Pore structure characterization through path-finding and Lattice Boltzmann simulation

Characterization of underground porous media parameters at the micro and macro scales is fundamental in geo- sciences. A thorough comprehension of flow phenomena requires analyses and observations at the micro scale through adoption of micromodels as representative as possible of real geological formations. In this paper we focus on the analysis of 2D binary images of real rock thin sections to characterize pore network geometry and to estimate effective porosity, pore size distribution and tortuosity with the aim of providing suitable information for designing micromodels. To this end, a geometrical analysis of the pore structure, based on the identifica- tion and characterization of the set of the shortest geometrical pathways between inlets and outlets pairs, was implemented. The geometrical analysis is based on a path-finding algorithm derived from graph theory. Results provided by geometrical analysis were validated against hydrodynamic numerical simulation via the Lattice Boltz- mann Method (LBM). Results show that the path-finding approach provides reasonable and reliable estimates of tortuosity and can be successfully applied for analyzing the distribution of effective pore radius, as well as for estimating the effective porosity.