The aim of this work is to develop a real-time tool for an optimal control of a PEMFC system. The control has to maximize the output energy production of the system by computing and choosing the best operating conditions. The control method is based on a pseudo 2D PEMFC model. This model describes multi-component gas transport in the gas diffusion layers and in the bipolar plate channels. It also describes water transport in the membrane. The effects of gas and water on the cell potential are computed.

Optimal fuel cell operating conditions are analysed through the model. In order to achieve high performances, water management in the PEMFC is one of the main critical issues to address. Excess humidification (flooding) and the lack of humidification (drying) are thus harmful to the performances of the PEMFC. In this work, liquid water appearance at the center of the cell along the channel direction is assumed to correspond to optimal humidification conditions. Indeed, according to the literature, this is a good compromise between flooding and drying. The best gas humidification conditions are therefore computed for a given current demand. The influences of operating parameters, such as temperature, stoichiometric ratios or pressures, are analysed as well. Finally, dead-end and flow-through modes of hydrogen supply are compared.  

It is shown that the optimum relative humidity is highly dependant on current, temperature and oxygen stoechiometric ratio. Dead-end mode leads to water equilibrium between cathode and anode and therefore yields a good anode humidification without any anode input water. In many cases, the optimum cathode inlet relative humidity is low due to internal humidification caused by water production. So, the energy required by the humidification auxiliaries can be minimized.  

Consequently, the optimal operating parameters of a fuel cell are well chosen by using the developed PEMFC model. By modelling auxiliaries and their energy consumption, the efficiency of a complete PEMFC system can be established. Using a multivariable method, the optimal operating parameters can be set for a given current or power demand. Thus, an appropriate real-time strategy is being developed to run the fuel cell system more efficiently. This approach will improve fuel cell competitiveness compared to other technologies.