New Method of Generating and Clustering Pore Network Model

Document Type : Original Article

Authors

1 Petroleum Engineering, Chemical & Petroleum Engineering, Sharif, Tehran, Iran

2 petroleum engineering, chemistry and petroleum engineering, Sharif, Tehran, Iran

Abstract

In this study, a method for generating a pore network model based on porous media static characteristics has been presented. The method with the aim of pore size, coordination number distribution and the porosity is developed to generate an irregular pore network model. Because the model uses probability functions, it has different answers. In order to optimize model, porosity is used as target function. This irregular model has all of the desired static properties such as pore size, coordination number distribution and porosity. The model coordination number can vary from 0 to 26. An effective pore network is extracted using a new clustering algorithm. Finally, the results of an extracted pore network from the CT scan image of a synthetic silica achieved from maximal ball algorithm were used as the model inputs. The outcome of the developed model was compared with the mentioned model and statically and structural agreement was found.

Keywords