1D and 3D numerical simulations in PEM fuel cells

The potential of fuel cells for clean and efficient energy conversion is generally recognized. The proton-exchange membrane (PEM) fuel cells are one of the most promising types of fuel cells. Models play an important role in fuel cell development since they enable the understanding of the influence of different parameters on the cell performance allowing a systematic simulation, design and optimization of fuel cells systems. In the present work, one-dimensional and three-dimensional numerical simulations were performed and compared with experimental data obtained in a PEM fuel cell. The 1D model, coupling heat and mass transfer effects, was previously developed and validated by the same authors [1,2]. The 3D numerical simulations were obtained using the commercial code FLUENT - PEMFC module. The results show that 1D and 3D model simulations considering just one phase for the water flow are similar, with a slightly better accordance for the 1D model exhibiting a substantially lower CPU time. However both numerical results over predict the fuel cell performance while the 3D simulations reproduce very well the experimental data. The effect of the relative humidity of gases and operation temperature on fuel cell performance was also studied both through the comparison of the polarization curves for the 1D and 3D simulations and experimental data and through the analysis of relevant physical parameters such as the water membrane content and the proton conductivity. A polarization curve with the 1D model is obtained with a CPU time around 5 min, while the 3D computing time is around 24 h. The results show that the 1D model can be used to predict optimal operating conditions in PEMFCs and the general trends of the impact on fuel cell performance of several important physical parameters (such as those related to the water management). The use of the 3D numerical simulations is indicated if more detailed predictions are needed namely the spatial distribution and visualization of various relevant parameters. An important conclusion of this work is the demonstration that a simpler model using low CPU has potential to be used in real-time PEMFC simulations. Copyright (C) 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.