Pore space reconstruction using multiple-point statistics

摘要: The reconstruction of porous media is great interest in a wide variety fields, including earth science and engineering, biology, medicine. To predict multiphase flow through geologically realistic it necessary to have three-dimensional (3D) representation the pore space. Multiple-point statistics were used, based on two-dimensional (2D) thin-sections as training images, generate 3D space representations. method was borrowed from geostatistical techniques that use pixel-based representations reproduce large-scale patterns. Thin-section images can provide multiple-point statistics, which describe statistical relation between multiple spatial locations. Assuming medium isotropic, image be generated preserves typical patterns void seen thin sections. tested Fontainebleau Berea sandstones for micro-CT scanning are available. predicts long-range connectivity structures (measured by local percolation probability) better than standard two-point methods. selection key issue discussed detail.