Simultaneous determination of capillary pressure and relative permeability curves from core-flooding experiments with various fluid pairs

Abstract

[1] Capillary pressure and relative permeability drainage curves are simultaneously measured on a single Berea Sandstone core by using three different fluid pairs, namely gCO 2/water, gN2/water and scCO 2/brine. This novel technique possesses many of the characteristics of a conventional steady-state relative permeability experiment and consists of injecting the nonwetting fluid at increasingly higher flow rates in a core that is initially saturated with the wetting phase, while observing fluid saturations with a medical x-ray CT scanner. Injection flow rates (0.5–75 mL/min) are varied so as to generate a large range of capillary pressures (up to 18 kPa), whereas fluid-pairs and experimental conditions are selected in order to move across a range interfacial tension values ( γ12=40−65 mN/m), while maintaining a constant viscosity ratio ( μw/μnw ≈30). Moreover, these experiments, carried out at moderate pressures ( P=2.4 MPa and T=50°C), can be compared directly with results for gas/liquid pairs reported in the literature and they set the benchmark for the experiment at a higher pressure ( P=9 MPa and T=50°C), where CO 2 is in the supercritical state. Contrary to some prior investigations, from these experiments we find no evidence that the scCO 2/brine system behaves differently than any of these other fluid pairs. At the same time, capillary pressure data show a significant (but consistent) effect of the different values for the interfacial tension. The fact that the three different fluid pairs yield the same drainage relative permeability curve is consistent with observations in the petroleum literature. Additionally, the observed end-point values for the relative permeability to the nonwetting phase ( kr,nw ≈0.9) and the corresponding irreducible water saturations ( Sw,irr ≈0.35) suggest that water-wet conditions are maintained in each experiment. The reliability of the measured relative permeability curves is supported by the very good agreement with data from the literature obtained on Berea Sandstone cores and with various gas/liquid pairs. The Brooks-Corey model is used to describe the capillary pressure data and the parameters derived from these matches provide a fair prediction of the relative permeability curves. It is shown that the apparent low end-point relative permeabilities to the nonwetting phase reported in previous experimental studies are caused by the low viscosity of CO 2 relative to water, rather than by the rock heterogeneity. In fact, the former controls the level of capillary pressure that can be achieved experimentally, thus restricting the applicability of some of the conventional methods to measure relative permeability curves for gas/liquid systems.