Start-up optimization of a CCGT power station using model-based gas turbine Control

The rapid growth of renewable generation and its intermittent nature has modified the role of combined cycle power stations in the energy industry, and the key feature for the operational excellence is now flexibility. Especially, the capability to start an installation quickly and efficiently after a shutdown period leads to lower operational cost and a higher capacity factor. However, most of existing thermal power stations worldwide are designed for continuous operation, with no special focus on an efficient start-up process. In most current start-up procedures, the gas turbine controls ensure maximum heat flow to the heat recovery steam generator, without feedback from the steam cycle. The steam cycle start-up controls work independently with as main control parameter the limitation of the thermal stresses in the steam turbine rotor. In this paper, a novel start-up procedure of an existing combined cycle power station is presented, and it uses a feedback loop between the steam turbine, the boiler and the gas turbine start-up controls. This feedback loop ensures that the steam turbine can be started up with a significant reduction in stresses. To devise and assess this start-up methodology, a flexible and accurate dynamic model was implemented in the Simulink environment. It contains >100 component blocks (heat exchangers, valves, meters and sensors, turbines, controls, etc.), and the mathematical component submodels are based on physical models and experimental correlations. This makes the model generally applicable to other power plant installations. The model was validated against process data related to the three start-up types (cold start, warm start, hot start). On this basis, the optimization model is implemented with feedback loops that control, for example, the exit temperature of the gas turbine based on the actual steam turbine housing temperature, resulting in a smoother heating up of the steam turbine. The optimization model was used to define the optimal inlet guide vanes position and gas turbine power output curves for the three types of start-up. These curves were used during real power station start-ups, leading to, for cold and warm starts, reductions in the start-up time of, respectively, 32.5% and 31.8%, and reductions in the fuel consumption of, respectively, 47.0% and 32.4%. A reduction of the thermal stress in the steam turbines is also achieved, thanks to the new start-up strategy.