A Generalized Method for Predicting Gas/Oil Miscibility
- A.S. Emanuel | R.A. Behrens | T.J. McMillen
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- September 1986
- Document Type
- Journal Paper
- 463 - 473
- 1986. Society of Petroleum Engineers
- 5.4 Enhanced Recovery, 5.2.2 Fluid Modeling, Equations of State, 5.3.4 Reduction of Residual Oil Saturation, 4.1.9 Tanks and storage systems, 5.4.4 Reduction of Residual Oil Saturation, 5.4.7 Chemical Flooding Methods (e.g., Polymer, Solvent, Nitrogen, Immiscible CO2, Surfactant, Vapex), 4.1.2 Separation and Treating, 5.2.1 Phase Behavior and PVT Measurements, 5.4.2 Gas Injection Methods, 5.3.2 Multiphase Flow, 4.1.5 Processing Equipment, 5.2 Reservoir Fluid Dynamics, 4.6 Natural Gas
- 0 in the last 30 days
- 393 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
An equilibrium-cell simulator provides reliable estimates of gas/oil minimum miscibility pressure (MMP) with a correlation of residual oil against capillary number. The basic flash calculations of the equilibrium cell are supplemented by a number of PVT and fluid-flow modeling techniques that provide a reasonable simulation of slim-tube performance. A PVT stability calculation has been incorporated as an option for systems where flash convergence is difficult. Interfacial tension (IFT) may be calculated either by the method of parachors or by a method based on critical-point scaling theory. An experimental correlation is presented for residual oil reduction with capillary number. The relative permeabilities are varied between the fully immiscible and fully miscible permeabilities are varied between the fully immiscible and fully miscible bounds as a function of capillary number. The slim-tube recovery results have been calibrated against 110 experimental runs for 21 different oils. Recovery by hydrocarbon gas is generally predicted within 10%; recovery by N2 and CO2 is generally predicted to within 15%. The program logic is designed to be most accurate in the miscible region. The simulator will typically give an a priori estimate of MMP within 10% and show the relative effects of changes in injection gas composition and system pressure. The simulator may be calibrated to an experimental recovery point. An oil-specific calibration gives much higher accuracy than the prediction that use the default parameters.
A standard procedure to determine miscibility pressure for a given reservoir oil displaced by an injection fluid is the slim-tube measurement. A system that is not miscible in all proportions or first-contact miscible may display a generated miscibility. The slim tube allows for a transition region between the gas and oil so that generated miscibility is possible. The concept of a slim-tube simulator was described by Metcalfe et al and Fussell et al as a simple, inexpensive model to predict the development of miscibility. A series of equilibrium flash cells with relative permeability rules was the basis of the model. The simulator was permeability rules was the basis of the model. The simulator was introduced because existing correlations were insufficient to predict miscibility.
Correlations of MMP
An early correlation for MMP still in wide use is that of Benham et al. The maximum C1 composition in a hydrocarbon gas injectant stream was specified as a function of temperature, pressure, oil C5+ molecular weight (MW), and injectant gas C2+ MW. The correlation is often conservative and does not treat such nonhydrocarbon gases as N2 or CO2. This work of Benham et al. was the pioneering work in MMP correlations. Since that time, numerous correlations have been created. There are many different MMP correlations for CO2 injection. Holm and Josendal used the methodology of Benham et al to create a similar correlation for miscibility, except that the injection gas was pure CO2. MMP was given as a function of C5 + MW and reservoir temperature. This correlation was extended to higher MW's by Mungan. Yellig and Metcalfe offered an empirical correlation based solely on the reservoir temperature with the bubblepoint as a limit on the MMP. Johnson and Pollin offered a correlation that could also be used with impure CO2 injection gases. The empirical correlation was developed for either live or stock-tank oils and requires the oil gravity and MW, reservoir temperature, and injection-gas composition. Alston et al also gave an empirically derived correlation for MMP for pure or impure CO2 streams. Their correlation was based on the oil volatile fraction, intermediate fraction, and the heavy (C5+) MW. The other necessary input parameters were temperature and injection-stream composition. Enick et al. made a correlation that was based on an equation of state (EOS) calculation of a pseudobinary system consisting of the pure or impure CO2 injection-gas stream and a normal alkane with the same MW as the C5+ fraction of the oil. The influence of oil composition on MMP for CO2 flooding is not well defined. Yellig and Metcalfe found no influence of oil composition on MMP, but this was contested by Holm and Josendal. Holm and Josendal later showed that increasing aromaticity decreases MMP. Silva and Orr recently concluded that the distribution of molecular sizes has the greatest effect on the development of miscibility and that effects of hydrocarbon structural variations are smaller. For a given molecular size distribution, however, solubility decreases as the structure varies from branched paraffins to normal paraffins to naphthenes and finally to aromatics. The side chains, particularly branched side chains, on ring structures help solubilize the molecules.
|File Size||912 KB||Number of Pages||11|