Comparing Centrifuge, Steady-State and Semi-Dynamic Methods for Relative Permeability and Capillary Pressure Determination: New Insights

Abstract

Relative permeability and capillary pressure are essential parameters for understanding multiphase flow in porous media and scenarios of production in oil or gas reservoirs. There are several experimental methods for determining the relative permeability curves: unsteady-state (USS), steady-state (SS), and semi-dynamic (SD) methods. Each method has advantages and weaknesses. Although the USS approach leads to fast data results, the interpretation neglects the capillary pressure effects and provides a limited amount of data points obtained after breakthrough. The SS method is time consuming but enables covering a wider range of saturation with data points if the test is well designed. The SD method may be more time consuming than the SS method but provides both relative permeability to the injected phase and capillary pressure. The relative permeability to the produced phase is then determined by numerical means. The main objective of this study was to compare the water-oil relative permeability curves obtained from the steady-state and semi-dynamic methods performed at reservoir conditions with live fluids. Carbonate core plugs of same rock type and same properties were selected for this experimental program. The samples were brought to the same irreducible water saturation at a constant brine-oil capillary pressure using a centrifuge before being dynamically aged with live oil. In addition to monitoring the average saturation using material balance (MB), a linear X-ray scanner was used for in-situ saturation monitoring (ISSM) along the core samples. The oil relative permeability from the SD method was simulated with fixed water relative permeability and capillary pressure by history-matching the oil production and the differential pressure signal. Two additional centrifuge tests on twin plug were performed in order to measure imbibition capillary pressure and oil relative permeability at pseudo-reservoir conditions. This comparative study shows that the SD method provides similar capillary pressure and oil relative permeability curves to those obtained by centrifuge methods. Even if all Kr curves are in an acceptable envelop, some differences are observed between SD and SS Kr curves: several investigative leads are given to explain this discrepancy. It is also shown that a better saturation method needs to be implemented, especially when dealing with heterogeneous rocks. While a more robust ISSM method is being tested at TOTAL, the results presented in this paper are very encouraging.