Residual Oil Saturation from Polymer Floods: Laboratory Measurements and Theoretical Interpretation

Abstract

Abstract A detailed analysis of the laboratory corefloods and their simulation matches, carried out earlier in our laboratory, show that, while a tertiary polymer flood cannot mobilize the waterflood residual oil, a secondary polymer flood can displace oil below the waterflood residual oil saturation observed in the same core. We show with a simplified pore-level modeling study that, when a viscoelastic polymer solution surrounds a mobile, funicular oil column in a chain of pores and slowly drains it, the breakage of the oil column into oil ganglia is delayed by the polymer's elasticity resisting the deformation of the oil/water interface. The residual oil saturations from waterflood and a subsequent (i.e., tertiary) polymer flood were measured in homogeneous Berea sandstone cores and in Antolini sandstone cores with small-scale heterogeneity, by Wreath (1989) and Wang (1995). For the latter, the distributions of porosity/permeability in the core samples were measured with mini-permeameter and CAT scan, and were employed as input for the history-matching simulations. The residual oil saturations from waterflood and a secondary polymer flood in the same core were also measured in Berea and Antolini cores. For both Berea and Antolini cores, a tertiary polymer flood did not reduce the waterflood residual oil saturation. For nine Antolini cores, a secondary polymer flood reduced the residual oil saturation below the waterflood value by 0.02–0.22 in saturation units. To investigate whether the differences observed for the heterogeneous Antolini cores are due entirely to the improved volumetric sweep or due also to the true reduction in residual oil saturation, history-matching simulations have been carried out by Lu (1994). The secondary polymer flood results could be matched only when its residual oil saturation value is lower than that for waterflood. Therefore, the higher oil recovery from the secondary polymer flood is the result of combination of improved sweep and the reduction in residual oil saturation. In order to model the delay of the oil column breakage into oil ganglia qualitatively, we consider a circular oil column surrounded by an annulus of a polymer solution in a circular tube. Assuming that the polymer solution can be modeled as a Maxwell fluid which has both viscous and elastic rheological character, the stability of the oil column is analyzed in terms of the fluid properties and the tube and oil column radii. While the interfacial tension between oil and polymer solution make a long oil column become unstable and break into oil droplets, the elasticity of the polymer solution resists the deformation of the interface. Calculation of the growth rate for the interfacial disturbance shows that, as the elastic modulus of the Maxwell fluid increases, the growth rate decreases, suggesting that the oil column could become thinner, thereby potentially decreasing the residual oil saturation.