Waterflood Relative Permeabilities in Composite Cores

Abstract

JPT Forum When cores longer than about 3 in. are desired for waterflood testing, it is common to assemble a composite core from a set of short core pieces. Long composite cores reduce the importance of capillary effects such that lower flooding rates can be used. Composite cores are also useful where end effects would invalidate results from tests on individual short cores. The assumptions made in using such a composite core for waterflood testing are that all the core sections:have nearly identical relative permeability curves,are homogeneous and isotropic, andhave nearly identical connate water and residual oil values. It is further assumed that there is good contact between sections and that the flooding rate is high enough to make capillary pressure effects unimportant in all core sections. When these assumptions are met, we can make several assertions based on analysis of the one-dimensional, two-phase flow equations. The water-saturation distributions throughout a flood are the same in a composite core as in a single homogeneous core having the average properties of the composite, no matter what ordering of core sections is used in the composite. Production histories are also the same for any ordering of core segments. Thus experimental waterflood relative permeability ratios may be quite accurate for any ordering. The individual-phase relative permeability curves calculated from composite waterflood data may be in error.