Numerical Simulation of the Behaviour of Enriched Gas Injected Alternately With Water In Horizontal Miscible Floods

Abstract

Abstract An investigation was made of the effects of solvent:water injection ratios, length of injection cycles and solvent volume injected on solvent bank movement and sweep efficiency when enriched gas is injected alternately with water in horizontal miscible floods. A two-dimensional three-phase reservoir simulator incorporating a nine-point finite-difference formulation was used in this study. Introduction The work reviewed here was undertaken to improve understanding of the performance of the enriched gas (solvent) water bank in horizontal hydrocarbon miscible floods. Of particular interest were:solvent-water bank propagation;the effect of varying soivent:water ratio;injection cycle size effects;the effects of solvent volume injected;zonal areal sweep;producing characteristics after solvent breakthrough; andinjection fluid distribution into a layered system. Assumptions and Limitations Basic data representative of an existing horizontal hydrocarbon miscible flood were used throughout the study. Basic reservoir parameters and the oil-water relative permeability curves are included as Table I and Figure I respectively. The solvent-oil relative permeability curves were derived from an assumed fractional flow equation used to simulate miscible displacement as noted on Figure 2. For the model work conducted, it was assumed that the oil-water relative permeability curve held for solvent-water as well. This assumption implies that oil and solvent have the same flow behaviour. The solvent-oil mixing zone is neglected. It was also assumed that there was no' hysteresis in the oil-water relative permeability curve. The formation studied is moderately oil-wet, resulting in hysteresis in the water relative permeability curve. There is negligible hysteresis in the oil relative permeability curve. As the results presented in this study are intended to be qualitative, these assumptions would not affect the conclusions. Three-phase relative permeability was handled by entering the water-oil relative permeability curve (Fig. 1) as a waterhydrocarbon phase curve. The relative permeability of the hydrocarbon phase was then divided between oil and solvent by the solvent?oil relative permeability curve (Fig. 2). Two reservoir simulators were used: an areal black oil and a radial coning black oil model. Both are two-dimensional and capable of handling three fluid phases. The black oil model was used in two configurations. A homogeneous I-D configuration was used to investigate the mechanism of alternate solvent-water injection and yield a qualitative understanding of the process. A' homogeneous 2--D areal configuration was us!'d to investigate areal sweep effects. The coning model was used to investigate cyclic production after solvent breakthrough and the effects of permeability contrast in injection wellbores on solvent:water ratio. The radial configuration was useful in simulating viscous effects near the wellbore and pressure changes in the inter-well area. Published data(1) have shown that the 9-point finite difference formulation is required to eliminate grid orientation effects under adverse mobility ratio conditions. The 9-point formulation was used for this study because an adverse mobility ratio existed during each cycle when the injection was switched from low-mobility water to high mobility solvent.