Abstract
Micro-CT scans provide visualization of core samples that can be used to quantify physical properties important for subsurface flow including, as examples, porosity and pore size distribution, pore connectivity and permeability, amongst many others. 3D scans can deliver a very high level of detail, but at a cost of many hours of scan time. This also implies enormous numerical datasets and associated computational processing load. It is therefore important to understand (1) the voxel resolution that is required to retain physical structure fidelity (e.g., pore size and tortuosity) and (2) the smallest sample size that provides reliable and representative transport calculations (e.g., directional permeability and connectivity). As an application example, we analyze a 3D sandstone sample from the Chinchilla area of the Surat Basin comprising a cuboid of 6.6 mm × 6.6 mm × 12.276 mm. The CT image at 6.6 μm voxel resolution generates 1.86 billion grid elements. The work then examines the effect on physical properties and flow simulation of coarsening the voxel resolution, and reducing the physical size of the sample. For each of the different cases relating to voxel size, the major physical properties are evaluated. For different sample sizes at constant (small) voxel size, the fluid flow is modeled using a lattice Boltzmann method (LBM) on multicore supercomputers. The LBM is efficient and convenient for application in parallel computing, which is required to meet the computational resources needs of 3D simulations with large datasets. Comparison of the results obtained at different resolution scales reveals the minimum representative elementary volume (REV) for voxels to characterize the physical features of porous media, and for sample sizes that effectively eliminate boundary effects, capturing and describing the transport phenomena that underpin reservoir analysis.