Energetic Macroscopic Representation Modeling and Control of a Low Temperature Fuel Cell System Fed by Hydrocarbons

Fuel cell systems (FCS) are considered to be upcoming technology for electrical power generation. They can be used for portable, stationary and transportation applications. Fuel cells are run on hydrogen. Hydrogen can be produced either by electrolysis using electricity from renewable energy or by converting conventional hydrocarbons into a hydrogen rich gas. Among others, a FCS incorporates two main components, the fuel processing unit and the fuel cell stack. The use of hydrocarbon fueled FCS as auxiliary power units (APU) in transportation applications is a possible entry market for this technology that utilizes the existing infrastructure of fuel supply. Hydrocarbon fueled FCSs are complex multi-domain systems combining aspects from various energetic regimes, like electro-chemical, electrical, pneumatical and thermal. These systems work only in a narrow and well defined range of operation. Therefore, this application requires a well adapted system control to respect these constraints. Classical control structure development often requires the transfer function of the system. This can be difficult or even impossible to derive for complex systems. Therefore, control structure development for complex multi-domain systems is often based on empirical observation and experience. It is desireable to find an approach that allows the development of a control structure based on system description. Such an approach will simplify control structure development and ensure that the control structure is adapted to the system needs. Model based control structure design is an approach that can meet these demands. This thesis presents a complete model of a low temperature FCS fueled by commercial diesel, and is well adapted for model based control development. The studied FCS provides 25 kW electric power, and at the same time the system waste heat is used for climatization. In chapter (2) several modeling methodologies are introduced. Each is evaluated