VGB PowerTech

Solar gas turbine (GT) systems provide the opportunity to utilize solar heat at a much higher temperature than solar thermal power plants based on steam turbine cycles. Therefore gas turbine technology has the potential to improve the efficiency of future solar thermal power plants. Nevertheless, to achieve mature technology for commercial application, further development steps are required. Knowledge of the operational behavior of the solar GT system is the basis for the development of the systems control architecture and safety concept. The paper addresses dynamic simulation of high solar share GT systems, which are characterized by primary input of solar heat to the gas turbine. To analyze the dynamic operating behavior, a model with parallel arrangement of the combustion chamber and the solar receiver was set up. By using the Heaviside step function, the system dynamics were translated into transfer functions which are used to develop controllers for the particular system configuration. Two operating conditions were simulated to test the controller performance. The first case is the slow increase and decrease of solar heat flow, as part of a regular operation. The second case is an assumed rapid change of solar heat flow, which can be caused by clouds. For all cases time plots of critical system parameters are shown and analyzed. The simulation results show much more complex system behavior compared to conventional GT systems. This is due to the additional solar heat source, large volumes and stored thermal energy as well as the time delay of energy transportation caused by the piping system.

Based on an overview of the political targets, market constraints, and system requirements, this chapter highlights the technical and economic challenges for the conventional fleet and concludes that conventional power plants will also have a significant role in delivering in the coming decades: on the 20/20/20 targets by co-combustion of biomass and by combined production of electricity and heat a substantial part of the European supply of district heating and together with nuclear and hydro: the residual load stability to the grid by compensating the intermittent production from the renewables security of supply when sun and wind are not available system services: black start, primary/secondary/tertiary control and voltage regulation of the grid. The key issue is for politicians, suppliers, and operators to develop jointly the technology and the electricity market to make this possible.

During the last decades the research within the field of local approach (LA) models lead to an increase of the predictive capabilities of these models and thus extended the range of applicability. Since these LA models were originally developed for cleavage fracture, they had to be adopted to account also for ductile fracture events in order to assess fracture behavior in the ductile-to-brittle transition (DBT) region. This article presents a case study, where a recently developed and validated LA model is applied to predict failure of a RPV under a realistic transient loading situation. Within this case study a loss of coolant accident is considered, where a postulated subclad flaw with 20mm depth in the nozzle corner is examined. Because the RPV steel in the nozzle corner region is ductile during a LOCA event, no failure would be predicted using original material properties. Therefore the base metal is simulated using unrealistic high embrittled material properties, i.e. irradiated at a fluence of 3.5E19n/cm2 and an irradiation temperature of TIRR = 150°C., which is to be a very conservative material condition. The reason to use this conservative material condition is to obtain a large shift in transition temperature and the necessary brittle material condition to allow the application of the LA model. The outcome of the LA simulation is compared to other constraint correction methods, like the T-Stress correction. It can be stated that: • the conventional method, which uses no constraint corrections, leads to very conservative results. • T-Stress and T0-corrections are less conservative than the conventional method. • the prediction results attained by means of the LA model describes the failure behavior well; a large safety margin between the different concepts can be shown.