Redefinition of stack efficiency and optimization of stack performance for PEMFCs

Kurzfassung

More and more PEMFC stacks become commercially available and system integrators have to benchmark the stacks for the use in a system. This comparison should be based on the efficiency of different stacks for the conversion of the fuel chemical energy to useful electrical energy. The commonly used stack electrical- and fuel efficiencies do not take into account the operating conditions of the examined stacks. As a consequence, the determined efficiency can strongly differ for two stacks, even when their energy conversion performance in a system would roughly be comparable. This will happen if one of the stacks is examined close to ambient conditions and the other one at high pressure and high stoichiometric values. In the first part of this work, we present a general redefinition of the PEMFC stack efficiency taking into account all power losses directly connected with the stack performance and the applied stack operating conditions. These are the stack fuel loss, the stack polarization- and reaction entropy- and enthalpy losses, and the theoretical losses for stack feed stream conditioning. The latter includes humidification and pressurization (or pumping) of the reactants as well as pumping of the coolant. Examples will be given which power losses are relevant for different applications and which are dominant. By the use of this new figure of merit, the efficiencies of two stacks can be compared to each other in a system-relevant way. In the second part of this work, a procedure for the optimization of the stack performance will be highlighted. The performance depends on the operating conditions characterized by 7 parameters: the temperature, the stoichiometric values, the relative humidity, and the gas pressures. The parameter effect on the performance is non-linear and synergistic. A separate optimization of each parameter is not meaningful. Therefore, a direct-search algorithm, the Nelder-Mead simplex, was used for the simultaneous optimization. The optimization process will be demonstrated and explained. Examples using different test output parameters will be shown. It will be demonstrated that an optimization for the stack voltage can easily result in operating conditions characterized by high pressure, high stoichiometric values and high humidity. These conditions also result in an increased power consumption of the balance-of-plant components. To consider these losses, the redefined stack efficiency introduced above can be used. We will present the results of an optimization process on the stack level which leads directly to an increase of the efficiency with a maximum of 4 %.