Rock-physics modeling by using particle swarm optimization algorithm

Abstract

Rock physics modeling is an important method used to study reservoir characteristics in the seismic method. Optimum rock physics modeling must consider the research area and the theory of the effective medium used. Rock physics modeling in real cases requires mineralogical composition data from XRD data and mineral elastic modulus values which are generally taken from the literature. However, XRD data is not continuous in every depth, as well as mineral's elastic modulus information which has different characteristics in each research area, so using literature data can cause inaccuracies. This study aims to estimate rock physics parameters such as elastic modulus of minerals and rock pore shape using the Particle Swarm Optimization (PSO) to solve this problem. In this study, Kuster-Toksöz's effective medium theory and Gassmann's relation were used to produce the effective moduli of saturated rock. There are five forward and inverse modeling scenarios in this experiment. The first scenario is a combination of Kuster-Toksöz and Gassmann's relation. The second and third scenario utilizes the Kuster-Toksöz equation with penny cracks and sphere inclusion. The fourth scenario is carried out only using the Gassmann relation, and the fifth scenario is cross modeling with different forward and inverse algorithms. The results of this study show good fitting of saturated rock's elastic moduli. The Kuster-Toksöz scenario with single inclusion best estimates rock physics parameters.