Numerical computation of elastic properties for porous rocks based on CT-scanned images using direct mapping method

Abstract

Abstract In recent years, computed tomography (CT) technologies have been applied to reconstruct reservoir rocks, in digital format, from CT-scanned images in oil-exploration- and production-related research and practice. The geometric, compositional, and mechanical properties of rock samples can be obtained from their digital counterparts through simple analysis or more involved computation. The geometric properties, such as size, distribution, and connectivity of pore space, and compositional properties, such as mineral and organic matter, can be determined rather accurately from digital rock without much difficulty. However, the computation of the mechanical properties for rock samples from their digital counterparts only became possible in very recent years because of the rapid rise of computer power. In this paper, idealized porous rock, synthetic rock, and real rock are employed to test the reliability of the predictions from numerical simulation. The meshing insensitivity of computing rock mechanical properties in digital rock is first demonstrated in a set of simple numerical simulations on idealized porous rocks, for which the computational meshes are generated arbitrarily and the numerical simulation results are compared against analytical solutions. For synthetic rock and Berea sandstone, the computational models are created through direct mapping of their CT-scanned images, and the numerical simulation results are checked with laboratory experimental data. This work confirms that the digital rock methodology has potential to become an accurate tool for measuring elastic mechanical properties for reservoir rocks.