Toward connectionist model for predicting bubble point pressure of crude oils: Application of artificial intelligence

Knowledge about reservoir fluid properties such as bubble point pressure (Pb) plays a vital role in improving reliability of oil reservoir simulation. In this work, hybrid of swarm intelligence and artificial neural network (ANN) as a robust and effective method was executed to determine the Pb of crude oil samples. In addition, the exactly precise Pb data samples reported in the literatures were employed to create and validate the PSO-ANN model. To prove and depict the reliability of the smart model developed in this study for estimating Pb of crude oils, the conventional approaches were applied on the same data set. Based on the results generated by PSO-ANN model and other conventional methods and equation of states (EOS), the PSO-ANN model is a reliable and accurate approach for estimating Pb of crude oils. This is certified by high value of correlation coefficient (R2) and insignificant value of average absolute relative deviation (AARD%) which are obtained from PSO-ANN outputs. Outcomes of this study could help reservoir engineers to have better understanding of reservoir fluid behavior in absence of reliable and experimental data samples.