Abstract
Phase behavior of CO2/Crude-oil mixtures which exhibit liquid/liquid (L/L) and liquid/ liquid/vapor (L/L/V) equilibria is examined. Results of single-contact phase behavior experiments for CO2/separator-oil mixtures are reported. Experimental results are interpreted using pseudoternary phase diagrams based on a review of phase behavior data for binary and ternary mixtures of CO2 with alkanes. Implications for the displacement process of L/L/V phase behavior are examined using a one-dimensional finite difference simulator. Results of the analysis suggest that L/L and L/L/V equilibria will occur for CO2/crude-oil mixtures at temperatures below about 120 degrees F (49 degrees C) and that development of miscibility occurs by extraction of hydrocarbons from the oil into a CO2-rich liquid phase in such systems.
Introduction
The efficiency of a displacement of oil by CO2 depends on a variety of factors, including phase behavior of CO2/crude-oil mixtures generated during the displacement, densities and viscosities of the phases present, relative permeabilities to individual phases, and a host of additional complications such as dispersion, viscous fingering, reservoir heterogeneities, and layering. It generally is acknowledged that phase behavior and attendant compositional effects on fluid properties strongly influence local displacement efficiency, though it also is clear that on a reservoir scale, poor vertical and areal sweep efficiency (caused by the low viscosity of the displacing CO2) may negate the favorable effects of phase behavior.Interpretation of the effects of phase behavior on displacement efficiency is made difficult by the complexity of the behavior of CO2/crude-oil mixtures. The standard interpretation of CO2 flooding phase behaviour, given first by Rathmell et al. is that CO2 flooding behaves much like a vaporizing gas drive, as described originally by Hutchinson and Braun. During a flood, vaporphase CO2 mixes with oil in place and extracts light and intermediate hydrocarbons. After multiple contacts, the CO2-rich phase vaporizes enough hydrocarbons to develop a composition that can displace oil efficiently, if not miscibly. The picture presented by Rathmell et al. appears to be consistent with phase behavior observed for CO2/ crudeoil mixtures as long as the reservoir temperature is high enough. Table 1 summarizes data reported for CO2/crude-oil mixtures. Of the 10 systems studied, all those at temperatures above 120 degrees F (50 degrees C) show only L/V equilibria while those below 120 degrees F exhibit L/L/V separations (Stalkup also reports two phase diagrams that are qualitatively similar to the other low-temperature diagrams but does not give temperatures). Thus, at temperatures not too far above the critical temperature of CO2 [88 degrees F (31 degrees C)], mixtures of CO2 and crude oil exhibit multiple liquid phases, and at some pressures L/L/V equilibria are observed.
It has not been established whether Rathmell et al.'s interpretation of the process mechanism can be extended to cover the more complex phase behavior of low-temperature CO2/crude-oil mixtures. In a recent paper, Metcalfe and Yarborough argued critical temperature CO2 floods behave more like condensing gas drives, whereas Kamath et al. concluded that an increase in the solubility of liquid-phase CO2 in crude oil at temperatures near the critical temperature of CO2 should cause more efficient displacements of oil by CO2.
SPEJ
P. 480^
Pilot test of CO2 flooding at Sarukawa oil field
