Exception Signal Improvement for More-Accurate Identification of High Tubing Pressure Well in Delta Field

Abstract

Abstract. IEMS (Integrated Exception Management System) is integrated solution for exception management which applies data science process to organize large amount of data and automatically generate exception signals that are prioritized for review and actions. The scope is currently focused on reduce Lost Production Opportunities (LPO), optimizing wells by reducing time taken for abnormal conditions to be identified and anticipated. One of problems in Delta field is high tubing pressure that occurs due to upset of fluid to flow caused by restriction from pipeline. Exception signal to identify anomaly tubing pressure using data tubing pressure from dynamometer is mostly end up as “no further action” or false signal. Which mean, criteria or data science process in IEMS to captured high tubing pressure is not robust enough. For more accurate identification of high tubing pressure wells, team has applied to alter the maximum tubing pressure to average tubing pressure to capture fluctuation tubing pressure. Additional criteria added to strengthen the criteria is using Artificial Intelligent. Machine learning will be labeling and  classification the dyno card to identify high tubing pressure identic with dyno shape with hump. The combination of these data, we expect to reduce false signal that related to resource’s manning hour and  reduce potential hidden wells down. Recommendation for wells after high tubing pressure exception signal is flushing the pipeline with water. Several flushing requests are rejected by operation team due to normal tubing pressure. To avoid discrepancy data tubing pressure, we recommend for regular calibration pressure gauge or pressure transmitter.