Sharif Digital Repository

In this study, coupled Generative Adversarial and Auto-Encoder neural networks have been used to reconstruct realizations of three-dimensional porous media. The gradient-descent-based optimization method is used for training and stabilizing the neural networks. The multi-scale reconstruction has been conducted for both sandstone and carbonate samples from an Iranian oilfield. The sandstone contains inter and intra-grain porosity. The generative adversarial network predicts the inter-grain pores and the auto-encoder provides the generative adversarial network result with intra-grain pores (micro-porosity). Different matching criteria, including porosity, permeability, auto-correlation... 

Fluid flow behavior in a porous medium is a function of the geometry and topology of its pore space. The construction of a three dimensional pore space model of a porous medium is therefore an important first step in characterizing the medium and predicting its flow properties. A stochastic technique for reconstruction of the 3D pore structure of unstructured random porous media from a 2D thin section training image is presented. The proposed technique relies on successive 2D multiple point statistics simulations coupled to a multi-scale conditioning data extraction procedure. The Single Normal Equation Simulation Algorithm (SNESIM), originally developed as a tool for reproduction of...