Displacement of oil by Solvent at High Water Saturation

Abstract

Abstract Displacements of laboratory oils by propane in long, consolidated sandstone cores in the presence of high water saturations have shown that oil recoveries approaching 100 percent may be realized by continuous water-propane injection, even for oil saturations close to residual oil. However, it was often necessary to inject many pore volumes of solvent to attain this high a recovery. Initial oil saturations were established by injecting water and oil at a constant ratio into the porous medium containing residual oil to a waterflood until a steady state was obtained. Propane and water were then injected in the same fixed ratio to displace the oil. These and other experiments indicate that in the presence of a high water saturation only part of the presence of a high water saturation only part of the oil is flowable. Part resides in locations that are blocked by water, and the oil in these stagnant locations is not flowable. This nonflowable oil, it is believed, can be recovered by molecular diffusion into the flowing propane of a water-propane displacement. Values for the saturation of hydrocarbon that is contained in the stagnant locations and values for the ratio of the longitudinal hydrodynamic-dispersion coefficient to displacement velocity were determined at various water saturations in the test sandstones. The data suggest that rock wettability may influence the stagnant saturation and that stagnant oil saturations may not be as large in reservoir rocks as they are observed to be in laboratory sandstones. Mass transfer between the flowing solvent and hydrocarbon components in the stagnant saturation was expressed by a first-order rate expression. Rough values for the mass transfer coefficients for the propane-trimethylhexane hydrocarbon pair were estimated from experiments. Computations using these values for mass transfer coefficients indicate that experiments in laboratory-size cores may show much poorer displacement efficiency than that which might actually occur in the field. Introduction Injection of water with light hydrocarbon solvents is a technique that may be used to partially control solvent mobility. The higher water saturation forced by water injection reduces the permeability to solvent flow, and the mobility of the solvent region is reduced relative to that of the oil-bank region. However, it also might be expected that this higher water saturation influences the microscopic unit displacement of oil by solvent to some degree. For example, as discussed by Thomas et al., two possible effects of high water saturation on the displacement mechanism come to mind. First, a miscible displacement in the presence of water is operating on a different pore-size distribution than if no water were present. Pore-size distribution and the dp term (product of the microscopic inhomogeneity factor and "effective" particle diameter) may considerably influence the magnitudes of transverse and longitudinal dispersion coefficients. Secondly, in a multiphase system the wetting phase may trap single pores or even isolate large fingers or dendrites of the nonwetting phase. The nonwetting phase in these dead-end pores or dendrites would be phase in these dead-end pores or dendrites would be nonflowing and might either be completely isolated by the wetting phase or might communicate with the flowing nonwetting fluid by diffusion through nonwetting fluid-filled pores. Aspects of miscible displacement in the presence of water have been investigated by a number of researchers. Fitzgerald and Nielson observed that the simultaneous injection of gasoline and water into a Berea sandstone core in a 1:2 ratio recovered only 36 percent of the Bradford crude oil left in the core after waterflooding, and that only 55 to 75 percent recoveries were obtained for simultaneous water-solvent injection into the core when it contained crude oil at connate water saturation. Moreover, these authors reported recoveries of only 60 to 80 percent when solvent alone was injected into the core to displace residual oil to a waterflood. Raimondi et al. injected ethyl benzene (oil) and water simultaneously into a Berea sandstone core to establish flowing oil and water saturations and then injected heptane (solvent) and water simultaneously into the core to miscibly displace the ethyl benzene. SPEJ p. 337