Image-Based Upscaling of Permeability in Opalinus Clay

This paper presents an approach for upscaling permeability in Opalinus Clay from pore scale to a coarser scale, where pores cannot be resolved and where the transport properties are controlled by the geometry of the clay matrix considered as a distinct phase. Microstructures reconstructed from multiscale image data are cellular models with different sizes of basic building blocks, the so-called voxels. Flow simulations were performed on the basis of the grid that is naturally inherent in the coarser-scale image data that were acquired by synchrotron X-ray computed tomography. Thereby, the spatial distribution of permeability at this coarser scale was determined from image data at the pore scale acquired by focused ion beam nanotomography. Then, coarse-scale microstructures with different clay matrix contents were used as input for flow simulations, which allowed predicting the vertical and horizontal bulk permeability at the mesoscale. In agreement with results of other workers, it turned out that vertical (cross-plane) permeability K-v decreases with increasing clay content, which can be explained in that the presence of coarse nonclayey grains increases permeability of the porous clay matrix when compared to a matrix consisting only of fine-grained shales. Mesoscale permeability anisotropy K-h/K-v increases from about 1 to 2 if the clay content increases from 0.2 to 0.9. The predicted permeability behavior is discussed and compared to results from experiments and other predictive models.