THE POLYMER ELECTROLYTE MEMBRANE FUEL CELL AS ELECTRIC

ENERGY SOURCE, STEADY STATE AND DYNAMIC BEHAVIOR

R. A. Costa1, J. R. Camacho2, S. C. Guimarães Jr.2, and C. H. Salerno2

1
FEB - Fundação Educacional de Barretos,
School of Electrical Engineering,
Av. Prof. Roberto Frade Monte, 389 – Aeroporto,
14783.226 - Barretos – SP – Brazil.
2
Universidade Federal de Uberlândia,
School of Electrical Engineering, Rural Electricity and Alternative Sources Lab.,
Av. João N. de Ávila, 2121 - 38400.902, Uberlândia, MG, Brazil.
e-mail: [email protected]

with the anode and, simultaneously, a flux of an oxidant

(in general the oxygen in the air) also fed through
Abstract - The main objective of this work is to give
information on the behavior of three small PEMFC (Polymer
channels, make the contact with the cathode. [1] The
Electrolyte Membrane Fuel Cell/Proton Exchange Membrane baking layers are responsible for the gas diffusion
Fuel Cell) prototypes under static and dynamic load conditions. process in the electrodes, with a thickness between 300 to
This is a fuel cell that holds promise in the use for energy in 400 µm, they are made of an electrical conductor
automotive and household applications. A computational model material, like porous carbon paper, a carbon based fabric
was developed to simulate the static and dynamic performance with a Teflon® layer that prevents water adherence and
of this particular type of fuel cell. This model is based on make possible the fast diffusion of gases. [2]
electrochemical equations and takes into consideration the
advantages and disadvantages of the device in order to generate
power. The model takes into consideration the operating and The process that takes place is a natural reaction,
design parameters of the materials, with the results being however due to its low speed it is necessary the use of a
compared with practical experiments. catalyst around the electrodes for the reaction to be
This research gives the possibility to infer, from steady state efficient, the element used as catalyst at the present is the
and dynamic studies, on the design of PEMFC’s of different platinum. Other two important elements in the fuel cell
sizes and also to develop a further research on the need for are, firstly the separating plates, to give mechanical
control such as of hydrogen and oxygen pressure and flow. The rigidity and to direct the flow of gas through the cell,
study of sizing of the fuel cell is also an invaluable asset due to secondly the refrigeration cells that have the function to
the low cost of the simulation. extract the heat of the exothermal reaction, the ratio used
for interspersing the refrigeration cells among fuel cells
Keywords – alternative energy, distributed generation, are used in a proportion that varies from 1:1 to 1:5.
dynamic studies, proton exchange, polymer electrolyte, steady
state.
A.. PEM fuel cells
1. Introduction The PEMFC (Polymer Electrolyte Membrane Fuel Cell /
Proton Exchange Membrane Fuel Cell) are fuel cells
Fuel cells are similar to batteries; the difference is that where the electrolyte is made of an organic polymer that
they work continuously producing the current through an has the characteristic of a good proton carrier when in
electrochemical natural reaction taking hydrogen based presence of a water solution.
fuel in the anode, oxygen in the cathode and producing
electricity, heat and vaporized water. Its electrochemical The only liquid existent in this kind of cell is water and,
operating principle is the opposite of the electrolysis. therefore, the corrosion problem is minimum. The water
management in PEMFC cells is extremely important. The
The basic physical structure of a fuel cell consists of an fact that the membrane needs to be always in water
electrolyte and polymeric membrane (with typical solution, will limit the temperature of the cell operation
thickness between 12 and 210 µm) that separates the to the water vaporization temperature, avoiding
porous anode and cathode from each other; they have membrane dryness. Not allowing energy co-generation
thickness between 5 and 15µm. The porous and thin from the thermal energy dissipated. The efficiency of the
electrodes, make it possible to assembly it together with whole process is around 50% to 60%.
the polymeric membrane, making an integrate set called
MEA - Membrane Electrode Assembly. In a typical fuel The main initial difficulty to the use of PEMFC cells was
cell, a flux of fuel rich in hydrogen (liquid or gaseous) is the high cost of the platinum necessary as a catalyst in
fed continuously through channels that keep it in contact
the electrodes. Initially the PEMFC cells used 4 mg of Among the main losses to be considered are: the
platinum per cm2 of membrane area that made expensive activation (Vatv) and concentration (Vcon) effects in the
the use of this kind of cell. Nowadays, with the anode and in the cathode, and the voltage drop due to the
technological development in the materials for the resistance of the cell (Vohm), each of them predominantly
electrodes it is possible the use of 0.15 mg\cm2 of occurs during an operational load (value of electrical
platinum, with the same electrical efficiency, making the current) of the cell, producing a typical behavior.
fuel cell less expensive and closer to the economical
viability.[8] The cell operational voltage (Vop) can be computed
through the open circuit voltage (E0), and internal voltage
The PEMFC’s have a robust design and are relatively
drops. Given by:
simple to build. The main characteristics are the low
operating temperature (≈80oC) that allows a fast response Vop = E 0 − (Vatv + Vohm + Vcon ) [V] (1)
when being turned on and off. To these properties can be
added the advantages of the direct utilization of air in the where Vatv are the activation losses, Vohm are the resistive
cathode and zero emission of COx and NOx, making this losses and Vcon are the concentration losses.
type of cell the most promising alternative for the
substitution of the combustion engines used in 1) Open circuit voltage or ideal voltage
transportation vehicles, and for the electrical energy
production in small and medium size stationary units The maximum value of produced voltage (open circuit
below 250kW.
voltage, reversible voltage or ideal voltage) in a fuel cell
is an specific amount from the global reaction that occurs
The number of fuel cells in the array determines the value
in the cell. In the case of a PEMFC cell, this can be
of the operation voltage in a fuel cell stack. The voltage
is also defined by the internal losses, materials, calculated by the energy difference between the initial
assembling methods, the current intensity imposed by the states of reagents (H2 + ½ O2) and the final result (H2O).
load and the whole operational process. Therefore, the The maximum voltage produced by the cell for the
electric power generated by the stack is defined by a reaction occurred (H2 + ½ O2 → H2O) will be calculated
series of assembling and operational parameters. through the Nernst expression:

2. Mathematical Modeling of the PEM Fuel E Nernst =

1
2F
{ ( ) ( ) } [V] (2)
∆ G + ∆ S. ( T0 + Tref ) + R.Top . ⎡ ln PH 2 + 1 2.ln PO 2 ⎤
⎣ ⎦

Cell Where F is the so-called Faraday constant (96.487 C),

∆G is the variation of Gibbs free energy (-237.165 J), ∆H
A mathematical model that presents correct answers for
the fuel cell is of paramount importance, since it allows and ∆S are respectively the variation of enthalpy (-
the development of electronic systems for feeding and to 285.823 J) and entropy (-163.2 J/K) in the reaction, R is
control for fuel cells without the need of a great number the gases universal constant (8,314 J/oK.mol), and Tref
of experimental prototypes. A fuel cell works basically as (298.15 oK) and Top are respectively the reference and
a current source. The voltage response for the cell will be operational temperatures in oK.
studied from the current variation required by the load.
The main operational parameters are: the operational By the application of constants to the Nernst maximum
electrical current (Iop), temperature (Top), and the voltage equation can be simplified to:
transportation of gases. With the main parameters
involved in the gas feeding system being: ( ) ( ) ( )
E0 = 1.229 − 0.85.10−3. Top − 298.15 − 4.31.10−3.Top . ⎡ln PH2 + 1 2.ln PO2 ⎤ [V] (3)
⎣ ⎦

PH 2 and PO2 – Partial pressure feeding for Hydrogen with this equation, it is made the relationship of the ideal
voltage with the main operational variables (operating
and Oxygen (in atm); temperature and the feeding pressures of hydrogen and
φ H 2 and φO2 – Relative humidity of Hydrogen and oxygen), the remaining variables are not significant to the
Oxygen (in percent); maximum operating voltage.
mH 2 , mO2 and m Air – mass flow of Hydrogen, Oxygen
2) Activation effects
and Air (dimensionless);
The activation voltage drop (Vatv) is dominant in the
λ H 2 , λO2 and λ Air – Stoichiometries of Hydrogen, initial stage of chemical reaction, that is, when the
Oxygen and Air (dimensionless). current density in the cell is low. Those are effects that
change with the increase in current density. The
activation voltage drop can be defined by the Tafel
A. PEM fuel cell with static load equation:

The electrical behavior of a PEM fuel cell is obtained ⎛ jop ⎞

R Top
Vatv = ln ⎜ ⎟ [V]. (4)
from the maximum voltage that can be produced by one α n F ⎝ j0 ⎠
cell (E0) and by the existing energy losses during normal
operation. Where R is the gases universal constant (8,314 J/K.mol),
n is the ratio between the number of moles involved in
this reaction for each mol of H2 (a typical value is 2), jop electrode/membrane (mol/cm3).
is the cell operation current density, j0 is the exchange
current density in the electrode (typical values are around Working mathematically in equations (3) and (7)
0.5 A/cm2 ), and α is the electronic transference current together, we can have for the cathode the following value
(around 0.4 for the cathode and 1.0 for the anode) that for the activation voltage[3]:
represents the proportion of energy available at the
electrode that is used in the electrochemical reaction.[13] ( )
Vatvc = ξ1 + ξ 2 .Top + ξ3 .Top .ln CO* 2 + ξ 4 .Top .ln ( i ) (8)
3) Ohmic effects where:
∆FEc
ξ1 = (9)
The cell ohmic resistance occurs due to two distinct nc .α c
factors: the resistance that the electrolyte (polymeric
membrane) offers to the ionic flux and the resistance ξ2 =
R
n c .F .α c
(
ln 4.F .k 0 + ) R
n c.F . α c
ln ( A ) +
R
n c.F
ln C*H2 (10) ( )
offered by the electrodes (Rm), and the resistance of gas
spreaders and separating plates to the flux of electrons CH* 2 - hydrogen concentration in the interface
(Rcj). We can consider that the ohmic losses (Vohm) in the
electrode/membrane (mol/cm3),
cell as:
(
Vohm = R m + R cj ⋅ i op ) [V] (5)
ξ3 =
R (1 − α c )
(11)
nc .F .α c
4) Concentration effects and
R
On the other hand, every time that one of the reagents is ξ4 = (12)
n c ⋅ F ⋅ αc
consumed by the electrochemical reaction in the cell
electrodes, the cell doesn’t have the capacity to keep the
2) Anode activation losses
initial concentration in the neighborhood of the electrode.
This reagent concentration effect gradient, close to the
With a more detailed equation (7), taking in
electrode produces a decrease in the cell electrical
consideration the effect of operational variables in the
voltage named concentration effect. The concentration
value of i0 at the anode as[5]:
voltage drop (Vcon) can be calculated by the equation:
( −∆FEa )
i0, a = na .F . A.k 0 .CH* 2 .e
R .Top
(13) [A]
RTop⎛ iop ⎞
Vcon = ln ⎜ 1 − ⎟ [V] (6) where: na –number of (6)moles of reagent at the anode;
nF ⎝ iL ⎠
∆FEa – reaction activation energy at the anode (J/mol); A
Where iop is the operational cell current and iL is the - anode area / cell useful area (cm2); αa -
electrode current limit. They both have different values to load transference coefficient at the anode (non
the cathode (iL,c) and anode (iL,a). The iL values will dimensional); k0 - reaction degree intrinsic
*
depend on the assembling conditions, as the cell area, constant(cm/s); C H 2 - hydrogen concentration in the
shape of channels that drive the gas internally in the cell, interface electrode membrane (mol/cm3);
gas spreader characteristics, type of membrane, and
etc.[4][6] Working mathematically in equations (4) and (6)
together, we can have for the cathode the following value
B. Activation losses for the activation voltage[4]:

1) Cathode activation losses ( )

Vatv ,c = ξ5 + ξ 6 .Top .ln CH* 2 + ξ 7 .Top .ln ( i ) [V] (14)
Observing equation (4) in detail, the consideration on the
where:
effect of operating variables in the value of i0 in the
∆FE a
cathode can be taken as[5]: ξ5 = (15)
⎛ −∆FEc ⎞ na .F .α a
⎜ ⎟
(
i0 c = n c .F . A.k 0. C *proton )
1− αc
(
. C *H2 O ) . (C )
αc *
O2
1− αc
.e
⎜ R.Top ⎟
⎝ ⎠ (7)
ξ6 =
(
− R.ln na .F . A.k 0 )
(16)
(7) na . F.α a
for the cathode, nc is the reagent number of moles, ∆FEc R
is the reaction activation energy (J/mol), A is the net area ξ7 = (17)
n a .F .α a
/ useful area of the cell (cm2), αc is the load transference
coefficient (not dimensional), k0 is the intrinsic constant
As previously stated, the activation voltage at the anode
reaction degree (cm/s), C*proton is the total concentration
*
will have lower values compared with the activation
of protons in the membrane (mol/cm3), CH 2O is the water voltage at the cathode, and can be neglected in the
* operating behavior of the PEM Fuel Cell fed with pure
concentration in the membrane (mol/cm3) and CO2 is the hydrogen.
Oxygen concentration in the interface
C. Resistive losses 1) Concentration losses at the cathod

To illustrate the empirics to obtain fuel cell operating The value for the current limit at the cathode (IL,c) will be
parameters, the first part of the resistive losses (Rohm) is adopted[13] through the following equation:
given by the polymeric membrane and the second part by
the electrodes, spreaders and other series components. R.Top . A.γ .Cion, m
iL, c = [A] (20)
nc .F .δ c
1) Polymeric Membrane Resistive Losses
where: nc - number of moles of the reagent in the
The resistance in Ohm for this membrane (Rm) which cathode; γ - membrane ionic conductance
represents the resistance for transference of protons (Siemens.cm2/mol); A – cathode area / cell useful area
through the solid membrane, and can be computed by the (cm2); δc – Nernst diffusion layer thickness at the cathode
two following equations: (cm); F- Faraday constant (96487 C/mol); R - gases
universal constant (8,314 J/K.mol); Top – cell operation
ρ m .l temperature (oK); Cion,m – membrane ionic concentration
Rm = [Ω ] (18)
A (mol/cm3).
Where ρm is the resistivity of the membrane (Ω.cm), l is 2) Concentration losses at the anode
the membrane thickness (cm) and A is the membrane
area, the electrode area. The resistivity values for the They will be simulated as in equation (20), and the value
Nafion®[7] membranes (produced by DuPont) can be for the current limit at the anode (IL,a) will be adopted
computed through the following equation: through the following equation:
⎡ ⎛ Top ⎞ ⎛ iop ⎞ ⎤
2 2.5
⎛ iop ⎞
181.6 ⎢1 + 0.03⎜ ⎟ + 0.0062.⎜ R.Top . A.γ .Cion, m
⎟ .⎜ ⎟ ⎥ iL, a = [A]
⎣⎢ ⎝ A⎠ ⎝ 303 ⎠ ⎝ A ⎠ ⎥⎦ na .F .δ a
(21)
ρm = [ Ω.cm ] [19]
⎡ ⎡
⎟⎥ ⎤
⎛ T −303⎞ ⎤
⎢4.18.⎜ op
⎢ λ − 0.634 − 3.⎛ iop ⎞ .e⎢⎣ ⎜⎝ Top ⎟⎠ ⎥⎦ ⎥ where: na - number of moles of the reagent in the anode;
⎢ ⎜ ⎟ ⎥
⎢⎣ ⎝A⎠ ⎥⎦
δa - Nernst diffusion layer thickness at the anode(cm).

For the Nafion® membrane, the typical thickness in each E. PEMFC cell in a dynamic load regimen
case is indicated in Table I.[8]
When a variation occurs in the load fed by a fuel cell, this
Table I – Typical thickness of Nafion membranes is the cause for a variation in the cell operating voltage.
Membrane Nominal Thickness Thickness Thickness
However, a transient regimen exists due to the
thickness when dry when wet when MEA capacitance generated in the polymeric membrane due to
(mm) (mm) (mm) (mm) an internal physical process called “double layer effect”.
Nafion 117 0,178 0,183 ± 0,208 ± 0,148 ± Taking into consideration the influence of this
0,003 0,005 0,002
Nafion 115 0,127 0,141 ± 0,161 ± 0,100 ± capacitance, the equivalent circuit for the PEMFC fuel
0,003 0,003 0,002 cell can be depicted in a simplified form, for constant
Nafion 113 0,089 0,091 ± 0,111 ± 0,075 ± pressure, as can be seen in Figure 1.
0,002 0,002 0,003
Nafion 112 0,051 0,050 ± 0,058 ± 0,040 ±
0,002 0,003 0,002

Therefore, λ is the ratio of the number of water moles for

each sulphonic group in the membrane. The calue of λ
indicates the membrane relative humidity, being
regulated by the feeding gases relative humidity, their
stoichiometries and assembling characteristics involved
in the efficiency of the gas transport and membrane Fig. 1. Simplified electrical circuit for the PEM fuel cell.
humidity maintenance. The literature indicates a value of
approximately 14 for λ corresponding to a membrane in Where Rohm, Ratv, and Rcon are respectively the
humidified air at 100%, and approximately to 22 the representation for the ohmic, the activation and the
value corresponding to a water immersed membrane[3]. concentration resistance; with Cdc corresponding to the
membrane capacitance due to the double layer effect. The
For the simulation, the value for λ will be calculated double layer effect arises when two different electrically
through an empirical equation with assembling charged materials are put in contact. Alternatively two
parameters being influenced by operating variables, things can happen, a concentration of charges in the
being the most important the values of feeding gases contact surface or charge transference from one material
relative humidity and their stoichiometries. to the other.

D. Concentration losses The equation that represents the voltage behavior during
transients can be expressed as:
 [ ]
V( t ) = V(∞) + V(0 + ) − V(∞) ⋅ exp(− t τ) (22)
Where V(∞) is the final value to which the voltage will
go after the load variation, and V(0+) is the voltage initial
value, before the variation of the load fed by the PEM
fuel cell. The time constant (τ) is a measure of the
necessary time for the transient analysis in RC circuits,
computed by the following equation:
τ = Rth .Cdc [s] (23)
and Rth is the Thevenin equivalent resistance of the
circuit seeing from the capacitance.

3. Methods Fig. 2. Typical elemental operating scheme for a PEM fuel cell
stack.
In order to develop this study, computational modeling
tools were built up to model the fuel cell behavior Table II – Design characteristics for the first prototype
through a detailed theoretical base taking into
Description Symbol Amount
consideration assembling and operating parameters. Area of the cell Acel 5 cm2
Number of cells Ncel 1
With this purpose; to build the fuel cell simulator besides Membrane - Nafion® 112
the operating parameters already mentioned (Top, Iop, Electronic transfer coefficient (cathode) αc 0.43
Electronic transfer coefficient (anode) αa 1.0
PH 2 , PO2 , φ H 2 , φO2 , λ H 2 , λO2 or λ Air ), the Reaction grade constant (cathode) k0c 2.7 cm/s
constructive parameters (electronic transference
coefficients, electrode reaction level constants, and the Figure 3 shows the behavior of the cell with the
physicochemical properties of electrodes and membrane), variation of temperature for the first prototype with
and the dimensioning parameters for the fuel cell stack constants described in Table II:
(number of cells, cell area and de membrane thickness)
were also considered. The simulation results are
compared with practical experiments.

The experimental tests were made through controlled

processes using different models and prototypes of
PEMFC fuel cell stack as the system shown in the basic
scheme of Figure 2.

Under normal operation a frequency inverter is used to

feed ac energy due to the voltage regulation showed by
the fuel cell. An electronically controlled resistive load
was used in order to make possible a better control of the
necessary electrical load during laboratory experiments.

4. Results Fig. 3. Behavior of a PEM fuel cell under static load with
A.. PEM fuel cell in steady-state load regimen the temperature.

Figure 3 shows the behavior of a cell with net area of Figure 4 shows the behavior of the cell with the variation
5cm2, fed by hydrogen and pure oxygen with feeding of pressure for the second prototype with constants
pressures of 1 atm, stoichiometries of 1.1, humidification described in Table III:
100% under operating temperatures of 40oC, 60oC and
Table III – Design characteristics for the second prototype
80oC in static load regimen, being the curves assembled
from the cell voltage response according to the current Description Symbol Amount
requested by the load. Similar curves can be obtained Area of the cell Acel 20 cm2
fixing the temperature and changing other operating Number of cells Ncel 1
Membrane - Nafion® 112
parameters. Electronic transfer coefficient (cathode) αc 0.43
Electronic transfer coefficient (anode) αa 1.0
Reaction grade constant (cathode) k0c 3.5 cm/s
 Practically, this substitution, although showing loss of
efficiency, it is advantageous for commercial projects,
avoiding the need of oxygen transport and storage, since
one of the major challenges for the PEM fuel cells is to
find an efficient way to store bulky amounts of gaseous
fuels like hydrogen and oxygen.

The decreasing efficiency is due to the lower oxygen

concentration in the composition of atmospheric air, and
to the consequent increase in the cathode concentration
effect (Figure 6). The ohmic voltage drop and activation
effect are not dependent on this parameter.

Fig. 4. Behavior of a PEM fuel cell under static load with the
gas pressure.

The third prototype was used for experimental and

dynamic simulation tests, it is a fuel cell stack built from
two cells, assembled in a test bench with main design
characteristics as shown Table IV.

Table IV – Design characteristics for the third prototype

Description Symbol Amount

Area of the cell Acel 80 cm2
Number of cells Ncel 2
Membrane - Nafion® 112
Electronic transfer coefficient (cathode) αc 0.43
Electronic transfer coefficient (anode) αa 1.0 Fig. 6 – Ohmic and activation losses when comparing feeding
Reaction grade constant (cathode) k0c 1.3 cm/s of pure oxygen and air.

A second reason for this variation is the concentration

B. Utilization of air instead of pure oxygen effect in the cathode. With the decrease in concentration
of the fuel gas, the maximum possible current in this
electrode decreases. Adding the fact that the experiment
has been made in a temperature relatively low (50ºC),
bring the cell response curve for high current densities
close to the saturation limit (oxygen concentration in the
interface membrane-electrode close to zero). This
phenomenon occurs in a region where the power is
decreasing, and could still be possible the use of the cell
stack in this operating condition.

Figure 7 shows also that the anode concentration effect is

minimal, and due to its very small amount, they are
irrelevant to the process.

Fig. 5 – Behavior of a PEM Fuel Cell for pure oxygen and air.

For this analysis it was used the second prototype with air
or oxygen as a cathode reagent with an operating
temperature of 50o C. The other cell parameters are the
ones specified in Table II. The values obtained
experimentally and through computational simulation are
indicated in Figure 5.

The result above shows the developed model efficiency

and the success in the collection of constructive
parameters, and also demonstrates the influence of the
use of atmospheric air and pure oxygen.
Fig. 7. Concentration effects when comparing feeding of pure
oxygen and air.
C) PEMFC in dynamic load regimen[4]

In Figures 8, 9, 10 and 11 it is shown the simulation of

the behavior of a fuel stack of two PEM fuel cells with
data given in Table III and operating under a temperature
of 70oC fed by hydrogen and pure oxygen with pressures
of 1 atm, stoichiometries of 1.1 and humidification
100%, where was imposed a current pulse from a load
that changes from 45A to 80A over a 5 seconds interval
(dynamic load variation). Similar studies were done
taking into consideration several operating
configurations.

All the obtained results in practical experiments were Fig. 10. Graph of power (W) against time (s).
compared with the corresponding computer simulation.

Fig. 11. Curve of efficiency (%) against time (s).

Fig. 8. Graph of imposed current (A) against time (s)
The PEM fuel cells are sensible to all operational
5. Discussion parameters. In the case of pressure values and flow of
gases fed, the effect was more expressive when the
The mathematical simulation reproduced accurately variations lead to critical situations of activation or
the behavior of the PEM fuel cells evaluated. The biggest concentration of gases in the electrodes.
difficult by far was the determination of constructive
parameters for the materials used in the assembling of the In the case of values of humidification of feeding gases,
cells. This happened due to the fact that the technology the effect was more expressive when the variations took
employed in this cells is still under development. Some to critical situations of polymeric membrane dryness.
materials were adopted from other industrial processes
(like the Nafion®) and materials developed The effect of operating temperature variation is felt all
experimentally with separating plates, spreaders and over the process, being also the one that is more difficult
electrodes. With the technological evolution of PEM fuel to control. All the energy produced by the reaction and
cells, and the development of proper, more efficient and not converted into electrical energy is dissipated as heat,
standardized materials those difficulties will be reduced so the cooling system must be very efficient in case of
considerably. large fuel stacks.

6. Conclusion
Mathematical simulation is a useful tool, allowing the
reproduction with good degree of fidelity the PEM fuel
cell behavior. Should be mentioned here the fact that
many of the assembling parameters used were not easy to
be obtained without the help of the prototype in the
laboratory.

The PEM fuel cells under scrutiny showed their

enormous potential producing electrical energy with an
impressive efficiency, between 40% and 60%, when
Fig. 9. Graph of voltage (V) against time (s).
compared for example, with internal combustion engines.
The byproducts of the whole energy generation process
are oxygen and hot water. It is also possible the use of the
generated hot water in domestic and industrial processes.
However it is not possible to use the water in co- Corporation, Reading, PA, 2000, for the U.S. Department
generation processes due to the relatively low of Energy, Office of Fossil Energy, West Virginia.
temperature (up to 90oC). The studies of static and [2] Thomas, S. and Zalbowitz, M. Fuel Cells – Green
dynamic behavior under variable loads showed the large Power, Los Alamos National Laboratory, U.S.
influence of variable operating parameters in the Department of Energy, Office of Advanced Automotive
performance of the cell. Due to this aspect it is necessary Technologies.
a rigid control system for this variables in order to [3] Corrêa, J. M.; Farret, A. F. and Canha, L. N. “An
improve the performance of the PEM fuel cell stack. Analysis of the Dynamic Performance of Proton
Exchange Membrane Fuel Cells Using an
As can be seen from this study, the coupling of a fuel cell Electrochemical Model”, 27th Annual Conference of
with an inverter to produce ac current is not as simple as IEEE Industrial Electronics Society, 2001, pp.141-146.
it seems. The fuel cell changes its parameters for each [4] Amphlett, J. C.; Mann, R. F.; Peppley, B. A.,
load or operating point. The inverter design is not simple Roberge, P. R. and Rodrigues, A. “A Practical PEM Fuel
since its characteristics should be in agreement with the Cell Model for Simulating Vehicle Power Sources”,
fuel cell variable parameters. Department of Chemistry and Chemical Engineering,
Royal Military College of Canada, Ontario, 1995,
The study of the dynamic behavior shows that the RC pp.221-226.
constant during the load increase is bigger than during [5] Rodrigues, A.; Amphlett, J. C.; Mann, R. F.; Peppley,
the load decrease, as can be seen in Figure 9; opposing to B. A. and Roberge, P. R., “Carbon Monoxide Poisoning
the fact that the resistance increases with the value of of Proton-Exchange Membrane Fuel Cell”, Department
current and temperature, this happening due to the of Chemistry and Chemical Engineering, Royal Military
decrease of holes and free electrons in the electrodes. College of Canada, Ontario, 1195, pp.768-773.
This fact shows the influence of parameters related to the [6] Balkin, A. R., “Modelling a 500W Polymer
feeding of gases in the final value of the cell RC constant. Electrolyte Membrane Fuel Cell”, Faculty of
The implementation of the fuel cell needs a very Engineering, University of Technology, Sydney,
sophisticated set of controls to operate in a variable load Australia, 2002.
condition. [7] Slade, S.; Campbell, S. A.; Ralph, T. R. and Walsh, F.
C. “Ionic Conductivity of an Extruded Nafion 1100 EW
The implementation of this model allows the assembling Series of Membranes”, Journal of The Electrochemical
of polymeric membrane hydrogen stacks of different Society, 2002, no: 49, pp. 1556-1564.
sizes, and also allows the increase on the knowledge of [8] Suzuki, T.; Murata, H.; Hatanaka, T.; Morimoto Y.;
effects of variations on the components over the Analysis of the Catalyst Layer of Polymer Electrolyte
performance in each of the cells. Fuel Cells, R&D Review of Toyota CRDL, volume 39,
número 3, 2003.
Anode concentration and activation effect proved to be [9] Benziger, J.B.; et al.; The auto humidification
less dependent of the type of feeding (air or pure O2) than Polymer Electrolyte Membrane Fuel Cell, Princeton
the cathode effects. Ohmic voltage drops have a non- University, Princeton, 2001.
linear dependency with current and can be smaller than [10] Passos, R. R.; Ticianelli, E.A.; Effects of the
the anode activation losses for operating current Operational Conditions on the Membrane and Electrode
densities. Properties of a Polymer Electrolyte Fuel Cell, Journal of
the Brazilian Chemical Society, number 4, pg. 483-489,
The developed computational simulator, demonstrated to 2002.
be an efficient and valuable tool to the development of [11] Slade, S.; et al.; Ionic Conductivity of an Extruded
polymeric membrane fuel cell stacks; making possible Nafion 1100 EW Series of Membranes, Journal of The
the verification of planned effects without the need of the Electrochemical Society, number 149, pg. 1556-1564,
extra design cost of assembling a new prototype. 2002.
[12] Pathapati, P.R.; Xue, X.; Tang, J.; A New dynamic
Acknowledgement model for predicting transient phenomena in PEM fuel
cell system, Elsevier Renewable Energy, 2004.
The authors wish to thank to UNITECH Ltd., represented [13] Atkins, P.W. and de Paula, J.; Physical Chemistry,
by Dr. Antonio César Ferreira, it’s Director of Research Oxford University Press, 7th Edition, January, 2002.
and Development, due to the possibility of developing
this work in direct contact with the PEM fuel cell.

The authors also would like to thank CAPES, CNPq and

FAPEMIG for the financial support.

References

[1] J. H. Hirschenhofer, D. B. Stauffer, R. R. Engleman,

M. G. Klett Fuel Cell Handbook, 5th Edition, Parsons