New techniques for the analysis of flexible operation of gas turbine based systems

Abstract

In the current European energy market, gas power plants are required to operate in cyclical modes to fill the gaps in renewable energy supply. Renewable sources have dispatch priority due to their relatively low variable operational costs. However, because of their high unpredictability, conventional power plants such as Combined Cycle Power Plants (CCPP) now operate with frequent load changes to fill the gaps in supply by participating in the balancing market. Substantial efforts to develop innovative solutions to the new challenges are invested by the commercial and research community, where investigation into improving understanding of complex part-load operation is of utmost techno-economical importance. To date, main techniques used to simulate part-load operation of CCPPs were developed in the late twentieth century and are based on cumbersome and iterative methods requiring initial approximation of variables. In the wake of recent large scale renewable power installations, these techniques are not effective enough to carry complex optimisation studies to adopt CCPPs to quickly evolving market conditions. A number of improvements have been proposed; however, these modified methods are not able to cope with the required complexity and flexibility of studying various component layout optimisations and their impact on techno-economic performance. The current work pursues a novel method for part-load performance estimation of CCPPs, which is less complex, more effective, and can be seamlessly applied to any further optimisation studies. Initially the technique has been developed based on binarycoded genetic algorithm. The method enables simulation of part-load performance without the need for making initial guess of variables, thus simplifying the procedure. The method has been validated against commercial software showing good agreement in the results. However, it has been concluded that the method does not provide a long term benefit to the research community because it is fundamentally based on search space iterations with unavoidable residual (error) in the solution, and requiring significant computational time. The complex optimisation studies conducted by other authors require a much simpler and flexible method. This led to the development of a novel Direct Solution Method (DSM), which provides a simple solution with zero residual without need for cumbersome iterations. The DSM has been validated against commercial software showing good agreement; thus proving to be a promising alternative to the existing techniques. To improve understanding of part-load gas turbine operation, a set of comprehensive maps have been developed. A Gas Turbine Operational Map allows study and visualisation of complex trade-offs arising from gas turbine load reduction strategies. The load change strategy will determine the life consumption of critical gas turbine components, which led to the development of a Life Consumption Map which takes into account low cycle fatigue and creep mechanisms.