(128c) Maintenance & Reliability - Predictive Analytics Using Process & Operating Data

Self-service analytics provides engineers and operators with new ways to monitor assets in order to improve availability and overall effectiveness. Cases are presented reflecting successes derived from applying an analytics platform to processes and the benefits received.

Asset performance in the petrochemical industry is increasingly an area of focus for operating companies. Firms are working to control short-term maintenance and operating costs; and reach the longer-term goal of operational excellence. The majority of the traditional data modelling exercises only provide a positive return on investment (ROI) for the 1% critical assets, since they are typically very engineering-intensive projects which require a skilled data scientist who is not necessarily familiar with the process.

Three cases related to asset reliability of particular interest for ethylene producers are presented in this paper: valve performance, heat exchanger efficiency, and compressor reliability. For valve performance, the issue of drifting valve characteristics, and valve wear and tear are tackled. For heat exchanger efficiency, we demonstrate how prediction of fouling can improve overall production. Finally, in the compressor case, the focus is on understanding, and predicting oil temperature alarms.

Using these examples, we will demonstrate how deploying self-service analytics helps engineers increase the reliability of operating assets in ethylene production by establishing baseline performances, identifying anomalies, and creating background monitors that act as an early warning system. For the company’s bottom line, predictive, self-service analytics reduces stress on assets, avoids unplanned downtime, and helps establish timely and properly planned and prioritized maintenance timetables.