Algebraic approach for the diagnosis of turbine cycles in nuclear power plants

According to plant operating staff's practical needs, authors proposed a diagnosis model to identify the performance degradation of steam turbine cycles in nuclear power plants (NPPs). The essential idea of this study is how to identify the intrinsically degraded component which causes electric loss. Authors found that there were not so many turbine cycle diagnosis applications in NPPs currently because of technical, financial, or social characteristics of the plant. So a great part of the diagnosis has been dependent on operating staff's experience and knowledge. However as economic competition becomes severe, the efficiency staffs is asking for reliable and practical advisory tools. For the solution of these shortcomings, authors proposed a simple and intuitive diagnosis concept based on the superposition rule of degradation phenomena, which can be derived by simple algebra and correlation analysis. Though the superposition rule is not so significant statistically, almost all of the performance indices under normal operation are fairly compatible with this model. Authors developed a prototype model of quantitative root-cause diagnosis and validated the background theory using the simulated data. The turbine cycle advisory system using this model was applied to Gori NPP units 3&4.