Transient Water Transport in Nafion Membranes Under Activity Gradients

Transient water transport experiments on Nafion of different thicknesses were carried out in the temperature range of 30 to 70 °C. These experiments report on water transport measurements under activity gradients in the time domain for liquid and vapour equilibrated Nafion membranes. Using a permeability test rig with a gated valve, the water crossover was measured as a function of time. The typical response is shown as a time dependent flux, and it shows the dynamic transport from an initially dry condition up to the final steady state. Contrarily to previous reports from dynamic water transport measurements, where the activity gradient across the membrane is absent; in this work, the membrane was subjected to an activity gradient acting as the driving force to transport water from an environment with higher water activity to an environment with lower water activity through the membrane’s structure. Measurements explored temperature and membrane thickness variation effect on the transient response. Results showed dependency on temperature and a slower water transport rate across the vapour-membrane interface than for the liquid-membrane interface. These measurements showed the transport dependency on water content at the beginning of the experiment when the membrane was in a close-to-dry condition suggesting a transport phenomenon transition due to a reached critical water content value. The new protocol for transient measurements proposed here will allow the characterization of water transport dependency on membrane water content with a more rational representation of the membrane-environment interface.

Pore Structure Modeling of Flow in Gas Diffusion Layers of Proton Exchange Membrane Fuel Cells

The gas diffusion layer (GDL) in a proton exchange membrane (PEM) fuel cell has a porous structure with anisotropic and non-homogenous properties. The objective of this research is to develop a PEM fuel cell model where transport phenomena in the GDL are simulated based on GDL’s pore structure. The GDL pore structure was obtained by using a scanning electron microscope (SEM). GDL’s cross-section view instead of surface view was scanned under the SEM. The SEM image was then processed using an image processing tool to obtain a two dimensional computational domain. This pore structure model was then coupled with an electrochemical model to predict the overall fuel cell performance. The transport phenomena in the GDL were simulated by solving the Navier-Stokes equation directly in the GDL pore structure. By comparing with the testing data, the fuel cell model predicted a reasonable fuel cell polarization curve. The pore structure model was further used to calculate the GDL permeability. The numerically predicted permeability was close to the value published in the literature. A future application of the current pore structure model is to predict GDL thermal and electric related properties.

RIB/Channel Effect on Cell Performance Under High Current Density Operation

Heat and water transport in polymer electrolyte membrane fuel cell (PEMFC) has considerable impacts on cell performance under high current density which is desired in PEMFC for automobiles. In this study, the impact of rib/channel, heat and water transport on cell performance under high current density was investigated by experimental evaluation of liquid water distribution and numerical validation. Liquid water distribution between rib and channel is evaluated by Neutron Radiography. In order to neglect the effect of liquid water in channel and the distribution of oxygen and hydrogen concentration distribution along with channel length, the differential cell was used in this study. Experimental results show that liquid water under channel was dramatically changed with Rib/Channel width. From numerical study, it is found that the change of liquid water distribution was strongly affected by temperature distribution between rib and channel. In addition, not only heat transport but also water transport through membrane also significantly affected cell performance under high current density operation. From numerical validation, it is concluded that this effect on cell performance under high current density could be due to the enhancement of back-diffusion of water through membrane.

Predicted Liquid Water Saturation in Unstructured Pore Networks Based on PEMFC GDL Porosity Profiles

An unstructured, two-dimensional pore network model is employed to describe the effect of through-plane porosity profiles on liquid water saturation within the gas diffusion layer (GDL) of the polymer electrolyte membrane fuel cell. Random fibre placements are based on the porosity profiles of six commercially available GDL materials recently obtained through x-ray computed tomography experiments. The pore space is characterized with a Voronoi diagram, and invasion percolation-based simulations are performed. It is shown that water tends to accumulate in regions of relatively high porosity due to the lower associated capillary pressures. It is predicted that GDLs tailored to have smooth porosity profiles will have fewer pockets of high saturation levels within the bulk of the material.

Simulation Comparison of PEMFC Micro-Cogeneration Units With Conventional and Innovative Fuel Processing

This paper deals with the performance comparison over simulated micro-cogeneration units based on polymer electrolyte membrane fuel cells (PEMFC or PEM), when the fuel is processed by means of two contrasting techniques. On the one hand with the use of conventional natural gas steam reforming (SR), and on the other, the adoption of an innovative palladium based membrane-reformer. After the definition of the plant layout, which reflects the results of previous studies and includes all the components of a 4 kW PEM for combined heat and power production, the comparison among the plant performances is carried out with two approaches: (i) using a in-house developed code (GS), able to calculate mass and energy balances, as well as a number of specific component parameters, already applied to a large variety of plant simulations, and (ii) using a commercial code (Aspen Plus®). The comparison allows to validate the simulated performance results as well as to evidence the advantages of the two approaches and to assess the effects of different simulation assumptions.

Study on Catalytic Reactions in Solid Oxide Fuel Cells With Comparison to Gas Phase Reactions in Internal Combustion Engines

Solid oxide fuel cells (SOFCs) have the attractive feature to be able to make use of hydrocarbon fuels in their operation by reforming the fuel into pure hydrogen, either internally or externally. This can open up for a smoother transition from the existing hydro-carbon economy toward a more renewable hydrogen economy. Since both SOFCs and internal combustion (IC) engines can make use of hydrocarbon fuels, it is of interest to examine the major differences in their utilization of the hydrocarbons and investigate how this type of fuel contributes to the power output of the respective systems. Thereby, various advantages and disadvantages of their reactions are raised. It was shown that even though there are fundamental differences between SOFCs and IC engines, both types face similar problems in their designs. These problems mostly include material design and operation management, but even problems related to the chemical reactions, e.g., carbon deposition for SOFCs and pollutant formation for IC engines.

Effect of MEAs Preparation Procedure on Their Performance in Room Temperature DMFC

In this work the effect of the MEA preparation techniques on the performance of DMFC was evaluated using three different methods of electrocatalyst deposition: i) catalyst coated membrane; ii) catalyst coated carbon paper; and iii) decal deposition. Optimization of the nafion content (5–15 wt. %) at anode and cathode sides of the MEA and the pressure (150–500 atm) were also performed. Activities of both supported and unsupported Pt and PtRu catalysts (Johnson Matthew) were compared in room temperature DMFC (RT-DMFC) using polarization curves. All MEAs prepared were also characterized by electrochemical (cyclic voltammetry, impedance spectroscopy) methods. It was shown that optimal nafion content is 5–10 wt. % at both anode and cathode sides, while the optimal pressure is in the 300–500 atm. range. The unsupported catalysts showed slightly higher power density at RT-DMFC (∼ 14 mW/cm2) as compared to the supported ones (∼10 mW/cm2) at the same Pt load. Variation of the wetness of MEAs upon mounting in DMFC allowed us to increase of the power density of RT-DMFC up to 32 mW/cm2.

A Dry-Borohydride/Injected-Hydrogen-Peroxide Fuel Cell

The design and testing of a 20-W (average power with short pulses to 45W) prototype fuel cell is presented. This cell is intended as an auxiliary power supply for a small robotic vehicle. The energy density exceeds 300 Watt-hour/kg. This cell is essentially a dry-borohydride/injected-hydrogen-peroxide fuel cell. This enables extremely long shelf life prior to use. The anode utilizes dry NaBH4 for storage while the cathode chamber is empty during storage. The initiation of cell operation is done by injection of the oxidizer, an aqueous H2O2 solution (stored in a separate container) to the cathode side of the fuel cell. The ionic conduction required for membrane operation is initially helped by the H2O content from the H2O2 solution. Once the electrochemical reaction starts, more water is generated as the reaction product and this continues to maintain a good ionic conductance over the run time of the cell. Continued operation is done with auxiliary fuel tanks to maintain very long run time when required. Once a run is over, the cell can be drain, flushed clean and returned to storage waiting for the next mission. The experimental details of such a cell stack are described in this paper.

CFD Modeling of a Catalytic Flat Plate Fuel Reformer for Hydrogen Generation

In this study, steam reforming of methane coupled with methane catalytic combustion in a catalytic plate reactor is studied using a two-dimensional mathematical model for co-current flow arrangement. A two-dimensional approach makes the model more realistic by increasing its capability to capture the effect of parameters such as catalyst thickness, reaction rates, inlet temperature and velocity, and channel height, and eliminates the uncertainties introduced by heat and mass transfer coefficients used in one-dimensional models. In our work, we simulate the entire flat plate reformer (both reforming side and combustion side) and carry out parametric studies related to channel height, inlet velocities, and catalyst layer thickness that can provide guidance for the practical implementation of such design. The operating conditions chosen make possible a comparison of the catalytic plate reactor and catalytic combustion analysis with the conventional steam reformer. The CFD results obtained in this study will be very helpful to understand the optimization of design parameters to build a first generation prototype.

Nonlinear Analysis of Two-Phase Instabilities in PEMFC Parallel-Channel Cathodes

Liquid water management is a critical issue in the development of proton exchange membrane (PEM) fuel cells. Liquid water produced electrochemically can accumulate and flood the microchannels in the cathodes of PEM fuel cells. Since the liquid coverage of the cathode can fluctuate in time for two-phase flow, the rate of oxygen transport to the cathode catalyst layer can also fluctuate in time, and this can cause the fuel cell power output to fluctuate. This paper will report experimental data on the voltage loss and the voltage fluctuations of a PEM fuel cell due to flooding as a function of the number of parallel microchannels and the air flow rate stoichiometric ratio. The data was analyzed to identify general scaling relationships between voltage loss and fluctuations and the number of channels in parallel and the air stoichiometric ratio. The voltage loss was found to scale proportionally to the square root of the number of channels divided by the air stoichiometric ratio. The amplitude of the fluctuations was found to be linearly proportional to the number of microchannels and inversely proportional to the air stoichiometric ratio squared. The data was further analyzed by plotting power spectrums and by evaluating the non-linear statistics of the voltage time-series.