Abstract
The continued expansion of fluctuating energy sources such as wind turbines and solar systems will increase the demand for more flexible operation modes of power plants. Especially steam turbines with all their components will have to sustain a higher amount of start-stop cycles in order to compensate for variations in wind and solar radiation. Besides the rotor, inner casings are an example for main steam turbine components which are strongly loaded by thermal cycles at each start and shut down procedure. A precise prediction of the attainable number of start-stop cycles enables a more flexible operation within the guaranteed lifetime. However, this would require time-consuming FE calculations for each power plant due to their specific steam parameters.
In this paper, a physics based surrogate model is discussed for a fast prediction of permissible start-stop cycles at plant specific steam parameters. The correlation between the physical properties from the surrogate model (wall temperature difference and the resulting stresses) and the attainable number of start-stop cycles from the FE model is determined. A validation with a different inner casing design within a usual wall temperature range confirms the high accuracy level of the surrogate model compared to uncertainties like material scatter or casting tolerances. With the provided approach typically a higher number of starts can be efficiently calculated in the bidding phase compared to assuming only one conservative value for each turbine type or size. Furthermore, the steam parameters can be optimized for increasing the number of starts to the required value without additional and time-consuming FE calculations.
Modeling of ultra super critical power plants integrated with the chilled ammonia process
