Resistivity and permeability anisotropy measured in laminated sands via digital core analysis

Many aeolian sandstone reservoirs contain significant volumes of recoverable hydrocarbons in intervals where the average lamina thickness is well below the resolution of any logging tool. The variability in petrophysical properties of the laminations increases uncertainties and in turn can lead to an underestimation of the hydrocarbon in place. To date estimates of the Archie exponents m and n in thinly laminated sand reservoirs have been based on simplified model structures. Here we illustrate an ability to visualize the anisotropy in aeolian sands at the pore scale via digital microtomographic imaging, and to measure the anisotropy in transport properties via direct calculation on the resultant images. In this study, 3D pore scale imaging of an aeolian core plug exhibiting fine scale laminae (laminations at the mm scale) is undertaken via high resolution micro-CT. The full 3D image is obtained at 5.6 micron resolution. The composite image is made up of a 2000 squared voxel cross section (7.5 mm squared) parallel to the bedding planes and a continuous 3 cm length perpendicular to the bedding plane (6,000 voxels). Strong variation in lamina porosity is observed along the length of the core and more than 30 distinct bedding planes are evident. Petrophysical calculations are performed on a 20003 subset exhibiting over 10 laminations. Permeability and resistivity are derived for individual lamina including estimation of Archie's cementation exponent m and saturation exponent n. A composite permeability and m and n are then calculated both parallel and perpendicular to the bedding planes across varying numbers of lamina. These upscaled permeabilities are in good agreement with experimental core data. The values of m and n are found to strongly depend on the relative volume fractions of the different laminae and the orientation of the conductivity measurements. Estimates of m parallel and perpendicular to the bedding planes based on simple averaging are in poor agreement with laboratory measurements for the same core material. Predictions of m for the multi-layered system based on idealized layering (Kennedy and Herrick, 2003) give good estimates. The resistivity exponent n based upon the imaged porosity increases when averaged across multiple laminations both perpendicular and parallel to the bedding plane. This is in contradiction to laboratory measurements. This highlights the need to include the contribution of pores below the current image resolution to accurately calculate the resistivity index. We use a radiographic method to estimate the total unresolved porosity of the sample across fine and coarse grained laminae. Including this estimate of the unresolved porosity to the formation resisitivity index leads to a realistic estimate of n.