<![CDATA[The increasing complexity of thermal power plant operations requires advanced digital systems capable of managing large-scale operational data to improve efficiency, reliability, and intelligent maintenance. The limitations of conventional Plant Information Management Systems (PIMS), including restricted data storage capacity, inadequate integration with Distributed Control Systems (DCS), and limited analytical capabilities, hinder effective operational decision-making. This study aimed to upgrade the PIMS infrastructure and develop an integrated heat consumption deviation analysis system supported by machine learning technology at PLTU Kendari-3. The research employed an applied engineering research design involving system diagnosis, hardware and software upgrades, DCS-PIMS integration, historical data migration, analytical module development, and post-implementation performance verification. The results showed that the upgraded PIMS successfully expanded historical data storage capacity from 90 days to more than five years, enabled real-time data integration, and provided automated heat consumption deviation analysis. Furthermore, the developed machine learning model achieved equipment fault prediction accuracy exceeding 85%. The implementation also contributed to an approximately 3% improvement in unit operating efficiency, a 5–8% reduction in maintenance costs, and an estimated 3–5-year extension of equipment service life. This study concludes that integrated PIMS modernization combined with intelligent analytics can significantly enhance thermal power plant operations and provide a practical framework for digital transformation toward intelligent operation and maintenance.]]>
Copyrights © 2026
