by Vishal Chawla

Reliance Power avoids equipment failure using predictive analytics

Oct 16, 2017

Reliance Power uses predictive analytics for condition monitoring in order to identify and prevent critical equipment failures.

For a power generation company, equipment failure can be very disruptive. For a unit having a typical 300MW coal-based plant, the cost of each unit tripping due to equipment failure can run into crores, in addition to other costs related to regeneration, maintenance, unit synchronization and unscheduled interchange (UI). It can also harm a company’s reputation if such problems compromise the operating results. To preempt such a disruption, Reliance Power has deployed various IT solutions, the most significant segment which is a predictive diagnostic software solution for condition monitoring used to identify and prevent critical equipment failures and thus extends error detection to diagnostic management for the company. Preventing equipment failure using analyticsThere are various IT constituents involved in the project. The most important one is a Condition Monitoring and Diagnostic System, which is a predictive analytics application that uses operating data to make informed maintenance decisions. The predictive analytics generated by the system provides early warnings and prescriptive advisories on equipment condition for better operational insights. Moreover, the company deploys a Distributed Control System (DCS) which had advanced diagnostic capabilities to improve control reliability and performance. Distributed Control System (DCS) uses controllers, which are connected to plant equipment sensors and actuators to capture the equipment parameter values. Another solution known as the Historian Application is deployed for taking the data of the identified critical parameters from Distributed Control System through OPC connectivity and storing it in a time series database.  The predictive analytics application has its own models developed for the power plant equipment based on equipment design, physics, and empirical regression. These models are customized as per the plant design and use the Historian data to generate early warning advisories. These advisories are notified to a team through a workflow for corrective and preventive action. The equipment in the plant which are covered under the system to prevent failure include steam turbine, power generator, condenser, boiler feed pumps, boiler feed booster pumps, condensate pumps, circulating water pumps, feed water heaters, air heaters, primary air fans, forced draft fans, induced draft fans and coal mills. How did predictive analytics benefit the company?Through innovative technological systems under its IT project, Reliance Power has improved its productivity and efficiency. The solution helps Reliance Power in making informed maintenance decisions using operating data and converts maintenance activities from a reactive mode to proactive mode.  The solution’s actionable information and alerts, delivered at the very outset of developing issues, provide Reliance Power the foresight and advance notice to perform necessary maintenance before these problems can compromise operating results. The innovative IT system has reduced maintenance costs through fewer trips and equipment degradation and has helped the company achieve commercial targets. In addition, it has helped in minimizing Unscheduled Interchange (UI) penalties by timely updating the generation declaration to the load dispatch center. Through early warnings for all critical failures, plant operability has been enriched.