Data based online operational performance optimization with varying work conditions for steam-turbine system

One of the most urgent problems facing China today is saving energy in the power industry. Moreover, because the steam-turbine system is a main part of the power unit, its operation is of great significance for energy saving. This paper presents an optimization model for steam-turbine systems based on a data-mining method from an operator's perspective. The main aim of the proposed methodology is to provide reference values of the independent variables for the operator to minimize heat consumption rate. These were determined based on fuzzy C-means clustering and statistical methods by mining the historical operation data resources with respect to varying load and ambient temperature. The proposed methodology application was implemented as an online optimization system for an on-duty steam-turbine system. The application results showed that the energy savings reached up to 79,000 GJ, which is remarkable. The reference values of variables were helpful for improving the steam-turbine system's performance.