Investigation of three-phase counter-current flow using X-ray computerized tomography and neuro-simulation modeling

This work examines the physics of three-phase counter-current fluid flow driven by gravity and capillary transport mechanisms through a combination of experimental measurements and neuro-simulation modeling. There have been only a few scientific reports in the literature addressing counter-current flow phenomena, especially under three-phase flow conditions. Counter-current flow driven by gravity and capillarity may occur in fractured reservoirs (both in matrix and fractures), in solution gas drive and gas cap reservoirs with high vertical permeability, and in some water-flooding operations. Three-phase flow experiments were conducted with an idealized system and the temporal and spatial saturation distributions during counter-current flow periods were acquired using computerized tomography. The experimental results were simulated using a conventional reservoir simulator. The simulation results provided the basis for training an artificial neural network through which the best relative permeability and capillary pressure functions that match the experimental data were obtained using a pattern recognition protocol.