Implementing ANFIS for prediction of reservoir oil solution gas-oil ratio

• ANFIS was designed for estimation of reservoir oil solution gas–oil ratio.
• Hybrid optimization technique gave more accurate results than backpropagation method.
• The developed models are more accurate than all other well-known published correlations.

Thorough knowledge of PVT properties of oil and gas reservoirs plays an important role in forecasting the phase behavior of oil reservoirs and designing appropriate actions for optimized production from them. Among these PVT properties, some have a determinative role in gas and oil equilibrium in the hydrocarbon reservoirs. In this study, a powerful computational intelligent model is designed to develop a reliable model for predicting amount of dissolved gas in oil at reservoir conditions as one of the most important PVT properties of reservoir oils. To achieve this model, different Adaptive Neuro-Fuzzy Inference System (ANFIS) models (by changing the training optimization algorithms) are designed. Moreover, prediction accuracy of the developed models has been compared with the number of well-known correlations in literature. The results show that the proposed model has a significantly improved performance in comparison with the other existing correlations.