COMBINATION OF A NONLINEAR STATIC AND A LINEAR DYNAMIC MODEL OF THE NETL HYPER SYSTEM

A nonlinear steady-state thermodynamic model was coupled with linearized dynamic transfer functions to achieve a dynamic description of the NETL HyPer Fuel Cell Gas Turbine (FC/GT) power plant. Nonlinear dynamic models insure accuracy in modeling steady-state behavior over a wide range of operation, but such models are often complex and difficult to implement in real-time using conventional control systems equipment. Conversely, the linearized models provide the ability to predict transient behavior upon which dynamic control systems can be constructed, but are valid only about a narrow operating point. In systems with one or two state variables, it is relatively straightforward to construct controllers that use gain scheduling schemes. But the HyPer system contains many coupled state variables and high degrees of nonlinearity. A method called Real-Time Piecewise Linear Dynamic Modeling (RPLDM) has been implemented to provide both modeling accuracy and real-time performance for the HyPer system over a multi-dimensional hypersurface. Both the nonlinear and the linear constituent models were constructed based on experimental data collected in tests performed on the HyPer system. The models presently consider only the cathode circuit of the fuel cell and contain a recuperated gas turbine system equipped with an electric generator, a simulated fuel cell cathode and various bypass valves for thermal management and system control. The key variables of air temperature, air pressure and mass flow to the cathode of the fuel cell and the turbomachinery have been predicted to within 2% of measured values. This paper presents the modeling technique and comparisons of the model output with experimental data.