Prediction of Riser Base Pressure in a Multiphase Pipeline-Riser System Using Artificial Neural Networks

Abstract

In the multiphase flow of oil and gas in pipeline-riser systems, reliable pressure measurements and monitoring is of utmost importance for flow assurance. These measurements are usually obtained using remote pressure measuring gauges and other devices. They are employed in the automatic slug flow control technique. However, these devices are quite expensive and often require calibration at intervals to guarantee accuracy and precision. There is therefore, the need for suitable alternatives. In this study, a feed-forward back propagation artificial neural network (ANN) for predicting riser base pressure in offshore pipeline riser systems is presented. A total of 16,870 experimental data sets were used to develop the ANN model. The results show near perfect predictions with an average mean square error of 0.00207197 and regression correlation coefficient, R values as high as 0.99919. The models obtained from this work can be pivotal to the development of data driven control of slug in pipeline-riser systems.