Effects of Gravity and Viscous Crossflow on Hydrocarbon Miscible Flood Performance in Heterogeneous Reservoirs

Abstract

SPE Member Abstract Injected solvent composition is a key parameter in the design of hydrocarbon miscible floods. Economic optimization of a field flood requires consideration of the effect of solvent composition on flood performance at the reservoir-scale. In this study, a compositional simulation approach was used to study the effects of permeability heterogeneity, fluid crossflow, and phase behavior on displacement performance in two-dimensional, fine-grid (x-z) cross-sections. Results of secondary and tertiary solvent flood simulations in various rock property fields indicate that floods with solvent enrichment at or below that required for development of multicontact miscibility in one-dimensional now can perform as well or better than floods with richer solvents. A significant change to the current practice of solvent enrichment selection is suggested based on analysis of the simulation results. Introduction Oil recovery in field-scale miscible floods depends on both local displacement efficiency and sweep efficiency. High local displacement efficiency is achieved in miscible floods when the interfacial tension between phases is reduced to zero. This high local displacement efficiency, however, may be offset by low macroscopic sweep efficiency due to fluid channeling. Fluid channeling due to permeability heterogeneity and gravity segregation can be severe in miscible floods because of the high adverse mobility ratios and large density contrasts between injected gas and reservoir oil. Injected solvent composition affects the values of fluid mobility, density, and interfacial tension that occur during a miscible flood. Hence, both local displacement efficiency and sweep efficiency will be affected by injected solvent composition. The effect of solvent composition on local displacement efficiency in one-dimensional flow is routinely evaluated as part of any miscible flood design. Oil recovery and effluent compositions from slimtube displacement experiments and/or numerical simulations of flow provide the basis for evaluating the minimum solvent enrichment level (MSE) for secondary solvent floods in one-dimension. In this paper, the MSE is defined as the enrichment level at which oil recovery on a graph of oil recovery at 1.0 pore volumes fluid injected (PVI) versus solvent enrichment levels off. Little economic benefit is derived from enriching the solvent above the MSE because little or no incremental recovery is obtained with additional enrichment Current industry practice specifies a solvent enriched to the MSE or beyond for use in field-scale floods. Inherent in this practice is the assumption that the MSE observed in one-dimensional flow is the optimum economic enrichment level in multidimensional flow at the reservoir-scale. At the reservoir scale, rock heterogeneity, fluid mixing due to viscous, gravity, and capillary forces, diffusion, dispersion, and the presence of mobile water are some factors that may cause the optimum economic enrichment level in field-scale floods to differ from that in one-dimensional flow. There is some evidence that local displacement efficiency is reduced in MCM displacements when viscous crossflow causes mixing of fluid between zones of fast and slow flow. Gardner and Ypma, for example, showed that viscous crossflow between zones of fast and slow flow causes two-phase flow in an otherwise MCM CO2 flood. Pande and Orr found that viscous crossflow caused two-phase flow in otherwise MCM displacements, for both a vaporizing CO2 gas drive and a condensing hydrocarbon gas drive, in a two-layer permeability field. Although local displacement efficiency was reduced, overall recovery was improved as fluid crossflow improved sweep efficiency by reducing fluid channeling through the high permeability layer. P. 839^