A new empirical correlation of MMP prediction for oil – impure CO2 systems

Abstract

CO2 flooding is a widely-used technique for enhancing oil production, with miscible CO2 flooding exhibiting higher recovery potential compared to both immiscible and near-miscible CO2 flooding. However, the problem of insufficient CO2 supply is increasingly critical in many reservoirs. When faced with a shortage of gas supply, the key to the successful implementation of miscible CO2 flooding lies in accurately predicting the MMP of impure CO2. This study leverages MMP data derived from both experimental and simulation results, utilizing both linear and nonlinear regression methodologies to formulate empirical correlations for MMP prediction. In addition, ridge regression is used to solve collinearity problem.Compared with former investigations, the impure CO2 correlation proposed in this study integrates the ratio of the specific injection gas mole fraction to the crude oil mole fraction as novel independent variables. This inclusion reflects the actual interaction between the specific injection gas and crude oil during the miscibility process. Furthermore, the correlation can be linearized to the greatest extent through this way and such linearization illustrates the fundamental concepts more distinctly, offering a more practical and comprehensible approach. This formula combines the prediction of the minimum miscible pressure with the interaction mechanism of the miscible process, and the dominant factors in the miscible process can be determined. In terms of reducing the minimum miscible pressure, the interaction between C4 + C5 in the injected gas and C1 + N2 in the crude oil component is the strongest, while the interaction between N2 in the injected gas and C2-6 in the crude oil was most significant in increasing the minimum miscible pressure by comparison.The correlations presented for impure CO2 yield an average absolute deviation of 2.18 % and sustain a relative deviation within ± 10 %, surpassing previously published correlations. This study highlights the accuracy and effectiveness of the developed correlations for estimating MMP for impure CO2. The suggested MMP correlation offers a practical and efficient means of determining MMP for the implementation of impure CO2 miscible flooding in the field.