MR41D-2671
Transport Properties of Carbonate and Sandstone Samples: Digital Rock Physics and Laboratory Measurements

Thursday, 17 December 2015
Poster Hall (Moscone South)
Abrar Ahmed Alabbad1, Jack P Dvorkin1 and Stanford Rock Physics , (1)Stanford University, Stanford, CA, United States
Abstract:
We examined six carbonate samples that included three pairs, each pair cut from the same core, normal and parallel to the bedding. We also examined three sandstone samples comprised of a pair, cur normal and parallel to the bedding, and one sample cut normal to the bedding. For each of these samples, we obtained dual energy digital images with the resolution approximately 0.004 mm, coarser than the pore-scale resolution. As a result, we did not resolve the pore structure. Still, these images provided us with the bulk density (ρb) and photoelectric factor (Pf) at each voxel in 3D. The Pf volumes were used to estimate the mineralogy at each voxel by partitioning the carbonate mineralogy between calcite and dolomite and partitioning the sandstone mineralogy between pure quartz and “dirty sandstone” (Plumb et al., 1991) for one scenario and between pure quartz and illite for the other scenario. From this mineralogy we obtained the grain (matrix) density (ρs) at each voxel. Next, by using ρb and ρs and assuming that the pores were filled with air, we computed the total porosity (ϕt) at each voxel from mass balance. Porosity thus obtained was used to estimate the electrical formation factor (F) at each voxel by assuming that F relates to ft according to Archie’s equation (Archie, 1942). We also computed the absolute permeability (k) at each voxel by assuming that k relates to ϕt according to the Kozeny-Carman equation (Carman, 1956). Next, by employing a Darcy simulator, these 3D resistivity and permeability volumes were used to compute the effective permeability and electrical formation factor of the whole samples along the three axial directions to assess the anisotropy of these transport properties. These computational results were compared to laboratory measurements. The computed effective bulk density, grain density, and porosity appeared to closely match the laboratory values. So did the formation factor. By selecting an appropriate grain size, we also achieved a match between the computed and measured permeability, although not as robust as for the electrical formation factor. Both the computational and laboratory results indicated that all samples were practically isotropic. This study is an example of using coarse resolution digital images to estimate the effective transport properties of rock.