Journal of Canadian Petroleum Technology, Vol.46, No.8, 19-25, 2007
Genetic algorithm (GA)-based correlations offer more reliable prediction of minimum miscibility pressures (MMP) between reservoir oil and CO2 or flue gas
Two new genetic algorithm (GA)-based correlations were proposed for more reliable prediction of minimum miscibility pressure (MW) between reservoir oil and CO2 or flue gas. Both correlations are particularly useful when experimental data are lacking and also in developing an optimal laboratory program, to estimate MMP. The key input parameters in a GA-based CO2-oil MMP correlation, relation, in order of their impact, were: reservoir temperature MW of C5+, and volatiles (C-1 and N-2) to intermediates (C-2-C-4, H,S and CO) ratio. This correlation, which has been successfully validated with published experimental data and compared to common correlations in the literature, offered the best match with the lowest error (5.5%) and standard deviation (7.4%). For a GA-based flue gas-oil MMP correlation, the MMP was regarded as a function of the injected gas solvency into the oil which, in turn, is related to the injected gas critical properties. It has also been successfully validated against published experimental data and compared to several correlations in the literature. It yielded the best match with the lowest average error and standard deviation (6.2%). Moreover, unlike other correlations, it can be used more reliably for gases with high N-2 (UP to 20%) and non-CO2 components (up to 78%), e.g., H2S, N-2, SOx, O-2, and C-1-C-4.