Numerical Modeling of Nmr Surface Relaxation in Porous Media with Improved Pore Surface Area Evaluation and Derivation of Interfacial Absorption Probability

In digital rock physics, surface relaxation of NMR decay on a 3D porous media is often simulated with Random-Walker methods, where Brownian walkers representing the magnetized nuclei are absorbed by the solid surface under an interfacial absorption probability. Previous simulations usually neglected the blocky pixel appearance of 3D pore space images, which results in over estimation of the surface area, particularly for unconventional reservoir rocks with large fractions of micropores, such as tight sandstones or shale. This leads to arbitrary adjustments of the interfacial absorption probability by tuning the value of the surface relaxivity strength (SRS) in order to match simulation results with experimental measurements. However, this arbitrary adjustment violates the fact that SRS is a rock dependent physical property. In this study a new interfacial absorption probability for NMR simulations is presented that honours the physical principles. This new absorption probability computation depends on accurate evaluation of the surface area, which is assured by a new approach of 3D pore space surface area evaluation. The new algorithm is verified by performing NMR Random-Walker simulation tests on both standard geometries and realistic 3D heterogeneous porous media. INTRODUCTION In Random-Walk NMR simulation the decrease of random walkers, which simulates the