Random Network Modeling of Tight Formations

Abstract

The objective if this study is investigating the capability of random network modeling techniques to predict fluid flow behaviour in unconventional formations. 3D random networks are constructed in order to represent Berea sandstone, synthetic oil-sand, and the Mesaverde formation located in north Wyoming, United States. The modified network modeling software solves the fundamental equations of single-phase and two-phase immiscible flow incorporating wettability and contact angle assuming a quasi-static displacement mechanism. Macroscopic properties network such as porosity, absolute permeability, and formation factor are calculated and compared to experimental data. The model is modified in order to implement the wetting phase trapping during the primary drainage process. A new methodology for reconstruction of digital porous media using experimental MICP data is introduced. Subsequently, immiscible two phase flow properties such as capillary pressure, relative permeability, and resistivity curves are predicted and compared to available experimental data.