On the prediction of solubility of alkane in carbon dioxide using a novel ANFIS-GA method

Recently attention to the enhancement of oil recovery processes increases because of increasing energy demand and problems of declination of reservoirs. One of the effective and applicable approaches in the enhancement of oil recovery is carbon dioxide injection which also has a positive effect on the environment. The carbon dioxide injection can enhance oil recovery by various approaches. The extraction of lighter components of crude oil is known as one of the effects of carbon dioxide injection. The solubility of hydrocarbons in carbon dioxide is known as one of the dominant factors which can influence the hydrocarbon extraction from crude oil. Due to the importance of this parameter, in the present paper, Adaptive neuro-fuzzy inference system (ANFIS) joint with Genetic algorithm (GA) was developed to forecast the solubility of hydrocarbon in carbon dioxide based on carbon number of alkane, carbon dioxide density, pressure, and temperature. The ANFIS-GA solubility values were compared with the real solubility values. The determined coefficients of determination for training and testing phases are 0.96741 and 0.91648, respectively. So it is obvious the model has great capacity in the prediction of solubility.