Appropriate relative permeability curves for two-phase flows through subsurface fractures remain unclear. We have conducted decane-water and nitrogen-water two-phase flow experiments and simulations on real variable-aperture fractures in rocks under confining stress. Experiments have been conducted on fractures for different combinations of rock type (granite or limestone), wettability (contact angle of water: 0° or 90°), and intrinsic fracture permeability (10-11 m2 or 10-10 m2) using different combinations of shear displacement (0 or 1 mm) and effective confining stress (1 or 40 MPa). It has been demonstrated that nonwetting phase relative permeability depends on capillary pressure, except at either a higher contact angle or higher intrinsic permeability (i.e., bigger aperture), where no influence of capillarity is expected from the Young-Laplace equation. In the absence of an influence of capillarity, relations between wetting and nonwetting phase relative permeabilities agree with that of the X-type relative permeability curves. In order to determine the relative permeability curves under the influence of capillarity, the experimental results have been analyzed by two-phase flow simulations of the aperture distributions of the fractures. It has been revealed that nonwetting phase relative permeability becomes zero, even at a small wetting phase saturation of approximately 0.3, while wetting phase relative permeability exhibits Corey-type behavior, resulting in ν-shaped relative permeability curves. Similar curves have been reported in the literature, but have not been demonstrated for real fractures. It has been revealed that the new ν-type and traditional X-type relative permeability curves are appropriate for describing two-phase flows through subsurface fractures.
ASJC Scopus subject areas