Composition Paths in Four-Component Systems: Effect of Dissolved Methane on 1D CO2 Flood Performance
- Wesley W. Monroe (Stanford U.) | Matthew K. Silva (New Mexico Petroleum Recovery Research Center) | L.L. Larson (New Mexico Petroleum Recovery Research Center) | Franklin M. Orr Jr. (Stanford U.)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- August 1990
- Document Type
- Journal Paper
- 423 - 432
- 1990. Society of Petroleum Engineers
- 4.3.4 Scale, 1.10 Drilling Equipment, 4.6 Natural Gas, 5.2.1 Phase Behavior and PVT Measurements, 5.4.9 Miscible Methods, 5.2 Reservoir Fluid Dynamics, 5.1.5 Geologic Modeling, 5.2.2 Fluid Modeling, Equations of State, 5.3.1 Flow in Porous Media, 5.4.7 Chemical Flooding Methods (e.g., Polymer, Solvent, Nitrogen, Immiscible CO2, Surfactant, Vapex), 2.5.2 Fracturing Materials (Fluids, Proppant)
- 1 in the last 30 days
- 396 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
Summary. New measurements of phase compositions and densities are reported for a quaternary system containing CO2, methane, butane, and decane and related ternary systems at 160 degrees F and 1,250 psia [71 degrees C and 8620 kpa]. Measured values of phase compositions and vapor densities agreed well with values calculated with the Peng-Robinson equation of state (PREOS), though calculated liquid densities were less accurate. Composition paths for two-phase flow of four-component mixtures were calculated with the method of characteristics and the equation-of-state (EOS) representation of the phase behavior. Analysis of resulting paths indicates why displacement efficiency in ID CO, floods is insensitive to the addition of dissolved methane to the oil displaced. Methane in the oil partitions so strongly into the more mobile vapor phase that a methane-rich bank forms at the leading edge of the transition zone. The displacing CO2 then encounters hydrocarbon mixtures without methane. Analysis indicates that high displacement efficiency is possible even when two-phase flow occurs throughout the displacement and that high recovery is possible even when a live oil is displaced below its bubblepoint pressure (BP), if the pressure is above the minimum miscibility pressure (MMP) for the same oil with all methane removed.
Understanding the role of phase behavior in the development of miscibility in CO2 floods, vaporizing gasdrives, and condensing gasdrives usually derives from analysis of composition paths represented on ternary or pseudoternary phase diagrams. Consideration of four-component systems, however, has been limited to the qualitative discussions of Deffrenne et al., Rathmell et al., and Stalkup. Mathematical analysis of such flows also has been based on ternary representations of the phase behavior of mixtures of the injected and in-place fluids. For example, Weige et al. calculated composition paths for enriched-gas drives and included the effects of volume change as components transferred between phases. Helfferich generalized the analysis of Weige et al. to systems containing an arbitrary number of components, but restricted consideration to those in which effects of volume change on mixing are negligible; however, Helfferich's examples dealt only with ternary systems. Applications of similar theory to ternary systems of interest in surfactant floods were presented by Larsons and Hirasaki. Dumor et al. extended the analyses of Welge et al. and Helfferich to describe condensing and vaporizing gasdrives for ternary systems in which volume change on mixing is important.
All these mathematical descriptions are for the case of ID flow in which the effects of dispersion are negligible. In such cases, it can be shown that compositions of fluids that form during oil displacement by CO2 do not pass through the two-phase region unless the oil composition lies within the region of tie-line extensions (from the liquid portion of the binodal curve) on a ternary diagram. Thus, the critical tie-line-the tie-line tangent to the binodal curve at the plait point-marks the boundary between oil compositions that develop miscibility and those that do not. According to the ID theory, the MMP is the pressure at which the oil composition lies on the critical tie-line, so that additional pressure increases shrink the two-phase region enough to move the oil composition outside the region of tie-line extensions. That theory applies equally well to systems containing an arbitrary number of components. In such systems, a surface in nc - 1 dimensions divides original oil compositions that develop miscibility from those that do not.
Experimental determination of the MMP, however, is rarely based on analysis of phase diagrams. Instead, NMP is obtained from measurements of the fraction of oil recovered in a slim-tube displacement at a given pressure. Several criteria have been proposed by which the MMP can be determined from displacement data. Most require that the recovery be nearly 100% (typically is greater than 90%) and that recovery increase only slightly in displacements at pressures greater than the MMP.
Numerous correlations have been offered that account for the effects of temperature and oil-composition variations on MMP. Several correlations do not account for the effect of the amount of light hydrocarbons, such as C1, in the oil. Yellig and Metcalf found, for example, that addition of C1 to an oil did not change MMP's appreciably. Neglect of the presence of light hydrocarbons is based on the assumption that such components volatilize and are transported ahead of the displacement front and do not affect miscibility development. Thus, the correlations include the provision that when enough light components are present to raise the oil's BP above the MMP predicted for the dead oil, the BP is taken to be the MMP. This provision is inconsistent with the analytical description of miscibility development. An oil's composition at its BP must be at the end of a tie-line. Hence, the composition must lie within the region of the tie-line extensions when the oil composition is plotted on a pseudoternary diagram. Thus, the definition of the MMP based on analysis of ID flow of a ternary system conflicts with part of the experimental evidence on MMP behavior as the amount of gas dissolved in the oil increases.
To resolve the inconsistency, we investigate the flow of a four-component CO2/hydrocarbon system. We report measured phase compositions and densities for a quaternary system containing CO2, C1, C4, and C10 and show that the PREOS reproduces the observed behavior with reasonable accuracy. We extend the calculations of Dumor et al. to four-component systems and then use the validated PREOS representation of the phase behavior to construct composition paths for the CO2/C1/C4/C10 system. The solutions obtained show that it is indeed possible to have high displacement efficiency in 1D flow even when the original oil composition does not lie outside the region of tie-line extensions.
Phase Composition and Density Measurements
Phase compositions and densities were measured in a continuous-multiple-contact (CMC) apparatus described previously. Measurements were made at 160 degrees F and 1,250 psia [71 degrees C and 8620 kPa] for the CO2/C1/C4/C10 system and for associated CO2/C1/C10 and CO2/C1/C4 systems. Data for the CO2/C4/C10 system were reported by Metcalf and Yarborough and Orr and Silva. The C1/C4/C10 system was studied by Reamer et al., and data are also available for the corresponding CO2/C4, CO2/C10, C1/C10, and C1/C4 binary systems.
Table 1 in Ref. 22 gives compositions reported previously 15 for the CO2/C4/C10 system and density data not previously reported. Fig. 1 in Ref. 22 compares these data with compositions measured by Metcalf and Yarborough. Also shown are compositions calculated with the PREOS.
|File Size||2 MB||Number of Pages||10|