Modelling and Numerical Analysis of the Centrifugal Compressor and its Operation in the Pressurised SOFC GT Process

Abstract: Small-scale power plants based on the SOFC-GT (Solid Oxide Fuel Cell-Gas Turbine) system are foreseen for future application such as APU units in flight transportation systems or stationary power production. Although these new systems offer high load flexibility and potentially high efficiency, there are several constraints referring to thermal/mechanical stress of the fuel cell as well as to a potential risk of aerodynamic instability of the GT compressor due to the different physical behavior of the SOFC and the GT. In order to better understand the different physical interaction of the SOFC and GT sub-components, one aim of this thesis is the numerical modelling of the compressor and its steady-state and transient system operation in a pressurized SOFC-GT system. Focus regarding the system modelling was set on the determination of compressor specification and its constraints during transient operation. For dynamic system modelling, a simulation program has been developed based on a benchmarked 1-D SOFC model and a modified low-order dynamic compressor system model, the Moore-Greitzer model. The transient compressor operation line was analyzed regarding the risk of surge during system load changes and shut down/start-up in dependence of different compressor system parameters. The risk of compressor surge was shown for abrupt load reduction at variable speed as well as for stepwise load reduction at constant speed. For reasonable physical results regarding the transient compressor operation, emphasis was set on the accurate modelling of the GT components by implementation of compressor characteristics either based on real component data or based on 1-D models. Validated 1-D and 3-D compressor design tools have been developed and used for a preliminary design of a compressor that matches the dynamic and steady-state requirements of the system analyzed in this thesis. Results of a 1-D parameter study have been benchmarked against 3-D results with satisfactory agreement between them. Postprocessed data from the validated CFD model have been utilized to improve the two-zone model prediction results.