Comparison between Different Activation Overvoltage Descriptions for Semiempirical Proton-Exchange Membrane Fuel Cell Models

Using parameter estimation along with semiempirical fuel cell models is a promising strategy to obtain simple yet accurate models for proton-exchange membrane fuel cells. This work compares a model using a common semiempirical description of the activation overvoltage with one based on an agglomerate model. The parameters are estimated by nonlinear regression and the resulting optimization problem is solved using a parallel implementation of the genetic algorithm. The results indicate that the model incorporating the agglomerate description offers parameters with more straightforward physical interpretations and achieves superior fits to the experimental polarization data, both for individual curve fitting and when multiple curves are fitted simultaneously. Moreover, an analysis of the overvoltages shows that it is more robust in modeling reactant depletion effects at high current densities, attesting to its ability to represent relevant phenomena reasonably. Thus, further usage of parameter estimation along with agglomerate models is recommended as a useful strategy to describe fuel cells with limited data availability.