The design of the flow field is highly responsible for the performance of the Proton Exchange Membrane Fuel Cell (PEMFC). In this study, pin type flow channel is numerically analyzed by arranging carbon made porous material in uniform and zigzag manner on the rib surface of the flow field. The study focuses on enhancing the performance of PEMFC by reducing liquid flooding in the interface between the rib and Gas Diffusion Layer (GDL). A single PEMFC having an active area of 25 cm 2, with three flow channel designs (conventional serpentine, pin type flow channel with 2 mm cubical porous inserts in zigzag and uniform pattern) are modeled for the numerical analysis. The effect of porosity of the carbon inserts on the cell performance is studied by varying its value from 0.6 to 0.9. The results show that the performance of the flow channel with zigzag and uniformly positioned porous inserts is more than the conventional serpentine flow channel by 20.36% and 16.87% respectively. The reason for this increase is the removal of the accumulated water from the rib surface due to the capillary action of the porous carbon inserts. This helps in eliminating the stagnant water regions under the rib and thereby helps in reducing liquid flooding.
With the increasing problem of environmental degradation due to the harmful emissions from the existing fuels and the shortage of the non-renewable fossil fuels, alternative fuels and non-polluting sources of energy are the need of the hour. Fuel cells are a viable option for the future as they are portable and have almost no harmful emissions [
In a typical PEMFC, hydrogen gas is used as a fuel and is passed through the anode region, whereas an oxidant either oxygen or air is passed through the cathode region. Water is obtained as a by-product at the cathode end due to the electrochemical reaction between hydrogen and oxygen. The water obtained as a by-product has to be precisely balanced for the better performance of PEMFC. At the cathode side, oxygen reduction reaction (ORR) occurs in the catalyst-membrane interface and due to this water is generated. For the hydration of the membrane, inlet fuel and oxidant are supplied in a fully humidified state. Due to this, the gases in the fuel cell may get oversaturated with the water vapour and get condensed to form liquid water, thereby the amount of water content in the cathode region increases. Apart from this, the cathode region also gets water through electro-osmotic drag, wherein the positively charged protons tend to attract and drag the water molecules along with them, when migrating from anode to cathode flow channels. The water transport also takes place due to pressure drop as well as concentration difference. Hence, managing the produced water towards the cathode end becomes important. The water content towards the cathode end must be judiciously maintained in order to achieve high efficiency. At present, the biggest problem researchers are facing is the water management at the cathode side. If the water removal rate is greater than the generation rate, the membrane gets dehydrated and therefore results in the poor performance due to excessive ohmic losses [
Different methods have been proposed for managing the water in an effective manner and the most productive method found is to effectively designing the flow channel [
For solving this problem of cathode flooding, different researchers have suggested different techniques. Nguyen [
A three-dimensional model with three flow field patterns (conventional serpentine, pin type flow channel with the adoption of porous carbon inserts in uniform and zigzag pattern) has been evaluated on an active area of 25 cm2. As the problem of water flooding occurs towards the cathode side of the PEMFC, a modification is made in the cathode flow channel only. The modification is done by inserting porous carbon inserts in the uniform and zigzag pattern on the rib surface of the pin type flow channel. By placing these porous carbon inserts in a uniform and zigzag manner on the rib surface, the pin type flow channel becomes similar to a serpentine flow channel (dimension of porous inserts 2 mm × 2 mm × 2 mm). Therefore, it gains the merits of the serpentine flow channel.
The commercially available Fluid Dynamics software FLUENT 14.5 which is based on a control volume approach was used to solve the various governing equations. 3-dimensional, steady state, laminar, double precision, serial processing modes were used for this simulation. A special add-on module called “Battery and Fuel Cell” was used for this study. In order to get accurate results, a grid independency test was done and finally the results were found grid independent at about 650,000 elements. In order to simplify the analysis, the gases used were considered to be ideal. The flow was taken as steady, laminar, incompressible, and the system was considered as isothermal at 325 K. The various thermo physical properties were considered as constant and the effect of gravity was neglected. Also, the GDL, the catalyst layer, carbon inserts and the membrane were considered to be isotropic.
Part | Length (mm) | Width (mm) | Height (mm) |
---|---|---|---|
GDL | 50 | 50 | 0.3 |
Catalyst Layer | 50 | 50 | 0.06 |
Membrane | 50 | 50 | 0.15 |
Serpentine/Pin Type Flow Channel | 50 | 50 | 2 |
Current Collector | 80 | 80 | 10 |
In this model, the anode and cathode channel inlet zone were set as mass flow inlet respectively, the anode and cathode channel outlet zone were set as pressure outlet, exhaust-fan, outlet-vent whereas the other left surfaces were set as wall. The operating conditions for all the three patterns were set at a pressure of 1 bar and temperature of 325 K. Water produced was assumed to be fully liquid (100% humidified) and the mass fractions towards the anode side were set at 0.8, 0.2, 0 for H2, H2O, O2 respectively [
The various equations which are solved in this study (discussed in sections 2.3 to 2.5) are taken from the Falcao et al. [
In this study Fuel Cell and Electrolysis Model is used to solve the two potential equations and they are as follows [
Reaction Parameters | Value | Units | Reference | |
---|---|---|---|---|
Open circuit voltage | 1.1 | V | [ | |
Reference anode concentration | 1 | k・mol・m−3 | [ | |
Anode charge transfer coefficient | 2 | [ | ||
Anode exchange current density | 10,000 | A・m−2 | [ | |
Reference cathode concentration | 1 | k・mol・m−3 | [ | |
Cathode charge transfer coefficient | 2 | [ | ||
Cathode exchange current density | 20 | A・m−2 | [ | |
Physical Parameters | Value | Units | Reference | |
GDL | Porosity | 0.5 | [ | |
Thermal conductivity | 10 | W・m−1・K−1 | [ | |
Density | 2719 | kg・m−3 | [ | |
Electrical conductivity | 5000 | ohm−1・m−1 | [ | |
Catalyst Layer | Porosity | 0.5 | [ | |
Surface/volume ratio | 200,000 | m−1 | [ | |
Thermal conductivity | 10 | W・m−1・K−1 | [ | |
Electrical conductivity | 5000 | ohm−1・m−1 | [ | |
Membrane | Thermal conductivity | 2 | W・m−1・K−1 | [ |
Dry membrane density | 1980 | kg・m−3 | [ | |
Equivalent weight | 1100 | kg・K−1・mol−1 | [ | |
Electrical conductivity | 1 × 10−16 | ohm−1・m−1 | [ |
where
if phase is solid,
if phase is membrane,
Source terms for Equations (1) and (2) are also termed as exchange current density and have the generic Tafel formulation [
where G, T, and F are the gas constant, the temperature of the fuel cell and the Faraday’s constant respectively.
The source terms
where
A saturation model is used in FLUENT for modeling the formation and transport of the liquid water. The following conservation equation governs the formation of the water (in liquid form) and its transport [
where
where
The capillary pressure
for
for
where
On comparing the peak power density for pin type flow channel with uniformly positioned porous carbon inserts with different porosities, it was found that there was not much effect on the cell performance with the porosity variation (peak power density for 0.9 porosity was 0.512 W/cm2 whereas for 0.6 porosity peak power density obtained was 0.511 W/cm2). Similarly, on comparing the peak power density for pin type flow channel with zigzag positioned porous carbon inserts with different porosities, it was found that there was not much effect on the cell performance with the porosity variation (peak power density for 0.9 porosity was 0.527 W/cm2 whereas for 0.6 porosity peak power density obtained was 0.526 W/cm2).
The performance of the uniform and zigzag pattern with porous carbon inserts is analyzed by varying the porosity from 0.6 to 0.9. Figures 2(a)-(d) show the P-I (Power Density-Current Density) and the V-I (Voltage-Current Density) curves for both the patterns used, at porosity equals to 0.6, 0.7, 0.8 and 0.9 respectively. The above mentioned figures show that the performance of zigzag pin type channel with porous inserts is better than the uniform pin type channel with porous carbon inserts since the plot for zigzag pattern is higher than the uniform pattern for all the different porosity used. Also,
Flow channel Type/pattern | Peak Power Density (W/cm2) | Current Density (A/cm2) | Voltage(V) |
---|---|---|---|
Conventional serpentine | 0.438336 | 0.73056 | 0.6 |
Uniform pin type flow channel with porous carbon inserts (porosity 0.9) | 0.512296 | 0.853824 | 0.6 |
Zigzag pin type flow channel with porous carbon inserts (porosity 0.9) | 0.527584 | 0.879306 | 0.6 |
increase in the values of power density for zigzag pattern over the uniform pattern by 2.98%. In the case of uniform pattern, the arrangement of porous inserts becomes localized, whereas for the zigzag pattern the arrangement is widespread. As a result, the water absorbed by the zigzag pattern in the interfacial region between the gas diffusion layer and the membrane becomes more globalized and hence eliminates the stagnant water more effectively. This effective water elimination results in the higher performance of the zigzag pattern.
Numerical studies have been done on the three flow patterns (conventional serpentine, pin type flow channel with the adoption of porous carbon inserts in zigzag and uniform pattern) in order to analyze the performance of PEMFC.
both the modified pin type flow patterns there is more water generation. The reason for this higher water generation rate is the effective distribution of the reactants throughout the cathode channel (as there is no water logging in both the modified flow pin type flow channel). This validates that the use of both the modified pin type flow patterns enhances the performance by reducing the water accumulation in the flow channel.
Analysis was done on 25 cm2 PEMFC with three flow channel configurations (conventional serpentine, pin type flow channel with the adoption of porous carbon inserts in zigzag and uniform pattern on the rib area) in order to study the cell performance. From the analysis, it was found that the uniform and zigzag pattern with porous inserts show 16.87% and 20.36% increase in performance when compared to the conventional serpentine flow channel. While comparing pin type channels (uniform and zigzag) with porous carbon inserts, zigzag pattern pin type flow channel showed 2.98% more performance than the uniform. It is also found that there is not much effect on the performance of the two new adopted designs with the variation in porosity level of the porous inserts from 0.6 to 0.9. The analysis also shows that the water absorption in case of the zigzag pattern flow channel is more than the uniform porous carbon insert channel pattern.
The work done for this paper was due to the support of the Computer Support Group Department, National Institute of Technology, Tiruchirappalli and Department of Automobile Engineering, PSG College of Technology, Coimbatore.