New Method of Pore Network Extraction from 3D CT scan Images of Rock

Reservoir rock geometry is one of the important factors in the analysis of static and dynamic properties. CT scan images are used to obtain the reservoir rock geometry. Because these images are not easy computable, the pore-throat network models are used to geometrical conversation. The process of extracting these models from CT images is done using statistical probabilistic image processing based methods, such as the modified maximal ball algorithm. In this study, new methods for extracting the pore network model and pore connection detection are presented. Throat spanning probability is a criterion for determining the presence of connections. In this method, the geometric static parameters of the rock, including the porosity, the porosity-based homogeneity index of rock, the distribution, and the average pore size& coordination number are calculated. Finally, using the image of a synthetic silica sample, the results of the proposed method are compared with the results of the modified maximal ball (MMB) method. Good agreement was found between the results of the models. Also, the pore network extracted from sandstone, carbonate and synthetic rocks and corresponding porosity-based homogeneity index are presented and discussed.