A New Methodology for Prediction of Bottomhole Flowing Pressure in Vertical Multiphase Flow in Iranian Oil Fields Using Artificial Neural Networks (ANNs)

Abstract In this paper, Artificial Neural Networks (ANN) are used to predict the bottom-hole flowing pressure in vertical multiphase flow. Two-phase flow of gas and liquids is commonly encountered in the production and transportation of oil and gas. Knowing the bottom-hole pressure (BHP) of a well and the productivity index (PI or J) can help predict the well potential during its life-cycle. In other words, well production monitoring can be performed, which is a key objective for oil production maximization and operational cost reduction. Different correlations considering different operating conditions and flow models were studied in order to find the most effective input parameters. ANN accuracy is highly dependent on the validity of the input and output data. After gathering the input and output data from selected southern Iranian oil fields, all the data were filtered with the help of existing models to eliminate the unreliable data. Then, 167 data sets were normalized and carefully imported into the ANN models. Different ANN models with different numbers of hidden layers and transfer functions were developed and tested, and the best one with the least error was chosen. The accuracy of the pressure predicted by the developed ANN model was improved by approximately five times as compared with existing correlations. To show the accuracy of the method, the results are compared with those obtained from the existing correlations. Accurate prediction of pressure drop in vertical multiphase flow is needed for effective design of tubing and optimum production strategies. Different kinds of two-phase flow correlations have been developed and are currently being used in industry. In addition to the limitations on the applicability of all existing correlations, they all fail to predict the desired accuracy of pressure drop predictions.