Open Access Journal

Journal of Power Technologies 93 (1) (2013) 27–36

journal homepage:papers.itc.pw.edu.pl

Electrochemical model for proton exchange membrane fuel cell systems

Abdellah Beicha∗,a , Radia Zaamoucheb
a
Jijel University, Faculty of Sciences and Technology
Department of Process Engineering 18000 Jijel, Algeria
b
University of Batna, Faculty of Sciences and Technology
Department of Mechanics 05000 Batna, Algeria

Abstract
This paper presents an electrochemical model for simulation and evaluation of the performance of a proton
exchange membrane (PEM) fuel cell. The results of the model are used to predict the efficiency and power
of the fuel cell as a function of operational parameters of the cell, like temperature, partial pressures and
membrane humidity. A one-dimensional mass transport model was also developed to investigate water trans-
port through the membrane electrode assembly (MEA). The model enables quantification of the water flux
corresponding to each of the three water transport mechanisms, such as diffusion, electro-osmotic drag and
convection. The influence of temperature on the fuel cell’s characteristics is more pronounced than the in-
fluence of partial pressures and membrane humidity. The effect of platinum loading on cell performance is
examined with Pt loadings of 0.18, 0.38 and 0.4 mg/cm2 .
Keywords: Fuel cell, Modeling, Simulation, Power, Membrane

1. Introduction are several studies regarding the influence of cer-

tain operating parameters on fuel cell characteris-
Non-polluting energy generation and other envi-
tics [8]. Many other models have been developed
ronmental issues have been driving an increasing de-
and reported in literature, but they did not focus on
mand over the last few years for new energy con-
the impact of individual operating parameters on fuel
version technologies. It is in this context that PEM
cell output characteristics.
fuel cell systems constitute a promising alternative
due their high efficiency and low impact on the envi-
ronment. A theoretical approach with mathematical
simulation is very effective for designing and analyz-
ing the performance of the PEM fuel cell.
Some work has been reported in the literature on This paper presents a simple model of a PEM fuel
steady-state fuel-cell modeling (e.g. [1–4], as well cell that can be used to analyze the impact of an in-
as dynamic modeling [5–7]). These studies are dividual fuel cell’s operating parameters on cell per-
mostly based on empirical equations and/or the elec- formance. Using this model, it is also possible to
trochemical reactions inside the fuel cell. Also there determine the activation loss parameters of the PEM
∗
Corresponding author
fuel cell. The model is well adapted for the PEM cell
Email addresses: [email protected] (Abdellah and incorporates the key physical and electrochemi-
Beicha∗, ), [email protected] (Radia Zaamouche) cal processes that happen in the cell.
 Journal of Power Technologies 93 (1) (2013) 27–36

2. Experimental Table 1: Parameters used in the simulation

2.1. Electrode preparation Parameters Value
Temperature, T 298 K
For the electrode preparation, a substrate made
Pressure, P 1 atm
from carbon cloth “A” with thickness 350 µm,
Electrode Area 25 cm2
platinum-supported carbon 10 wt% Pt/C, activated
Membrane thickness, lm 117 µm
carbon black, and Teflon (PTFE) 60 wt% were used.
Diffusion layer thickness, 350 µm
Two different platinum loadings were used in prepa-
ld
ration of the electrodes. One electrode was with plat-
PO2 2 atm
inum loading 0.18 mg Pt/cm2 ; while the second was
PH2 1 atm
with platinum loading 0.38 mg Pt/cm2 . The com-
Diffusion layer electronic 5 Ω−1 cm−1
position of the catalyst layer was 70 wt% Pt/C and
conductivity, σd
30 wt% Teflon (PTFE). Detailed electrode prepara-
Diffusion coefficient of 1.6×10−4 m2 /s
tion can be found in reference [9].
H2 into H2 O, DH2 , H2 O
2.2. Operation of a single cell in fuel cell test appa- Diffusion coefficient of 3.11×10−5 m2 /s
ratus O2 into H2 O, DO2 , H2 O
During the operation of PEM fuel cells, the fol- Diffusion coefficient of 3×10−10 m2 /s
lowing processes take place in the electrode: (i) the H2O into membrane,
reactant gases diffuse through the porous backing DH2 O, m
layer; (ii) at the gas-electrolyte interface, the gases Equivalent weight, EW 1100 g/mol
dissolve and then diffuse to the electrolyte-electrode Dry membrane density, 2020 Kg/m3
interface; (iii) electro-catalytic reaction inside the ρdry
catalyst layer precedes the gas adsorption at the elec-
trode surface; (iv) ionic transport occurs in the elec-
A schematic diagram of a PEM fuel cell is shown in
trolyte, but electronic transport takes place in the
Fig. 1.
electrode.
The hydrogen diffuses through the electrode until
Oxygen and hydrogen were passed through hu-
it reaches the catalytic layer of the anode where it
midifiers before being fed into the cell cells. Hy-
reacts to form protons and electrons, as shown below
drogen was fed into the anode at a flow rate of
in the reaction:
140 ml/min and 1 atm. Oxygen entered the fuel cell
through the cathode at a flow rate of 380 ml/min
and 2 atm. The electrons generated from the an- H2 → 2H + + 2e− (1)
ode were connected to a digital multimeter, with
an external variable resistance to measure the cur- The protons are transferred through the membrane
rent and voltage produced by the cell. Electric con- to the catalytic layer of the cathode. Electrons pass
ductivity was measured by using a resistivity meter through an external electric circuit to the cathode. On
(Loresta-GP MCP-T600). The specific resistance of the cathode side of the fuel cell, protons, electrons
the gas diffusion layer composed of 70 wt% Pt/C and and oxygen react in the following reaction:
30 wt% PTFE was measured to be 0.21 Ω·cm. The
parameters of the fuel cells used in the simulation are 1
2H + + 2e− + O2 → H2 O + heat (2)
listed in Table 1. 2
Hence, the overall chemical reaction of the PEM
3. PEM fuel cell modeling
fuel cell is
3.1. Basic fuel cell operation
A PEM fuel cell consists of a membrane through 1
H2 + O2 → H2 O + heat + energy (3)
which hydrogen ions diffuse from anode to cathode. 2

— 28 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Cell reversible voltage. The fuel cell’s output volt-

age is determined by the cell’s reversible voltage,
which arises from the potential difference produced
by the chemical reaction and several voltage losses
that take place inside the cell. The fuel cell’s re-
versible voltage is a function of temperature and par-
tial pressures of reactants and product, as is shown in
the following equation:

E = ∆G + ∆S
 
2F 2F
T − T re f +
h
i (6)
+
RT
1
2F
ln P H2 2
ln P O 2

where ∆G is the change in the Gibbs free energy;

F is the constant of Faraday; ∆S is the change of
the entropy; R is the universal constant of the gases;
while PH2 and PO2 are the partial pressures of hy-
drogen and oxygen, respectively. Variable T denotes
the cell operation temperature and T re f the reference
temperature. Using the standard pressure and tem-
Figure 1: Schematic of a PEM fuel cell
perature (SPT) values for ∆G, ∆S and T re f , Eq. (6)
can be simplified to [10]:
3.2. Model formulation
3.2.1. Electrochemical model E = 1.229 − 0.85 −3
h × 10
(T − 298.15) (7)
The output voltage of a single cell can be defined +4.31 × 10−5 T ln PH2 + 12 ln PO2

i
as the result of the following [5, 10]:
Activation voltage drop. Activation polarization is
related to the energy barrier that must be overcome
Vcell = E − Vactivation − Vohmic − Vconcentration (4) to initiate a chemical reaction between reactants. At
low current draw, the electron transfer rate is slow
In the equation above, E is the thermodynamic po- and a portion of the electrode voltage is lost in order
tential of the cell and it represents its reversible volt- to compensate for the lack of electro-catalytic activ-
age; Vactivation is the voltage drop due to the activa- ity. Expression for activation losses is given by:
tion of the anode and cathode, a measure of the volt-
age drop associated with the electrodes; Vohmic is the
!
RT i
ohmic voltage drop, a measure of the ohmic voltage Vactivation = ln (8)
2ψF i0
drop resulting from the resistances of the conduction
of protons through the solid electrolyte and the elec- ψ is electron transfer coefficient, and is unit less.
trons through its path; and Vconcentration represents the This value describes the proportion of the electrical
voltage drop resulting from the reduction in concen- energy applied that is harnessed in changing the rate
tration of the reactants gases or, alternatively, from of an electrochemical reaction. It is this value that
the transport of mass of oxygen and hydrogen; and differs from one material to another. i represents the
fuel cell’s output power density is given by: cell’s current density, whereas i0 is exchange current
density. i0 is the value on the Tafel plot when the
Pcell = Vcell i (5) current begins to move away from zero.
Due to higher anode exchange current density,
Each of the terms of Eq. (4) is discussed and mod- cathode activation losses are significantly higher so
eled separately in the subsections that follow. anode activation losses are negligible. The value of

— 29 —
 Journal of Power Technologies 93 (1) (2013) 27–36

cathode exchange current density also depends on

operating parameters, as is shown by:
!
2.46βF
i0c = 2Fkc exp (9)
RT
β and kc are, respectively, symmetry factor and
factor related to reaction speed.

Ohmic voltage drop. Ohmic voltage drop (or

“Ohmic polarization”) occurs due to resistive losses
in the cell. These resistive losses occur within the
electrolyte (ionic), in the electrodes (electronic and
ionic), and in the terminal connections in the cell
(electronic). Since the stack plates and electrolyte
obey Ohm’s law, the amount of voltage lost in order
to force conduction varies mostly linearly through- Figure 2: Mass transport through MEA
out this region. This is the working region of the fuel
cell. while the product accumulates at a greater rate than
it can be removed. Ultimately, these effects inhibit
Vohmic = i (rion + rel ) (10) further reaction altogether and the cell voltage drops
The following expression for the ionic resistance to zero. Expression for the fuel cell’s concentration
is used [10]: losses is given by:
!
RT i
  2  Vconcentration = − ln 1 − (14)
181.6 1 + 0.03i + 0.062 303
T
i2.5 2F iL
rion = lm h  i (11) iL represents limiting current density. This param-
(λ − 0.634 − 3i) exp 4.18 T −303
T eter describes the maximum current density that can
flow through the electrode.
where lm is the thickness of the membrane. The
parameter λ is influenced by the membrane prepara- 3.2.2. Mass transport
tion [11] and it can be related to relative humidity of
Molar flux. At anode channel:
the membrane Φ by the following expression [12]:
Molar flux of hydrogen,

i
λ = 0.043 + 17.8Φ − 39.8Φ2 + 36.0Φ3 (12) N H2 = (15)
2F
The electronic resistance can be written as: At membrane:
Molar flux of H + ,
2ld
rel = (13)
σd i
NH + = (16)
where ld is diffusion layer thickness and σd is dif- F
fusion layer electronic conductivity. At cathode channel:
Concentration or mass transport voltage drop. Mass Molar flux of oxygen,
transport or concentration polarization results when i
the electrode reactions are hindered by mass trans- NO2 = − (17)
4F
fer effects. In this region, the reactants become con-
sumed at greater rates than they can be supplied, At interface membrane-anode:

— 30 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Membrane transport.
NHa 2 O = NHm2 O (18) ρdry λ
cH2 O = (28)
EW
At interface membrane-cathode:
P
Φ = ya,c
H2 O (29)
P sat
i
NHc 2 O = NHm2 O + NHprod
2O
and NHprod
2O
= (19) The vapor pressure is given by
2F
!
P sat 5096.23
Diffusion of species. At anode channel: = exp 13.67 − (30)
P T
Water transport in the membrane is given by:
yH2 + yaH2 O = 1 (20)
0.114λi dcH2 O
At cathode channel: NHm2 O = − DH2 O,m (31)
F dz
λ (z) − λa 1 − exp (km z)
yO2 + ycH2 O = 1 (21) = (32)
λc − λa 1 − exp (km lm )
The Stefan-Maxwell equation in one dimension where
can be written as:
0.114 × EW × i
At anode channel: km = (33)
ρdry DH2 O,m F
From Eqs. (28), (31) and (32), one can get NHm2 O
dyH2 1 h   i as:
= yH2 NHa 2 O + NH2 − NH2 (22)
dz cDH2 ,H2 O
λc − λa
" #
0.114i
where c = RT
P NHm2 O = λa + (34)
F 1 − exp (km lm )
Eq. (22) can be solved by using the following
boundary conditions: 4. Results and discussions

P sat Concentration polarization results from restric-

H2 O = RHa
yae (23) tions on the transport of the fuel gases to the reaction
p
sites. This usually occurs at high current, because
the forming of product water and excess humidifi-
yeH2 = 1 − yae
H2 O (24) cation blocks the reaction sites. This polarization is
also affected by the physical restriction of the trans-
At cathode channel:
fer of oxygen to the reaction sites on the cathode side
of the fuel cell. Concentration polarization can be
dyO2 1 h   i reduced by using thinner electrodes, which shortens
= yO2 NHc 2 O + NO2 − NO2 (25) the path of the gas to the sites [13, 14]. Jordan et
dz cDO2 ,H2 O
al. [14] observed a dramatic change in slope of the
That can be solved by using the following bound- voltage versus current density plot using air oxidant.
ary conditions: This change, indicative of a diffusion-limited reac-
tion, was not so apparent when pure oxygen is used
P sat as the oxidant. The same behavior was also observed
H2 O = RHc
yce (26) by other researchers [15]. This is consistent with the
p
experimental data showed in Fig. 2. Hence, in pa-
rameter estimation the authors neglected the effect
yeO2 = 1 − yce
H2 O (27) of concentration polarization drop.

— 31 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Pt loading 0.38 mg/cm2 and the commercial on E-

TECK with Pt loading 0.4 mg/cm2 . The cell voltage
in E-TECK dropped slightly faster than the former
electrode. Also, the value of i0 is decreased from
4.95×10−8 A/cm2 (prepared electrode with Pt load-
ing 0.38 mg/cm2 ) to 2.86×10−8 A/cm2 (E-TECK
electrode with Pt loading 0.4 mg/cm2 ). This may
be attributed to the fabrication process in the prepa-
ration of electrode with Pt loading 0.38 mg/cm2 ,
which creates a better particle distribution of electro-
catalyst. The placing of platinum in the catalyst
layer can be improved through the spraying tech-
nique [9]. The estimated values of i0 are lower than
those reported by Amphlett et al. [16], which are
107.6×10−8 A/cm2 at 298 K. From these results, it
can be concluded that Pt loading is not alone in af-
fecting the value of i0 .
Figure 3: Fuel cell voltage as a function of the cell’s current Typically, the value of lies in a very narrow range
density at T = 298 K, data: N Pt loading = 0.18 mg/cm-2, of about 0.1 to 0.5 [17]. The exchange current den-
 Pt loading = 0.38 mg/cm2 and • E-TECK electrode with sity constant varies over a wide range, and thus has
Pt loading = 0.4 mg/cm2 [9]; model: (dashed line) Pt load- a dramatic effect on the performance of fuel cells at
ing = 0.18 mg/cm2 , (solid line) Pt loading = 0.38 mg/cm2 and
low current densities. Hence, it is vital to design fuel
(dotted line) E-TECK electrode with Pt loading = 0.4 mg/cm2
cells with high exchange current densities.
Table 2: Variation of the parameters ψ and i0 with platinum
loading at T = 298K 4.2. Effect of temperature on the performance of the
Parameters Fabricated Fabricated E-TECK
PEM fuel cell
electrode 0.18 electrode 0.38 electrode 0.4 Fuel cell voltage as a function of the cell’s current
mg Pt/cm2 mg Pt/cm2 mg Pt/cm2 density is shown in Fig. 4. In Fig. 5 the cell’s power
ψ 0.13 0.28 0.3 density is shown as a function of current density for
i0 , A/cm2 2.94×10−9 4.95×10−8 2.86×10−8 two different temperatures. It can be seen from these
figures that the fuel cell’s efficiency is low and that
4.1. Determination of the activation loss parameters a significant part of the theoretical output voltage is
lost because of various losses inside the cell. Also
The experimental data of the cell voltage versus
noticeable is that an increase in fuel cell operating
current density (Fig. 3) for the two fabricated elec-
temperature will cause an increase in the cell’s output
trodes and the commercial electrode E-TECK with
voltage and power. An increase in temperature from
Pt loadings 0.18, 0.38 and 0.4 mg/cm2 , respectively,
298 K to 353 K gives a 28% increase in fuel cell volt-
are fitted to the present model using a nonlinear least
age (Fig. 4). The reason for this is that higher tem-
squares method. The characteristics of fabricated
peratures improve mass transfer within fuel cells and
and commercial electrodes are listed in Table 1. The
results in a net decrease in cell resistance, thereby
activation loss parameters, and i0 , are determined and
improving the reaction rate.
listed in Table 2. Both parameters depend on Pt load-
ing. The increase of Pt loading will cause an increase
4.3. Effect of the partial pressures
of and i0 . This can be attributed to the increase in ac-
tive sites for hydrogen adsorption. Change in output power related to increase in par-
An inconsistency is observed between the po- tial pressures is shown in Fig. 6. It can be seen that
larization curves for the prepared electrode with the power increase due to reactant pressure increase

— 32 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Figure 5: Fuel cell power density as a function of the

Figure 4: Fuel cell voltage as a function of the cell’s current cell’s current density for various temperatures at Pt load-
density for various temperatures at Pt loading = 0.38 mg/cm2 ; ing = 0.38 mg/cm2 ; model: (solid line) 298 K and (dashed line)
model: (dotted line) 298 K, (solid line) 313 K and (dashed line) 353 K
353 K
ter by convection from the anode towards the cath-
is 1.1%, which is smaller than the change caused by ode.
temperature variations. Description of the mass transport in the electrodes,
and in particular in the membrane, is necessary for
the modeling of the heat and electron transfer in the
4.4. Membrane humidity
fuel cell. Indeed, protonic resistance is strongly de-
Sufficient gas stream humidification is essential pendent on the hydration state of the membrane and
for PEM fuel cell operation since water molecules this is responsible for a considerable release of heat.
move with the hydrogen ions during the ion ex- The observed variation of concentrations in the gas
change reaction. Insufficient humidification water mixture on the electrodes shows their weak influence
dehydrates the membrane and can lead to cracks or on the transport of water. The variations of the molar
holes in the membrane. A decrease in membrane hu- fraction of water in the electrodes (anodic and ca-
midity from 100% to 30% decreases fuel cell voltage thodic) are given in Figures 9 and 10. These vari-
by 1.8% (Fig. 7). Reducing membrane humidity can ations are observed to be more important in the ca-
result in slightly slower electrode kinetics, including thodic electrode than in the anodic one. It should be
electrode reaction and mass diffusion rates. noted that these variations are slightly dependent on
Analysis of the profiles of water content in the the current density.
membrane can provide some recommendations for These simulations show that attention must be
the humidification of gases. The profiles of water paid to temperature, whereas reactant pressures and
content in the membrane are given in Fig. 8. This membrane humidity have a less significant influence
figure shows the influence of transport by electro- on output voltage and power.
osmosis on the humidification of the membrane.
With low current densities, the profile of water con- 5. Conclusion
tent in the membrane approaches the linear profile
imposed by diffusive flow. In contrast, for high cur- Using the present PEM fuel cell model, the authors
rent densities, electro-osmotic flow tends to drag wa- analyzed the influence of fuel cell operating param-

— 33 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Figure 6: Fuel cell voltage as a function of the cell’s cur- Figure 7: Fuel cell voltage as a function of the cell’s cur-
rent density for various fuel cell reactants partial pressure at rent density for various membrane humidity at T = 298 K and
T = 298 K and Pt loading = 0.38 mg/cm2 ; model: (dashed Pt loading = 0.38 mg/cm2 ; model: (solid line) Φ = 100%,
line) PH2 = 0.995×105 Pa and PO2 = 0.606×105 Pa, (solid line) (dashed line) 50% and (dotted line) Φ = 30%
PH2 = 1.01×105 Pa and PO2 = 1.01×105 Pa

[2] A. Rowe, X. Li, Mathematical modeling of pro-

eters (temperature, partial pressures and membrane ton exchange membrane fuel cells, Journal of Power
humidity) on fuel cell performance. Temperature Sources 102 (1–2) (2001) 82–96. doi:10.1016/S0378-
was found to have a significant influence on output 7753(01)00798-4.
voltage and power. The influence of partial pressures [3] D. Bevers, M. Wohr, K. Yasuda, K. Oguro, Simula-
and membrane humidity is less significant. tion of a polymer electrolyte fuel cell electrode, Jour-
nal of Applied Electrochemistry 27 (1997) 1254–1264.
Both the electron transfer coefficient and exchange doi:10.1023/A:1018488021355.
current density are platinum loading dependent. The [4] G. Maggio, V. Recupero, L. Pino, Modeling poly-
exchange current density constant varies signifi- mer electrolyte fuel cells: an innovative approach,
cantly with platinum loading and thus has a dramatic Journal of Power Sources 101 (2) (2001) 275–286.
doi:10.1016/S0378-7753(01)00758-3.
effect on fuel cell performance at low current den-
[5] J. Amphlett, R. Mann, B. Peppley, P. Roberge, A. Ro-
sities. This parameter is crucial for the purpose of drigues, A model predicting transient responses of pro-
PEM fuel cell design. ton exchange membrane fuel cells, Journal of Power
Sources 61 (1–2) (1996) 183–188. doi:10.1016/S0378-
Acknowledgments 7753(96)02360-9.
[6] J. Hamelin, K. Agbossou, A. Laperrière, F. Laurencelle,
The authors are grateful for financial support from T. Bose, Dynamic behavior of a pem fuel cell stack for
the ministry for higher education and scientific re- stationary applications, International Journal of Hydro-
search through national programs of research. gen Energy 26 (6) (2001) 625 – 629. doi:10.1016/S0360-
3199(00)00121-X.
[7] H. P. van Bussel, F. G. Koene, R. K. Mallant, Dy-
References namic model of solid polymer fuel cell water manage-
ment, Journal of Power Sources 71 (1–2) (1998) 218–222.
[1] R. Cownden, M. Nahon, M. Rosen, Modelling and doi:10.1016/S0378-7753(97)02744-4.
analysis of a solid polymer fuel cell system for trans- [8] Z. Zhang, X. Huang, J. Jiang, B. Wu, An improved
portation applications, International Journal of Hydro- dynamic model considering effects of temperature and
gen Energy 26 (6) (2001) 615–623. doi:10.1016/S0360- equivalent internal resistance for pem fuel cell power
3199(00)00126-9. modules, Journal of Power Sources 161 (2) (2006) 1062–

— 34 —
 Journal of Power Technologies 93 (1) (2013) 27–36

Figure 10: Variation of mole fraction of water at cathode for

i = 0.01; 0.04; 0.1 A/cm2

Figure 8: Variation of membrane humidity for i = 0.01; 0.1; [12] T. Springer, T. Zawodzinski, S. Gottesfeld, Polymer elec-
0.2 A/cm2 trolyte fuel cell model, Journal of Electrochemical Soci-
ety 138 (8) (1991) 2334–2342.
[13] J. H. Hirschenhofer, Fuel cell status 1994, IEEE AES Sys-
1068. doi:10.1016/j.jpowsour.2006.05.030. tems Magazine (1994) 10–15.
[9] R. Chebbi, A. Beicha, W. Daud, R. Zaamouche, Surface [14] L. Jordan, A. Shukla, T. Behrsing, N. Avery, B. Mud-
analysis for catalyst layer (pt/ptfe/c) and diffusion layer dle, M. Forsyth, Diffusion layer parameters influenc-
(ptfe/c) for proton exchange membrane fuel cells systems ing optimal fuel cell performance, Journal of Power
(pemfcs), Applied Surface Science 255 (12) (2009) 6367– Sources 86 (1–2) (2000) 250–254. doi:10.1016/S0378-
6371. doi:10.1016/j.apsusc.2009.02.017. 7753(99)00489-9.
[10] R. F. Mann, J. C. Amphlett, M. A. Hooper, H. M. Jensen, [15] J. Song, S. Cha, W. Lee, Optimal composition of poly-
B. A. Peppley, P. R. Roberge, Development and applica- mer electrolyte fuel cell electrodes determined by the
tion of a generalised steady-state electrochemical model ac impedance method, Journal of Power Sources 94 (1)
for a pem fuel cell, Journal of Power Sources 86 (1–2) (2001) 78 – 84. doi:10.1016/S0378-7753(00)00629-7.
(2000) 173–180. doi:10.1016/S0378-7753(99)00484-X. [16] J. Laminie, A. Dicks, Fuel cell systems explained, 2nd
[11] M. G. Nguyen, R. White, A water and heat management Edition, Wiley & Sons, England, 2005.
model for proton-exchange-membrane fuel cells, Journal [17] J. C. Amphlett, R. M. Baurnet, R. F. Mann, Perfor-
of Electrochemical Society 140 (8) (1993) 2178–2186. mance modeling of the ballard mark iv solid polymer
electrolyte fuel cell, Journal of Electrochemical Society
142 (1) (1995) 1–15.

Nomenclature
β symmetry factor
λ water content
λa water content
λc water content
Φ relative humidity,
ψ electron transfer coefficient
Figure 9: Variation of mole fraction of water at anode for
i = 0.01; 0.04; 0.1 A/cm2 ρdry dry membrane density, kg/m3

— 35 —
 Journal of Power Technologies 93 (1) (2013) 27–36

σd diffusion layer electronic conductivity, rion ionic resistance, Ω·cm2

A/(V·cm)
RHa relative humidity of anode
DH2 ,H2 O diffusion coefficient of H2 in H2 O, m2 /s
RHc relative humidity of cathode
DH2 O,m diffusion coefficient of H2 O in membrane,
m2 /s T temperature, K

DO2 ,H2 O diffusion coefficient of O2 in H2 O, m2 /s V voltage, V

E fuel cell’s reversible voltage, V Vactivation activation voltage drop, V

EW Equivalent weight, g/mol Vconcentration concentration voltage drop, V

F Faraday constant, C Vohmic ohmic voltage drop, V

I current density, A/cm2 yae

H2 O mole fraction of H2 O at entrance of anode

i0c cathode exchange current density, A/cm2 yaH2 O mole fraction of H2 O at anode

i0 exchange current density, A/cm2 yce

H2 O mole fraction of H2 O at entrance of cathode

iL limiting current density, A/cm2 ycH2 O mole fraction of H2 O at cathode

kc factor related to reaction speed, A/(cm2 ·C) y H2 mole fraction of H2

ld diffusion layer thickness, cm yeH2 mole fraction of H2 at entrance of anode

lm membrane thickness, cm yO2 mole fraction of O2

NH + molar flux of H + , mol/(m2 ·s) yeO2 mole fraction of O2 at entrance of cathode

NHa 2 O molar flux of H2 O at membrane-anode inter- z position, m

face, mol/(m2 ·s)

NHc 2 O molar flux of H2 O at membrane-cathode inter-

face, mol/(m2 ·s)

NHm2 O molar flux of H2 O in membrane, mol/(m2 ·s)

NHprod
2O
molar flux of produced water, mol/(m2 ·s)

NH2 molar flux of H2 , mol/(m2 ·s)

NO2 molar flux of O2 , mol/(m2 ·s)

Pcell fuel cell’s power density, W/cm2

P H2 partial pressure of hydrogen, atm

PO2 partial pressure of oxygen, atm

R universal gas constant, 8.314 J/(K·mol)

rel electronic resistance, Ω·cm2

— 36 —