Technologies of Predictive Diagnostic Condition Monitoring in Yeongheung Power plant

摘要

This paper introduces the background and major cases of how Yeongheung achieves its goal of stable power supply and cost reduction through pre-failure real-time monitoring and preventive activities of the two newly built thermal power plants. These days, the issues of reducing operating costs and process operations variability, response to demand uncertainty and global competition from low cost source has changed the environment of power plant industry. The rapidly changing environment of power plant industry has pursued operational excellence and started to develop new ways to achieve it. By reason of that, many power plants over the world have adopted various operation monitoring systems and predictive diagnosis solutions by utilizing both hardware and software to effectively monitor and manage their plants. Youngheung Power Plant is one of the Korea's largest thermal power plants, accounting for 19.6% of the total power consumption of capital area. We operate 1 to 4 units, which have a capacity of 3,340 MW. Recently, two new units with a capacity of 1,740 MW have been added for the stable power supply to the load center of the capital area. New power plants are strongly demanded for the reliability and stability. This is because equipment failures could lead to the shutdown of the power plant, and finally it results in interrupting high-quality power supply. As a response to this, the new power plants, Yeongheung unit 5 and 6, installed integrated real time monitoring system in order to identify the critical changes of major components. It allows us to continuously monitor the system by converting facilities' variation into meaning data and trend. We also deployed predictive maintenance solution, which helps to take appropriate actions in advance by detecting the abnormal condition of plants in index form. This system learns the historical plant data during the normal condition to build expected model, and generates early warning when there are some deviations between real time value and expected value. In addition, we utilize operational marginal system to monitor the plant shutdown related signals’ operating range from normal to plant shutdown status in radar charts. We constructed an PMD(Predictive Monitoring & Diagnostic) center to strengthen the integrated on-site monitoring of new units, and this center has discovered core values with these systems.