Description of PEM Fuel Cells System

Diego Feroldi and Marta Basualdo

Abstract This chapter provides a description of polymer electrolyte membrane

(PEM) fuel cell-based systems and different modeling approaches. First, it shows the
structure of a single cell, the advantages and disadvantages of this type of fuel cell, the
expressions of the generated voltage and the efficiency, and the generic structure of
a generation system based on PEM fuel cell. Second, the chapter provides a review
of the principal models presented in the literature to describe the behavior of the
system. Different types of PEM fuel cell models are presented, focusing on dynamic
models suitable for control purposes. Particularly, this chapter describes in detail the
dynamic model used as a base to represent the system in the subsequent chapters
of the book. Then, the described model is used to study the optimal operation of a
fuel cell at different loads, showing the benefits of an optimal operation in terms of
hydrogen reduction and greater peak power.

1 Introduction

The polymer electrolyte membrane fuel cell (PEMFC), also known as proton
exchange membrane fuel cell, takes its name from the type of electrolyte: a poly-
meric membrane with high proton conductivity when the membrane is conveniently

D. Feroldi (B) M. S. Basualdo

CAPEG-CIFASIS-(CONICET-UNR-UPCAM),
27 de Febrero 210 bis, S2000EZP Rosario, Argentina
e-mail: [email protected]
M. S. Basualdo
UTN-FRRo, Zeballos 1341, S2000BQA Rosario, Argentina
e-mail: [email protected]
D. Feroldi
DCC-FCEIA-UNR, Pellegrini 250, S2000BTP Rosario, Argentina

M. S. Basualdo et al. (eds.), PEM Fuel Cells with Bio-Ethanol Processor Systems, 49
Green Energy and Technology, DOI: 10.1007/978-1-84996-184-4_2,
Springer-Verlag London Limited 2012
50 D. Feroldi and M. Basualdo

Fig. 1 Three-dimensional schematic diagram of a fuel cell

hydrated [1]. At the moment, the most common polymer used in these types of cells
is the Nafion developed by Du Pont (USA), which is fabricated with chemically
stabilized perfluorosulfonic acid copolymer [2].

1.1 Basic PEM Fuel Cell Structure

Basically, the physical structure of a PEMFC consists of seven components,

according to Fig. 1 [3]: feeding channels, diffusion layer, and catalytic layer in the
anode; membrane; catalytic layer, diffusion layer, and feeding channels in the cath-
ode. The PEMFC combines in a very compact unit the electrodes and the elec-
trolyte. This structure, well known as membrane electrode assembly (MEA), is not
thicker than a few hundred microns. It is the heart of the fuel cell and is fed with
hydrogen and oxygen,1 generating electrical power with a power density of around
1 W cm2 [5].
The polymeric solid electrolyte forms a thin electronic insulator and a barrier
for gases between both electrodes, allowing fast proton transport and high current

1 Usually, the fuel cell is fed with atmospheric air instead of pure oxygen. The oxygen mole fraction

in atmospheric air is 0.21 [4].

Description of PEM Fuel Cells System 51

density. The solid electrolyte has the advantage, as opposed to those of liquid type,
that allows the FC to operate in any spatial position [1].
The electrodes consist of a catalytic layer of great superficial area on a substratum
of coal, permeable to gases. Electrocatalyst materials are necessary to obtain a good
operation, increasing the speed of the chemical reaction. In this way, the gases can
react with a lower energy of activation, allowing the reaction to take place at a lower
temperature [3]. The electrocatalyst used in PEMFC is platinum, which is one of the
major drawbacks of this technology because of its high cost.
However, there are research advances of high temperature PEM fuel cells
(HT-PEMFCs) in several fields because there are several reasons for operating at
temperatures above 100 C [6]. First, the electrochemical kinetics for the reactions in
cathode and anode are enhanced. Second, the water management issue can be simpli-
fied because there would be no liquid water. Third, the cooling system is simplified
due to the increased temperature gradient between the fuel cell stack and the coolant.
Fourth, the waste heat can be exploited using cogeneration. Fifth, the tolerance to CO
is increased allowing the use of lower quality reformed hydrogen. Unfortunately, the
area of HT-PEMFCs is incipient and still needs much research to be implemented in
commercial applications.

1.2 Advantages and Disadvantages of PEM Fuel Cells

The main advantage of PEM fuel cells is their high efficiency compared with other
energy conversion devices [7]. This allows the efficiency of a fuel cell vehicle using
direct-hydrogen FC2 to be twice that in a gasoline vehicle [8, 9]. Moreover, unlike
the internal combustion engines where the efficiency is maximum with the highest
loads, the FC efficiency is also high with partial loads. This is advantageous because
in typical driving conditions, like urban and suburban scenarios, most of the time the
vehicle is demanding a small fraction of the nominal FC power [10]. Thus, an FC
vehicle will be working mostly at high efficiencies. At the same time, using direct-
hydrogen FC, the local emissions problem in densely urban areas can be eliminated.
Another important advantage of PEMFC, in contrast to other types of fuel cells,
is the low operation temperature (below 80 C) [11], allowing to reach the operation
point quickly. In addition, the cost of the materials is smaller than for the high
temperature fuel cells (except the catalyst, which is based on platinum) and their
operation is safer. All these characteristics turn PEMFC particularly appropriate
for applications in vehicles. Nevertheless, it is necessary to use better, and more
economic, catalyst so that the reaction occurs at lower temperatures.
The main disadvantage of fuel cells is their high cost and the high production cost
of hydrogen. Hydrogen is preferred because of the fast electrochemical reaction, and

2Direct-hydrogen FC refers to an FCS that is directly fed with hydrogen from a pressurized tank
opposite to the case where the hydrogen is produced with an on-site reformer.
52 D. Feroldi and M. Basualdo

its high specific energy.3 Nevertheless, as it was mentioned, hydrogen is not a primary
fuel. Usually, it is produced from hydrocarbon reforming or water electrolysis [1, 13].
The use of electrolysis is advisable especially when some type of renewable energy
is used, avoiding fossil fuel use. It is expected that the cost of fuel cells and hydrogen
will diminish with the progress in technology. Thus, hydrogen has the possibilities
of becoming an alternative to fossil fuels with the joint use of renewable energies.

1.3 Fuel Cell Voltage

The standard potential E 0 is a quantitative measurement of the maximum cell poten-

tial, i.e., the open circuit voltage. For a hydrogenoxygen cell, in which there is a
transfer of two electrons by each water molecule, E 0 = 1.229 V if the produced
water is in liquid state and E 0 = 1.18 V if the produced water is in gaseous state
[13]. These values correspond to normalized conditions: cell temperature (Tfc ) equal
to 298.5 K and partial pressures of oxygen ( pO2 ) and hydrogen ( pH2 ) equal to 1 atm.
In the work of Amphlett et al. [14], the following expression of E 0 is given, depend-
ing on the temperature and the reactant partial pressures, which are used in several
models of fuel cells e.g., [4, 15]:

E 0 = 1.229 8.5 104 (Tfc 298.5)

 
  1  
+ 4.3085 105 Tfc ln pH2 + ln p O2 . (1)
2

However, in practice the cell potential is significantly lower than the theoretical
potential because there are some losses even when no external load is connected.
Moreover, when a load is connected to the fuel cell, the voltage in the terminals
decreases still more due to a number of factors, including polarization losses and
interconnection losses between cells. The main voltage losses in a fuel cell are the
following [1, 13, 16]:
Activation loss The activation losses vact are caused by the slowness of the reaction
that takes place on the surface of the electrodes. A proportion of the generated voltage
is lost in maintaining the chemical reaction that transfers electrons from the negative
electrode toward the positive electrode. This phenomenon is strongly nonlinear, and
more important, it is at low current densities.
Fuel crossover and internal currents These energy losses result from the waste
of fuel passing through the electroslyte and from electron conduction through the
electrolyte. In PEMFC, the fuel losses and internal current are small and their effects
are usually negligible.

3 The Specific Energy of hydrogen at ambient pressure is 8890 W h kg1 , meanwhile the corre-
sponding one of petrol is 694 W h kg1 . However, the necessary volume for its storage is greater.
The energy density of petrol is 500 W h dm3 , meanwhile for hydrogen (300 bar) is 55 W h dm3
[12].
Description of PEM Fuel Cells System 53

1 0.6
Potential
Power density
0.5
0.9

Power density [W cm2]

0.4
Potential [V]

0.8
0.3
0.7
0.2

0.6
0.1

0.5 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
2
Current density [A cm ]

Fig. 2 Polarization curve showing the cell potential and the power density versus the cell current
density at pH2 = pO2 = 2.28 atm and Tfc = 353 K

Ohmic loss The ohmic losses vohm are caused by the resistance to the transport of
electrons through the electrodes and the different interconnections, and also to the
passage of ions through the electrolyte. The behavior of vohm is approximately linear
with the current density.
Concentration loss The concentration losses vconc are caused by the diffusion of
ions through the electrolyte which produces an increase in the concentration gradient,
diminishing the speed of transport. The relation between the voltage of the cell and
the current density is approximately linear upto a limit value, beyond which the losses
grow quickly.
Therefore, the fuel cell voltage of a simple cell can be expressed as

vfc = E 0 vohm vact vconc . (2)

A typical polarization curve showing the potential and power density as a function
of the current density is shown in Fig. 2. This curve is obtained using the fuel cell
model developed in [4]. In practice, a succession of cells are connected in series in
order to provide the necessary voltage and power output, constituting a Fuel Cell
Stack System.

1.4 Theoretical and Real Fuel Cell Efficiency

The efficiency of any energy conversion device is defined as the ratio between the
useful energy output and the energy input. In a fuel cell, the useful energy output is the
generated electrical energy and the energy input is the energy content in the mass of
hydrogen supplied. The energy content of an energy carrier is called the Higher Heat-
ing Value (HHHV ). The HHHV of hydrogen is 286.02 kJ mol1 or 141.9 MJ kg1 .
This is the amount of heat that may be generated by a complete combustion of
1 mol or 1 kg of hydrogen, respectively. The HHHV of hydrogen is experimentally
54 D. Feroldi and M. Basualdo

determined by reacting a stoichiometric mixture of hydrogen and oxygen in a steel

container at 25 C. If hydrogen and oxygen are combined, water vapor emerges at high
temperatures. Then, the container and its content are cooled down to the original 25 C
and the HHHV is determined by measuring the heat released between the identical
initial and final temperatures. On the contrary, if the cooling is stopped at 150 C, the
reaction heat is only partially recovered (241.98 kJ mol1 or 120.1 MJ kg1 ). This is
known as the lower heating value (HLHV ) of hydrogen [17].
Assuming that all the Gibbs free energy of hydrogen,4 G, can be converted into
electrical energy, the maximum possible (theoretical) efficiency of a fuel cell is [16]
HHV = G/ HHHV = 237.34/286.02 = 83%. (3)
However, the HLHV is used very often to express the fuel cell efficiency to com-
pare it with the internal combustion engine, whose efficiency has traditionally been
expressed using the fuel HLHV . In this case, the maximum theoretical fuel cell
efficiency results in a higher number:
LHV = G/ HLHV = 228.74/241.98 = 94.5%. (4)
If both G and HLHV in (3) are divided by 2F, where 2 is the number of
electrons per molecule of H2 and F is the Faraday number, the fuel cell efficiency
may be expressed as a ratio of two potentials:
G
G 1.23
HHV = = 2F
HHHV
= = 83%, (5)
HHHV 1.482
2F
HHHV
where G
2F = 1.23V is the theoretical cell potential, and 2F = 1.482V is the
potential corresponding to the HLHV , or thermoneutral potential.
In this section, we analyzed the theoretical fuel cell efficiency. However, as
explained in the previous sections, in a real fuel cell the efficiency is quite lower
and also depends on the fuel cell current. The fuel cell efficiency FC can also be
defined as the ratio between the power produced and the power of hydrogen con-
sumed [16]:
Vfc Ifc
PFC HHHV Ifc Vfc
HHV = = = , (6)
PH2 2F 1.482
Vfc Ifc
PFC HLHV Ifc Vfc
LHV = = = , (7)
PH2 2F 1.254

where Vfc is the generated voltage and Ifc is the fuel cell current. Thus, the FC
efficiency is related to the actual voltage, which is related to the fuel cell current
through the polarization curve.

4 The Gibbs free energy is used to represent the available energy to do external work. The changes
in Gibbs free energy G are negative, which means that the energy is released from the reaction,
and varies with both temperature and pressure [18].
Description of PEM Fuel Cells System 55

Fig. 3 Efficiency curve of a fuel cell system

Moreover, in a real system it is necessary to incorporate some auxiliary systems

which consume a fraction of the generated power. As a result, the efficiency of the
fuel cell system, fcs , is even lower than that expressed in Eq. 6:
 
Pnet Pfc Paux Paux
fcs = HHV = HHV = HHV 1 , (8)
Pfc Pfc Pfc
where Pnet is the net power output, Pfc is the fuel cell power, and Paux is the power
consumed by the auxiliary components, which include, in particular, the air com-
pressor. The efficiency curve of a fuel cell system of 50 kW modelled in ADvanced
VehIcle SimulatOR (ADVISOR)5 is shown in Fig. 3.

1.5 Generic Structure of a Fuel Cell-Based Power

Generation System

In order to be able to produce energy, it is necessary to integrate the fuel cell stack
with other components to form a fuel cell-based power generation system. A generic
scheme showing the interrelation between the main components of the power gen-
eration system is presented in Fig. 4. These components can be divided into the
following subsystems [18]:

5 ADVISOR is a toolbox developed by the national renewable energy laboratory with the aim of
analyzing the performance and fuel economy of conventional, electric, and hybrid vehicles [19, 20].
56 D. Feroldi and M. Basualdo

Fig. 4 General scheme of a FCS oriented to automotive applications

1.5.1 Reactant Flow Subsystem

The reactant flow subsystem consists of the hydrogen and air supply circuits. The
objective is to supply the adequate reactant flow to ensure fast transient response
and minimal auxiliary power consumption. The hydrogen supply circuit is generally
composed of a pressurized tank with pure H2 connected to the anode through a
pressure-reduction valve and a pressure-controlled valve, meanwhile, the air supply
circuit is generally composed of an air compressor which feeds the cathode with
pressurized air from the atmosphere. The anode output is generally operated in dead-
ended mode and a purge valve in the anode output is periodically opened to remove
the water and accumulated nitrogen gas. In case the anode output is not closed it
is possible to reinject the out-flowing hydrogen into the anode input. On the other
hand, the cathode output is normally open through a fixed restriction. The cathode
air supply will be studied in Chap. 3 where we propose to close the cathode output
with a controlled valve.

1.5.2 Heat and Temperature Subsystem

The heat and temperature subsystem includes the fuel cell stack cooling system and
the reactant heating system. The thermal management of the fuel cell is critical since
the performance depends strongly on the temperature. The stack temperature control
can be done using a fan or a water refrigeration subsystem.
Description of PEM Fuel Cells System 57

1.5.3 Water Management Subsystem

The objective of the water management subsystem is to maintain an effective hydra-

tion of the polymer membrane and an adequate water balance, because the fuel cell
performance is also strongly dependent on membrane hydration. Both the air and
the hydrogen, are usually humidified before entering the fuel cell with humidifiers in
both circuits. The water that leaves the cathode can be recovered in a water separator
and reinjected in the humidifiers through a pump.

1.5.4 Power Conditioning Subsystem

Fuel cells generate an unregulated DC voltage which drops off when the current
increases according to the polarization curve. In general, some power conditioning
actions are necessary to supply the load properly. Such actions make necessary the
use of DC/DC regulators and/or inverter regulators.

1.5.5 Power Management Subsystem

The power management subsystem controls the power drawn from the fuel cell stack.
If no energy storage devices are used, the full load must be supplied by the fuel cell
and no power management is necessary. However, if an energy storage system is
included, such as batteries or supercapacitors, it is necessary to implement a power
management between these two power sources. A review of fuel cell hybrid systems
is described in the following section and a detailed study is done in Chap. 4.

1.6 Modeling of Polymer Electrolyte Membrane Fuel Cell Systems

Validated mathematical models provide powerful tools for the development and
improvement of fuel cell-based systems. Mathematical models can be used to
describe the fundamental phenomena that take place in the system to predict the
behavior under different operating conditions and to design and optimize the control
of the system.
The fuel cell system models describe quantitatively the physical and electrochem-
ical phenomena that take place in to the cells. The models can be divided into two
groups [3]: empirical models and mechanistic models. Most of the empirical models
are focused on the prediction of the polarization curve, which is used to characterize
the electrical operation of the FC, by means of empirical equations. The following
empirical equation developed by Kim et al. [21] is used to calculate the voltage (E)
at different current densities (J), fitting experimental data at several temperatures,
pressures, and oxygen compositions in the cathode gas mixture:
E = E 0 b logJ R J m exp(n J ), (9)
58 D. Feroldi and M. Basualdo

where E 0 is the thermodynamic open-circuit voltage. The exponential term char-

acterizes the mass-transport region of the polarization curve, the zone where the
increase in slope of the pseudo-linear region and the subsequent rapid fall-off of the
cell potential. The parameter n has more pronounced effects than the parameter m
in this region. The terms E 0 and b yield the electrode kinetic parameters for oxygen
reduction in the cell. R represents the resistance, predominantly ohmic, and, slightly,
the charge-transfer resistance of the electro-oxidation of hydrogen.
Later, Squadrito et al. [22] used an empirical approach to account the mass trans-
port limitation,
 modifying the model of Kim et al. by replacing the last term of
(9) with m J n ln (1 J/Jlim ) where m and n are empirically determined constants
and Jlim is the limiting current density obtained by fitting experimental data. The
equation describes experimental data over the full range of current density taking
into account possible mass transport limitations. The empirical equation was used to
fit experimental data obtained from a single cell, showing good agreement between
theoretical and experimental data.
One example of an empirical model widely used is the model developed by
Amphlett et al. [14], which incorporates as much empirical properties as mecha-
nistic to obtain a transient model able to predict the FC voltage as a function of
the current, the FC temperature and the hydrogen and oxygen partial pressures for
a 35-cell 5 kW PEM fuel cell. The model works well for the experimental system
in study but can be generalized to any other system recalculating the values of the
model.
On the other hand, the mechanistic model considers the fundamental phenomena
in detail such as, heat and mass transport, forces, and electrochemical processes.
Mechanistic modeling (single and multi-domain) has been utilized to study a wide
range of phenomena including polarization effects (activation, ohmic and concentra-
tion overpotentials), water management, thermal management, CO kinetics, catalyst
utilization and flow field geometry [23]. All these models are parametric in the sense
that they account for the cell performance of various input parameters, typically
temperature, pressure and humidity.
Many of the mechanistic models are one-dimensional where only the direction
across the fuel cell is taken into account. Some of the first works dealing with this
type of models are [2426]. Then two-dimensional and three-dimensional models
were developed. Some two-dimensional models describe the plane perpendicular
through the flow channels [27, 28] while others simulate the fuel cell along the flow
channels [29, 30]. Within the three-dimensional models can be mentioned the model
developed by Nguyen et al. [31], which is a three-dimensional model of a PEM fuel
cell with serpentine gas channels, and the model developed by Natarajan [32].
Some of the earlier works focused in the humidification problem, addressing the
humidification requirements of the inlet gases to maintain the water balance [33].
Fuller and Newman [34] also addressed the water and thermal management, similar
to Nguyen and White [29] who studied different forms of gas humidification and
water management, analyzing the cell performance. The model of Whr et al. [35]
showed that for fuel cell stacks water management becomes even more difficult and
Description of PEM Fuel Cells System 59

is strongly related to thermal management. The temperatures of the inner cells of the
stack are higher than the outer cells resulting in membrane dehydration.
Another approach to fuel cell modeling is the use of equivalent circuits to represent
the system behavior. In fact, one of the most attractive aspects of the Electrochemical
Impedance Spectroscopy as a tool for investigating the electrical and electrochemical
properties of materials and systems, is the direct connection that often exists between
the behavior of a real system and that of an idealized model circuit consisting of dis-
crete electrical components [36]. The procedure typically consist in the comparison
or fitting the impedance data to an equivalent circuit, which is representative of the
physical processes taking place in the system under investigation. There are analogies
between the circuit elements and the electrochemical processes, so that the results
of data fitting can be more easily converted into physical understanding.
Many authors have studied the modeling approach based on equivalent elec-
trical circuits, consisting of an arrangement of different electrical components to
have the same frequency response as that obtained by electrochemical impedance
spectroscopy tests [3740]. Some works present equivalent circuits using electri-
cal elements, such as resistances, capacitances or inductance. However, other works
use additional distributed elements that represent electrochemical or mass and ionic
transport phenomena. For example, Warburg impedance represents the impedance of
one-dimensional distributed diffusion of a species in an electrode. Another example
is a constant phase element, used for describing a distributed charge accumulation
on rough irregular electrode surfaces.
However, there are some problems in using analogies to describe electrochemical
systems: one of them is that different equivalent circuits obtain the same frequency
response, another problem is the overlapping frequency response of different phe-
nomena and the dependence of the electrochemical phenomena on the operating
conditions (temperature, current, pressure, etc.).
So far we have seen different types of models that study the cell behavior in both
stationary and transient states. Nevertheless, there are few dynamic models suitable
for control purposes. In the work of Pukrushpan et al. [4] a dynamic model for
PEMFCS that is suitable for the control study has been developed. The model captures
the transitory behavior of the air compressor, the gases filling dynamics (in the
cathode and anode), and the effect of the membrane humidity. These variables affect
the cell voltage and, therefore, the efficiency and the output power. The polarization
curve in this model is a function of the hydrogen and oxygen partial pressures, the
stack temperature and the membrane water content. This allows to evaluate the effect
of variations of oxygen concentration and membrane humidity in the output voltage,
which are necessary to make the control during transitory operation.
Another practical model oriented to control was developed by Del Real et al. [41]
where the model parameters have been adjusted specifically for a 1.2 kW Ballard
stack, which is considered a benchmark as it is widely used by research groups in the
PEM fuel cell field. The model can predict both steady and transient states subject to
variable loads (including flooding and anode purges), as well as the system start-up.
The proposed model methodology is accurate since the simulated results show good
agreement, compared to experimental data from the Ballard stack.
60 D. Feroldi and M. Basualdo

On the other hand, a dynamic model which incorporates the effects of charge
double layer capacitance, the dynamics of flow and pressure in the anode and cath-
ode channels and mass/heat transfer transient features in the fuel cell body has been
presented [42]. This dynamic model can predict the transient response of cell volt-
age, temperature of the cell, hydrogen/oxygen out-flow rates and cathode and anode
channel temperatures/pressures under sudden change in load current. The simulation
results are analyzed and compared to benchmark results, reporting that a good agree-
ment is found between tests and simulations. Similarly, a dynamic electrochemical
model of a grid independent fuel cell power plant is presented in [43]. The model
includes the methanol reformer, the PEM stack and the power conditioning unit. The
model is used to predict the output voltage when subjected to rapid changes in a
residential load connected to it, showing a high degree of accordance. More recently,
a semi-empirical dynamic model for stack voltage, based on experimental investiga-
tion, was presented [44]. The proposed model can predict the transient response of
stack voltage under step change in current with good agreement between tests and
simulations.
The model developed by Pukrushpan et al. [4], which has been employed in several
works for control purposes [4551], is utilized in this book as a base to represent the
behavior of a generic fuel cell system. The model is described in detail in Sects. 1
and 2.

1.7 Description of the Fuel Cell System Model

The model developed by Pukrushpan et al. [4] is used in several chapters of this book
as a base to characterize the dynamic behavior and performance of PEM fuel cell
systems. This model contains four main subsystems that interact with each other:

FC voltage subsystem
Membrane hydration subsystem
Cathode flow subsystem
Anode flow subsystem

The spatial variation of the parameters is not considered and, thus, they are treated
as lumped parameters. On the other hand, the time constants of the electrochemical
reactions are in the order of magnitude of 1019 s despite another work [52] argues
that this constants are significatively lower (109 s). In any case, all the literature
agrees in the fastness of the electrochemical reactions, as remarked in [53] and [42].
Thus, for control purpose, these time constants can be assumed negligible compared
to other constants much slower: temperature (102 s) and dynamics of volume filling
(101 s).
In this book, this base model has been modified to adapt it to the proposed control
structures described in Sect. 2. In the original model there is only one control variable:
the compressor motor voltage. In this book, it is proposed to add an extra variable,
the throttle opening area in the cathode output ( At ), adding a control valve in the
Description of PEM Fuel Cells System 61

cathode output. The advantages of this new configuration are covered in detail in
Sect. 2.3.
Usually, in model-based control, it is necessary to find simplified models, derived
from the complete ones or from experimental data, to be used as inner models into
the controller. Here, the step response is utilized to obtain a simplified model that is
used as internal model to predict the future process response in the control strategy
implemented in Sect. 2. The main advantage of this simplified model is that it is
easily obtainable through experimental data with good agreement with respect to the
original nonlinear system in the considered operating point.

1.8 Principal Equations in the Fuel Cell System Model

The model described in the previous section has the following governing equations
where the mass of air in the supply manifold, the masses of oxygen, nitrogen and
water in the cathode and the masses of hydrogen and water in the anode are defined
using the principle of mass conservation [45]:
dm sm
= Wcp Wsm,out , (10)
dt
dm O2 ,ca
= WO2 ,ca,in WO2 ,ca,out WO2 ,rct , (11)
dt
dm N2 ,ca
= WN2 ,ca,in WN2 ,ca,out , (12)
dt
dm w,ca
= Wv,ca,in Wv,ca,out Wv,ca,gen + Wv,m , (13)
dt
dm H2
= WH2 ,an,in WH2 ,an,out WH2 ,rct , (14)
dt
dm w,an
= Wv,an,in Wv,an,out Wv,m Wl,an,out , (15)
dt

where
Wcp is the compressor flow,
Wsm,out is the outlet mass flow,
WO2 ,ca,in is the mass flow rate of oxygen gas entering the cathode,
WO2 ,ca,out is the mass flow rate of oxygen leaving the cathode,
WO2 ,rct is the mass flow rate of oxygen reacted,
WN2 ,ca,in is the mass flow rate of nitrogen gas entering the cathode,
WN2 ,ca,out is the mass flow rate of nitrogen gas leaving the cathode,
Wv,ca,in is the mass flow rate of vapor entering the cathode,
Wv,ca,out is the mass flow rate of vapor leaving the cathode,
Wv,ca,gen is the rate of vapor generated in the fuel cell reaction,
62 D. Feroldi and M. Basualdo

Wv,menbr is the mass flow rate of water across the fuel cell membrane,
Wl,ca,out is the mass flow rate of liquid water leaving the cathode,
WH2 ,an,in is the mass flow rate of hydrogen gas entering the anode,
WH2 ,an,out is the mass flow rate of hydrogen gas leaving the anode,
WH2 ,rct is the rate of hydrogen reacted,
Wv,an,in is the mass flow rate of vapor entering the anode,
Wv,an,out is the mass flow rate of vapor leaving the anode,
Wv,m is the mass flow rate of water transfer across the fuel cell membrane, and
Wl,an,out is the rate of liquid water leaving the anode.

The voltage of a fuel cell stack consisting of n fuel cells is given as

vst = n vfc , (16)

where the voltage of a single fuel cell is defined as

vfc = E vact vohm vconc (17)

with E being the open circuit voltage and vact , vohm and vconc being the activation,
ohmic and concentration overpotentials, respectively. By fitting experimental data
to the phenomenological model equations, the open circuit voltage and the three
overpotentials are respectively defined as

E = 1.229 0.85 103 (Tfc Tamb )

 
1
+ 4.3085 105 Tfc ln(1.01325 pH2 ) + ln(1.01325 pO2 ) . (18)
2

The activation voltage is

vact = v0 + va (1 ec1 i ) (19)

with

v0 = 0.279 8.5 104 (Tfc Tamb ) + 4.308 105 Tfc

    
pCa psat (Tfc ) 1 0.1173( pCa psat (Tfc ))
ln + ln , (20)
1.01325 2 1.01325

va = (1.618 105 Tfc + 1.618 102 )

p 2
O2
+ psat (Tfc )
0.1173
p 
+ (1.8 104 Tfc 0.166)
O2
+ psat (Tfc )
0.1173
4
(5.8 10 Tfc + 0.5736), (21)
Description of PEM Fuel Cells System 63

and

c1 = 10. (22)

The ohmic voltage is

vohm = i Rohm (23)

with the fuel cell electrical resistance

tm
Rohm = , (24)
m
the membrane conductivity
  
1 1
m = (b11 m b12 ) exp b2
303 Tfc
b11 = 5.139 103 , b12 = 3.26 103 , b12 = 350 (25)

and
 c3
i
vconc = i c2 (26)
i max

with

p 
(7.16 104 Tfc 0.622) 0.1173
O2

+ psat (Tfc )





+(1.45 103 Tfc + 1.68) for p O2 + psat (Tfc ) < 2 atm
0.1173
c2 =
p  (27)

(8.66 105 Tfc 0.068) 0.1173
O2
+ psat (Tfc )





+(1.6 104 T + 0.54) for p O2 + p (T ) 2 atm
fc 0.1173 sat fc

and

i max = 2.2, c3 = 2. (28)

The governing equations for the supply manifold pressure and the return manifold
pressure are respectively defined using the energy conservation principle and the
standard thermodynamics relationships as follows:
d psm Ra
= (Wcp Tcp Wsm,out Tsm ), (29)
dt Vsm
d prm Ra Trm
= (Wca,out Wrm,out ), (30)
dt Vrm

where Vsm is the supply manifold volume, Vrm is the return manifold volume, Tsm
is the supply manifold air temperature, Trm is the return manifold air temperature,
64 D. Feroldi and M. Basualdo

Tcp is the temperature of the air leaving the compressor, Ra is the air gas constant,
and is the air specific heat ratio.
To express the governing equations in terms of the states, the closure relations
Eqs. 3138 are used. The supply manifold outlet air rate Wsm,out is related to the
supply manifold pressure psm and the cathode pressure pca via the linearized nozzle
equation:

Wsm,out = ksm,out ( psm pca ), (31)

where ksm,out is the supply manifold outlet orifice constant.

The inlet oxygen, nitrogen, and cathode vapour mass flow rates, WO2 ,in , WN2 ,in ,
and Wv,ca,in are related to the cathode inlet air mass flowrate, the inlet air humidity and
the mass fraction of oxygen and nitrogen in dry air using the ideal gas relations. The
outlet oxygen, nitrogen and cathode vapour mass flow rates, WO2 ,out , WN2 ,out , and
Wv,ca,out , are likewise related to the outlet air mass flowrate, the outlet air humidity
and the mass fraction of the oxygen and nitrogen in dry air at the cathode outlet using
the ideal gas relations. The reacted oxygen and hydrogen and generated water vapor
(in the cathode) mass flow rates, WO2 ,rct , WH2 ,rct , and Wv,ca,gen , are related to the
fuel cell current:
n Ifc
WO2 ,rct = MO2 , (32)
4F
n Ifc
WH2 ,rct = MH2 , (33)
2F
n Ifc
Wv,ca,gen = Mv , (34)
2F

where the constants 4 and 2 in the denominators denote the of electrons involved in
the oxidation and the reduction half-reactions respectively, MO2 is the molar mass
of oxygen, MH2 is the molar mass of hydrogen, Mv is the molar mass of vapor, and
F is the Faraday constant.
The water mass flowrate through the membrane, Wv,m , is defined using the hydra-
tion model. The outlet hydrogen and water masses are assumed to be zero, that is,
hydrogen is assumed to react completely in the anode, while water generated by the
oxidation half-reaction is assumed to be transported via electro-osmosis through the
membrane towards the cathode.
The governing equation for the rotational speed of the compressor is defined by
the power conservation principle as

dcp
Jcp = cm cp , (35)
dt

where Jcp is the combined inertia of the compressor and the motor (kg. m2 ), cp
is the compressor speed (rad/sec), cm is the compressor motor torque input (Nm)
Description of PEM Fuel Cells System 65

calculated in Eq. 36, and cp is the torque required to drive the compressor (Nm)
calculated in Eq. 37.
The compressor motor torque cm is related to the compressor motor voltage Vcm
and the compressor motor rotational speed cp by the static motor equation:
kt  
cm = cm Vcm kv .cp , (36)
Rcm
where kt , Rcm , and kv are motor constants and cm is the motor mechanical efficiency.
The steady state compressor torque cp is related to the supply manifold pressure,
the compressor motor rotational speed and the compressor air flowrate Wcp via the
thermodynamic relations
  
C P Tatm psm ( 1)/
cp = 1 Wcp , (37)
cp cp patm

where C p is the air specific heat, cp is the compressor efficiency, and Tatm and patm
are the atmospheric temperature and pressure, respectively.
The air temperature in the compressor, Tcp , is defined using basic thermodynamic
relations
  
Tatm psm ( 1)/
Tcp = Tatm + 1 . (38)
cp patm

The air temperature in the supply manifold, Tsm , is obtained from m sm , psm and
Vsm using the ideal gas law. The cathode outlet air flowrate Wca,out is related to
the cathode pressure and return manifold pressure via a linearized nozzle equation
analogous to Eq. 31. The return manifold outlet air flowrate Wrm,out is defined using
a non-linearized nozzle relation as discussed in Sect. 2.2, while the return manifold
air temperature Trm is considered to be constant and equal to the temperature of the
fuel cell stack.
The flowrate of water through the membrane is controlled by two transport phe-
nomena: electroosmotic drag of water molecules by the protons and back-diffusion
from the cathode towards the anode. The transport phenomena is defined as
 
i (Cv,ca Cv,an )
Wv,m = Mv Afc n n d Dw , (39)
F tm
where the electroosmotic drag coefficient is given as

n d = 0.00292m + 0.05m 3.4 1019 (40)

and m is the mean water content in the membrane. The water content is defined as

0.043 + 17.81ai 39.85ai + 36.0ai , 0 < ai 1
2 3

ui = 14 + 1.4(ai 1), 1 < ai 3 (41)


(i = m, an, ca)
66 D. Feroldi and M. Basualdo

where the water vapour activity is defined as

xv,i pi pv,i
ai = = (i = an, ca). (42)
psat,i psat,i

The average water vapour activity in the membrane is defined as

aan + aca
am = (43)
2
and the water diffusion coefficient is given as
  
1 1
DW = D exp 2416 104 (44)
303 Tfc

with the preexponential term

6

10 , m < 2


106 [1 + 2(m 2)], 2 m 3
D = (45)

106 [3 1.67(m 3)], 3 < m < 4.5



1.25 106 , m 4.5

The water concentration at the membrane surfaces on anode and cathode sides in
Eq. 39 is a function of the membrane water content
m,dry
Cv,i = i (i = an, ca), (46)
Mm,dry

where m,dry (kg/cm3 ) is the membrane dry density and Mm,dry (kg/mol) is the
equivalent weight.

2 Auxiliary Equipment and System Modeling

While the fuel cell operates with oxygen as reactant in the cathode, it is more practical
to use oxygen from air. Air is mainly composed of nitrogen (78.084%), oxygen
(20.946%), and argon (0.9340%). The effect of using air instead of pure oxygen is
a reduction of approximately 50 mV in the cell voltage [16]. Additionally, there is a
reduction in the fuel cell efficiency because of the power consumption to pump the
oxygen and almost four times that of nitrogen.
In a hydrogen-air fuel cell system the air is supplied by a fan, a blower or a
compressor. In any case, there is an electric motor with a power consumption that
implies a reduction in the efficiency of the fuel cell system. The net power output,
Pnet , that is actually available is the fuel cell power, Pfc , less the power consumed
by the ancillary components, Paux , which includes the compressor or the blower:
Description of PEM Fuel Cells System 67

Pnet = Pfc Paux . (47)

The principal consumption between the ancillary equipments is that corresponding

to the electric motor that runs the compressor or the blower. Besides, it depends
strongly on the operating conditions. Thus, it can be assumed that the consumption
of the ancillary equipments is the consumption of this electric motor.
When the fuel cell operates at high pressures, the power output is higher. Thus,
the most common scenario is a fuel cell where the cathode is supplied with air
by a compressor. However, the power consumption of the compressor increases
significantly with the pressure. Thus, when the compressor consumption is taken
into account, there is a trade-off between pressure and efficiency.
The operation of the fuel cell system at high pressures increases the generated
voltage as a result of the increase in the cathode oxygen partial pressure and anode
hydrogen partial pressure (see Eq. 1). Especially, an increase in the cathode pressure
produces an increase in the supply manifold pressure and thus, an increase in the
pressure ratio across the compressor and in the compressor power consumption
contributing to a reduction in the system efficiency. The power consumed by the air
compressor is
 1
 
C p Tatm psm
Pcp = 1 Wcp , (48)
cp patm

where Wcp is the compressor air flow rate, Pcp is the compressor power, Tatm is the
inlet air temperature in the compressor, cp is the compressor efficiency, psm is the
supply manifold pressure, C p = 1,004 J kg1 K1 is the specific heat capacity of
air, and = 1.4 is the ratio of the specific heat of air. Actually, the power consumed
by the electric motor is higher because of the mechanical and electric inefficiencies:

Pcomp
PEM = , (49)
mec EM

where mec is the compressor mechanical efficiency and EM is the efficiency of

the electric motor. In Chap. 3 will be seen how this power consumption affects the
efficiency of the system. Besides, a solution from the point of view of control will
be studied in detail.

3 Optimal Operation of the FCS

The FCS model is useful to study the optimal operation of an FCS, especially at low
loads. An adequate operation produces important benefits, increasing the system
efficiency in terms of hydrogen reduction and allowing a greater peak power. In
a direct-hydrogen FCS with the cathode supplied with air through a compressor,
the air supply subsystem has a crucial role in the improvement of the performance
68 D. Feroldi and M. Basualdo

of the system [10]. In fact, there are two external variables that have greater impact
on the polarization curve: the air pressure and the air stoichiometry.
The air pressure and the air stoichiometry control the oxygen partial pressure in
the catalytic layer of the cathode, which determines the cathode polarization and,
therefore, the efficiency. In [54], it is also stated the importance of the air pressure
control to improve the FCS efficiency. The efficiency improvement for a given load is
based on a trade-off between the increase of the air pressure and the air stoichiometry,
and the increase of the parasitic compressor power.
In the work of Friedman and Moore [55], it is shown that an FCS can be optimized
to obtain high peak power and high efficiency over a broad range of output powers.
The key to obtain this objective is to vary the pressure and the air flow. Based on
this result, it can be concluded that an FCS must be operated to the greater possible
pressure and air stoichiometry. Nevertheless, if the energy necessary to compress the
air is considered, the result is different: for a fixed air flow, the compressor power
consumption increases significantly when the pressure is increased. This means that
it is possible to find an optimal combination of pressure and air flow.
A similar conclusion is stated in [45]. In this work, it is assumed that the FCS
net power Pnet can be approximately defined as the difference between the power
produced by the FCS, Pfcs , and the consumed compressor power Pcm . For each
load current Ist , an air flow increment increases the cathode pressure and, therefore,
increases the oxygen partial pressure, increasing the FCS voltage. This also leads to
an increment in the oxygen excess ratio in the cathode, O2 . The initial increase in
the oxygen excess is translated into an FCS power increase and a net FCS power.
Nevertheless, if a limit O2 is exceeded, an excessive compressor power is produced
so that the net power falls.
Following this analysis and using the model, it is possible to find the optimal rela-
tion among maximum net power, Pmax net , optimal oxygen ratio, opt , optimal supply
O2
opt opt
manifold air pressure, pSM , and optimal compressor motor voltage, vCM , with the
stack current, Ist [45]. These relations are shown in Eqs. 5053 and also in Fig. 5.
net
Pmax = 2.99 104 Ist2 + 0.271Ist 0.871 (50)

O2 = 2.7633 106 Ist2 15.931 104 Ist + 2.7331 (51)

opt

opt
pSM = 2.9996Ist2 + 736.4Ist + 8.932 104 (52)

vCM = 1.36 103 Ist2 + 1.17Ist + 14.3

opt
(53)

The Eq. 51 expresses that at each level of the stack current there is an optimal
value of O2 at which Pnet has the maximum value. This optimal value in the oxygen
excess ratio is achieved by applying a compressor voltage according to Eq. 53. The
Eqs. 5053 are obtained by solving the model equations previously shown under
steady-state condition and the analysis corresponds to the typical fuel cell operating
conditions (fuel cell temperature Tfc = 353 K and relative humidity Ca = 1).
Description of PEM Fuel Cells System 69

60 2.8

50 2.6
Pnet [kW]

opt
2
O
40 2.4
max

30 2.2

20 2
100 150 200 250 100 150 200 250
Ist [A] Ist [A]

5
x 10
2.5 250

2
popt [kPa]

200

Vopt [V]
1.5

CM
SM

150
1

0.5 100
100 150 200 250 100 150 200 250
I [A] Ist [A]
st

net , , p opt opt opt

Fig. 5 Optimal relations of Pmax O2 SM , and vCM as a function of the stack current Ist

4 Conclusions

This chapter presents the main concepts about PEM fuel cell systems, showing the
structure of these systems, the advantages and disadvantages, and, particularly, the
expressions corresponding to the fuel cell voltage and efficiency. Also, different
approaches for modeling are presented. Accordingly, a review of PEM fuel cell
model has been done. Fuel cell models may be classified into one of three categories:
analytic, semi-empirical or mechanistic (theoretical). Among these models, a model
widely used in the literature for control purposes is described in detail. Using this
model, the influence of the auxiliary equipment in the system efficiency is analyzed
and the model is also used to study the optimal operation of a PEM fuel cell system,
finding an optimal compressor voltage for each current load that maximizes the net
output power. Similarly, this dynamic model is used as a base model in the other
chapters of this book.

References

1. Laughton MA (2002) Fuel cells. Power Eng J 16:3747

2. Tang H, Peikang S, Jiang SP, Wang F, Pan M (2007) A degradation study of Nafion proton
exchange membrane of PEM fuel cells. J Power Sources 170(1):8592
3. Yao K, Karan K, McAuley K, Oosthuizen P, Peppley B, Xie T (2004) A review of mathematical
models for hydrogen and direct methanol polymer electrolyte membrane fuel cells. Fuel Cells
4(12):329
4. Pukrushpan JT, Peng AGH (2002) Modeling and control for PEM fuel cell stack system.
American Control Conference, Proceedings of the 2002, 4:31173122
70 D. Feroldi and M. Basualdo

5. DArco S, Ianuzzi D, Pagano M, Tricoli P (2005) Design criteria optimizing the use of fuel cell
source in electric power system. In: Proceedings of the 16th IFAC World Congress, Prague
6. Zhang J, Xie Z, Zhang J, Tang Y, Song C, Navessin T, Shi Z, Song D, Wang H, Wilkinson DP,
Liu Z-S, Holdcroft S (2006) High temperature pem fuel cells. J Power Sources 160(2):872891
7. Carrette L, Friedrich K, Stimming U (2001) Fuel cells-fundamentals and applications. Fuel
Cells 1(1):539
8. Rajashekara K (2000) Propulsion system strategies for fuel cell vehicles. Fuel Cell Technol for
Vehicles 1:179187
9. Jeong K, Oh B (2002) Fuel economic and life-cycle cost analysis of a fuel cell hybrid vehicle.
J Power Sources 105:5865
10. Friedman D, Moore R (1998) PEM fuel cell system optimization. In: Proc Electrochem Soc
27:407423
11. EG & G Technical Services, Inc. Science Applications International Corporation (2002) IN:
U.S. Department of Energy, Fuel Cell Handbook, 6th edition, Morgantown, West Virginia,
USA
12. Flipsen S (2006) Power sources compared: the ultimate truth? J Power Sources 162:927934
13. Larminie J, Dicks A (2003) Fuel Cell Systems Explained, 2nd edn Wiley, New York, USA
14. Amphlett J, Mann R, Peppley B, Roberge P, Rodrigues A (1996) A model predicting transient
responses of proton exchange membrane fuel cells. J Power Sources 61:183188
15. Correa J, Farret F, Canha L, Simoes M (2004) An electrochemical-based fuel-cell model suitable
for electrical engineering automation approach. Ind Electron, IEEE Trans on 51(5):11031112
16. Barbir F (2005) PEM fuel cells: theory and practice. Elsevier, Burlington, MA, USA
17. Bossel U (2003) Well-to-wheel studies, heating values, and the energy conservation principle.
IN:Ulf Bossel (ed) European Fuel Cell Forum, October 2003, Oberrohrdorf, Switzerland 1:15
18. Pukrushpan J (2003) Modelling and control of fuel cell systems and fuel processors. PhD
thesis, University of Michigan, Ann Arbor, Michigan, USA. http://www-personal.umich.edu/
annastef/FuelCellPdf/pukrushpan_thesis.pdf
19. Wipke K, Cuddy M, Burch S (1999) Advisor 2.1: a user-friendly advanced powertrain simula-
tion using a combined backward/forward approach. IEEE Trans on Vehicular Technol 48:1751
1761
20. Markel T, Brooker A, Hendricks T, Johnson V, Kelly K, Kramer B, OKeefe M, Sprik S, Wipke K
(2002) Advisor: a system analysis tool for advanced vehicle modeling. J Power Sources
110:255266
21. Kim J, Lee SM, Srinivasan S, Chamberlin CE (1995) Modeling of proton exchange membrane
fuel cell performance with an empirical equation. J Electrochem Soc 142:2670
22. Squadrito G, Maggio G, Passalacqua E, Lufrano F, Patti A (1999) An empirical equation for
polymer electrolyte fuel cell (PEFC) behaviour. J Appl Electrochem 29(12):14491455
23. Cheddie D, Munroe N (2005) Review and comparison of approaches to proton exchange
membrane fuel cell modeling. J Power Sources 147(12):7284
24. Bernardi DM, Verbrugge MW (1991) Mathematical model of a gas diffusion electrode bonded
to a polymer electrolyte. AIChE J 37(8):11511163
25. Bernardi DM, Verbrugge MW (1992) A Mathematical model of the solid-polymer-electrolyte
fuel cell. J Electrochem Soc 139:2477
26. Springer TE, Zawodzinski TA, Gottesfeld S (1991) Polymer electrolyte fuel cell model.
J Electrochem Soc 138(8):23342342
27. He W, Yi JS, Van Nguyen T (2000) Two-phase flow model of the cathode of PEM fuel cells
using interdigitated flow fields. AIChE J 46(10):20532064
28. Natarajan D, Van Nguyen T (2001) A two-dimensional, two-phase, multicomponent, tran-
sient model for the cathode of a proton exchange membrane fuel cell using conventional gas
distributors. J Electrochem Soc 148:A1324
29. Nguyen TV, White RE (1993) A water and heat management model for proton-exchange-
membrane fuel cells. J Electrochem Soc 140:2178
Description of PEM Fuel Cells System 71

30. Wang ZH, Wang CY, Chen KS (2001) Two-phase flow and transport in the air cathode of proton
exchange membrane fuel cells. J Power Sources 94(1):4050
31. Nguyen PT, Berning T, Djilali N (2004) Computational model of a PEM fuel cell with serpentine
gas flow channels. J Power Sources 130(12):149157
32. Natarajan D, Van Nguyen T (2003) Three-dimensional effects of liquid water flooding in the
cathode of a PEM fuel cell. J Power Sources 115(1):6680
33. Bernardi DM (1990) Water-balance calculations for solid-polymer-electrolyte fuel cells. J Elec-
trochem Soc 137:3344
34. Fuller TF, Newman J (1993) Water and thermal management in solid-polymer-electrolyte fuel
cells. J Electrochem Soc 140:1218
35. Wohr M, Bolwin K, Schnurnberger W, Fischer M, Neubrand W, Eigenberger G (1998) Dynamic
modelling and simulation of a polymer membrane fuel cell including mass transport limitation.
Int J Hydrogen Energy 23(3):213218
36. Barsoukov En, Macdonald JR (2005) Impedance spectroscopy: theory, experiment, and appli-
cations. Wiley, Hoboken, New Jersey, USA
37. Andreaus B, McEvoy AJ, Scherer GG (2002) Analysis of performance losses in polymer
electrolyte fuel cells at high current densities by impedance spectroscopy. Electrochimica Acta
47(1314):22232229
38. Ciureanu M, Mikhailenko SD, Kaliaguine S (2003) PEM fuel cells as membrane reactors:
kinetic analysis by impedance spectroscopy. Catal Today 82(14):195206
39. Macdonald DD (2006) Reflections on the history of electrochemical impedance spectroscopy.
Electrochimica Acta 51(89):13761388
40. Primucci M, Ferrer L, Serra M, Riera J (2008) Characterisation of fuel cell state using Elec-
trochemical Impedance Spectroscopy analysis. Simposium Ibrico de Hidrgeno, Pilas de
Combustible y Bateras Avanzadas (HYCELTEC)
41. Alejandro J, Real D, Arce A, Bordons C (2007) Development and experimental validation of
a pem fuel cell dynamic model. J Power Sources 173(1):310324
42. Pathapati PR, Xue X, Tang J (2005) A new dynamic model for predicting transient phenomena
in a pem fuel cell system. Renew Energy 30(1):122
43. El-Sharkh M, Rahman A, Alam M, Byrne P, Sakla A, Thomas T (2004) A dynamic model
for a stand-alone PEM fuel cell power plant for residential applications. J Power Sources
138:199204
44. An improved dynamic voltage model of pem fuel cell stack (2010) Int J Hydrogen Energy
35(20):1115411160 Hyceltec 2009 Conference
45. Grujicic M, Chittajallu KM, Law EH, Pukrushpan JT (2004) Model-based control strategies
in the dynamic interaction of air supply and fuel cell. Proc Inst Mech Eng, Part A: J Power
Energy 218(7):487499
46. Vahidi A, Peng A (2004) Model predictive control for starvation prevention in a hybrid fuel
cell system. Am Control Conf, Proc 2004 1:834839
47. Varigonda S, Pukrushpan JT, Stefanopoulou AG (2003) American Institute of Chemical Engi-
neers. Challenges in fuel cell power plant control: the role of system level dynamic models.
Am Inst Chem Eng 1:101110
48. Gelfi S, Stefanopoulou AG, Pukrushpan JT, Huei Peng (2003) Dynamics of low-pressure and
high-pressure fuel cell air supply systems. In: Am Control Conf, 2003. Proceedings of the 2003
3:20492054
49. Caux S, Lachaize J, Fadel M, Shott P, Nicod L (2005) Energy management of fuel cell system
and supercaps elements. In: Proceedings of the 16th IFAC World Congress, Prague
50. Serra M, Aguado J, Ansade X, Riera J (2005) Controllability analysis of decentralized linear
controllers for polymeric fuel cells. J Power Sources 151:93102
51. Feroldi D, Serra M, Riera J (2007) Performance improvement of a pemfc system controlling
the cathode outlet air flow. J Power Sources 169(1):205212
52. Zenith F, Skogestad S (2007) Control of fuel cell power output. J Process Control 17(4):333
347
72 D. Feroldi and M. Basualdo

53. Ceraolo M, Miulli C, Pozio A (2003) Modelling static and dynamic behaviour of proton
exchange membrane fuel cells on the basis of electro-chemical description. J Power Sources
113(1):131144
54. Yang W, Bates B, Fletcher N, Pow R (1998) Control challenges and methodologies in fuel cell
vehicles development. Fuel Cell Technol for Vehicles 1:249256
55. Friedman D (1999) Maximizing direct-hydrogen PEM fuel cell vehicle efficiency-is hybridiza-
tion necessary? SAE Int 1:265272
http://www.springer.com/978-1-84996-183-7