Pang-Chieh Sui

Pore Scale Simulation of Transport and Electrochemical Reactions in Reconstructed PEMFC Catalyst Layers

A mesoscale simulation is developed to simulate transport and electrochemistry in a small section of a proton exchange membrane fuel cell (PEMFC) cathode catalyst layer. Oxygen, proton, and electron transport are considered in the model. Many simulations are run with a wide variety of different parameters on stochastically reconstructed microstructures with a resolution of 2 nm. Knudsen diffusion plays an important role in limiting the transport of oxygen through the catalyst layer. Using larger carbon spheres in the catalyst layer increases the effective diffusivity of oxygen through the catalyst layer. The effective proton conductivity increases when larger spheres are used, a normal distribution of spheres is used, or a higher overlap tolerance is used. Increasing the overlap tolerance or overlap probability results in an increase in the effective electron conductivity. When electrochemical reactions are considered in a part of the catalyst layer that is close to the gas diffusion layer, the critical parameter that determines oxygen consumption is the carbon sphere radius. Oxygen consumption at a given carbon volume fraction is larger in microstructures containing spheres with smaller radii, because there is more surface area available for electrochemical reactions.

Transport phenomena in fuel cells: from microscale to macroscale

Fuel cells have emerged as one of the most promising energy conversion technologies to help mitigate pollution and greenhouse gas emissions. This relatively young and rapidly evolving technology offers scope for innovation in both computational modelling and design. The operation of a fuel cell depends on the optimised regulation of the flow of reactant gases, product water, heat and charged species in conjunction with reaction kinetics. These strongly coupled processes take place over a broad range of length and time scales, and in diverse structures and materials. This gives rise to a fascinating and challenging array of transport phenomena problems. This paper provides an overview of these transport phenomena in polymer electrolyte membrane fuel cells, and a critical discussion of computational strategies to resolve processes in key components: polymer electrolyte membrane, porous gas diffusion electrodes and microchannels. The integration of the various transport phenomena and components into a CFD framework is illustrated for single fuel cells and for manifolding and gas distribution in a stack. Multi-scale strategies and the coupling of CFD based models to multi-variable optimisation methods are also discussed and illustrated for catalyst layers. The paper closes with a perspective on some of the pacing items toward achieving truly functional computational design tools for fuel cells.

Analysis of Water Transport in Proton Exchange Membranes Using a Phenomenological Model

An investigation of water transport across the membrane of a proton exchange membrane fuel cell is performed to gain further insight into water management issues and the overall behavior of a representative phenomenological model. The model accounts for water transport via electro-osmotic drag and diffusion and is solved using a finite volume method for a one-dimensional isothermal system. Transport properties including the water drag and diffusion coefficients and membrane ionic conductivity are expressed as functions of water content and temperature. An analytical solution based on a generalized form of the transport properties is also derived and used to validate the numerical solutions. The effects of property variations on the water flux across the membrane and on the overall membrane protonic conductivity are analyzed. The balance between transport via electro-osmotic drag and diffusion depends not only on operating conditions, such as current density and relative humidity at the membrane boundaries, but also on design parameters, such as membrane thickness and membrane material. Computed water fluxes for different humidity boundary conditions indicate that for a thick membrane (e.g., Nafion 117), electro-osmotic drag dominates the transport over a wide range of operating conditions, whereas for a thin membrane (e.g., Nafion 112), diffusion of water becomes equally important under certain humidification conditions and current densities. Implications for the resolution of membrane transport in CFD-based models of proton exchange membrane fuel cells are also discussed. DOI: 10.1115/1.1895945

Modelling and Simulations on Mitigation Techniques for Carbon Oxidation Reaction Caused by Local Fuel Starvation in a PEMFC

Experimental studies have shown that carbon oxidation reaction (COR) can be induced due to local fuel (H2) starvation, as well as during startup and shutdown procedures, e.g. [1]. It is also found that carbon corrosion is important to performance degradation of PEMFCs [2]. When COR occurs, the carbon support in the catalyst layer is reduced in volume and connectivity, which causes Pt particles to become isolated and less effective. The catalyst layer then suffers from reduction of electrochemical area (ECA), which leads to performance degradation. In the literature, several effective techniques for mitigating COR have been reported: (1) Use catalysts , e.g. Pt-Ir/C, that are favorable for the oxygen evolution reaction (OER); (2) Lower O2 membrane permeability that reduces O2 cross-over to anode; (3) Use of corrosion resistant carbon support materials, e.g. graphitized carbon black; (4) Increase of proton conductivity the catalyst layer. To the best of our knowledge, there have been few works published thus far on modelling and numerical simulation on these mitigation techniques, and quantitative information of these mitigation techniques is lacking. The objective of the present study is to investigate numerically on the mitigation techniques for the COR caused by local fuel starvation. The simulation procedure is based on our previous work on the simulation of COR caused by local starvation [4]. In essence the numerical procedure solves the conservation equations of gas species in a computational domain that consists of an MEA and gas channels of a unit cell. A portion of the domain has a prescribed zone with low gas diffusivity to mimic blockage by liquid water, which creates a local starvation region. Electrochemical reactions including ORR, HOR, OER and COR are considered in the catalyst layers. Figures 1(a) and 1(b) compare simulation results with high and low concentration of OER favorable catalyst in the catalyst layer. The concentration of the OER favorable catalyst is adjusted by changing the OER volumetric exchange current density in the computation. From Fig. 1(a) & (b), one can see for both cases the total reversal current is similar but the contribution of COR current differ significantly. The COR current density is much lower when a high concentration of OER favorable is used. This indicates that the COR reaction is inhibited due to the OER reaction. It is noted that OER reaction is believed to have minimum impact to cell performance and durability. Therefore, if the reversal current caused by local starvation is dominated by OER current, cf. Fig. 1(a), degradation caused by COR during local fuel starvation can be mitigated. Figure 2 summarizes numerical results for the four COR mitigation techniques. One can see that both OERfavorable catalyst and corrosion resistant carbon support materials are effective ways to mitigate carbon corrosion caused by local fuel starvation. In conclusion, based on our general local starvation model, four mitigation techniques are evaluated numerically and the effects of these mitigation techniques are determined quantitatively. It is found that using OER favorable catalyst and using corrosion resistant carbon support materials are the most effective mitigation techniques.

CFD and Flow Network Analysis of Manifolding in a PEMFC

Near-uniform flow distribution in a fuel cell stack is essential to stack performance and overall system efficiency. The gradients induced by the non-uniformity of the flow within each of the unit cells also have a significant impact on stack durability. In typical configurations, the oxidant and fuel are fed into a stack through manifolds and then enter each unit cell through secondary inlet port. After flowing through the unit cells, the spent gases as well as possible liquid water then enter the outlet header to leave the stack. The objective of this paper is to develop a practical model to predict cell-to-cell flow distribution in a proton exchange membrane fuel cell (PEMFC) stack. The flow distribution is first simulated using a computational fluid dynamics (CFD) tool, CFD-ACE+, in a 3D computational domain for single-phase gas flows. The simulations use a domain encompassing the flow from the inlet header through an array of unit cells to the outlet header. The CFD simulations show that in the outlet header, the flow injected from the unit cells to the header changes the flow pattern considerably, which results in a reduced cross section area for the flow in the axial direction. A circulation zone is seen near the low velocity end of the header, which may potentially become a region where liquid water accumulates. Increasing static pressure along the flow direction is observed in the inlet header. The simulated results are validated and found to be in good agreement with experimentally measured pressures in a fuel cell stack. Based on the observations in the CFD simulations, a flow network model is developed to provide quick estimates of the flow distribution as a function of stack dimensions including header and unit cell geometry. In essence, the flow network model solves for the pressure at each junction of the unit cell and the header. Three fitting parameters are introduced to account for effects of surface roughness of the headers, reduced effective header area in the outlet header, and pressure drop in the unit cell. The flow network model is shown to capture the characteristics of pressure variation and flow distribution obtained in the CFD simulations. The flow network model can effectively match experimental data and be used as a fast tool for initial design of a PEMFC stack.Copyright

Design and Optimization of the Gas Channels of a PEMFC Using CFD-Based Simulation

The flow field plate of a proton exchange membrane fuel cell (PEMFC) functions as electron conductor and provides the pathway for oxidant and fuel to reach the membrane electrode assembly (MEA). CFD-based simulation tools can be effective in designing and optimization of flow field plates as they cab fully account for the complexity and coupling of various transport phenomena as well as the 3-D geometry. The objective of this paper is to report on the development of such a simulation platform and on its application to investigate the impact of several geometric parameters on fuel cell performance and detailed distribution of transport processes. The simulation tool is built upon a commercial computational fluid dynamics (CFD) code, CFD-ACE+, along with supporting software and script codes to automate the design workflow. A 3-D, straight channel model with material properties and model parameters validated with experimental data is used as the baseline for the present study. The workflow includes automated grid generation, model setup and job execution. Parametric study is performed for geometric parameters including (1) Channel width versus land area width (2) Channel height (3) Channel pitch and length, as well as material parameters including (4) Porosity and (5) Electrical conductivity of the gas diffusion layer (GDL). Among these parameters, it is found that predicted cell performance is most sensitive to the channel/land width ratio and to the anisotropy of the GDL property. When isotropic properties are used for the GDL, the predicted cell performance decreases with increasing channel/land width ratio. This is because the current distribution in the MEA is dictated by electrical conduction through the GDL and increasing channel width causes current to peak underneath the land area, which in turn increases ohmic losses. When the in-plane electrical conductivity is reduced, the effect of mass transfer on the current distribution becomes comparable to electron transfer and the predicted trend line of cell performance shows an optimum value as a function of the channel/land width ratio. The CFD based design tool developed in the present work has the advantage of providing more reliable prediction than methods based on reduced dimensionality or simplified transport models.Copyright

Numerical Analysis of Water Transport in PEM Fuel Cell Membranes Using a Phenomenological Model

A numerical investigation on the water transport across the membrane of a proton exchange membrane fuel cell is carried out to gain insight into water management issues, which are crucial to the efficient operation of such fuel cells. The transport equation of water content based on a phenomenological model, which includes an electro-osmotic drag term and a diffusion term, is solved using the finite volume method for a 1-D configuration with the assumption of a uniform temperature distribution. Transport properties including the drag coefficient and diffusion coefficient of water in the membrane and the ionic conductivity of the membrane are expressed as functions of water content and temperature. The effects on the water flux across the membrane and on overall membrane protonic conductivity due to variations of these properties are studied. The numerical results show that water transport in the membrane is mainly determined by the relative strength of electro-osmotic drag and diffusion, which are affected by operating conditions such as current density and relative humidity at the membrane surface, and design parameters such as membrane thickness and membrane material. Computed water fluxes for different humidity boundary conditions indicate that for a thick membrane, e.g. Nafion 117, electro-osmotic drag dominates transport over a wide range of operating conditions, whereas for a thin membrane, e.g. Nafion 112, diffusion of water becomes equally important under certain conditions. Implications of the one-dimensional investigation on comprehensive CFD based modelling of proton exchange membrane fuel cell are also discussed.Copyright

Effect of Channel Geometry on the Dynamics of a Water Droplet in a Microchannel

The objective of the present study is to investigate the effects of the microchannel geometry on the dynamic behaviour of liquid water emerging from a pore into a microchannel of a cross gas flow. The flow characteristics are resolved using the volume-of-fluid (VOF) method in conjunction with an interface tracking technique. A microchannel with dimensions of a typical proton exchange membrane fuel cell (PEMFC) gas channel (a square cross section of 250 μm in width) and a pore of 50 μm in diameter on the bottom wall is adopted as the baseline case. Simulations for microchannels of different cross sections, including trapezoid, upside-down trapezoid, triangle, rectangle, and rectangle with a arch bottom wall, are performed and the results are compared with the baseline case. The evolution of liquid water includes stages identified as emergence, growth, deformation, detachment, and remove. The simulations show that the cross section of the microchannel has significant impacts on the dynamics of the water droplet. The detachment time and diameter and the remove time of the water droplet are found to be in this order: triangle < trapezoid < rectangle with arch bottom wall < rectangle < upside-down trapezoid. The present study will advance our understanding in the transport of liquid water in a PEMFC where water is produced in the catalyst layer and flows through the pores of the porous electrode to the gas channel.Copyright

A Pore Scale Model for the Transport Phenomena in the Catalyst Layer of a PEM Fuel Cell

In a proton exchange membrane fuel cell (PEMFC), the catalyst layer is a porous medium made of carbon-supported catalysts and solid electrolyte, and has a thickness in the order of 10 μm. Within this layer, complex transport phenomena take place: transport of charged species (H+ , electrons and ionic radicals), non-charged species (gaseous H2 O, O2 , H2 , N2 and liquid water) and heat transfer occur in their own pathways. Furthermore, phase change of water and physiochemical/electrochemical reactions also take place on phase boundaries. These transport process take place in an intertwined network of materials having characteristic length scale ranging from nano-meters to micro-meters. The objective of the present study is two-fold, i.e., to develop a rigorous theoretical framework based on which the transport in the micro-structural level can be modelled, and to construct a pore scale model that resolves the geometry of the phases (carbon, ionomer and gas pores) for which direct numerical simulation can be performed. The theoretical framework is developed by employing the volume-averaging techniques for multi-phase porous media. The complete set of the conservation equations for all species in all phases are derived and every interfacial transport is accounted. The problem of model closure on the terms in the transport equations is addressed by the pore-scale model reported in the present study. A 3-D pore-scale model is constructed by a solid model that consists of packing spherical carbon particles and simulated ionomer coating on these carbon aggregates. The index system of the pore-scale model allows easy identification of volumetric pathway, interfaces and triple phase boundaries. The transport of charged and non-charged species is simulated by solving the equations based on first principle in the entire representative element volume (REV) domain. The computational domain contains typically several million cells and a parallelized, iterative solver, GMRES, is employed to solve the coupled transport with complex geometries. Computational results based on the pore-scale model show that the effective transport properties of the species are strongly affected by the micro-structure, e.g. morphology and phase-connectivity. Further simulations and investigation on the coupling effects of the transport are underway. Combination of the proposed theoretical framework and pore-scale model will lay a foundation for the construction of multi-scale modelling of the PEMFC catalyst layer. On the one hand, the pore-scale model helps close the macroscopic volume-averaged equations in the framework. On the other hand, the pore-scale model provides a platform to include microscopic or atomistic simulations.Copyright