upscaling methodology similar to that developed in this study for IBRP computations and integration with core analysis. Using this methodology IBRP offers deeper understanding of building blocks of the upscaled-properties measured by core analysis. IBRP also offers the ability to measure/compute relative permeabilities that are nearly physically impossible to measure on core and the ability to construct digital rocks that allow evaluation of complete suites of rocks and their properties.