Application of miscibility calculations to gas floods

Local displacement efficiency from gas injection is highly dependent on the minimum miscibility pressure (MMP) or minimum miscibility enrichment (MME). The values for these design parameters depend in turn on the displacement mechanisms, vaporizing, condensing, or a combination of the two known as a condensing/vaporizing (CV) drive. Analytical methods, which are inexpensive and quick to use, have been developed to estimate MMP’s for complex fluid characterizations. This thesis presents a simplified and robust method for MMP or MME calculation and quantification of the displacement mechanism. The calculations are also applied to develop new correlations for CO2 floods. The approach relies on finding key crossover tie lines for a dispersion-free displacement using method of characteristic theory (MOC). The new method, however, differs from published methods by significantly reducing the number of equations and unknown parameters, and by providing a fast and robust method that can avoid trivial and false solutions. We demonstrate the improvements by calculation of the MMP and MME for a variety of gas/oil systems and also give new analytical solutions for constant K-value systems that give insight into the nature of the false solutions. A method also based on MOC theory is presented to quantify the fraction of a multicomponent gas flood that is vaporizing or condensing as the pressure or gas enrichment is increased. We quantify the displacement mechanism for any number of oil or gas components by calculating the displacement path lengths along ruled surfaces bounded by these key tie lines. Several multicomponent fluid characterizations are considered. The results show that as the pressure or enrichment is increased condensation occurs at the expense of vaporization. We also show by numerical simulation that the sensitivity of the local displacement efficiency to dispersion depends on the condensing fraction of the displacement. The analytical method is also applied to the displacement of multicomponent oil by CO2. Example calculations were performed for a variety of reservoir fluids. New correlations are also generated for more accurate MMP prediction for CO2 floods. In addition, a new lumping scheme for psuedoization is proposed and applied for CO2 floods so that compositional reservoir simulation can be used in field scale where the effect of dispersion is significant.