Modeling of transport phenomena in proton-exchange-membrane fuel cell
Date of Issue2010
School of Mechanical and Aerospace Engineering
Fuel cells are electrochemical devices whose performance is closely related to the transport of reactants (oxygen and hydrogen) and products (water). These transport processes are coupled with electrochemistry and further complicated by phase change, porous media (gas diffusion electrodes) and a complex microstructure. Several relatively simple yet representative models have been presented to better understand the fundamental of transport phenomena in PEMFC and to investigate the impact of various operating and design parameters on the performance to guide the optimal design of a fuel cell. These covered almost all major transport phenomena, including water and proton transport; electrochemical reaction; transport of electrons; heat generation within the cell; multi-component mass transport in the gas flow channels. A simple 1D model is presented and validated using in-house experimental data on a laboratory-scale single cell with 2-pass serpentine flow field. Results showed that the measured polarization curves generally agree well with the predicted results in the lower and medium current densities but significant discrepancy exist in the high current densities. Predicting the cell performance at high current density is a tough and challenging task. Simple 1D model is not able to provide the necessary insights on the spatial and temporal variation of the performance-related parameters in the cell, posing limit on the predictable range. A multi-dimensional model is, therefore, required for studying the complex interaction among these parameters and is also useful for one to use it to guide the design of a fuel cell. A 2D single-phase model for mass, momentum, species, charge, and energy transport coupled with a phenomenological membrane model and agglomerate model in the cathode catalyst layer for a PEMFC has been derived, discussed and validated with in-house experimental data obtained from a single cell with porous flow field. Two GDL cases (single layer GDL and CFGDL) have been used to calibrate and validate the models. Close agreement between predicted results and experimental data was obtained for the two cases in the entire current density range. In order to study the impact of these variations on stack performance, an extended mathematical model comprising two single cells and a coolant plate is considered. Several parameters are perturbed for one of the cells while the other is operating at base case conditions, an extension to larger stack units is also discussed. A simple extension model to account for two-phase flow has been derived, implemented and the associated key parameters (E*, ΔE, s*) are discussed to study ‘curve-back’ phenomena that appeared in the polarization curve. Partial and full flooding have been simulated, which include localized liquid water accumulation and uniform liquid water accumulation. A series of parametric studies have been conducted to identify the parameters (e.g. operating, geometric, intrinsic and extrinsic) that are sensitive or insensitive to the fuel cell performance. All the models have showed their predictive capabilities and various level of accuracy when validated with the experiment results and they are useful to guide the design of a fuel cell.