Cost and durability of polymer electrolyte fuel cells (PEFCs) remain the major hurdles in the way of their commercialization. To optimize cell performance at high current densities for low loading cells, mathematical modeling provides an ideal framework. A computational model with accurate implementation of fuel cell physics is useful in optimizing the cell parameters for improved performance and low cost, while saving time and cost associated with experimental studies.

The aim of current work is to develop a detailed computational model for PEFC simulation under varying operating conditions. A computational model has been developed in COMSOL to simulate a two-dimensional cross section of PEFC. The model accounts for multicomponent diffusion, gas and liquid convection, electronic and protonic conduction, and water transport in the membrane. A novel agglomerate model with non-first-order double trap (DT) kinetics and local film resistance has been implemented in the numerical model for cathode oxygen reduction reaction (ORR)¹. The DT kinetics helps in accounting for the variable Tafel slope. The two-dimensional model has been used to model cells at high stoichiometry where the concentration and relative humidity vary only marginally over gas channels. ^2,3

While the two-dimensional model is sufficient for differential-cell conditions, a three-dimensional model is required for modeling all possible cell conditions and more complex flowfield geometry. Since a full three-dimensional model is computationally expensive, a 2+1D approach is proposed as shown in Figure 1. The fuel cell is divided into several sections along the flow path of reactants. Each section is modeled using the developed COMSOL model as a 2-D slice, as shown in Fig 1 (b). Using the simulated results in the 2-D cross section, total reactant consumption and production within the 3D slice are calculated. The inlet concentrations, relative humidity, and flow rates for the next 2-D slice are then computed using mass and energy balances. While this approach has limitations in accounting for gradients in the z-direction (flow direction), nevertheless it can account for concentration and humidity changes along it.

In this paper, such changes will be simulated and compared to segmented-cell data for PEFCs including both straight channel and serpentine cells. In addition, the possibility of defects in thicknesses and loadings will be theoretical explored. The extension of the model to include liquid flow in the channels and droplets will be discussed. Changes in humidity are expected to be even more limiting in hydroxide-exchange-membrane fuel cells due to the stronger dependence of hydroxide conduction with relative humidity. The model will be adapted for these systems to explore how changes down the channel impact overall cell performance.

Acknowledgements

The authors would like to thank Dusan Spernjak at Los Alamos National laboratory for providing segmented cell data. The work was funded under the Fuel Cell Performance and Durability Consortium (FC-PAD), by the Fuel Cell Technologies Office (FCTO), Office of Energy Efficiency and Renewable Energy (EERE), of the U.S. Department of Energy under contract number DE-AC02-05CH11231. In addition, we thank helpful discussions with Mike Ulsh, Guido Bender, and Adam Phillips.

References

1. L. M. Pant and A. Z. Weber, Journal of The Electrochemical Society, 164, E3102 (2017).

2. M. L. Perry, R. M. Darling, L. M. Pant and A. Weber, in 231st Electrochemical Society Meeting, New Orleans (2017).

3. I. V. Zenyuk, P. K. Das and A. Z. Weber, Journal of The Electrochemical Society, 163, F691 (2016).