Phase Equilibria in the SACROC Oil/CO2, System

Abstract

Summary. Phase-equilibrium measurements on SACROC oil/CO, mixtures show this system to display liquid/vapor (L/V) criticality at a composition of 60 mol% Co, with a pressure of 2,560 psia [17.65 MPa] and a temperature of 131deg.F [328.15 K]. In the vicinity of the critical region. liquid quality lines show three-and four-fold pressure multiplicities for a given composition. These multiplicities are a consequence of two effects:the solvent power of supercritical CO, to extract liquid-phase hydrocarbons at pressures below the saturation pressures of the overall mixtures andthe sensitivity of the L/V phase distribution to changes in pressure when the saturation pressure of the overall mixture is in the vicinity of its critical point. On a CO2-free basis, retrograde liquid produced by flashing CO2-enriched reservoir oil at 3,000 psia and 131deg.F [20.68 MPa and 328.15 K] was found to be similar in composition to other west Texas oils. Under conditions of repeated CO, contact, this retrograde liquid was found to revaporize like other west Texas oils. Introduction Commercial-scale CO2 flooding is currently being carried out in west Texas. Knowledge of the phase behavior of the oil/CO2 mixture is important in understanding the recovery process. Such data are especially important in the application of compositional simulators to predict the performance of CO2 flooding processes. Phase-equilibrium data are presented in this paper for oil from the Phase-equilibrium data are presented in this paper for oil from the SACROC Unit, located in Scurry County, TX, and producing from the Canyon formation. Experimental data include both static measurements (first contact) to define the pressure/composition, p/x, diagram and multiple-contact measurements (CO2 cycling) to examine the revaporization of retrograde liquid. This system was previously studied by Simon et al. Their results show that this system has a L/V critical point. L/V criticality is qualitatively different phase behavior than the liquid/liquid immiscibility observed by Turck et al. for all but one of a group of other west Texas reservoir oils. That one L/V critical system, however, was not studied extensively above the critical pressure Study of the SACROC oil/CO2 system offers opportunity to gain additional insight into critical phase behavior and the effect of such behavior on the flooding process. Thus, the primary objectives of this work were to corroborate the reported L/V critical behavior in this system and to study that behavior in more detail. These ad-ditional studies provided insight into the role of supercritical extraction in the development of a miscible CO2 flood. A second objective was to study the revaporization characteristics of retrograde liquids likely to be produced during CO2 flooding. A multiple-contact, or gas-cycling, experiment was carried out to aid in determining these characteristics. Retrograde fluids would be produced during operation at reservoir pressures greater than the L/V critical once the saturation concentration of CO2 was exceeded. The hydrocarbon portion of these fluids would be significantly different from that of the reservoir oil, and its susceptibility to recovery by CO2 is unknown. Measurement Procedure Mixtures of the subject oil and CO2 were studied under static and multiple-contact conditions. The static measurements consisted of a series of constant-composition volumetric expansion (CCVE) experiments to determine phase distributions across the composition range from the subject reservoir oil to mixtures containing up to 73 mol% CO2. These data were subsequently cross-plotted to form the p/x diagram. A multiple-contact experiment was carried out to provide phase distributions, equilibrium phase compositions, and phase densities for a retrograde fluid as the hydrocarbon content of that fluid was changed by the addition Of CO2- In the multiple-contact experiment, the mixture is equilibrated at specified flash conditions, the vapor phase is discarded, and the residual liquid is recontacted with fresh CO2. This process simulates the trailing edge of the CO2 flood. Following Jacoby and Yarborough the reservoir or base oil was prepared by recombination of primary separator oil and a synthetic separator gas. The molar recombination ratio was selected to approximate the overall fluid composition and bubblepoint reported by Simon et al. The composition of the recombined reservoir oil and those of the separator oil and synthetic gas are shown in Table 1, which also shows a comparison of the recombined oil used in this investigation with that of Simon et al. The oil compositions are in good agreement for recombined fluids prepared by different laboratories. A breakdown of the C7, fraction is given in Table 2. This distribution was obtained by temperature-programmed gas chromatography (GC) and provides concentrations of narrow fractions with the n-paraffin component as the highest-boiling species in the fraction. For the lumped C7. fraction, the molecular weight, as determined by benzene freezing-point depression, and the specific gravity are also shown in Table 1. CCVE Measurements. The recombined oil was charged to a windowed, variable-volume PVT cell." This vessel was used in all the phase-distribution observations reported here. CO2 was added to phase-distribution observations reported here. CO2 was added to achieve target concentrations across the composition range. At each composition, a CCVE experiment was carried out. The CCVE experiments were begun by raising the cell pressure well above the mixture saturation pressure. The mixture was then agitated for several hours before any pressure/volume, p/i, data were obtained. An isothermal P/v traverse was carried out to establish the saturation pressure and the phase distribution for each mixture. After each pressure and the phase distribution for each mixture. After each reduction in pressure, time was allowed for the formation of a second phase, if one was stable at that condition. The phase volumes were recorded at each pressure step. The process of equilibration followed by pressure reduction was continued until the vapor phase occupied at least 80 vol% or until the system pressure had fallen to about 500 psia [3.45 MPa]. The overall composition of each mixture was confirmed by a chromatographic analysis. Fluid samples for compositional analysis were caught in dimple valves that allowed the fluid to be introduced directly to the carrier gas of the GC system. SPERE P. 103