Effects of Reservoir Heterogeneity on Chemically Enhanced Oil Recovery

Abstract

Summary. A three-dimensional (3D), multiphase, multicomponent micellar/polymer flooding simulator has been used to investigate the process performance under different conditions of reservoir heterogeneity process performance under different conditions of reservoir heterogeneity (stratifications) and mobility ratios. The significance of crossflow has been studied by relating chemical flood recoveries to the effective length-to-thickness ratio and the transverse dispersion number of the reservoir. A long-standing question in micellar/polymer flooding concerns whether small, high-concentration chemical slugs are preferred to large lower-concentration slugs. This aspect of slug size vs. slug concentration has been studied in 3D in the presence of heterogeneity. Finally, chromatographic separation of surfactant and alcohol has been investigated and discussed. It is found that a salinity gradient can be tailored to minimize this separation tendency and that this in fact is the single most important factor, even when compared with adsorption and partitioning. This is a highly favorable result that can and should be exploited to improve the practical design of micellar/polymer floods. Introduction The principal mechanisms enhancing oil recovery in a micellar/ polymer flood aremiscibility between the slug and neighboring polymer flood are (1) miscibility between the slug and neighboring fluids,ultralow interfacial tension (IFT), and hence reduced residual oil saturation, andoil swelling or solubilization. The capability for mobility control and the technical feasibility of the process make micellar/polymer flooding a potential secondary or process make micellar/polymer flooding a potential secondary or tertiary recovery process under a wide range of reservoir conditions. In the design and optimization of such a process, however, one is confronted with a complex displacement mechanism, associated adsorption, chromatographic separation of components, and several intricate interactions of fluid and reservoir rock properties. In fact, several of the key phenomena affecting oil recovery in a micellar flood are strongly coupled and thus need to be considered simultaneously. This coupling makes laboratory investigation of the sensitivity of the process with respect to any specific parameter difficult. Numerical simulation constitutes an integral part of any design or scale-up study for a process as complicated as micellar/polymer flooding. Several authors have examined by numerical simulation the complex compositional effects that occur during displacement of oil with surfactants and polymers. The efficiency of the oil displacement has been studied as a function of slug size, polymer drive size, surfactant and oil concentrations in the slug, mobility ratios, adsorption, IFT, phase types, and electrolyte gradient. Most of these earlier studies, however, have been limited to a one-dimensional (lD) homogeneous medium. Another important aspect in micellar/polymer flooding is the chromatograptiic separation of surfactant and cosurfactants as they traport through the porous medium. Until recently, investigations with regard to the separation phenomenon have been limited because of inadequate representation of phase behavior. The phase behavior has been represented in the past on ternary planes, for example by lumping together two of the components in a fixed ratio. The most common approach has been to lump together surfactant and cosurfactant as chemical or active mixture. This way of lumping surfactant and cosurfactant together leads to thermodynamic inconsistencies because, under such conditions, middle-phase points are no longer unique and the optimal salinity changes with overall concentrations. Other kinds of ternary planes have been used (for example, fixed surfactant to hydrocarbon), but once again, they enable representation of only binodal surfaces and not tie-lines or tie-triangles because phase compositions do not lie in the same triangles as the overall composition points. Hence, phase behavior of such systems has to be represented on 3D tetrahedric diagrams. In this paper, we used a recently reported 3D multiphase, multi-component micellar/polymer flooding simulators to investigatthe efficiency of oil displacement, transverse dispersion, mobility ratio, and chromatographic separation of surfactants and cosurfactants under various operating strategies and different conditions of reservoir heterogeneity. This is the first time a 3D compositional simulator has been used to make a study of this kind with realistic process and reservoir parameters, and several significant process and reservoir parameters, and several significant conclusions emerge. In particular, we studied the impact of heterogeneities on the overall process performance and the significance of crossflow. Chemical flood recoveries have been related to the effective length-to-thickness ratio and the transverse dispersion number of the reservoir. A long-standing question in surfactant flooding concerns whether small, high-concentration chemical slugs are preferred to large, lower-concentration slugs. We studied this aspect of slug size in 3D and made comparisons between large, dilute vs. small, concentrated slugs in a highly heterogeneous medium. Furthermore, we have found that a salinity gradient can be tailored to the separating tendency of surfactant and cosurfactant, and that this in fact is the single most dominant factor, even when compared with adsorption and partitioning of the amphiphilic species. This is a highly favorable result that can and should be exploited to improve the practical design of micellar/polymer floods. Model Description Multicomponent, multiphase flow in permeable media occurs as transport of chemical species in multiple homogeneous phases under the influence of four predominant forces: viscous, gravity, dispersion (diffusion), and capillary forces. The conservation equation for each species applies at each point in the medium, including the stationary phase. The general conservation equation for Component K can be written as ..........................................(1) SPERE P. 479