A STUDY OF PROCESS ANOMALIES LEADING TO PLANT TRIPS AND THE EFFECTIVENESS OF A SOFTWARE TOOL IN TRIP PREVENTION

Abstract

In the hydrocarbon process industry, financial losses resulting from plant trips can be significant. In this study, reliability loss events or trips in 2018 at a sizeable hydrocarbon process plant are analysed to determine if they were preventable. Of particular interest is those loss events that exhibited early signs of process anomalies—such as temperature, pressure, and flow—at least twelve hours before the event. Using the concept of ‘reliability incident pyramid’ where plant process anomalies outnumber production slowdown and trip events, a software tool is deployed to help process engineers detect the anomalies. The idea is that troubleshooting these anomalies before they escalate to plant trips would reduce the number of trips and prevent production losses. Post-deployment plant process data for 2019 are analysed to determine if process engineers could pick up on the anomalies indicated by the software and intervene to prevent potential plant trips. Results indicate that in conjunction with vigilance by the process engineers, such software can reduce the number of reliability loss events in a hydrocarbon process plant. Keywords: Process anomalies, reliability incidents, loss events, reliability incident pyramid, CUSUM, EWMA, SPC