Implementing radial basis function neural network for prediction of surfactant retention in petroleum production and processing industries

Chemical flooding is an effective way to gain higher oil recovery as part of a tertiary oil recovery scheme. There are several variables contribute in surfactant retention in petroleum production including type of rock, pH, chemical structure of surfactant, salinity of formation water, acidity of oil, mobility, microemulsion viscosity, and cosolvent concentration. Although different theoretical studies on the mechanisms of surfactant retention are reported in the literature there is little research on the development of an accurate and effective model for prediction of surfactant retention in petroleum production. In this study, radial basis function was developed based on experimental dynamic surfactant retention data. The experimental data include a wide range of conditions. Results of the modeling study showed that the developed model is very accurate and robust in prediction of actual surfactant retention data. In addition, the comparison between the proposed model in this study and available models in literature showed the superiority of this model.