to express required ct-scan resolution for porosity and saturation calculations in terms of average grain sizes

Despite advancements in specifying 3d internal microstructure of reservoir rocks, identifying some sensitive phenomenons are still problematic particularly due to image resolution limitation. discretization study on such ct-scan data always has encountered with such conflicts that the original data do not fully describe the real porous media. as an alternative attractive approach, one can reconstruct porous media to generate pore space representations. the reconstructed structures are then used for simulations using some sort of discretization. in this paper, it is examined the effect of discretization on porosity and saturation calculations in porous media models. some 3d boolean models of random overlapping spheres of fixed and variable diameters in three dimensions are used. the generated models are then discretized over 3d grids with different number of voxels. the porosity can be calculated and saturation of the discretized models are then compared with the analytical solutions. the results show that when meshgrid sizes are 8% of smallest grains, porosity is calculated with 95% precision. in addition to that, meshgrid sizes of 5% and 3% of average grain diameter are adequate to calculate non-wetting and wetting phase saturations with at least 95% precision. this helps in choosing the optimum voxel size required in imaging for efficiently use of available computational facility.