Effect of Geometry on Pore-Scale Simulation of Single-Phase Fluid Flow

Abstract

Accurate estimation of physical rock properties is a crucial step in the process of designing a reservoir model for development reasons and to optimize hydrocarbon production. Absolute permeability is one of those properties. Digital rock physics have evolved as a promising new technology to complement traditional laboratory experiments on core plugs. Background and theory on digital rock physics regarding numerical predictions of absolute permeability through single-phase incompressible fluid flow are presented. In this thesis, two-dimensional pore structure images of different geometries are created with the programming language Python to study fluid flow behavior in different pore structure geometries. The material simulator GeoDict is used to simulate single-phase incompressible fluid flow in the different pore geometry images. The work in this thesis a theoretical approach and the motivation is to look at absolute permeability behavior as the pore throat radius is decreased relative to the pore radius. The ratio of pore radius to pore throat radius is defined as the F ratio. Among the pore geometries studied in this thesis, are: circular, triangular, diamond and ellipsoidal pores. It is observed that the numerically predicted absolute permeability versus increasing F ratio for all pore structure geometries follows a power function trendline. In addition, it is found that triangular pores show the largest variations in predicted permeability as the F ratio varies. On the other hand, the highest permeability with least variation is observed for ellipsoidal pores when fluid flows in the direction of the largest ellipsoidal radius. Interpretation of numerically predicted permeability results are based on the theoretical approach of this study. For future studies, investigation of permeability behavior in geometries that are comparable to real rocks can be studied. Thus, the results are more applicable to fluid flow problems in realistic porous media. In addition, it is suggested that permeability behavior of more complex pore geometries should be studied.