Development of Operation Optimization Technology for Gas Turbine Combined-Cycle Power Plants

In response to Korea’s national energy-transition policy and greenhouse-gas reduction targets, this study develops advanced operation-optimization technologies for gas-turbine combined-cycle power plants (CCGTs). Most domestic CCGTs rely heavily on foreign original equipment manufacturers (OEMs) for gas turbines and control systems, resulting in limited data accessibility, performance degradation, and increasing maintenance costs. To overcome these constraints, a standardized and OEM-independent operation-optimization framework was established, aiming to enhance plant efficiency, availability, and reliability while supporting autonomous operation capabilities. (1) Operation management: Artificial-intelligence (AI)–based performance models were developed for performance-impact assessment, part-load performance prediction, power-generation forecasting, and compressor-performance optimization. These models demonstrated measurable improvements in plant efficiency and heat-rate reduction. (2) Monitoring and diagnostics: A hybrid approach combining AI and eplainable AI (XAI) techniques was implemented to construct an intelligent anomaly-detection and prediction model that supports early fault identification and proactive maintenance. (3) Operation-history management: A unified database and visualization system were designed to integrate equipment operational records and enable life-cycle performance tracking. All developed solutions were implemented within a cloud-based intelligent big-data platform capable of real-time data acquisition, storage, and analysis. The platform provides an open architecture that integrates AI-driven optimization services and supports cross-plant scalability. The proposed technologies enable autonomous real-time monitoring, diagnosis, and optimization of domestic CCGTs, contributing to reduced O&M cost, improved asset protection, and lower power-generation cost. Furthermore, standardization of AI- and data-driven diagnostic methodologies is expected to strengthen the global competitiveness of Korea’s gas-turbine and digital power-plant-operation technologies.