Identification of optimal operating strategy of direct air-cooling condenser for Rankine cycle based power plants

摘要

Direct air-cooling condenser has attracted significant attention in the last decade due to the employment of Rankine-cycle based power plants from renewable (e.g., concentrated solar) or traditional (e.g., coal) heat sources in water-scarce areas. The optimal operating strategy of direct air-cooling condenser to maximize net power gain under given plant status and boundary conditions is rather complicated due to strong impacts from the steam turbine subsystem and varying ambient conditions. This paper aims at determining, for various boundary conditions, the optimal operating fan frequency and the corresponding back pressure of a typical large-scale air-cooled coal-fired power plant via accurate off-design models of both the turbine subsystem and air-cooling condenser, which are derived by combining aggregated physical equations and real operating data. Several data pre-processing techniques, e.g., quasi steady-state selection, are employed first to improve the data quality. Then, the processed data are divided into two parts for the performance characterization of involved equipment and the accuracy testing of the derived models, respectively. The results show that good agreement has been achieved between the prediction of the established models and the real operating data within a wide range of load factor (50-100%), and ambient temperature (10-30 degrees C). To maximize the plant profit, practical and quantitative operating guidelines of the air fans have been derived, which are further employed to examine current operating strategy of the air-cooling condenser of the considered power plant. It is found that with a load factor over 85%, even the full-load operation of all equipped air fans cannot deliver the theoretical optimal back pressure for the steam turbine subsystem, indicating potential benefits of enlarging the condenser for high operating loads. The proposed identification procedure can be easily implemented as an online monitoring and supervision system to practically assist the optimal plant operation.