Analysis of Rock Microstructure Using High- Resolution X-ray Tomography

In previous work we have described and validated robust techniques for partitioning the pore space of a porous material into simple regions, thereby allowing it to be represented by a network of simple building blocks. The primary purpose of this analysis is the generation of a pore-throat network for modelling immiscible multi-phase fluid displacements. However, combining our partitioning algorithms with recently developed methods for studying complex and disordered networks, and geometric studies of the building blocks themselves, results in powerful tools for characterising rock microstructure. We extend this approach by applying the analysis to both the pore space and its complement, the grain space. Partitioning the grain space has particular meaning for clastic rocks where it is equivalent to identifying individual grains. Studying the two interpenetrating networks that result gives us a more complete description of the material, and in particular, gives us the chance to study causal relationships linking microstructure to macroscopic properties. We apply this methodology to a selection of clastic and carbonate rock images from the library of samples imaged at the ANU X-Ray micro-CT facility. For the clastic samples, we are able to assess whether differences in porosity and connectivity are the result of variations in grain size, grain shape or packing efficiency. For all samples, we look at the robustness and usefulness of several characterisation measures.