To see a rock in a grain of sand

Permeability is perhaps one of the most important yet elusive reservoir properties. Unlike porosity, it correlates poorly with the elastic properties of rocks and, as a result, cannot be mapped remotely. Rather, it has to be directly measured in the lab on core plugs extracted from a well. In such lab experiments, a pressure gradient is applied to a fluid-filled rock sample, the fluid's volume flux is recorded, and the absolute permeability is calculated from Darcy's law by relating the flux to the pressure gradient. Physical experiments may be augmented or even partially replaced by numerical experiments provided that a numerical simulation accurately mimics the physical process. Successful numerical “measurements” of permeability have been reported by Bakke and Oren (1997), Keehm et al. (2001), Knackstedt et al. (2004), and others. It appears that the absolute (and even relative) permeability can be recovered with reasonable accuracy from 3D micro-images of the pore space where the geometrical features of the pore-space structure relevant to fluid flow are carefully resolved and digitally represented in the computer. Such images are available from high-resolution X-ray tomography, also known as CT-scanning. CT-scan produces multiple closely spaced 2D slice-images of a rock that, when stacked together, represent the 3D volume. However, high-resolution scanning devices are still prohibitively expensive and the scanning time is too long to be practically useful in massive numerical experimentation. An alternative is a statistical reconstruction of 3D volume from a 2D slice. A flat thin section, where the pore space appears in a single color because it is impregnated with dyed epoxy and is easily distinguishable from the grains (Figure 1), can be used for this purpose. Thin sections are relatively easy and cheap to prepare either from core plugs or cuttings. Figure 1. A sandstone thin section. The pore space is filled with epoxy …