Pore to pore validation of pore network modelling against micromodel experiment results

摘要

Pore network modelling (PNM) has been widely used to study the multiphase flow and transport in porous media. Although a number of recent papers discussed the PNM validation on core-scale parameters such as permeability, relative permeability and capillary pressure; quantitative predictive potential of PNM on pore by pore basis has rarely been studied. The aim of this paper is to present a direct comparison between PNM simulations and corresponding micro-model experiments at the same scale and the same geometry. A number of well-defined and constrained two-phase flow in porous medium experimental scenarios were utilized to validate the physics solving part in PNM (filling rules, capillary and viscous pressure). This work validates that a dynamic pore network flow solver can predict two-phase flow displacements for these experiments for drainage situations at both pore and plug scales. A glass-etched micro-model is used to quantify the accuracy of a dynamic PNM solver on pore and core levels. Two-phase drainage micro fluidic experiments at different flow conditions are performed on micro-models. PNM simulations are performed on the same pattern and flow conditions as used in micro-model experiments. The two-phase distribution extracted from experiment images is registered onto rsults of PNM simulations for direct pore to pore comparison. Pore-scale matching level is found at around 75 % for all three test cases. The matching level of core-scale parameters such as S (w c) and oil-phase permeability varies from case to case; the relative error to micro-model experiment measurements varies from 15 to 60 %. Possible reasons leading to discrepancies on core-scale parameters are discussed: missing considerations during validation of the combination of uncertainty in both simulator input parameters and experiments are seen as the principal factors.