Network model for the study of multiphase flow behavior in porous media

Abstract

In designing a displacement recovery process, it is desirable to have a method of predicting how the relative permeability, capillary pressure, re-sidual oil saturation and hence the displacement efficiency will be governed by the controllable variables, i.e., viscosity ratios, interfacial tensions, contact angles and flooding rates. This paper describes a simulation procedure in which a network model for the reservoir rock is generated by the Monte-Carlo technique, using pore size distribution and structural parameters appropriate to the particular reservoir. Applying existing hydrodynamic and thermodynamic knowledge about two-phase flow in pore spaces, the displacement mechanism in the porous medium is simulated. Progress of the flood is followed pore-space by pore-space over the entire network model and data on the effluents from the model boundary provide information on the performance. The model has been used to study displacement behavior in unconsolidated sand packs. The model results simulate the relative permeability behavior of porous media quite satisfactorily but rather semi-quantitatively.