You are on page 1of 85

FLUENT 6.

3 Fuel Cell Modules Manual

September 2006

Copyright c 2006 by Fluent Inc. All rights reserved. No part of this document may be reproduced or otherwise used in any form without express written permission from Fluent Inc.

Airpak, FIDAP, FLUENT, FLUENT for CATIA V5, FloWizard, GAMBIT, Icemax, Icepak, Icepro, Icewave, Icechip, MixSim, and POLYFLOW are registered trademarks of Fluent Inc. All other products or name brands are trademarks of their respective holders. CHEMKIN is a registered trademark of Reaction Design Inc. Portions of this program include material copyrighted by PathScale Corporation 2003-2004.

Fluent Inc. Centerra Resource Park 10 Cavendish Court Lebanon, NH 03766

Contents

Preface 1 PEM Fuel Cell Model Theory 1.1 1.2 1.3 1.4 1.5 1.6 Introduction

UTM-1 1-1 1-1 1-3 1-6 1-7 1-8

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Electrochemistry Modeling

Current and Mass Conservation . . . . . . . . . . . . . . . . . . . . . . . Liquid Water Formation, Transport, and its Eects . . . . . . . . . . . . Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Transient Simulation of PEM Fuel Cells . . . . . . . . . . . . . . . . . . 1-10 2-1 2-1 2-2 2-2 2-3 2-3 2-4 2-5 2-6 2-7

2 Using the PEM Fuel Cell Model 2.1 2.2 2.3 2.4 2.5 2.6 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Geometry Denition for the PEM Fuel Cell Model . . . . . . . . . . . . Installing the PEM Fuel Cell Model . . . . . . . . . . . . . . . . . . . .

Loading the PEM Fuel Cell Module . . . . . . . . . . . . . . . . . . . . . Setting Up the PEM Fuel Cell Module . . . . . . . . . . . . . . . . . . . PEM Fuel Cell Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6.1 2.6.2 2.6.3 2.6.4 2.6.5 2.6.6 2.6.7 2.6.8 Specifying the PEM Model Options . . . . . . . . . . . . . . . . Specifying the PEM Model Parameters . . . . . . . . . . . . . . Specifying Anode Properties . . . . . . . . . . . . . . . . . . . .

Specifying Membrane Properties . . . . . . . . . . . . . . . . . . 2-11 Specifying Cathode Properties . . . . . . . . . . . . . . . . . . . 2-12 Setting Advanced Properties for the PEM Fuel Cell . . . . . . . 2-16 Specifying the Solution Controls . . . . . . . . . . . . . . . . . . 2-20 Reporting on the Solution . . . . . . . . . . . . . . . . . . . . . . 2-21

c Fluent Inc. July 19, 2006

CONTENTS

2.7 2.8 2.9

Modeling of Current Collectors . . . . . . . . . . . . . . . . . . . . . . . 2-22 PEM Fuel Cell Boundary Conditions . . . . . . . . . . . . . . . . . . . . 2-23 Solution Guidelines for the PEM Fuel Cell Model . . . . . . . . . . . . . 2-24

2.10 Postprocessing the PEM Fuel Cell Model . . . . . . . . . . . . . . . . . . 2-24 2.11 User-Accessible Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 2-26 2.11.1 Compiling the Customized PEM Fuel Cell Source Code . . . . . 2-29

2.12 IV-Curve Calculations Using the Text Interface . . . . . . . . . . . . . . 2-30 3 SOFC Fuel Cell Model Theory 3.1 Introduction 3.1.1 3.1.2 3.1.3 3.2 3.3 3.4 3.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 3-1 3-3 3-3 3-4 3-4 3-5 3-6 3-6 3-7 3-7

Motivation for Consideration of the Electric Field

Overview of the Electric Field Model . . . . . . . . . . . . . . . . Overview of the Electrochemical Model . . . . . . . . . . . . . .

The SOFC Modeling Strategy . . . . . . . . . . . . . . . . . . . . . . . . Modeling Stacked Fuel Cells . . . . . . . . . . . . . . . . . . . . . . . . . Modeling Fluid Flow, Heat Transfer, and Mass Transfer . . . . . . . . . Modeling Current Transport and the Potential Field . . . . . . . . . . . 3.5.1 3.5.2 Treatment of the Activation Overpotential Cell Potential . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.6

Modeling Electrochemical Reactions . . . . . . . . . . . . . . . . . . . . 3-15 4-1 4-1 4-1 4-2 4-8

4 Using the Solid Oxide Fuel Cell Model 4.1 4.2 4.3 4.4 4.5 Installing the Solid Oxide Fuel Cell Model . . . . . . . . . . . . . . . . . Loading the Solid Oxide Fuel Cell Module . . . . . . . . . . . . . . . . . Solid Oxide Fuel Cell Module Set Up Procedure . . . . . . . . . . . . . . Setting the Parameters for the SOFC Model . . . . . . . . . . . . . . . .

Setting Up the Electrode-Electrolyte Interfaces . . . . . . . . . . . . . . 4-10 4.5.1 4.5.2 Setting Up the Anode Electrode-Electrolyte Interface . . . . . . 4-10

Setting Up the Cathode Electrode-Electrolyte Interface . . . . . 4-11

4.6

Setting Up the Electric Field Model Parameters . . . . . . . . . . . . . . 4-12

ii

c Fluent Inc. July 19, 2006

CONTENTS

4.7 4.8

Setting Up the Tortuosity Parameters . . . . . . . . . . . . . . . . . . . 4-13 Setting Up the Activation Parameters . . . . . . . . . . . . . . . . . . . 4-14

c Fluent Inc. July 19, 2006

iii

CONTENTS

iv

c Fluent Inc. July 19, 2006

Using This Manual


The Contents of This Manual
The FLUENT Fuel Cell Modules Manual tells you what you need to know to model polymer electrolyte membrane (PEM) fuel cells or solid oxide fuel cell (SOFC) fuel cells with FLUENT. In this manual, you will nd background information pertaining to the models, a theoretical discussion of the models used in FLUENT, and a description of using the models for your CFD simulations.

Typographical Conventions
Several typographical conventions are used in this manuals text to facilitate your learning process.

An informational icon ( An warning icon (

) marks an important note.

! ) marks a warning.

Dierent type styles are used to indicate graphical user interface menu items and text interface menu items (e.g., Iso-Surface panel, surface/iso-surface command). The text interface type style is also used when illustrating exactly what appears on the screen or exactly what you need to type into a eld in a panel. The information displayed on the screen is enclosed in a large box to distinguish it from the narrative text, and user inputs are often enclosed in smaller boxes. A mini ow chart is used to indicate the menu selections that lead you to a specic command or panel. For example, Dene Boundary Conditions... indicates that the Boundary Conditions... menu item can be selected from the Dene pull-down menu, and display grid indicates that the grid command is available in the display text menu. The words before the arrows invoke menus (or submenus) and the arrows point from a specic menu toward the item you should select from that menu. In this

c Fluent Inc. July 19, 2006

UTM-1

Using This Manual

manual, mini ow charts usually precede a description of a panel or command, or a screen illustration showing how to use the panel or command. They allow you to look up information about a command or panel and quickly determine how to access it without having to search the preceding material. The menu selections that will lead you to a particular panel are also indicated (usually within a paragraph) using a /. For example, Dene/Materials... tells you to choose the Materials... menu item from the Dene pull-down menu.

Mathematical Conventions
Where possible, vector quantities are displayed with a raised arrow (e.g., a, A). Boldfaced characters are reserved for vectors and matrices as they apply to linear algebra (e.g., the identity matrix, I). The operator , referred to as grad, nabla, or del, represents the partial derivative of a quantity with respect to all directions in the chosen coordinate system. In Cartesian coordinates, is dened to be + + k x y z appears in several ways: The gradient of a scalar quantity is the vector whose components are the partial derivatives; for example, p= p p p + + k x y z

The gradient of a vector quantity is a second-order tensor; for example, in Cartesian coordinates, (v) = + + k x y z vx + vy + vz k

This tensor is usually written as



vx x vy x vz x vx y vy y vz y vx z vy z vz z

UTM-2

c Fluent Inc. July 19, 2006

Using This Manual

The divergence of a vector quantity, which is the inner product between and a vector; for example, v = vx vy vz + + x y z
2

The operator , which is usually written as Laplacian; for example,


2

and is known as the

T =

2T 2T 2T + + 2 x2 y 2 z

T is dierent from the expression ( T )2 , which is dened as ( T) =


2

T x

T + y

T + z

Technical Support
If you encounter diculties while using FLUENT, please rst refer to the section(s) of the manual containing information on the commands you are trying to use or the type of problem you are trying to solve. The product documentation is available from the online help on the documentation CD, or from the Fluent Inc. User Services Center (www.fluentusers.com). If you encounter an error, please write down the exact error message that appeared and note as much information as you can about what you were doing in FLUENT. Then refer to the following resources available on the Fluent Inc. User Services Center (www.fluentusers.com): Installation and System FAQs - link available from the main page on the User Services Center. The FAQs can be searched by word or phrase, and are available for general installation questions as well as for products. Known defects for FLUENT - link available from the product page. The defects can be searched by word or phrase, and are listed by categories. Online Technical Support - link available from the main page on the User Services Center. From the Online Technical Support Portal page, there is a link to the Search Solutions & Request Support page, where the solutions can be searched by word or phrase. The User Services Center also provides online forums, where you can discuss topics of mutual interest and share ideas and information with other Fluent users, and the ability to sign up for e-mail notications on our latest product releases.

c Fluent Inc. July 19, 2006

UTM-3

Using This Manual

Contacting Technical Support


If none of the resources available on the User Services Center help in resolving the problem, or you have complex modeling projects, we invite you to call your support engineer for assistance. However, there are a few things that we encourage you to do before calling: Note what you are trying to accomplish with FLUENT. Note what you were doing when the problem or error occurred. Save a journal or transcript le of the FLUENT session in which the problem occurred. This is the best source that we can use to reproduce the problem and thereby help to identify the cause.

UTM-4

c Fluent Inc. July 19, 2006

Chapter 1.

PEM Fuel Cell Model Theory

This chapter presents the theoretical background for the PEM fuel cell modeling capabilities in FLUENT. Section 1.1: Introduction Section 1.2: Electrochemistry Modeling Section 1.3: Current and Mass Conservation Section 1.4: Liquid Water Formation, Transport, and its Eects Section 1.5: Properties Section 1.6: Transient Simulation of PEM Fuel Cells

1.1

Introduction
The PEM Fuel Cell (PEMFC) module is provided as an addon module with the standard FLUENT licensed software. A special license is required to use the PEMFC module. A fuel cell is an energy conversion device that converts the chemical energy of fuel into electrical energy. A schematic of a polymer electrolyte membrane (PEM) fuel cell is shown in Figure 1.1.1.

c Fluent Inc. July 19, 2006

1-1

PEM Fuel Cell Model Theory

load

Hydrogen ows into the fuel cell on the anode side. It diuses through the porous gas diusion layers and comes in contact with the catalyst layer. Here it forms hydrogen ions and electrons. The hydrogen ions diuse through the polymer electrolyte membrane at the center, the electrons ow through the gas diusion layer to the current collectors and into the electric load attached. Electrons enter the cathode side through the current collectors and the gas diusion layer. At the catalyst layer on the cathode side, the electrons, the hydrogen ions and the oxygen combine to form water. In the PEM fuel cell model in FLUENT, two electric potential elds are solved. One potential is solved in the membrane and catalyst layers. The other is solved in the catalyst layers, the diusion layers, and the current collectors. Surface reactions on the porous catalyst region are solved and the reaction diusion balance is applied to compute the rates. Based on the cell voltage that you prescribe, the current density value is computed. Alternatively, a cell voltage can be computed based on a prescribed average current density.

1-2


Gas Channel (H 2) 2H2 H+ + 4H + 4e H+ Electrolyte Membrane + O + 4H + 4e 2 2H2O

Cooling Channel(s) Anode Collector Anode Gas Diffusion Layer Anode Catalyst Layer

Cathode Catalyst Layer Cathode Gas Diffusion Layer Cathode Collector Cooling Channel(s)


Gas Channel (air)

Figure 1.1.1: Schematic of a PEM Fuel Cell

c Fluent Inc. July 19, 2006

1.2 Electrochemistry Modeling

1.2

Electrochemistry Modeling
At the center of the electrochemistry is the computation of the rate of the hydrogen oxidation and the rate of oxygen reduction. In the FLUENT PEM model, these electrochemical processes are treated as heterogeneous reactions that take place on the catalyst surfaces inside the two catalyst layers on both sides of the membrane. Such detailed treatment has been used by other groups ([1], [2], and [8]). The driving force behind these reactions is the surface over-potential: the dierence between the phase potential of the solid and the phase potential of the electrolyte/membrane. Therefore, two potential equations are solved for in the PEM model: one potential equation (Equations 1.2-1) accounts for the electron transport e through the solid conductive materials (i.e., the current collectors and solid grids of the porous media); the other potential equation (1.2-2) represents the protonic (i.e., ionic) transport of H + . The two potential equations read,

(sol sol ) + Rsol = 0 (mem mem ) + Rmem = 0 where R = = = electrical conductivity (1/ohm-m) electric potential (volts) volumetric transfer current (A/m3 )

(1.2-1) (1.2-2)

The following gure illustrates the boundary conditions that are used to solve for sol and mem . There are two types of external boundaries. Those through which there passes an electrical current and those through which there passes no current. As no protonic current leaves the fuel cell through any external boundary, there is a zero ux boundary condition for the membrane phase potential, mem , on all outside boundaries. For the solid phase potential, sol , there are external boundaries on the anode and the cathode side that are in contact with the external electric circuit and only through these boundaries passes the electrical current generated in the fuel cell. On all other external boundaries there is a zero ux boundary condition for sol . On the external contact boundaries, we recommend to prescribe xed values for sol (potentiostatic boundary conditions). If the anode side is set to zero, the (positive) value prescribed on the cathode side is the cell voltage. Specifying a constant ux (say on the cathode side) means to specify galvanostatic boundary conditions.

c Fluent Inc. July 19, 2006

1-3

PEM Fuel Cell Model Theory


mem n mem =0 n sol =0 n

=0 sol = 0

mem =0 n sol =0 n

? Gas Channel (H2 )

  

Cooling Channel(s) Anode Collector Anode Gas Diusion Layer Anode Catalyst Layer Cathode Catalyst Layer Cathode Gas Diusion Layer Cathode Collector Cooling Channel(s)

2H2 4H + + 4e H+ Electrolyte Membrane


H+

?
O2 + 4H + + 4e 2H2 O

?  
Gas Channel (air)

 

6
sol = Vcell or
sol = constant n mem =0 n

Figure 1.2.1: Boundary Conditions for sol and mem

The transfer currents, or the source terms in Equations 1.2-1 and 1.2-2, are non-zero only inside the catalyst layers and are computed as: For the solid phase, Rsol = Ran (< 0) on the anode side and Rsol = +Rcat (> 0) on the cathode side. For the membrane phase, Rmem = +Ran (> 0) on the anode side and Rmem = Rcat (< 0) on the cathode side. The source terms in Equations 1.2-1 and 1.2-2 are also called the exchange current density (A/m3 ), and have the following general denitions:

Ran = Rcat =

ref jan

[H2 ] [H2 ]ref [O2 ] [O2 ]ref

an

ean F an /RT ecat F an /RT


cat

(1.2-3) (1.2-4)

ref jcat

e+an F cat /RT + ecat F cat /RT

1-4

c Fluent Inc. July 19, 2006

1.2 Electrochemistry Modeling

where j ref [ ],[ ]ref F = = = = = volumetric reference exchange current density (A/m3 ) local species concentration, reference value (kgmol/m3 ) concentration dependence (dimensionless) transfer coecient (dimensionless) Faraday constant (9.65 107 C/kgmol)

The above equation is the general formulation of the Butler-Volmer function. A simplication to this is the Tafel formulation that reads,

Ran = Rcat =

ref jan

[H2 ] [H2 ]ref [O2 ] [O2 ]ref

an

ean F an /RT
cat

(1.2-5) (1.2-6)

ref jcat

ecat F cat /RT

By default, the Butler-Volmer function is used in the FLUENT PEM model to compute the transfer currents inside the catalyst layers. The driving force for the kinetics is the local surface over-potential, , also known as the activation loss. It is generally the dierence between the solid and membrane potentials, sol and mem . The gain in electrical potential from crossing from the anode to the cathode side can then be taken into account by subtracting the open-circuit voltage Voc on the cathode side. an = sol mem cat = sol mem Voc From Equations 1.2-1 through 1.2-8, the two potential elds can be obtained. (1.2-7)

(1.2-8)

c Fluent Inc. July 19, 2006

1-5

PEM Fuel Cell Model Theory

1.3

Current and Mass Conservation


The following reactions occur, respectively, at the anode and the cathode: H2 2H + + 2e O2 + 4H + + 4e 2H2 O The volumetric source terms for the species equations (kg/m3 -s) and energy equation (W/m3 ) are given in Equations 1.3-11.3-4.

SH2 = SO2 = SH2 0 =

Mw,H2 Ran 2F Mw,O2 Rcat 4F

(1.3-1) (1.3-2) (1.3-3)

Mw,H2 0 Rcat 2F

Additional volumetric sources to the energy equation implemented in the FLUENT PEM model include ohmic heating, heat of formation of water, electric work and latent heat of water. Sh = I 2 Rohm + hreaction + Ran,cat + hphase (1.3-4)

The electrochemical reactions that take place inside the catalyst layers are considered heterogeneous reactions that take place on the catalyst surfaces in the porous media. Therefore, the species concentrations of hydrogen and oxygen in the rate calculation, Equations 1.2-3 and 1.2-6, are the surface values. The reactions are treated as surface reactions in the two catalyst layers, and it is assumed that the diusive ux of any reacting species is balanced by its rate of production. Di Mw,i (yi,surf yi,cent )r = Ran,cat nF where Di r yi,surf yi,cent mass diusivity of species i (m2 /s) specic reacting surface area of the catalyst layer, or surface-to-volume ratio (1/m) = mass fraction of species i at the reacting surface = mass fraction of species i at the cell center = average distance between the reaction surfaces and the cell center (m) = = (1.3-5)

1-6

c Fluent Inc. July 19, 2006

1.4 Liquid Water Formation, Transport, and its Effects

The left hand side of Equation 1.3-5 represents the diusive ux at the reacting surface and the right hand side represents the rate of mass generation. The average distance from the cell-center to the reacting surface is estimated as = 1/r . Equation 1.3-5 is used to obtain the surface values of H2 and O2 concentrations, applying a Newtonian solution procedure. These surface, or wall, values are then used to compute the rates in Equations 1.2-3 through 1.2-6.

1.4

Liquid Water Formation, Transport, and its Effects


Since PEM fuel cells operate under relatively low temperature (<100 C), the water vapor may condense to liquid water, especially at high current densities. While the existence of the liquid water keeps the membrane hydrated, it also blocks the gas diusion passage, reduces the diusion rate and the eective reacting surface area and hence the cell performance. To model the formation and transport of liquid water, FLUENT uses a saturation model based on [7],[5]. In this approach, the liquid water formation and transport is governed by the following conservation equation for the volume fraction of liquid water, s, or the water saturation, ( l s) + t l V l s = r w (1.4-1)

where the subscript l stands for liquid water, and rw is the condensation rate that is modeled as, rw = cr max (1 s) Pwv Psat Mw,H2 0 , [sl ] RT (1.4-2)

where rw is added to the water vapor equation, as well as the pressure correction (mass source). This term is not applied inside the membrane. The condensation rate constant is hardwired to cr = 100s1 . It is assumed that the liquid velocity, Vl , is equivalent to the gas velocity inside the gas channel (i.e., a ne mist). Inside the highly-resistant porous zones, the use of the capillary diusion term allows us to replace the convective term in Equation 1.4-1: ( l s) + t l Ks3 dpc s = rw l ds

(1.4-3)

Depending on the wetting phase, the capillary pressure is computed as a function of s (the Leverett function),

cosc (1.417(1 s) 2.12(1 s)2 ( K )0.5 cosc (1.417s 2.12s2 + 1.263s3 ) ( K )0.5

pc =

+ 1.263(1 s)3 ) c < 90 c > 90

(1.4-4)

c Fluent Inc. July 19, 2006

1-7

PEM Fuel Cell Model Theory where is the porosity, is the surface tension (N/m2 ), c is the contact angle and K the absolute permeability. Equation 1.4-1 models various physical processes such as condensation, vaporization, capillary diusion, and surface tension. The clogging of the porous media and the ooding of the reaction surface are modeled by multiplying the porosity and the active surface area by (1 s), respectively.

1.5

Properties
The gas phase species diusivities are given by Di = 1.5 (1 s)rs Di0 p0 p
p

T T0

(1.5-1)

where Di0 is the mass diusivity of species i at reference temperature and pressure (P0 , T0 ) [8]. These reference values and the exponents (p , t ) as well as the exponent of pore blockage (rs ) are dened in the PEM user dened functions (UDF) as,

p0 T0 p t rs

= = = = =

101325 N/m2 300 K 1.0 1.5 2.5

The electrolyte membrane of the fuel cell is modeled as a porous uid zone. Properties such as membrane phase electrical conductivity, water diusivity, and the osmotic drag coecient are evaluated as functions of the water content, using various correlations as suggested by [9]. To capture the relevant physics of the problem, various properties of the membrane are incorporated into the model as default options. You can, however, directly incorporate your own formulations and data for these properties by editing the functions dened in the provided source code le called pem user.c and compiling the code yourself. For more information, see Section 2.11: User-Accessible Functions. Membrane Phase Electric Conductivity mem = (0.514 0.326) e1268( 303 T )
1 1

(1.5-2)

where is the water content. Two model constants, and are introduced in FLUENT for generality. Equation 1.5-2 becomes the original correlation from [9] when = = 1.

1-8

c Fluent Inc. July 19, 2006

1.5 Properties

Osmotic Drag Coecient d = 2.5 Back Diusion Flux


dif Jw f =

22

(1.5-3)

m Mh20 Dl Mm

(1.5-4)

where m and Mm are the density and the equivalent weight of the dry membrane, respectively. Membrane Water Diusivity Dl = f ()e2416( 303 T ) Water Content The water content, , that appears in the preceding property computations are obtained using Springer et als correlation [9],
1 1

(1.5-5)

= 0.043 + 17.18a 39.85a2 + 36a3 (a < 1) = 14 + 1.4(a 1)(a > 1) here a is the water activity that is dened as, a= Water Vapor Pressure Pwv + 2s Psat

(1.5-6)

(1.5-7)

The water vapor pressure is computed based upon the vapor molar fraction and the local pressure, Pwv = xH2 O P Saturation Pressure The saturation pressure is calculated, in terms of atm, as, (1.5-8)

log10 Psat = 2.1794 + 0.02953(T 273.17) 9.1837 105 (T 273.17)2 + 1.4454 107 (T 273.17)3

(1.5-9)

c Fluent Inc. July 19, 2006

1-9

PEM Fuel Cell Model Theory

It is noted here that in [9], water activity is dened on the basis of total water or super-saturated water vapor. With phase change being invoked in the present two-phase model, 2s is added to the original formulation as suggested by [3].

1.6

Transient Simulation of PEM Fuel Cells


Dynamics response of PEM fuel cells to changes in operating conditions as a function of time can be modeled in the PEMFC module. For example, a change in the cell voltage or current density, or inlet mass ow rates at the anode and/or the cathode. The procedure for setting up and solving transient PEMFC problems are the same as that used for a normal FLUENT transient problem as discussed in the FLUENT Users Guide. Assuming that the time scales associated with the electric elds are much smaller than those associated with the ow and thermal elds, the steady-state equations are retained for the two electric potentials, (i.e., Equations 1.2-1 and 1.2-2). Transient terms in all other equations such as momentum transport, energy transport, species transport, liquid water transport, and membrane water content equations are activated.

1-10

c Fluent Inc. July 19, 2006

Chapter 2.

Using the PEM Fuel Cell Model

The procedure for setting up and solving PEM fuel cell problems is described in detail in this chapter. Please refer to the following sections for more information: Section 2.1: Introduction Section 2.2: Geometry Denition for the PEM Fuel Cell Model Section 2.3: Installing the PEM Fuel Cell Model Section 2.4: Loading the PEM Fuel Cell Module Section 2.5: Setting Up the PEM Fuel Cell Module Section 2.6: PEM Fuel Cell Modeling Section 2.7: Modeling of Current Collectors Section 2.8: PEM Fuel Cell Boundary Conditions Section 2.9: Solution Guidelines for the PEM Fuel Cell Model Section 2.10: Postprocessing the PEM Fuel Cell Model Section 2.11: User-Accessible Functions Section 2.12: IV-Curve Calculations Using the Text Interface

2.1

Introduction
The FLUENT PEM Fuel Cell (PEMFC) model is comprised of several user-dened functions (UDFs) and a graphical user interface. The potential elds are solved as userdened scalars. The liquid water saturation, s, and the water content, , are also solved as user-dened scalars. The electrochemical reactions occurring on the catalyst are modeled through various source terms while other model parameters are handled through the user interface. The PEMFC model can be used in parallel FLUENT as well.

c Fluent Inc. July 19, 2006

2-1

Using the PEM Fuel Cell Model

2.2

Geometry Denition for the PEM Fuel Cell Model


Due to the fact that there are a number of dierent physical zones associated with the fuel cell, the following regions must be present in the fuel cell mesh: Anode ow channel Anode gas diusion layer Anode catalyst layer Membrane layer Cathode catalyst layer Cathode gas diusion layer Cathode ow channel The following zones have to be identied, if present in the fuel cell mesh: Anode current collector Cathode current collector Coolant channel

2.3

Installing the PEM Fuel Cell Model


The PEM Fuel Cell (PEMFC) model is provided as an addon module with the standard FLUENT licensed software. A special license is required to use the PEMFC model. The module is installed with the standard installation of FLUENT in a directory called addons/fuelcells2.2 in your installation area. The PEMFC model consists of a UDF library and a pre-compiled scheme library, which needs to be loaded and activated before calculations can be performed.

2-2

c Fluent Inc. July 19, 2006

2.4 Loading the PEM Fuel Cell Module

2.4

Loading the PEM Fuel Cell Module


The PEM Fuel Cell (PEMFC) module is loaded into FLUENT through the text user interface (TUI). The module can only be loaded after a valid FLUENT case le has been set or read. The text command to load the addon module is define models addon-module A list of FLUENT addon modules is displayed: FLUENT Addon Modules: 1. MHD Model 2. Fiber Model 3. PEM Fuel Cell Model 4. SOFC Fuel Cell Model 5. Population Balance Model Enter Module Number: [1] 3 Select the PEMFC model by entering the module number 3. During the loading process a scheme library containing the graphical and text user interface, and a UDF library containing a set of user dened functions are loaded into FLUENT.

2.5

Setting Up the PEM Fuel Cell Module


The following describes an overview of the procedure required in order to use the PEM fuel cell model in FLUENT. 1. Start FLUENT. 2. Read the case le. 3. Scale the grid. 4. Use the PEM Model panel to dene the fuel cell model parameters. 5. Dene material properties. 6. Set the operating conditions. 7. Set the boundary conditions. 8. Start the calculations. 9. Save the case and data les. 10. Process your results.

c Fluent Inc. July 19, 2006

2-3

Using the PEM Fuel Cell Model

The PEM Model panel greatly simplies the input of parameters and boundary conditions, but it does not replace the boundary conditions interface. Therefore it is a good policy, to start the set up with the PEM Model panel and do the nishing steps for boundary conditions afterwards.

2.6

PEM Fuel Cell Modeling


In order to set PEM fuel cell model parameters and assign properties to the relevant regions in your fuel cell, you need to access the PEM fuel cell graphical user interface (the PEM Model panel) using: Dene Models PEMFC Here, you can identify the relevant zones for the current collectors, ow channels, gas diusion layers, catalyst layers, and the membrane. You can specify the following inputs using the PEM Model panel. Optional inputs are indicated as such. 1. Enable either the single-phase or the multi-phase PEM model. 2. Set the appropriate options for the PEM model (optional). 3. Set the various parameters for the PEM model. 4. Select the appropriate zones and specify the properties on the anode side. 5. Select the appropriate zones and specify the properties of the membrane. 6. Select the appropriate zones and specify the properties on the cathode side. 7. Provide input for advanced features such as contact resistivities, coolant channel properties, or stack management settings (optional). 8. Set solution controls such as under-relaxation factors (optional). 9. Provide input to assist reporting (optional).

2-4

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

2.6.1

Specifying the PEM Model Options

The Model tab of the PEM Model panel allows you to turn on or o various options when solving a PEM fuel cell problem.

Figure 2.6.1: The Model Tab of the PEM Model Panel

Several PEM model options are available in the Model Tab of the PEM Model panel including: The Joule Heating option takes into account ohmic heating. This option includes the I 2 R term in the energy source term from Equation 1.3-4 in the calculations. The Reaction Heating option takes into account the heat generated by the chemical reactants. This option includes the hreaction term in the energy source term from Equation 1.3-4 in the calculations. This accounts for the latent heat of formation of water vapor. The Butler-Volmer Rate option (the default) is used to compute the transfer currents inside the catalyst layers. If this option is turned o, the Tafel approximation (Equation 1.2-6) is used. The Membrane Water Transport option takes into account the transport of water across the membrane. The Multiphase option takes into account multiphase calculations. Use this option if you are solving for approximate liquid transport in the gas diusion layer of the fuel cell. Nearly all options are turned on by default. You may wish to override the default values, depending on the problem you wish to model. For instance, if you are not concerned with the heat generated due to chemical reaction, then you may want to turn o the Reaction Heating option.

c Fluent Inc. July 19, 2006

2-5

Using the PEM Fuel Cell Model

2.6.2

Specifying the PEM Model Parameters

You can use the Parameters tab of the PEM Model panel to specify the electrochemistry parameters for the PEM model, reference diusivities for the reactants, among other model parameters.

Figure 2.6.2: The Parameters Tab of the PEM Model Panel

There are various parameters under Electrochemistry in the PEM Model panel. For both the anode and the cathode, you can also set the following parameters or leave the default values.
ref ref The Ref. Current Density corresponds to jan and jcat , the reference exchange current density from Equation 1.2-3 and Equation 1.2-4.

The Ref. Concentration corresponds to the reference concentration ([H2 ]ref and [O2 ]ref ) with units of 1 kgmol/m3 (see Equation 1.2-3 and Equation 1.2-4). The Concentration Exponent corresponds to , the concentration dependence from Equation 1.2-3. The Exchange Coecient corresponds to , the transfer coecient from Equation 1.2-3. The Open-Circuit Voltage corresponds to Voc in Equation 1.2-8. Moreover, the following parameters can also be set here: The Reference Diusivities correspond to Di from Equation 1.3-5, the species mass diusivity. The Saturation Exponent for Pore Blockage corresponds to rs from Equation 1.5-1 for multiphase PEM calculations.

2-6

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

2.6.3

Specifying Anode Properties

You can use the Anode tab of the PEM Model panel to specify zones and properties of the current collector, the ow channel, the diusion layer, and the catalyst layer for the anode portion of the PEM fuel cell.

Specifying Current Collector Properties for the Anode

Figure 2.6.3: The Anode Tab of the PEM Model Panel With Current Collector Selected

1. Select the Anode tab of the PEM Model panel. 2. Select Current Collector under Anode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones.

c Fluent Inc. July 19, 2006

2-7

Using the PEM Fuel Cell Model

Specifying Flow Channel Properties for the Anode

Figure 2.6.4: The Anode Tab of the PEM Model Panel With Flow Channel Selected

1. Select the Anode tab of the PEM Model panel. 2. Select Flow Channel under Anode Zone Type. 3. Select a corresponding zone from the Zone(s) list.

2-8

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

Specifying Diffusion Layer Properties for the Anode

Figure 2.6.5: The Anode Tab of the PEM Model Panel With Diusion Layer Selected

1. Select the Anode tab of the PEM Model panel. 2. Select Diusion Layer under Anode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones. 5. Specify a value for the Porosity. 6. Specify a value for the Viscous Resistance. 7. Specify a value for the Contact Angle for multiphase PEM calculations (c inEquation 1.4-4).

c Fluent Inc. July 19, 2006

2-9

Using the PEM Fuel Cell Model

Specifying Catalyst Layer Properties for the Anode

Figure 2.6.6: The Anode Tab of the PEM Model Panel With Catalyst Layer Selected

1. Select the Anode tab of the PEM Model panel. 2. Select Catalyst Layer under Anode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones. 5. Specify a value for the Porosity. 6. Specify a value for the Viscous Resistance. 7. Specify a value for the Surface-to-Volume Ratio ( in Equation 1.3-5). This is the specic surface area of the catalyst medium. 8. Specify a value for the Contact Angle for multiphase PEM calculations (c inEquation 1.4-4).

2-10

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

Specifying Boundary Conditions for the Anode


For each case of the anodes current collector, diusion layer, and catalyst layer, you assign a solid material and/or set the porosity and the viscous resistance. These settings represent setting a boundary condition. With the Collective Boundary Conditions option turned on (the default setting), this boundary condition is applied to all selected zones in the Zone(s) list. If you want to set the boundary conditions for each zone individually using the corresponding Boundary Conditions panel, you should turn o the Collective Boundary Conditions option.

2.6.4

Specifying Membrane Properties

You can use the Membrane tab of the PEM Model panel to specify zones and properties of the membrane portion of the PEM fuel cell.

Figure 2.6.7: The Membrane Tab of the PEM Model Panel

1. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all membrane zones as a group. 2. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. 3. Specify a value for the Equivalent Weight (Mm in Equation 1.5-4). 4. Specify a value for the Protonic Conduction Coecient ( in Equation 1.5-2). This is used to calculate the membrane phase electric conductivity. 5. Specify a value for the Protonic Conduction Exponent ( in Equation 1.5-2).

c Fluent Inc. July 19, 2006

2-11

Using the PEM Fuel Cell Model

Specifying Boundary Conditions for the Membrane


When you assign a solid material to the membrane, you are setting a boundary condition. With the Collective Boundary Conditions option turned on (the default setting), this boundary condition is applied to all selected zones in the Zone(s) list. If you want to set the boundary conditions for each zone individually using the corresponding Boundary Conditions panel, you should turn o the Collective Boundary Conditions option.

2.6.5

Specifying Cathode Properties

You can use the Cathode tab of the PEM Model panel to specify zones and properties of the current collector, the ow channel, the diusion layer, and the catalyst layer for the cathode portion of the PEM fuel cell.

Specifying Current Collector Properties for the Cathode

Figure 2.6.8: The Cathode Tab of the PEM Model Panel With Current Collector Selected

1. Select the Cathode tab of the PEM Model panel. 2. Select Current Collector under Cathode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones.

2-12

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

Specifying Flow Channel Properties for the Cathode

Figure 2.6.9: The Cathode Tab of the PEM Model Panel With Flow Channel Selected

1. Select the Cathode tab of the PEM Model panel. 2. Select Flow Channel under Cathode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group.

c Fluent Inc. July 19, 2006

2-13

Using the PEM Fuel Cell Model

Specifying Diffusion Layer Properties for the Cathode

Figure 2.6.10: The Cathode Tab of the PEM Model Panel With Diusion Layer Selected

1. Select the Cathode tab of the PEM Model panel. 2. Select Diusion Layer under Cathode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones. 5. Specify a value for the Porosity. 6. Specify a value for the Viscous Resistance. 7. Specify a value for the Contact Angle for multiphase PEM calculations (c inEquation 1.4-4).

2-14

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

Specifying Catalyst Layer Properties for the Cathode

Figure 2.6.11: The Cathode Tab of the PEM Model Panel With Catalyst Layer Selected

1. Select the Cathode tab of the PEM Model panel. 2. Select Catalyst Layer under Cathode Zone Type. 3. Select a corresponding zone from the Zone(s) list. If you are modeling a fuel cell stack, then you must pick all zones of a particular type as a group. 4. Select a Solid Material from the corresponding drop-down list. Solid materials can be customized using the Materials panel. Note that for the Electrical Conductivity, you can only choose a constant value in the Materials panel. The solid electrical conductivity value is the diusivity of the solid phase potential in the solid zones. 5. Specify a value for the Porosity. 6. Specify a value for the Viscous Resistance. 7. Specify a value for the Surface-to-Volume Ratio (s in Equation 1.3-5). This is the specic surface area of the catalyst medium. 8. Specify a value for the Contact Angle for multiphase PEM calculations (c in Equation 1.4-4).

c Fluent Inc. July 19, 2006

2-15

Using the PEM Fuel Cell Model

Specifying Boundary Conditions for the Cathode


For each case of the cathodes current collector, diusion layer, and catalyst layer, you assign a solid material and/or set the porosity and the viscous resistance. These settings represent setting a boundary condition. With the Collective Boundary Conditions option turned on (the default setting), this boundary condition is applied to all selected zones in the Zone(s) list. If you want to set the boundary conditions for each zone individually using the corresponding Boundary Conditions panel, you should turn o the Collective Boundary Conditions option.

2.6.6

Setting Advanced Properties for the PEM Fuel Cell

You can use the Advanced tab of the PEM Model panel to specify the contact resistivity for any material interface in the geometry, set parameters for coolant channels, and dene fuel stack units for managing stacks of fuel cells.

Setting Contact Resistivities for the PEM Fuel Cell

Figure 2.6.12: The Advanced Tab of the PEM Model Panel for Contact Resistivities

1. Select the Advanced tab of the PEM Model panel. 2. Select Contact Resistivity under Advanced Setup. 3. Select any number of corresponding interfaces from the Available Zone(s) list. These zones are face zones over which a jump in electrical potential is caused by imperfect conduction. 4. Specify a value for the Resistivity for each specied zone. 5. To simplify the input, you can choose to use the resistivity value of the rst selected zone for all others as well by turning on the Use First Value for All option.

2-16

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

Setting Coolant Channel Properties for the PEM Fuel Cell

Figure 2.6.13: The Advanced Tab of the PEM Model Panel for the Coolant Channel

1. Select the Advanced tab of the PEM Model panel. 2. Select Coolant Channel under Advanced Setup. 3. Select any number of corresponding zones from the Zone(s) list. 4. Specify a value for the Density. 5. Specify a value for the Heat Capacity. 6. Specify a value for the Thermal Conductivity. 7. Specify a value for the Viscosity. 8. To enable the coolant channel, turn on the Enable Coolant Channel(s) option. Amongst other settings, this will change the mixture to include the coolant species, which is otherwise absent.

c Fluent Inc. July 19, 2006

2-17

Using the PEM Fuel Cell Model

Managing Stacks for the PEM Fuel Cell

Figure 2.6.14: The Advanced Tab of the PEM Model Panel for Stack Management

The FLUENT PEM fuel cell model allows you to model fuel cell stacks as well as individual fuel cells. In the Advanced tab of the PEM Model panel, you can dene fuel cell units for each fuel cell in a stack. A fuel cell unit consists of all zones of a single fuel cell in the stack.

If you are only modeling a single fuel cell, then you do not need to set anything for Stack Management in the Advanced tab of the PEM Model panel.

1. Select the Advanced tab of the PEM Model panel. 2. Select Stack Management under Advanced Setup. 3. Since a fuel cell unit consists of all zones of a single fuel cell in the stack, select the corresponding zones from the Zone(s) list. 4. Create a new fuel cell unit by clicking the Create button. The new fuel cell is listed under Fuel Cell Unit(s) with a default name. 5. Edit a preexisting fuel cell unit by selecting it in the Fuel Cell Unit(s) list. The zones in this fuel cell unit are automatically selected in the Zone(s) list. You can then modify the zones that comprise the fuel cell unit and/or change its name in the Name eld and click Modify to save the new settings. 6. Remove a preexisting fuel cell unit by selecting it in the Fuel Cell Unit(s) list and clicking the Delete button.

2-18

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

7. If your model contains many zone names, you can use the Match Zone Name Pattern eld to specify a pattern to look for in the names of zones. Type the pattern in the text eld and click Match to select (or deselect) the zones in the Zones list with names that match the specied pattern. You can match additional characters using * and ?. For example, if you specify wall*, all surfaces whose names begin with wall (e.g., wall-1, wall-top) will be selected automatically. If they are all selected already, they will be deselected. If you specify wall?, all surfaces whose names consist of wall followed by a single character will be selected (or deselected, if they are all selected already). For example, in a stack there are many fuel cells, say 10 - 100, each having at least 9 zones (current collector, gas channel, diusion layer, and catalyst layer for both anode and cathode and a membrane). Additionally, there may be coolant channels, and it may be that for mesh construction reasons each of these physical zones is made up of more than one mesh zone. Even for small stacks, you can easily end up having hundreds of cell zones in a FLUENT mesh. Therefore, you may want to consider numbering the fuel cells in a stack and to use the assigned fuel cell number in the names of the mesh zones. When you set up your stacked fuel cell case, you would use the Match Zone Name Pattern eld to pick all the zones belonging to a single fuel cell in the stack, rather than scrolling through the potentially very long list and selecting them manually.

c Fluent Inc. July 19, 2006

2-19

Using the PEM Fuel Cell Model

2.6.7

Specifying the Solution Controls

You can use the Controls tab of the PEM Model panel to inuence the solution process.

Figure 2.6.15: The Controls Tab of the PEM Model Panel

The saturation source term rw in Equation 1.4-2 usually requires under-relaxation. You can change the default value for the under-relaxation factor by changing the value for Saturation Source. The water content, , in Equation 1.5-6 also may need under-relaxation. You can change the default value for the under-relaxation factor by changing the value for Water Content. If you are only interested in the basic ow eld throughout the fuel cell, you can turn o the Electrochemistry option in order to suppress most eects of the PEM fuel cell model. To turn o all eects of the PEM fuel cell model, you should also turn o the Membrane Water Transport and Multiphase options in the Model tab of the PEM Model panel.

2-20

c Fluent Inc. July 19, 2006

2.6 PEM Fuel Cell Modeling

2.6.8

Reporting on the Solution

You can use the Reports tab of the PEM Model panel to set up parameters that will be useful in reporting data relevant to the fuel cell.

Figure 2.6.16: The Reports Tab of the PEM Model Panel

The Membrane-Electrode-Assembly Projected Area eld requires the projected area of the Membrane Electrolyte Assembly (MEA) and is only used to calculate the average current density. The assembly consists of the membrane and the catalyst layers above and below the membrane. The value of the projected area can be computed from the Projected Surface Areas panel. Reports Projected Areas... The External Contact Interface(s) elds requires the face zones that act as external contact surfaces for the anode and the cathode. These inputs are used to report cell voltage. For potentiostatic boundary conditions, this is the dierence between the provided values, but for galvanostatic boundary conditions, the cell voltage is part of the solution.

c Fluent Inc. July 19, 2006

2-21

Using the PEM Fuel Cell Model

2.7

Modeling of Current Collectors


In previous versions of FLUENT, user-dened scalar (UDS) equations could only be solved in uid zones. This restriction is now removed. As a result, the PEMFC module allows users to model current collectors as solid, as well as uid zones. One advantage of using solids as the current collector is that the convergence of the species equations are not hindered by the potentially skewed mesh inside the current collectors. If uid zones are used to model solid current collectors, FLUENT automatically set velocities to zero and cuts o species transport into these zones. If solid zones are used, however, you need to activate the solution of the electric potential (UDS-0) in these solid zones (see the separate FLUENT Users Guide for details). The value of the Electric Conductivity for the solid material needs to be assigned in the Materials panel.

Figure 2.7.1: The Electric Conductivity Field in the Materials Panel.

Note that the UDS Diusivity should be set to user-defined (cond::fuelcells2.2). Do not use the defined-per-uds option.

For more information on the user-dened scalar diusivity, see the separate FLUENT Users Guide.

2-22

c Fluent Inc. July 19, 2006

2.8 PEM Fuel Cell Boundary Conditions

2.8

PEM Fuel Cell Boundary Conditions


The following boundary conditions need to be dened for the PEM fuel cell simulation based on your problem specication: Anode Inlet Mass ow rate Temperature Direction specication method Mass fractions (e.g., h2, and h2o). The coolant must be set to zero if coolant channels are enabled. UDS-2 (Water Saturation) must be set to 0 Cathode Inlet Mass ow rate Temperature Direction specication method Mass fractions (e.g. o2, h2o, and n2). The coolant must be set to zero if coolant channels are enabled. UDS-2 (Water Saturation) must be set to 0 Coolant Inlet (if any) Mass ow rate Temperature Direction specication method Coolant mass fraction set to 1 UDS-2 (Water Saturation) must be set to 0 Pressure Outlets (all) Realistic backow conditions. Terminal Anode Temperature (or ux if known) UDS-0 (electric potential) set to ground voltage Terminal Cathode Temperature (or ux if known)

c Fluent Inc. July 19, 2006

2-23

Using the PEM Fuel Cell Model

UDS-0 (electric potential) is set to the voltage of the cathode (if solving to constant voltage), or the UDS-0 (electric potential) ux is set to the current density in A/m2 (SI units) (if solving for constant current). Note that the sign of the UDS-0 ux on the cathode side is negative.

2.9

Solution Guidelines for the PEM Fuel Cell Model


For potentiostatic boundary conditions, after initialization, steady state solutions are calculated easily for cell voltages close to the open-circuit voltage. The same can be said for galvanostatic boundary conditions and low electric current. By lowering the cell voltage or by raising the average electric current, you can calculate subsequent stationary solutions. In the event of convergence problems, it is recommended to change the multigrid cycle to F-cycle with BCGSTAB (bi-conjugate gradient stabilized method) selected as the stabilization method for the species and the two potential equations. For the species and the user-dened scalar equations, it may be necessary to reduce the termination (criterium) of the multigrid-cycles to 1 103 . For stack simulations, the termination criterium may be reduced to 107 for the two potential equations. Also, it may be useful to turn o Joule Heating and Reaction Heating in the PEM Model panel (in the Model tab) for the rst few (approximately 5-10) iterations after initializations. This allows the two electric potentials to adjust from their initial values to more physical values, avoiding the possibility of extreme electrochemical reactions and electric currents that would in turn adversely impact the solution.

2.10

Postprocessing the PEM Fuel Cell Model

You can perform post-processing using standard FLUENT quantities and by using userdened scalars and user-dened memory allocations. By default, the FLUENT PEM fuel cell model denes several user-dened scalars and user-dened memory allocations, described in Table 2.10.1 and Table 2.10.2. Table 2.10.1: User-Dened Scalar Allocations UDS UDS UDS UDS 0 1 2 3 Electric Potential (solid phase potential) (Volts) Protonic Potential (membrane phase potential) (Volts) Water Saturation (liquid saturation) Water Content

2-24

c Fluent Inc. July 19, 2006

2.10 Postprocessing the PEM Fuel Cell Model

Table 2.10.2: User-Dened Memory Allocations UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM UDM 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 X Current Flux Density (A/m2 ) Y Current Flux Density (A/m2 ) Z Current Flux Density (A/m2 ) Current Flux Density Magnitude (A/m2 ) Ohmic Heat Source (W/m3 ) Reaction Heat Source (W/m3 ) Overpotential (Volts) Phase Change Source (kg/m3 -s) Osmotic Drag Coecient Liquid Water Activity Membrane Water Content Protonic Conductivity (1/ohm-m) Back Diusion Mass Source (kg/m3 -s) Transfer Current (A/m3 ) Osmotic Drag Source (kg/m3 -s)

You can obtain this list by opening the Execute On Demand panel and pulling down the Function drop-down list. Dene User-Dened Execute On Demand... and access the execute-on-demand function called list pemfc udf. Alternatively, you can view the listing that appears when you rst load your PEM fuel cell case, or you can type list pemfc udf in the text user interface and the listing will appear in the console window.

When you load older PEM fuel cell cases into FLUENT 6.3, and you are monitoring a UDS using volume or surface monitors, make sure you re-visit the corresponding monitors panel (e.g., the Dene Volume Monitor or the Dene Surface Monitor panel) to make sure that the correct UDS name is used for the appropriate monitor.

c Fluent Inc. July 19, 2006

2-25

Using the PEM Fuel Cell Model

2.11 User-Accessible Functions


As noted in Section 1.5: Properties, you can directly incorporate your own formulations and data for the properties of the fuel cell membrane using the pem user.c source code le. The following listing represents a description of the contents of the pem user.c source code le: real heat apportionment factor(cell t c, Thread *t): A fraction of the energy released in the chemical reaction for the formation of water is released as heat. This function gives this fraction of the maximal available energy of this reaction. real Get P sat(real T): Returns the value of the water vapor saturation pressure as a function of temperature (Equation 1.5-9). real Water Activity(real P, real T, cell t c, Thread *t): Returns the value of water activity (Equation 1.5-7). real Water Content (real act): Returns the value of the membrane water content at the membrane catalyst interface (Equation 1.5-6). real Osmotic Drag Coefficient(real P, real T, cell t c, Thread *t): Returns the value of the osmotic drag coecient (Equation 1.5-3). real Membrane Conductivity(real lam, cell t c, Thread *t): Returns the value of the membranes protonic conductivity (Equation 1.5-2). real Water Content Diffusivity(real lam, real T, real mem mol density, cell t c, Thread *t): Returns the value of the water content diusivity in the membrane(Equation 1.5-5). real Gas Diffusivity(cell t c, Thread *t, int j spe): Returns the value of the gaseous species diusivities in the channels, gas diusion layers and catalysts (Equation 1.5-2). real Saturation Diffusivity(real sat, real cos theta, real porosity, cell t c, Thread *t): Returns the value of diusivity of the liquid saturation. It comprises the term l from Equation 1.4-1. real Anode AV Ratio(cell t c, Thread *t): Returns the value of the surfaceto-volume ratio s used in Equation 1.3-5 for the anode catalyst. Ks3 dpc l ds (2.11-1)

2-26

c Fluent Inc. July 19, 2006

2.11 User-Accessible Functions

real Cathode AV Ratio(cell t c, Thread *t): Returns the value of the surfaceto-volume ratio s used in Equation 1.3-5 for the cathode catalyst. real Anode J TransCoef(cell t c, Thread *t): Returns the value of the anode ref reaction reference current density ([H2j]an )an used in Equation 1.2-3. ref real Cathode J TransCoef(cell t c, Thread *t): Returns the value of the cathj ref ode reaction reference current density ([O2 ]cat)cat used in Equation 1.2-4. ref real Open Cell Voltage(cell t c, Thread *t): Returns the value of the opencircuit voltage Voc used in Equation 1.2-8. void Set UDS Names(char uds[n uds required][STRING SIZE]): Used to rename user dened scalars (UDSs). Note that the units of the user dened scalars cannot be changed. void Set_UDS_Names(char uds[n_uds_required][STRING_SIZE]) { strncpy(uds[0], "Electric potential", STRING_SIZE-1); strncpy(uds[1], "Protonic potential", STRING_SIZE-1); strncpy(uds[2], "Water saturation", STRING_SIZE-1); strncpy(uds[3], "Water content", STRING_SIZE-1); } If you want to change the names of UDSs, change the second argument of the strncpy functions, recompile and link the module as with any modication to pem user.c. Note that STRING SIZE is xed in pem.h and should not be changed.

When you load older PEM fuel cell cases into FLUENT 6.3, and you are monitoring a UDS using volume or surface monitors, make sure you re-visit the corresponding monitors panel (e.g., the Dene Volume Monitor or the Dene Surface Monitor panel) to make sure that the correct UDS name is used for the appropriate monitor.

c Fluent Inc. July 19, 2006

2-27

Using the PEM Fuel Cell Model

void Set UDM Names(char udm[n udm required][STRING SIZE]): Used to rename user dened memory (UDMs). Note that the units of user dened memory cannot be changed. void Set_UDM_Names(char udm[n_udm_required][STRING_SIZE]) { strncpy(udm[ 0], "X Current Flux Density", STRING_SIZE-1); strncpy(udm[ 1], "Y Current Flux Density", STRING_SIZE-1); strncpy(udm[ 2], "Z Current Flux Density", STRING_SIZE-1); strncpy(udm[ 3], "Mag Current Flux Density", STRING_SIZE-1); strncpy(udm[ 4], "Ohmic heat source", STRING_SIZE-1); strncpy(udm[ 5], "Reaction heat source", STRING_SIZE-1); strncpy(udm[ 6], "Overpotential", STRING_SIZE-1); strncpy(udm[ 7], "Phase change source", STRING_SIZE-1); strncpy(udm[ 8], "Osmotic drag coeff", STRING_SIZE-1); strncpy(udm[ 9], "Activity liquid water", STRING_SIZE-1); strncpy(udm[10], "Membrane water content", STRING_SIZE-1); strncpy(udm[11], "Protonic conductivity", STRING_SIZE-1); strncpy(udm[12], "Back diffusion mass source", STRING_SIZE-1); strncpy(udm[13], "Transfer current", STRING_SIZE-1); strncpy(udm[14], "Osmotic drag source", STRING_SIZE-1); } If you want to change the names of UDMs, change the second argument of the strncpy functions, recompile and link the module as with any modication to pem user.c. Note that STRING SIZE is xed in pem.h and should not be changed.

When you load older PEM fuel cell cases into FLUENT 6.3, and you are monitoring a UDM using volume or surface monitors, make sure you revisit the corresponding monitors panel (e.g., the Dene Volume Monitor or the Dene Surface Monitor panel) to make sure that the correct UDM name is used for the appropriate monitor.

2-28

c Fluent Inc. July 19, 2006

2.11 User-Accessible Functions

2.11.1

Compiling the Customized PEM Fuel Cell Source Code

This section includes instructions on how to compile a customized PEM fuel cell userdened module. Note that you can also refer to the le INSTRUCTIONS-CLIENT that comes with your distribution (see addons/fuelcells2.2).

It is assumed that you have a basic familiarity with compiling user-dened functions (UDFs). For an introduction on how to compile UDFs, please refer to the separate UDF manual.

You will rst want to use a local copy of the fuelcells2.2 directory in the addons directory before you recompile the PEM fuel cell module.

Compiling the Customized Source Code Under UNIX


1. Make a local copy of the fuelcells2.2 directory. Do not create a symbolic link. 2. Change directories to the fuelcells2.2/src directory. 3. Make changes to the pem user.c le. 4. Edit the makele. You must edit the makefile located in the src/ directory and make sure that the FLUENT INC variable correctly refers to the current FLUENT installation directory. Be careful not to leave any trailing spaces when you make your changes. 5. Change directories to the fuelcells2.2/ directory. 6. Issue the following make command: make -f Makefile-client FLUENT_ARCH=your_arch where your arch is lnx86 on LINUX, or ultra on the Sun operating system, etc.

Compiling the Customized Source Code Under Windows


1. Open Visual Studio .NET at the DOS prompt. 2. Make sure that the $FLUENT INC environment variable is correctly set to the current FLUENT installation directory. 3. Make a local copy of the fuelcells2.2 folder. Do not create a shortcut. 4. Enter the fuelcells2.2\src folder. 5. Make changes to the pem user.c le.

c Fluent Inc. July 19, 2006

2-29

Using the PEM Fuel Cell Model

6. Return to the fuelcells2.2 folder. 7. Issue the following command in the command window: nmake /f makefile_master-client.nt

2.12

IV-Curve Calculations Using the Text Interface

A text user interface (TUI) exists for the PEM Fuel Cell Model. The TUI allows text commands to be typed directly in the FLUENT console window where additional information can be extracted and processed for more advanced analysis. For valid case and data les, there are two text commands available to assist in the IV-curve calculation. These commands are set-stack-voltage (aliased as ssv) and set-stack-current-density (aliased as ssc), available from the PEM Fuel Cell text command menu: /define/models/pemfc/. For fuel cells, you either prescribe the voltage and obtain the total current delivered by the fuel cell as a result, or you specify the total current (via ux boundary conditions multiplied by the area) and obtain the voltage as part of the solution. The details of this IV-relation are specic for each single fuel cell and depend on mass and heat transport, electrochemistry and inlet conditions, outlet conditions, operating conditions, and any other parameter or material property involved in the calculation. The IV-curve is important for applications, because its product is the power delivered by the system. As described earlier in this manual, you would start a new simulation from fairly static conditions, i.e., high voltage/low current (which implies low species transport and low heat generation). After convergence, you typically may be interested in solutions for new electric boundary conditions, i.e., either for a new cell/stack voltage or current. In such cases, simply going to the Boundary Conditions panel and changing the value of the electric potential (uds-0) boundary condition, typically allows only small changes, most notably for stacks. Otherwise the solution will not converge. This is where the set-stack-voltage and set-stack-current-density commands are important. In addition to changing the boundary conditions (either to a prescribed voltage or current density), these commands process the current data in order to estimate the solution for the new boundary conditions. Because these commands modify the data, you are prompted to save your data, if you have not already done so. Before going into details of the commands, here are some general remarks about electric potential boundary conditions. For xed voltage boundary conditions, both external contacts have a xed value for the electric potential (uds-0). The anode value will typically be zero, but it does not have to be. The cathode value will be larger than the anode value and the dierence (Vcathode -Vanode ) is the positive cell/stack voltage.

2-30

c Fluent Inc. July 19, 2006

2.12 IV-Curve Calculations Using the Text Interface

For a xed current boundary condition, one external contact has to have a xed value and the other ux boundary conditions. As described earlier in the manual, typically, the anode will have a xed (zero) value, and the cathode will be oating, however, you can also set the cathode to a xed zero potential, yielding a oating negative anode potential. The set-stack-voltage command sets the eective stack voltage, i.e., the dierence (Vcathode -Vanode ). For xed voltage boundary conditions for the previous solution, boundary conditions on both boundaries are of type xed value and then the cathode value will be changed accordingly. In the case of xed current boundary conditions for the previous solution, the ux boundary condition will be changed to a xed value boundary condition, and the value adjusted accordingly with respect to the other xed value boundary condition. The set-stack-current-density command sets the current density on one boundary A A to the desired value. Note that the input will be in cm2 , not m2 as you would normally have to enter in the Boundary Conditions panel. The reason for this is that average current A densities reported in the text command interface are also in cm2 , and this makes it easier to choose the conditions you would like to prescribe next. Also, ux boundary conditions entered in the Boundary Conditions panel would have to have a positive sign on the anode side, and a negative sign on the cathode side. The input for the text interface command is just a positive number, signs are automatically accounted for. For xed current boundary conditions for the previous solution, the set-stack-current-density command changes the respective ux boundary condition accordingly. In the case of xed voltage boundary conditions for the previous solution, the cathode side is chosen to be changed from a xed value to a ux boundary condition with the new ux. The two commands may be mixed in an IV-curve calculation. For the type of boundary condition setups currently described in this manual, boundary condition changes will consistently happen on the cathode side. However, if anode ux boundary conditions had been chosen initially, switching to xed voltage boundary conditions by set-stack-voltage command and then back to xed current boundary conditions by the set-stack-current-density command will then have ux boundary conditions on the cathode side. In this case, using the set-stack-current-density command exclusively will preserve the anode ux boundary condition setting.

c Fluent Inc. July 19, 2006

2-31

Using the PEM Fuel Cell Model

2-32

c Fluent Inc. July 19, 2006

Chapter 3.

SOFC Fuel Cell Model Theory

This chapter presents an overview of theory and equations for solid oxide fuel cell (SOFC) modeling capabilities in FLUENT. Section 3.1: Introduction Section 3.2: The SOFC Modeling Strategy Section 3.3: Modeling Stacked Fuel Cells Section 3.4: Modeling Fluid Flow, Heat Transfer, and Mass Transfer Section 3.5: Modeling Current Transport and the Potential Field Section 3.6: Modeling Electrochemical Reactions

3.1

Introduction
The Solid Oxide Fuel Cell (SOFC) module is provided as an addon module with the standard FLUENT licensed software. A special license is required to use the SOFC module. A fuel cell is an energy conversion device that converts the chemical energy of fuel into the electrical energy. A schematic of a solid oxide fuel cell (SOFC) is shown in Figure 3.1.1.

c Fluent Inc. July 19, 2006

3-1

SOFC Fuel Cell Model Theory


Cathode Flow Channels

Cathode Current Collector Cathode Interlayer

Anode Interlayer Anode Current Collector

As noted in [4], a solid oxide fuel cell is typically composed of an anode, cathode, and an electrolyte. Multiple fuel cells can be connected together, or stacked, using electrical interconnects. The electrolyte material must be solid, i.e., non-porous, and exhibit a high ionic conductivity. All components of the fuel cell must have similar thermal expansion in order to minimize thermal stresses, which may cause cracking and de-lamination during thermal cycling. In addition, the components must be chemically stable in order to limit chemical interactions with other cell components. A solid oxide fuel cell works by having electrically conducting porous ceramic electrodes attached on each side of an ionically conducting ceramic material. At the cathode/electrolyte/gas interface, also known as the triple phase boundary, oxygen is reduced to oxygen ions. The oxygen ions are conducted through the oxygen vacancies in the electrolyte to the anode side. At the anode/electrolyte/gas interface, oxygen ions combine to react with hydrogen at the anode electrode to form water and release electrons. The electrons travel through an external circuit to a load and back to the cathode electrode to close the circuit.

3-2


2e O2 2e

Cathode Electrode

Electrolyte


2e H 2O H2 2e H 2O

Anode Electrode


Anode Flow Channels

Figure 3.1.1: Schematic of a Solid Oxide Fuel Cell

c Fluent Inc. July 19, 2006

3.1 Introduction

The FLUENT SOFC model provides the following features: Local electrochemical reactions coupling the electric eld and the mass, species, and energy transport. Electric eld solution in all porous and solid cell components, including ohmic heating in the bulk material. The ability to handle H2 and combined CO/H2 electrochemistry. Inclusion of tortuosity for porous regions The ability to treat an arbitrary number of electrochemical cells arrayed as a stack. Signicant geometric exibility for treating planar, tubular, and other nonstandard SOFC conguration.

3.1.1

Motivation for Consideration of the Electric Field

Electrical conduction throughout the entire fuel cell assembly can cause a signicant change to electrochemical operation of the fuel cell. Normally, the fuel cell is thought of as an idealized planar or cylindrical surface of reaction. Local species concentrations, pressure, and temperature inuence the distribution of current along the electrolyte. In addition, ohmic losses other than those at the electrolyte inuence the current density as well. As electrolyte thicknesses are reduced and the electrolyte resistance becomes less dominate, these external resistive losses become important. When the electrode resistance become signicant compared to that of the electrolyte, you can expect the size and location of the current collector to have an impact on the current density. This can certainly become the case for widely-spaced current collectors or thin electrodes.

3.1.2

Overview of the Electric Field Model

The existing functionality of solving a user dened scalar in FLUENT has been extended in order to solve for the potential eld in the solid oxide fuel cell. First, additional functionality has been added to the SOFC user-dened function that enables the solution of a user dened scalar in solid zones. Second, to handle the electrolyte interfaces and contact resistance interfaces, a more general ux computation needs to be dened and incorporated into the FLUENT UDS solution (see Section 3.5.2: Treatment of the Electrolyte Interface).

c Fluent Inc. July 19, 2006

3-3

SOFC Fuel Cell Model Theory

3.1.3

Overview of the Electrochemical Model

The local Nernst potential is a function of the composition of the ow immediately adjacent to the electrolyte in the anode and cathode. The activation and ohmic losses in the electrolyte is a function of the current density. Likewise, species and energy uxes are determined by the current density. Therefore, any external eects can have a tangible eect not only on the current density distribution but also on the transport behavior of the fuel cell. In the electrochemical model, the two sides of the electrolyte is separated by a local potential dierence. This dierence is determined by three physical eects: Nernst potential, ohmic losses in the electrolyte, and activation losses at the electrode-electrolyte interfaces. V = N iele tele act (i) (3.1-1)

where i is the local current density. For the idealized case with uniform contact between the electrodes and highly conductive current collectors, you can specify a cell voltage and solve for the current density distribution using this relation.

3.2

The SOFC Modeling Strategy


To model solid oxide fuel cells (SOFC), you need to perform the following: Capture the uid ow, heat transfer, and the mass transfer in the ow channels and in the porous anode and cathode electrodes. Model the transport of the current and the potential eld in the porous electrodes and in the solid conducting regions. Model the electrochemical reactions that take place at the electrolyte/electrode/gaseous species interface.

3-4

c Fluent Inc. July 19, 2006

3.3 Modeling Stacked Fuel Cells


Local species concentration and temperature

FLUENT
Species Momentum Energy Electric Potential Field

SOFC Model
Nernst Voltage Current Distribution and Overpotentials at Electrolyte Electric Potential Field Boundary Conditions

Species and heat flux at the boundaries

Figure 3.2.1: How the SOFC Model Works in FLUENT

3.3

Modeling Stacked Fuel Cells


Typically in fuel cell modeling, the cell current at a specied cell voltage is calculated. When modeling stacked fuel cells with this approach, you must insure the conservation of charge is satised at each cell in the stack.. To simplify stack modeling, the FLUENT SOFC model calculates the cell voltage at a specied cell current. Specifying the fuel cell current as an input insures that the total current ux through each cell in a stack is conserved.

c Fluent Inc. July 19, 2006

3-5

SOFC Fuel Cell Model Theory

3.4

Modeling Fluid Flow, Heat Transfer, and Mass Transfer


All aspects of uid ow, heat transfer, and mass transfer in the ow channels and porous electrodes are handled by FLUENT. The default multicomponent diusion model in FLUENT is used to calculate the mass diusion coecient of species i in the mixture. To account for the eect of porosity on the multicomponent mass diusion coecient Dij,e = Dij (3.4-1)

where is the porosity and is the tortuosity (i.e., the average path length over the actual length).

3.5

Modeling Current Transport and the Potential Field


Solving for three-dimensional electrical conduction is directly analogous to the calculation of heat transfer. The potential eld throughout the conductive regions is calculated based on the conservation of charge. i=0 where i = (3.5-2) (3.5-1)

and is the electrical conductivity and is the electrical potential. Therefore, the governing equation for the electric eld is the Laplace equation: ( ) = 0 The electric eld potential calculation combines the following attributes: Ohmic losses in all the conducting materials, including the electrolyte, electrodes, and current collectors. Contact resistance at the appropriate interfaces. Ohmic heating through conduction materials as the result of ohmic losses, and the current density throughout the domain. (3.5-3)

3-6

c Fluent Inc. July 19, 2006

3.5 Modeling Current Transport and the Potential Field

3.5.1

Treatment of the Activation Overpotential

Since the Butler-Volmer relation for activation overpotential is nonlinear, it is useful to build a locally linearized form. This allows for a more complete coupling of the electrochemistry and ultimately better model stability and improved convergence. Using a previous value for the local current density, iold , a Taylor series expansion can be constructed:

act (i) o 1 (i iold ) o = act (iold ) act 1 = |i=iold i

(3.5-4) (3.5-5) (3.5-6)

With this approximation, the formula for the voltage jump across the electrolyte becomes: V [N o + 1 iold ] i[ele tele + 1 ] (3.5-7)

The rst term in brackets is now a constant. The second term is linear with local current density. Here, the activation overpotential is included as a direct reduction of the voltage and as an additional eective resistance. During the calculation, this form allows the activation overpotential to respond to changes in local current density. Even though the Butler-Volmer relation is nonlinear, the linearized form becomes more accurate as the solution reaches convergence and is exact at convergence.

3.5.2 Cell Potential


The electrode reactions are assumed to take place in a single step. The charge transfer reaction acts as the rate limiting step for the electrode reactions. Oxygen is electrochemically reduced at the triple phase boundary at the cathode electrode: 1 O2,cathode + 2e O2 2 (3.5-8)

Oxygen is electrochemically re-oxidized at the triple phase boundary at the anode electrode: 1 O2 O2,anode + 2e 2 (3.5-9)

c Fluent Inc. July 19, 2006

3-7

SOFC Fuel Cell Model Theory

In the absence of an electrical load, the oxygen activity on both sides of the electrolyte is xed and given by their respective chemical potentials. Under equilibrium, the electromotive force, or reversible cell voltage, is given by the Nernst equation: ideal = RT pO2,cathode ln 4F pO2,anode (3.5-10)

If is present at the anode electrode, then the cell reaction becomes: H2 + O2 H2 O + 2e At equilibrium, the cell voltage is given by the Nernst equation:
1 2 RT pH2 pO2 = + ln 2F pH2 O

(3.5-11)

ideal

(3.5-12)

The cell potential measured at equilibrium (i.e., no load), is called the open circuit voltage. The open circuit voltage should be equivalent to the Nernst potential at no load, unless there is leakage across the electrolyte. When the external circuit is closed, then cell voltage drops due to polarization losses at the electrodes. The electric eld and the electrochemistry interact solely at the electrolyte interface. FLUENT treats the electrolyte interface as an impermeable wall. The potential eld must have a jump condition applied to the two sides of this wall to account for the eect of the electrochemistry. To closely couple the electrochemical behavior to the potential eld calculation, you need to include all of the electrochemical eects into this jump condition. It encapsulates the voltage jump due to Nernst, the voltage reduction due to activation, the Ohmic losses due to the resistivity of the electrolyte, and a linearized for voltage reduction due to activation. This interface condition relates the potential on the anode side and the cathode side of the electrolyte and has the following form: cell = jump s where jump = ideal ele act,a act,c (3.5-14) (3.5-13)

where ele represents the ohmic overpotential of the electrolyte, and act,a act,c represent the activation overpotential of the anode and the cathode. s represent ohmic losses in the solid conducting regions. ideal represents the Nernst potential. FLUENT solves the Laplace equation for the potential eld by subdividing the computational domain into control volumes and enforcing ux conservation on each cell. For

3-8

c Fluent Inc. July 19, 2006

3.5 Modeling Current Transport and the Potential Field

cells away from the electrolyte boundary, the ux from a cell to its neighbor is given by Ohms law: i = (3.5-15)

This gradient is approximated using the numerical dierence of the values of in each of the cells. At the electrolyte interface, Ohms law along with the relation for voltage dierence must be used. In that case, for normal ux, there are three separate voltage relations. That coupling the anode computational cell center voltage to the anode/electrolyte face voltage, the cathode cell center voltage to the cathode/electrolyte face voltage, and the electrochemistry jump condition:

anode centroid = anode wall ianode Danode cathode centroid = cathode wall icathode Dcathode anode wall = cathode wall i(t)e

(3.5-16) (3.5-17) (3.5-18)

where Danode and Dcathode are the distances from the electrode computational cell center to the electrolyte. Together, these three linear equations can be solved to compute the normal current ux, i, based on the values of the electric eld in the anode and cathode computational cells. By incorporating this linearized form of the electrochemistry model in this way, the solution of the electric eld directly includes variations in species concentrations via Nernst variations, resistive response of the electrolyte, and an approximate response to changes in activation losses.

Treatment of the Electrolyte Interface


Allowing the solution of the electric eld in the solid and the inclusion of electrolyte and contact resistance interfaces requires non-standard modications to be made to the FLUENT UDS solution procedure. FLUENT imposes a ux balance between computational cells that share a common face. It is the calculation of this ux that is the cornerstone of the solution procedure and where the special interface modications are to be made. This section presents an overview of how FLUENT manages the unique conditions at the electrolyte interface. For two neighboring computational cells, C0 and C1 , that share a face F , you can write formulae for the ux (or current density). Let the value of the potential eld at the cell centroids be denoted as 0 and 1 . Also, let the values of the potential eld at the face centroids be f0 and f1 . Note that though both values are located at the same physical location, the presence of contact resistance and/or electrolytic reactions can

c Fluent Inc. July 19, 2006

3-9

SOFC Fuel Cell Model Theory

cause a discontinuity in the potential eld. These two face values represent the values of the potential on each side of the discontinuity. Each computational cell can possess its own electrical conductivity, 0 and 1 . The face can have a potential jump, V , and a contact resistance, Rc . If we assume Cartesian cells for simplicity, the current ux from the C0 centroid to the face centroid is: i= 0 Aface (f0 0 ) dx0 (3.5-19)

The current from the face centroid to the C1 centroid is: i= 1 Aface (1 f1 ) dx1 (3.5-20)

Now, at the discontinuity, the face values of potential change discontinuously. The current can be deduced from the following relation: f1 = f0 + V iRc This can be rearranged into a form similar to the previous two equations: i= 1 (f0 f1 + V ) Rc (3.5-22) (3.5-21)

You can dene connection parameters to simplify these: i = 0 (f0 0 ) where 0 = and i = 1 (1 f1 ) where 1 = and i = c (f0 f1 + V ) (3.5-27) 1 Aface dx1 (3.5-26) (3.5-25) 0 Aface dx0 (3.5-23) (3.5-24)

3-10

c Fluent Inc. July 19, 2006

3.5 Modeling Current Transport and the Potential Field

where c =

1 Rc

(3.5-28)

These equations can then be manipulated to solve for the current and eliminate the face potential values: i= 0 1 c (0 1 + V ) 0 1 + 0 c + 1 c (3.5-29)

This denition of the current ux can then be used in FLUENT to build the system of linear equations to solve for the potential eld. This formulation is valid for any interface that has a contact resistance and voltage jump. It is trivial to set the voltage jump to zero for the case of pure contact resistance, as would occur between electrochemically inert conductive layers. For cases where no contact resistance is present, the above formula is valid if you take the limit as c . i= 0 1 (0 1 + V ) 0 + 1 (3.5-30)

Inclusion of these general ux computations into the FLUENT UDS solution scheme provides the ability to directly compute jump and contact resistance interfaces. Treatment of the Energy Equation at the Electrolyte Interface For an incompressible ow, the energy equation that FLUENT solves for within each computational cell is given by the following: (E) + t

(v(E + p)) =

ke T

hj Jj + ( e v) + Sh

(3.5-31)

where Sh is the volumetric source or sink of energy and where E =h and h=


j

p v2 + 2

(3.5-32)

Yj hj

(3.5-33)

c Fluent Inc. July 19, 2006

3-11

SOFC Fuel Cell Model Theory

In all electrically conducting zones (e.g., electrodes, current collectors, interconnects), ohmic heating, i2 Rohmic , is added to the energy equation as a source term. In other words, Sh = i2 Rohmic (3.5-34)

In addition, the energy equation needs treatment at the electrode-electrolyte interface to account for the heat generated or lost as the result of electrochemistry and the overpotentials (i.e., activation overpotential and ohmic loss through the electrolyte).
anode side

50 % into cell source term

energy balance

electrolyte interface

50 % into cell source term

cathode side

Figure 3.5.1: Energy Balance at the Electrolyte Interface

The total energy balance on the electrolyte interface is computed by enumerating the enthalpy ux of all species, including the heat of formation (sources of chemical energy entering the system), and then subtracting o the work done (leaving the system) which is simply the local voltage jump multiplied by the local current density. What remains is the waste heat due to irreversibilities. For hydrogen reaction, the balance would be Q = hH2 + hO2 hH2 0 iV (3.5-35)

where Q is the heat generation (W) and h is the total enthalpy of species (J/s) composed of the sensible enthalpy in addition to the enthalpy of formation.

3-12

c Fluent Inc. July 19, 2006

3.5 Modeling Current Transport and the Potential Field

The heat of formation is


T

hH2 = mH2 [

Tref

Cp dT + h0 ]

(3.5-36)
Q . V olume

The source term is then added in the cell energy equation by taking Sh =

One half of this value is applied as a source term to the energy equation of the anode computational cell adjacent to the electrolyte and the other half is applied as a source term to the energy equation for the cathode cell adjacent to the electrolyte. The equal distribution of the heat generation/destruction is purely arbitrary. Note that by using the work term, the eect from all overpotentials are taken into account.

Ohmic Overpotential
Ohmic polarization involves ionic losses through the electrolyte, electrical resistance in the conducting porous electrodes and solid collectors. The ohmic polarization also includes the electrical resistance at the interface of the current collectors and the electrodes or the electrodes and the membrane (i.e., the contact resistance). ohmic = i R (3.5-37)

Activation Overpotential
The general electrochemical reaction is, according to [6],
N j

aj Aj ne

(3.5-38)

where aj is the stoichiometric coecient of species j, Aj is the chemical species, and n is the number of electrons. The reaction rate is: r=
a nF i = ka e RT nF

ci pi kc e
i

c nF RT

c i qi

(3.5-39)

where is the voltage, ka and pi are the rate constant and the reaction order for the anodic direction, kc and qi are the rate constant and the reaction order for the cathodic direction, a is the anodic transfer coecient, c is the cathodic transfer coecient, and n is the number of electrons that are released.

c Fluent Inc. July 19, 2006

3-13

SOFC Fuel Cell Model Theory

At equilibrium, the forward and the backward reaction rates are the same, therefore:
a nF i0 = ka e RT 0 nF

ci pi = kc e
i

c nF RT

0 i

c i qi

(3.5-40)

where i0 is the exchange current density. The reaction rate (i.e., current) can be written in terms of the exchange current density i0 to obtain the Butler-Volmer formulation [6]: i = i0 e
a n(0 )F RT

c n(0 )F RT

(3.5-41)

The activation overpotential is the energy lost due to the slowness of electrochemical reactions at the anode and the cathode electrodes. act = 0 Using this relation, the Butler-Volmer equation can be written as: i = i0e e where i0e = i0,ref (Yj ) (3.5-44)
a nact F RT

(3.5-42)

c nact F RT

(3.5-43)

with i0,ref being the exchange current density at the reference condition, (Yj ) is the mole fraction and is the concentration exponent. Given values for a and c . the full version of the Butler-Volmer equation can be solved using the Newton method, therefore nding the activation overpotential at the anode (act,a ) and the cathode (act,c ).

3-14

c Fluent Inc. July 19, 2006

3.6 Modeling Electrochemical Reactions

3.6

Modeling Electrochemical Reactions


The rate of species production and destruction is: S= ai (g mole/m2 /sec) nF (3.6-1)

where S is the source or sink of the species (molar ux), a is the stoichiometric coecient, i is the current density (A/m2 ), n is the number of electrons per mole of fuel, and F is the Faraday constant. Using the local current information, the FLUENT SOFC model applies species uxes to the electrode boundaries. By convention [6], the current density is positive when it ows from the electrode into the electrolyte solution. The current densities are positive at the anodes and negative at the cathodes. The reaction at the cathode electrode is: 1 O2,cathode + 2e O2 2 or SO 2 = ( 1 )(i) i 2 = 2F 4F (i) i = 2F 2F (3.6-2)

(3.6-3)

2 SO =

(3.6-4)

The reaction at the anode electrode is: H2 + O2 H2 O + 2e i 2F i 2F (3.6-5)

SH2 =

(3.6-6)

2 SO =

(3.6-7)

SH2 O =

(1)(i) i = 2F 2F

(3.6-8)

c Fluent Inc. July 19, 2006

3-15

SOFC Fuel Cell Model Theory

3-16

c Fluent Inc. July 19, 2006

Chapter 4.

Using the Solid Oxide Fuel Cell Model

The procedure for setting up and solving solid oxide fuel cell problems is described in detail in this chapter. Please refer to the following sections for more information: Section 4.1: Installing the Solid Oxide Fuel Cell Model Section 4.2: Loading the Solid Oxide Fuel Cell Module Section 4.3: Solid Oxide Fuel Cell Module Set Up Procedure Section 4.4: Setting the Parameters for the SOFC Model Section 4.5: Setting Up the Electrode-Electrolyte Interfaces Section 4.6: Setting Up the Electric Field Model Parameters Section 4.7: Setting Up the Tortuosity Parameters Section 4.8: Setting Up the Activation Parameters

4.1

Installing the Solid Oxide Fuel Cell Model


The Solid Oxide Fuel Cell (SOFC) model is provided as an addon module with the standard FLUENT licensed software. A special license is required to use the SOFC model. The module is installed with the standard installation of FLUENT in a directory called addons/sofc1.2 in your installation area. The SOFC module consists of a UDF library and a pre-compiled scheme library, which needs to be loaded and activated before calculations can be performed.

4.2

Loading the Solid Oxide Fuel Cell Module


The Solid Oxide Fuel Cell (SOFC) module is loaded into FLUENT through the text user interface (TUI). The module can only be loaded after a valid FLUENT case le has been set or read. The text command to load the addon module is define models addon-module

c Fluent Inc. July 19, 2006

4-1

Using the Solid Oxide Fuel Cell Model

A list of FLUENT addon modules is displayed: FLUENT Addon Modules: 1. MHD Model 2. Fiber Model 3. PEM Fuel Cell Model 4. SOFC Fuel Cell Model 5. Population Balance Model Enter Module Number: [1] 4 Select the SOFC model by entering the module number 4. During the loading process a scheme library containing the graphical and text user interface, and a UDF library containing a set of user dened functions are loaded into FLUENT.

4.3

Solid Oxide Fuel Cell Module Set Up Procedure


The following describes an overview of the procedure required in order to use the SOFC model in FLUENT. 1. Start FLUENT. You must start FLUENT in 3d double-precision mode. Note that the SOFC model is only available in 3d. 2. Read the case le. File Read Case... 3. Scale the grid. Grid Scale... 4. Dene various model parameters for the simulation. (a) Open the Solver panel. Dene Models Solver... i. Enable Pressure Based under Solver. ii. Enable Implicit under Formulation. iii. Enable Steady under Time. iv. Enable Absolute under Velocity Formulation. (b) Open the Energy panel and enable the Energy option (if it is not already enabled). Dene Models Energy...

4-2

c Fluent Inc. July 19, 2006

4.3 Solid Oxide Fuel Cell Module Set Up Procedure

(c) Open the Viscous Model panel and enable the Laminar option. Dene Models Viscous... (d) Open the Species Model panel. Dene Models Species Transport & Reaction... i. Enable the Species Transport option. ii. Enable the Diusion Energy Source option under Options. iii. Enable the Full Multicomponent Diusion option under Options. iv. Enable the Thermal Diusion option under Options. (e) Set the parameters for the SOFC model. Dene Models SOFC SOFC Model Control Panel... i. Set the Current Underrelaxation Factor to a value of either 0.3 or 0.4. ii. Set the Electrochemical and Electrical Parameters according to your problem specication. (f) Set the anode interface components for the SOFC model. Dene Models SOFC Anode Interface Setup... i. Specify a zone in the Zone(s) list. ii. Enable Anode Interface if it is applicable. (g) Set the cathode interface components for the SOFC model. Dene Models SOFC Cathode Interface Setup... i. Specify a zone in the Zone(s) list. ii. Enable Cathode Interface if it is applicable. (h) Set the tortuosity parameters for the SOFC model. Dene Models SOFC Tortuosity Setup... i. Specify a zone in the Zone(s) list. ii. Turn on Enable Tortuosity if it is applicable. (i) Set the parameters for the electric eld model. Dene Models SOFC Electric Field Model Input Panel... i. Specify up to 5 conductive regions. ii. Specify up to 3 contact surfaces. iii. Specify a voltage tap surface. iv. Specify a current tap surface.

c Fluent Inc. July 19, 2006

4-3

Using the Solid Oxide Fuel Cell Model

(j) Set the activation parameters. Dene Models SOFC Activation Panel... i. Enter values for the Constant Exchange Current Densities for the anode and the cathode. These are the i0 values in the Butler-Volmer equation for the anodic and cathodic reactions. By default, the Anode Exchange Current Density is set to 1000 Amps and the Cathode Exchange Current Density is set to 100 Amps. ii. Enter values for the Mole Fraction Reference Values. These are the species concentrations at which the exchange current densities were taken. These are used to adjust the i0 values of reactant species that are depleted. By default, the H2 Reference Value is set to 0.8 moles/moles, the 02 Reference Value is set to 0.21 moles/moles, and the H2O Reference Value is set to 0.2 moles/moles. iii. Enter values for the Stoichiometric Exponents by setting values for the H2 Exponent, the H20 Exponent, and the O2 Exponent (defaulted to 0.5). These are the stoichiometric factors in the electrochemical reaction equation. They are used as exponents as part of the i0 scaling. iv. Enter values for the Butler-Volmer Transfer Coecients by setting values for the Anodic Transfer Coecient and the Cathode Transfer Coecient for both the anode reaction and the cathode reaction (defaulted to 0.5). These are the alpha values in the Butler-Volmer equation. They represent the forward and backward rates of reaction at both the anode and cathode. v. Enter a values for the Temperature Dependent Exchange Current Density by turning on the Enable Temperature Dependent I 0 option and setting values for A and B. These two coecients allow the i0 for the cathode to vary as a function of temperature. 5. Dene material properties. (a) Dene a user-dened scalar. i. Open the User-Dened Scalars panel. Dene User-Dened Scalars... ii. Change the Number of User-Dened Scalars to 1. iii. Indicate none as the Flux Function and click OK to close the User-Dened Scalars panel. (b) Dene 12 user-dened memory locations. i. Open the User-Dened Memory panel. Dene User-Dened Memory...

4-4

c Fluent Inc. July 19, 2006

4.3 Solid Oxide Fuel Cell Module Set Up Procedure

ii. Change the Number of User-Dened Memory Locations to 12. iii. Click OK to close the User-Dened Memory panel. (c) Dene user-dened function hooks. i. Open the User-Dened Function Hooks panel. Dene User-Dened Function Hooks... ii. Change the Adjust function to adjust function. iii. Click OK to close the User-Dened Function Hooks panel. (d) Create or re-dene new solid materials as appropriate for the anode, the cathode, and the electrolyte according to your problem specication.

Note that since the FLUENT SOFC model does not currently support the shell conduction model, you cannot use it to take into account the transversal conductive heat inside the membrane.

(e) Edit the mixture-template mixture material. i. In the Materials panel, click Fluent Database... to open the Fluent Database Materials panel. ii. In the Fluent Database Materials panel, make a copy of the h2 uid material. iii. In the Materials panel, change the Material Type to mixture and click Edit... for the Mixture Species. iv. In the Species panel, arrange the materials under Selected Species in the following order: h2o, o2, h2, and n2. v. In the Materials panel, change the Thermal Conductivity and the Viscosity to ideal-gas-mixing-law. vi. Change the Viscosity to ideal-gas-mixing-law. vii. Change the Mass Diusivity to user-dened and select diusivity::sofc1.2 as the corresponding user-dened function. viii. Change the UDS Diusivity to user-dened and select E Conductivity::sofc1.2 as the corresponding user-dened function. ix. Retain the default values for the other parameters and click Change/Create. 6. Set the operating conditions. Dene Operating Conditions... Retain the default values and click OK in the Operating Conditions panel.

c Fluent Inc. July 19, 2006

4-5

Using the Solid Oxide Fuel Cell Model

7. Set the boundary conditions. Dene Boundary Conditions... Dene the conditions at the anode, the cathode, the interface between the anode and current collector, the interface between the cathode and the current collector, the anode inlet, and the cathode inlet boundaries according to your problem specication. Note that sources only need to be hooked to the mass, species, and energy equation in a single uid zone in order for the FLUENT SOFC model to function properly (but can be hooked to all zones if you so choose). FLUENT does not rely on cell or thread referencing in the sources in order to cover the solution domain. 8. Set the multigrid control parameters. Solve Controls Multigrid... (a) Set the cycle type for h2, h2o, and o2 to V-cycle. For serial calculations, set the cycle type for Energy and User-dened Scalar-0 to W-cycle or F-cycle. For parallel calculations, select the F-cycle option for both. (b) Set the Max Cycles to 50. (c) Retain the default values for the rest of the parameters. 9. Dene the convergence criteria. Solve Monitors Residuals... In the Residual Monitors panel, set the Convergence Criterion for all equations to 1e-08. 10. Initialize the ow eld. Solve Initialize Initialize... In the Solution Initialization panel, retain the default values for all parameters and click the Init button. 11. In the Dene SOFC Model Parameters panel, turn on the FLUENT SOFC model by activating the Enable SOFC Model option, the Enable Surface Energy Source option, the Enable Species Source option, and the Disable CO Electrochemistry option. If the electrolyte resistivity changes as a function of temperature, then turn on the Enable Electrolyte Conductivity Submodel option. If there is CO in the fuel line, then you should turn o the Disable CO Electrochemistry option. 12. Run the simulation until all residuals are decreasing. 13. Turn on the Enable Volumetric Energy Source option and continue the simulation until convergence is achieved. 14. Save the case and data les.

4-6

c Fluent Inc. July 19, 2006

4.3 Solid Oxide Fuel Cell Module Set Up Procedure

15. Perform post-processing using standard quantities and by using the user-dened memory allocations. By default, the FLUENT SOFC model denes the following user-dened memory allocations: Table 4.3.1: User-Dened Memory Allocations UDM-0 Current Density (A/m2 ) UDM-1 Nernst voltage (Volts) UDM-2 Activation Overpotentail (Volts) UDM-3 Energy Source due to Ohmic Heating in Conduction Zones (W/m3 ) UDM-4 x Component of the Current Density (A/m2 ) UDM-5 y Component of the Current Density (A/m2 ) UDM-6 z Component of the Current Density (A/m2 ) UDM-7 Voltage jump UDM-8 Electrolyte Resistivity (Ohm-m) UDM-9 Eective Resistance UDM-10 Anode Activation Polarization UDM-11 Cathode Activation Polarization

Note that UDM-7 and UDM-8 are the linearized values that FLUENT uses to solve the potential eld and electrochemical coupling. They are necessary to the calculations but do not contain any real physical meaning. Note that UDM-10 and UDM-11 contain the disaggregated activation polarizations at the anode and cathode. The current density in UDM-0 (Equation 3.5-29) is a conservative ux of electricity (A/m2 ). Performing an area integral over the electrolyte surface will sum to the total current (A). That value is computed at the electrolyte faces during the electric eld solution. The values in UDM-4 - UDM-6 contain cell-centered values of the current density vector. These are not, and cannot be, conservative, so depending on the material conductivity and the geometric conguration, these can sometimes unavoidably produce values that do not match the UDM values. The same issues exist when interpolating velocity values to obtain the mass uxes.

c Fluent Inc. July 19, 2006

4-7

Using the Solid Oxide Fuel Cell Model

4.4

Setting the Parameters for the SOFC Model


You can specify the general settings for the SOFC model using the Dene SOFC Model Parameters panel. Dene Models SOFC SOFC Model Control Panel...

Figure 4.4.1: The Dene SOFC Model Parameters Panel

From this panel, you can set the various parameters for the SOFC model such as total system current, electrolyte thickness, electrolyte resistivity, etc. The Enable Electrolyte Conductivity Submodel option allows the ionic conductivity (or resistivity) of the electrolyte to change as a function of temperature. At the moment, there is one correlation that provides ionic conductivity (or resistivity) of the electrolyte as function of temperature. 0.3685 + 0.002838e resistivity = 100
10300 T

(4.4-1)

Note that this is valid only for temperatures ranging from 1073 K to 1373 K.

4-8

c Fluent Inc. July 19, 2006

4.4 Setting the Parameters for the SOFC Model

By turning o the Enable Surface Energy Source option, FLUENT excludes the heat addition due to electrochemistry and all the reversible processes. This option should be turned on at all times. The Enable Volumetric Energy Source option includes the ohmic heating throughout the electrically conducting zones. This option is typically turned on when the potential eld solution is close to convergence. The Disable CO Electrochemistry is enabled if there is carbon monoxide (CO) in the fuel line and if you want to include the CO in the electrochemistry. Since the calculations are very sensitive to large current uctuations early in the solution process, it is recommended to use 0.3 or 0.4 for the Current Underrelaxation Factor for a more eective solution. The leakage current is the total amount of current due to the leakage of oxidizer to the fuel side (through the electrolyte) and the electric current across the electrolyte due to any short circuit. You can specify a value for the leakage current under Leakage Current Density. The Converge to Specied System Voltage is used when the you want to specify a system voltage instead of system current as an input.

c Fluent Inc. July 19, 2006

4-9

Using the Solid Oxide Fuel Cell Model

4.5

Setting Up the Electrode-Electrolyte Interfaces


You can apply specic settings for both your anode and your cathode interfaces using the FLUENT SOFC model..

4.5.1

Setting Up the Anode Electrode-Electrolyte Interface

You can specify the anode settings for the SOFC model using the Anode Interface Setup panel. Dene Models SOFC Anode Interface Setup...

Figure 4.5.1: The Anode Interface Setup Panel

In the Anode Interface Setup panel, you select the surface that represents the anode electrode interface with the electrolyte from the Zone(s) list, turn on the Anode Interface option and click Apply. You can do this for as many zones as you need. When you are nished setting up the anode interface, click the OK button.

4-10

c Fluent Inc. July 19, 2006

4.5 Setting Up the Electrode-Electrolyte Interfaces

4.5.2

Setting Up the Cathode Electrode-Electrolyte Interface

You can specify the cathode settings for the SOFC model using the Cathode Interface Setup panel. Dene Models SOFC Cathode Interface Setup...

Figure 4.5.2: The Cathode Interface Setup Panel

In the Cathode Interface Setup panel, you select the surface that represents the cathode electrode interface with the electrolyte from the Zone(s) list, turn on the Cathode Interface option and click Apply. You can do this for as many zones as you need. When you are nished setting up the cathode interface, click the OK button.

c Fluent Inc. July 19, 2006

4-11

Using the Solid Oxide Fuel Cell Model

4.6

Setting Up the Electric Field Model Parameters


You can set up the details of the electric eld model by opening the Dene Electric Field Model Parameters panel. Dene Models SOFC Electric Field Input Panel...

Figure 4.6.1: The Dene Electric Field Model Parameters Panel

Here, you can designate conductive regions and specify their conductivity, assign boundaries to be contact surfaces and specify the contact resistances, as well as specify which surfaces are grounded and which surfaces exhibit a current. The Voltage Tap Surface is the surface that is grounded (i.e., V = 0). The Current Tap Surface is the surface that current is being drawn from (i.e., i is assigned a value).

4-12

c Fluent Inc. July 19, 2006

4.7 Setting Up the Tortuosity Parameters

4.7

Setting Up the Tortuosity Parameters


In fuel cell modeling, tortuosity is thought of as an eective diusive path length factor. You can specify tortuosity settings for the SOFC model using the Tortuosity Setup panel. Dene Models SOFC Tortuosity Setup...

Figure 4.7.1: The Tortuosity Setup Panel

In the Tortuosity Setup panel, you select an appropriate zone from the Zone(s) list, turn on the Enable Tortuosity option and click Apply. You can do this for as many zones as you need. When you are nished setting up the cathode interface, click the OK button. You can assign tortuosity values to any porous zones in your simulation. In a porous zone, the mass diusion coecient is reduced as follows due to the porosity eect: De = porosity D tortousity (4.7-1)

There typically are no standard means of measuring tortuosity as it either needs to be measured experimentally or tuned to match other experimental data. Tortuosity value may typically be in the range of 2 to 4, although you can use much higher values.

c Fluent Inc. July 19, 2006

4-13

Using the Solid Oxide Fuel Cell Model

4.8

Setting Up the Activation Parameters


You can specify activation settings for the SOFC model using the Dene Activation Parameters panel. Dene Models SOFC Activation Panel...

Figure 4.8.1: The Dene Activation Parameters Panel

In the Dene Activation Parameters panel, you can set the anode and cathode exchange current density, the anode and cathode mole fraction, the stoichiometric coecients, the Butler-Volmer transfer coecients, and the temperature-dependent exchange current density. You can specify a value for the exchange current density at the anode (the default value is 1000 Amps) using the Anode Exchange Current Density eld. Likewise, you can specify a value for the exchange current density at the cathode (the default value is 100 Amps) using the Cathode Exchange Current Density eld. You can also specify Mole Fraction Reference Values for the fuel cell reactants in the Dene Activation Parameters panel. By default, the reference value for H2 is 0.8, the reference value for O2 is 0.21, and the reference value for H2 O is 0.2. The Butler-Volmer transfer coecients can be set in the Dene Activation Parameters panel as well. These coecients are the a and c from Equation 3.5-41 for both the anode and the cathode reactions. i = i0 e
a n(0 )F RT

c n(0 )F RT

(4.8-1)

4-14

c Fluent Inc. July 19, 2006

4.8 Setting Up the Activation Parameters

Remember that has anodic and cathodic values at both the cathode and the anode. By default, the value of is set to 0.5 because of the nearly universal assumption that there is a symmetric balance between the forward and backward reactions. In most cases, these default values will be sucient. If you nd yourself changing the Butler-Volmer transfer coecients, or if you have some other rate-limiting reaction in your fuel cell simulation, you may also want to consider changing the exponents for the stoichiometric coecients for the fuel cell reactants. These exponents can be specied in the Dene Activation Parameters panel. By default, the exponent values for H2 , O2 , and H2 O are 0.5. The Enable Temperature Dependant I 0 option allows the exchange current density to change as a function of temperature in an exponential fashion i0 = AeBT where A and B are constants provided by the user. (4.8-2)

c Fluent Inc. July 19, 2006

4-15

Using the Solid Oxide Fuel Cell Model

4-16

c Fluent Inc. July 19, 2006

Bibliography
[1] J. Divisek A.A. Kulikovsky and A.A. Kornyshev. 146(11):39813991, 1999. J. Electrochemical Society,

[2] J. V. Cole and S. Mazumder. J. Electrochemical Soc., 150:A1510, 2003. [3] Brandon M. Eaton. One dimensional, Transient Model of Heat, Mass, and Charge Transfer in a Proton Exchange Membrane. M.S. Thesis, 2001. [4] Parsons Inc. EG&G Services and Science Applications International Corporation. Fuel Cell Handbook (5th Edition). 2000. [5] J.H. Nam and M. Karviany. Eective Diusivity and Water-Saturation Distribution in Single- and Two-layer PEMFC Diusion Medium. Int. J. Heat Mass Transfer, 2003. [6] J. S. Newman. Electrochemical Systems. Prentice Hall, Englewood Clis, New Jersey, 1973. [7] T. V. Nguyen. Electrochemical Soc. Proceedings, 99(14):222241. [8] C.Y. Wang Sukkee Um and K.S. Chen. J. Electrochemical Society, 147(12):44854493, 2000. [9] T.A. Zawodzinski T.E. Springer and S. Gottesfeld. J. Electrochemical Soc., 138.

c Fluent Inc. July 19, 2006

Bib-1