A computational intelligence based approach for the analysis and optimization of a crude oil desalting and dehydration process

Water and salt associated with oil production can cause considerable operational problems. Therefore, desalting/dehydration plants are often installed in crude oil production units to remove water-soluble salts from an oil stream. The performance of the desalting/dehydration process depends on various process parameters interacting with each other. These parameters include concentration of demulsifying agents, heating, wash water, salt concentration, and rate of mixing with wash water. In this study, the performance of the desalting/dehydration process was evaluated by calculating the salinity and water cut efficiencies that are expected to depend on the values of these five process parameters. The work concentrated on modeling and optimizing the performance of the desalting/dehydration process system. It was an attempt to develop and apply an artificial neural network (ANN) as a modeling technique for simulating and optimizing the desalting/dehydration process system. ANNs were selected due to their potential for modeling highly nonlinear relationships of the parameters involved in the desalting/dehydration process system. The neural network model predictions were compared with the actual observations, and the results were shown to be consistent. The prepared neural network model was then used to optimize the performance of the process. A composite objective function for measuring the performance of the desalting process was used in conjunction with the prepared ANN within an optimization model. The outcome of this research will help in improving oil production operations and therefore in lowering the cost per barrel produced.