Development of an artificial neural network model for predicting minimum miscibility pressure in CO 2 flooding

Abstract

This paper presents the development of an artificial neural network (ANN) model for the prediction of pure and impure CO 2 minimum miscibility pressures (MMP) of oils. The pure CO 2 MMP of a reservoir fluid (live oil) is correlated with the molecular weight of C 5+ fraction, reservoir temperature, and concentrations of volatile (methane) and intermediate (C 2–C 4) fractions in the oil. The impure CO 2 MMP factor, F imp, is predicted by correlating the concentration of contaminants (N 2, C 1, H 2S and SO 2) in CO 2 stream and their critical temperatures. The F imp is a correction factor to the MMP of pure CO 2. The advantage of using the ANN model is evaluated by comparing the measured MMP values with the predicted results from the ANN models as well as those from other statistical methods. The developed ANN models are able to reflect the impacts on CO 2 MMP of molecular weight of C 5+ fraction, reservoir temperature, and solution gas in the oil. The ANN model of impure CO 2 MMP factor can distinguish the effects on MMP of different contaminants in the CO 2 stream. It can also be used to predict the CO 2 MMP of a reservoir oil and the level of contaminants in the CO 2 stream which can be tolerated for a miscible injection.