We consider a method for the determination of the starting pressure gradients of oil and water in two-phase flow in low-permeability reservoirs. Moreover, we obtain the relative permeability curves of oil and water according to water saturation for the interpretation of oil-water two-phase flow.

In the case of two-phase Darcy flow in reservoirs, in order to obtain the relative permeability curves of oil and water according to water saturation, the JBN (Johnson, Bossler and Naumann) method is improved universally which is based on the analysis of data of two-phase unsteady displacement. The method may be improved in the case of low velocity non-Darcy flow in the low-permeability reservoirs, too. Therefore, in the context, we present a new method by which we can determine the starting pressure gradient of oil and water in two-phase non-Darcy flow in the low-permeability reservoirs. Based on these starting pressure gradients, we obtain relative permeability curves of oil and water according to water saturation by the process of data from two-phase unsteady displacement experiment using the Genetic Algorithm. In order to do for this, when oil and water flow simultaneously, the water content function in the condition of oil-water two-phase flow is obtained from the equations of motion of oil and water, and the equation of determining relative permeabilities using experimental data is presented. Based on this, the starting pressure gradients of oil and water are determined using search method of the Genetic Algorithm with powerful self-adaptability, and the curve of the relative permeability is plotted.

The method can be improved effectively to determine the starting pressure gradients and relative permeabilities of oil and water in two-phase seepage in the low-permeability reservoirs and to study the numerical simulation and the reservoir engineering in these reservoirs.