In recent year, the performance enhancement of turbine cycle in nuclear power plants is being highlighted because of worldwide deregulation environment. Especially the first target of operating plants became the reduction of operating cost to compete other power plants. It is known that overhaul interval is closely related to operating cost Author identified that the rapid and reliable performance tests, analysis, and diagnosis play an important role in the control of overhaul interval through field investigation. First the technical road map was proposed to clearly set up the objectives. The controversial issues were summarized into data gathering, analysis tool, and diagnosis method. Author proposed the integrated solution on the basis of intelligent data mining techniques. For the reliable data gathering, the state analyzer composed of statistical regression, wavelet analysis, and neural network was developed. The role of the state analyzer is to estimate unmeasured data and to increase the reliability of the collected data. For the advanced performance analysis, performance analysis toolbox was developed. The purpose of this tool makes analysis process easier and more accurate by providing three novel heat balance diagrams. This tool includes the state analyzer and turbine cycle simulation code. In diagnosis module, the probabilistic technique based on Bayesian network model and the deterministic technique based on algebraical model are provided together. It compromises the uncertainty in diagnosis process and the pin-point capability. All the modules were validated by simulated data as well as actual test data, and some modules are used as industrial applications.
We have a lot of thing to be improved in turbine cycle in order to increase plant availability. This study was accomplished to remind the concern about the importance of turbine cycle and to propose the solutions on the basis of academic as well as industrial needs.