Title: Note on Using Singular Value Decomposition to Solve for Single-Phase Flow and Identify Isolated Clusters in Heterogeneous Pore Networks

Abstract: Diluted pore networks or networks that are obtained directly from image analysis are commonly used to evaluate flow through complex porous media. These networks are often subjected to singular points and regions, that is, isolated sub-networks that are not connecting the inlet and outlet. These regions impair the invertibility of the modified Laplacian matrix and affect the accuracy and robustness of pore network models. Searching and eliminating these singularities is commonly done by preconditioning or searching algorithms like breadth-first search or depth-first search. Here, we propose using singular value decomposition (SVD) to simultaneously solve for single-phase flow and locate isolated regions. We demonstrate the performance of the method for networks obtained from images of a partially saturated medium, and for diluted networks close to the percolation threshold.