Abdolhossein Hemmati Sarapardeh - PhD student of Petroleum Engineering, Amirkabir University of Technology,
Mohammad Hossein Ghazanfari - PhD of Chemical Engineering, Sharif University of Technology
Shahab Ayatollahi - PhD of Chemical Engineering, Sharif University of Technology
Mohsen Masihi - PhD of Chemical Engineering, Sharif University of Technology

Multiple contact miscible floods such as CO2 injection has been considered as an attractive enhanced oil recovery technique especially for conventional reservoirs. A key parameter in design of CO2 miscible floods is minimum miscibility pressure (MMP), which is normally determined through expensive and time consuming laboratory experiments. Therefore, developing a quick and reliable model for prediction of MMP is inevitable. In this communication, a new reliable model on the basis of feed forward artificial neural network is presented. The input parameters of the developed model are reservoir temperature, reservoirs oil composition, and injected gas composition. Statistical and graphical error analyses have been employed to compare the developed model with the available models for the prediction of MMP. The results indicated that the developed model is more reliable and accurate over existing models, in a wide range of thermodynamic and process conditions. Finally, the leverage approach, in which the statistical hat matrix, Williams plot, and the residuals of the model results leads to identification of the likely outlier has been performed. The results showed that only two experimental MMP data points located out of the applicability domain of the proposed model and therfore, the developed model was found could be reliable enough for the prediction of CO2- reservoir oil MMP.

Minimum miscibility pressure; Intelligent aproach; Leverage approach; CO2 flooding