2856
12, DECEMBER 2009
Power Balance of a Hybrid Power Source in a Power
Plant for a Small Propulsion Aircraft
Elena Bataller-Planes, Nieves Lapeña-Rey, Jonay Mosquera, Fortunato Ortı́, Jesús Á. Oliver, Member, IEEE,
Óscar Garcı́a, Member, IEEE, Félix Moreno, Jorge Portilla, Yago Torroja, Miroslav Vasić, Santa Concepcion Huerta,
Matı́as Trocki, Pablo Zumel, Member, IEEE, and José A. Cobos, Member, IEEE
Abstract—This paper analyzes two different architectures for
a hybrid power source comprising a polymer electrolyte mem-
brane fuel cell and a Li ion battery. The hybrid power source feeds
the propulsion motor of an all-electrical aircraft, the Boeing Fuel
Cell Demonstrator. Unregulated and regulated hybrid power ar-
chitectures are examined. The regulation is achieved by means of a
controllable series boost converter (SBC) connected in series with
the fuel cell. Both architectures have been simulated, implemented,
and tested in the Boeing Fuel Cell Demonstrator airplane.
Index Terms—Fuel cell, hybrid power source, power balance,
series boost converter (SBC).
I. INTRODUCTION
HE need for CO2 footprint reduction to limit its effect on
T the climate change, the shortage of fossil fuels, and the
increased demand for energy are driving the research in new
and clean sources of energy, higher efficiency processing, and
saving of energy. The combination of these new sources with
the potential improvement in the energy processing provided
by advanced power electronics systems can provide significant
reduction of CO2 emissions and reduce the dependence of fossil
fuels.
The industry is aware of this potential, and a lot of research
resources and work have been involved on the analysis and
evaluation of fuel cells in many applications ranging from dis-
tributed generation to mobile applications [1], [2], [4]–[8]. In
Manuscript received July 17, 2008; revised September 26, 2008 and January
15, 2009. Current version published December 28, 2009. Recommended for
publication by Associate Editor M. G. Simoes. This paper was presented in part
at the IEEE 39th Power Electronics Specialists Conference, Rhodes, Greece,
June 16, 2008.
E. Bataller-Planes, N. Lapeña-Rey, J. Mosquera, and F. Orti are
with the Environmentally Progressive Air Transport Department, Boe-
ing Research and Technology Europe, Boeing Company, Madrid
28042, Spain (e-mail: elena.bataller@gmail.com; nieves.lapena@boeing.com;
jonay.mosquera@boeing.com; fortunato.d.orti@boeing.com).
J. A. Oliver, Ó. Garcı́a, F. Moreno, J. Portilla, Y. Torroja, M. Vasić,
S. C. Huerta, M. Trocki, and J. A. Cobos are with the Depar-
tamento de Automática, Ingenierı́a Electrónica e Informática, Escuela
Técnica Superior de Ingenieros Industriales (ETSII), Universidad Politécnica
de Madrid (UPM), Madrid 28006, Spain (e-mail: jesusangel.oliver@
upm.es; o.garcia@upm.es; jorge.portilla@upm.es; yago.torroja@upm.es;
miroslav.vasic@upm.es; conihuerta@etsii.upm.es; matiastrocki@hotmail.com;
joseantonio.cobos@upm.es).
P. Zumel is with the Universidad Carlos III de Madrid, Departamento de
Ingenieria, Electronica, Madrid 28911, Spain (e-mail: pzumel@ing.uc3m.es).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPEL.2009.2022943
0885-8993/$26.00 © 2009 IEEE
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO.
the transportation sector, this research has been focused mainly
in the automotive industry.
In the field of fuel cell hybrid electric vehicles, a comparison
of the train drive efficiency of electric vehicles (EV), hybrid
electric vehicles (HEV), and fuel cell vehicles (FCV) through
efficiency maps has been performed in [1]. An analysis of the
current situation of advanced electrical power systems in elec-
tric, hybrid electric, and fuel cell vehicles is given in [2]. Mean-
while, the research for the application of fuel cells in aircrafts
has been mainly focused on auxiliary power units (APUs) and
on unmanned aerial vehicles (UAV) [4], [5]. In this paper, the
focus is placed on the analysis of two architectures for a manned
full electrical aircraft.
The Boeing Fuel Cell Demonstrator is a manned full electrical
aircraft that flies level with only the energy of a polymer elec-
trolyte membrane (PEM) fuel cell. It is an example, not only of
an all-electrical aircraft, but also of an environmentally friendly
aircraft, because the only exhaust product is water vapor.
The aircraft’s platform is the Austrian HK36 Super Dimona
motor glider, a two-seater Jar 22 certified aircraft with a max-
imum takeoff weight of 770 kg, a wing span of 16.3 m, and a
maximum lift to drag ratio of 27. The original motor glider was
slightly modified to accommodate all different systems (fuel cell
system, fuel system, battery, power electronics and control, elec-
tric motor and drive, thermal management system, etc.). After
all different systems were installed onboard, the demonstrator
weighs 870 kg, including the pilot.
The airplane has two levels of power demand: a high-power
demand for take off and climb, and a low-power demand for
cruise (see Fig. 1). Since fuel cells have relatively low specific
power values, the fuel cell of the Boeing Fuel Cell Demonstrator
Airplane was sized for the cruise power demand. The additional
power required for take off and climb is provided by a Li ion
battery, which has higher power density but lower energy den-
sity than the fuel cell (for sufficiently high energy storage re-
quirements) [3]. Therefore, the Boeing Fuel Cell Demonstrator
Airplane has a hybrid power source comprising a PEM fuel cell
and a Li ion battery.
The objective of this paper is the analysis of two electri-
cal architectures for this hybrid power source: unregulated and
regulated.
II. POWER PLANT DESCRIPTION
The airplane is equipped with a dc brushless motor consisting
of an inverter and an electrical motor that drives a variable-pitch
BATALLER-PLANES et al.: POWER BALANCE OF A HYBRID POWER SOURCE IN A POWER PLANT FOR A SMALL PROPULSION AIRCRAFT
Fig. 1. Simplified power profile for the flight mission.
propeller. The motor power is proportional to the torque gener-
ated by the motor and the rotational speed set by the propeller
pitch angle. The pilot controls the torque magnitude by means
of the throttle lever and the rotational speed by setting it in the
propeller pitch controller. The throttle lever input is processed
by a controller (the power management and distribution control
board) that sends the torque command to the motor inverter. The
rotational speed is controlled by the propeller control unit.
During take off and climb, both the battery and the fuel cell
supply power to the electric motor. During the cruise, the battery
is disconnected and the fuel cell supplies all the power required
to maintain a straight level flight. The battery is not recharged
in flight.
The fuel cell system’s maximum output power is 24 kW gross.
The output voltage of the fuel cell system varies following its
polarization curve (V versus I) from approximately 320 V at
open circuit voltage (OCV) to 215 V at full power.
The battery can supply a maximum continuous power in the
range of 50–75 kW. The battery voltage depends on its state
of charge and current. The nominal voltage is 237 V, the OCV
when the battery is fully charged is 264 V and the battery can
discharge down to approximately 215 V.
The propulsion motor consumes approximately 40 kW dur-
ing take off and climb (approximately 7 min). The total power
delivered by the sources during take off and climb is approxi-
mately 43 kW. Additional loads include the fuel cell auxiliary
equipment and small actuators, controllers, and other electrical
equipment. The detailed descriptions of the power sources and
the airplane’s subsystems are given in [6] and [7].
III. PROPOSED ARCHITECTURES
Among the different architectures that could be used for the
Boeing Fuel Cell Demonstrator Airplane [2], [8]–[12], two have
been finally selected due to their simplicity (architecture 1) and
the power management capability (architecture 2).
In the first architecture (see Fig. 2), the hybrid power source
comprises the fuel cell and the battery connected in parallel to
a distribution bus by means of serial diodes.
In the second architecture (see Fig. 3), a converter is added
in series with the fuel cell to increase the voltage of this branch.
This converter is named series boost converter (SBC). A similar
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
2857
Fig. 2. Power architecture with unregulated hybrid power source.
architecture has already been presented and analyzed for a low-
power (135 W) fuel cell application by [8] but using a converter
in parallel with the fuel cell. The SBC is composed of six com-
mercial converters in parallel, in a master–slave configuration.
Individual serial diodes are added to each converter. The serial
boost converter, comprising the converters and the serial diodes,
is bypassed by a power diode. When the SBC is disabled, the
current flows through the bypassing diode.
The advantage of using a dc–dc converter in series with the
fuel cell rather than in parallel, to control the power balance,
is that the power handled by the dc–dc converter (PSBC ) is a
fraction of the fuel cell power (PFC ) given by
VBUS − VFC
PSBC = PFC . (1)
VBUS
The serial connection of the dc–dc implies lower losses, and
therefore, lower size and weight of the converter. In the demon-
strator airplane, the fraction of the fuel cell power processed
by the dc–dc converter is 15% (3 kW) in the worst-case sce-
nario. Additionally, the SBC acts as an active filter for the low-
frequency current harmonics of the inverter that can negatively
impact the behavior of the fuel cell [10].
A. Power Balance for an Unregulated Architecture
The bus voltage and the power delivered by each source is
a function of the demanded load and the state of charge of the
battery. The simplified power flight profile of the Boeing Fuel
Cell Demonstrator Airplane is shown in Fig. 1. During the high
power demand phase, more than half of the power is drawn from
the battery. As the battery discharges, the battery, bus, and fuel
cell voltages decrease. Therefore, the fuel cell power increases
(the lower the voltage of the fuel cell, the higher its current and
its power), whereas the battery power decreases.
With this architecture, the fuel cell is operated below its max-
imum power output capabilities in all conditions. The power
drawn from the battery and/or the fuel cell depends only on
the V–I characteristics and the internal resistance, showing low
design flexibility. Therefore, the battery and the fuel cell of the
demonstrator have been chosen to have similar V–I characteris-
tics in order to use most of the SOC of the battery.
2858 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO. 12, DECEMBER 2009

Fig. 3. Power architecture with regulated hybrid power source.

B. Power Balance for a Regulated Architecture

There are multiple options, depending on the objectives of Fig. 4. PMAD box.
the application, to control the power flow between the sources
and the load. A fuzzy logic algorithm is proposed in [14] to
tections, and a control board. The dimensions of the box are 147
maximize the global energy efficiency. A multiobjective control
× 276 × 438 mm. The PMAD box is in charge of distributing
algorithm is presented in [15] for the power converter based on
the power from the sources to the loads and of managing the
a state diagram that regulates the power flow between the fuel
power flow and protections. The power flow and protections
cell, the battery, and the load. In this case, the objective is to
management comprises the following:
minimize the energy supplied by the battery in order to reduce
1) the control of the wake-up/shut-down sequences;
its size.
2) the protection of the different components;
The power balance can be achieved by regulating the output
3) the interface with the pilot through the instrumentation
voltage of a converter (SBC) connected in series with the fuel
panel;
cell to the distribution bus (see Fig. 3). The output voltage of this
4) the regulation of the power balance between the fuel cell
converter sets the bus voltage within the limits of its operating
and the battery;
range.
5) the command of the propulsion motor by means of the
For a given bus voltage, the SBC allows the fuel cell to work
throttle input and the configuration and state of the elec-
at a lower voltage, i.e., at a higher power. In this way, the battery
trical system.
works at a lower power level and lasts longer. Additionally, the
The regulation of the power balance between the sources can
battery is discharged at a lower current, which normally results
be enabled or disabled to obtain the regulated and unregulated
in a higher useful capacity.
architectures, respectively. To disable this regulation, the SBC
The output voltage of the converter is regulated through a
is disabled, and the fuel cell current would flow through the
control loop in such a way that, for a given load, the power
bypassing diode (see Fig. 3). When using the regulated architec-
delivered by the fuel cell, and so its current, is maximized.
ture, the SBC is only enabled when both sources are connected
to the bus.
IV. SYSTEM DESCRIPTION
Both proposed architectures were implemented in the Boeing B. PMAD Control Board
Fuel Cell Demonstrator Airplane within the power management
The PMAD box is controlled by an internal control board. The
and distribution box (PMAD).
heart of the PMAD control board is an field programmable gate
array (FPGA) that implements the digital control loops that are
A. PMAD Functionalities
in charge of the SBC and throttle regulations, and the controls
This box (see Fig. 4) is the integrative element of the airplane that are in charge of the sequences, protections, and interface
electrical system, and is mainly composed of the dc–dc convert- with the pilot. The control board (see Fig. 5) comprises sensing
ers of the SBC (several modules in parallel), dc–dc converters and analog interface circuitry (sensors signal conditioning, dis-
to generate the low voltage buses (24 and 14 V), the power con- play signal conditioning, SBC command reference, and throttle
tactors of the architecture, electromagnetic interference (EMI) interface circuit), digital circuitry (analog-to-digital converters
filters, a filter for the inverter noise based on coils and a power (ADCs), digital-to-analog converters (DACs), FPGA, Flash and
capacitor, current and voltage sensors, distribution buses, pro- PROM memories, clocks, digital input/output (I/O) interface,

Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
BATALLER-PLANES et al.: POWER BALANCE OF A HYBRID POWER SOURCE IN A POWER PLANT FOR A SMALL PROPULSION AIRCRAFT
Fig. 5. PMAD control board.
and solid-state power controllers (SSPCs) state interface), and
power circuitry (power supplies conditioning, contactors com-
mands circuit and SSPCs). The FPGA receives the analog and
digital information from the sensors (current and voltage mea-
surements) and the pilot (switches and throttle lever position),
processes them, and commands the power contactors, the solid-
state power controllers, the LEDs and the digital displays on the
instrumentation panel, the enabling and command of the SBC,
and the torque request to the propulsion motor inverter.
C. Sequences and Protections
The start-up and shut-down procedures are controlled through
a state machine, which also protects the sources and the propul-
sion motor inverter.
A low-voltage, low-power battery feeds the PMAD control
board, and afterwards, allows the start of the Li ion battery by
powering its internal controller. The battery is connected to the
distribution bus through a classical soft start (see Fig. 2), gradu-
ally charging the capacitors connected to the bus (approximately
8 mF). Once the capacitors are charged, the battery is connected
to the bus through a power contactor. When the bus is energized,
the fuel cell balance of plant gets powered and allows the start of
the fuel cell system. The fuel cell stacks are directly connected
to the bus through an additional power contactor.
The state machine monitors the electrical variables to detect
electrical faults and uses the same power contactors to protect the
sources and the inverter against overcurrents, negative currents
in the sources, fuel cell current steps, undervoltages, overvolt-
ages, and regenerative operation of the motor. The low-power
lines are protected against overcurrents mainly by SSPCs.
D. Control Loops
The output voltage of the SBC is regulated through a control
loop that maximizes the fuel cell power, and therefore, the fuel
cell current in such a way that the required battery capacity is
minimized.
In order to determine the desired fuel cell and battery currents,
the load current is compared to the maximum fuel cell current
(see Fig. 6). If the load current is lower than the maximum
fuel cell current, the hybrid system is forced to work with the
fuel cell supplying all the demanded current and the battery
supplying no current. Otherwise, the hybrid system is forced to
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
2859
work with the fuel cell delivering its maximum current and the
battery delivering the rest of the load current.
The design of the control loops was based on the frequency
responses obtained through the ac analysis performed by the
simulator of the system under different operating conditions.
Since the dynamic behavior of the main components (battery,
fuel cell, dc–dc converters, and load) was accounted in their
models the simulator became a powerful tool for the design of
the loops. An example of the dc–dc converter reference voltage
(SBC_CMD) to the battery current transfer function used for
the design of the current control loop is shown in Fig. 7.
The control loop was first tested using the PEM fuel cell,
with a laboratory power supply replacing the battery and a pro-
grammable load unit replacing the propulsion motor and in-
verter. In this test, the fuel cell works at 20 A if the load is equal
or higher than 20 A. Otherwise, the control makes the fuel cell
supply the demanded current.
At the beginning, the load current is lower than 20 A, there-
fore, the SBC output voltage is set so the fuel cell delivers all
the current and the laboratory power supply delivers no current.
Afterwards, a positive current step is applied and the control
increases the SBC output voltage to make the fuel cell deliver
20 A and the laboratory power supply deliver the rest of the load
current (see Fig. 8).
Once the battery discharges down to the point where the min-
imum output voltage of the SBC is reached, the bus voltage is
fixed by the fuel cell. Since the battery continues discharging, it
is not longer able to supply the difference between the continu-
ous loads power and the maximum fuel cell power, so the fuel
cell would have to operate over its capabilities. To avoid this, a
control law that decreases the main load power (the propulsion
motor) was implemented in the throttle control of the demon-
strator airplane.
The throttle control sends a proportional torque command to
the propulsion motor inverter taking into account the throttle
lever position and the state of the system. If the throttle demand
implies the propulsion motor power being higher than the power
allocated in the sources, the control automatically reduces the
torque command to reach a balance between electrical sources
and electrical loads. This is done by means of two control loops
that compare the measured sources currents with their maximum
allowable currents, respectively.
Additionally, the throttle control processes the torque com-
mand through a slew ramp limiter in order to obtain smooth
power change rates. Since fuel cells have relatively slow dy-
namics [16], [17], or in other words, fast transient loads can
have a negative impact on the stacks [18], the slew ramp limiter
avoids exceeding the maximum allowable rate of power change
per unit time.
The SBC control maximizes the fuel cell current and the
throttle control can decrease both the fuel cell or the battery
currents when reducing the motor torque command and so the
motor power. Since, these controls can influence the sources
currents, there is a link between the SBC control and the fuel
cell current control loop of the throttle control (see Fig. 6),
so both cannot work at the same time. A logic state machine
decides when each one should operate. The throttle limitation
2860
Fig. 6. Control loops.
Fig. 7. Simulated SBC control to battery current frequency response.
would start operating once the SBC output voltage decreases
down to its minimum while the load is demanding a constant
power.
V. SIMULATIONS AND TESTING
Both architectures were simulated prior to implementing
them in the electrical system. The model for the hybrid power
system was developed in SIMPLORER and the model for the
control was developed in Simulink (MATLAB) to account for
its digital hardware implementation effects (resolution and fi-
nite arithmetic). The cosimulation of the whole system, power
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO. 12, DECEMBER 2009
architecture and control, allowed the evaluation of the system in
all different operative conditions and faults. The model was also
useful to fine-tune the controls and protections before setting
up the tests bench. The fuel cell and the battery were modeled
based on experimental data. The polarization curve was used for
the fuel cell steady behavior and results of small-signal source
impedance tests were used to fit the fuel cell dynamic behavior.
Fig. 9 represents a typical normalized polarization curve of a
PEM fuel cell [19] where it can be seen that the output voltage
from open circuit to full load can vary in almost a 2:1 ratio.
The model of the battery to perform the simulations is based
on [13] that accounts for the variation of the output voltage as
BATALLER-PLANES et al.: POWER BALANCE OF A HYBRID POWER SOURCE IN A POWER PLANT FOR A SMALL PROPULSION AIRCRAFT
Fig. 8. Validation of SBC control loop (CH1 is the bus voltage, CH2 is the
SBC output voltage, CH3 is the power supply current, CH4 is the fuel cell
current).
Fig. 9. Typical normalized fuel cell polarization curve.
a function of the state of charge and the effect of the rate of
discharge. The discharge curves of the battery cells at different
constant currents used to build the battery model are in [7].
The models of all individual components (fuel cell, battery,
dc–dc converters, motor, SSPCs, remote controlled contactors,
etc.) were developed and validated prior to their simulation of the
global system. These models where used to design the control
loops implemented in the FPGA.
In order to validate the system model, the simulation results
were cross-checked with test results of the take off and climb of
a typical flight mission, which provide the fastest transients in
the system. The hybrid power source only works during take off
and climb (the first 7 min approximately), that are the critical
parts of the flight mission, when a high power level is demanded.
In cruise, the battery is disconnected and the fuel cell supplies
all the energy to the horizontal fly.
After the implementation, both architectures were tested by
reproducing the power profile of the flight mission. All electrical
parameters were recorded along with the PMAD temperatures.
The tests were done statically either by replacing the propeller
of the demonstrator by a controllable hydraulic brake or using
the propeller with the aircraft braked.
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
2861
Fig. 10. Measured power values with unregulated hybrid source during take
off.
Fig. 11. Measured currents with unregulated hybrid source during take off.
In the first case, the measurements were done by voltage and
Hall effect current probes, were conditioned by acquisition cards
and recorded in a computer. Data were acquired at 0.17 Hz.
In the second case, the acquisitions were taken in the PMAD
box. Measurements were done by internal Hall effect current
and voltage probes, and were conditioned and recorded by the
PMAD control board. A Flash memory controlled by the FPGA
of the PMAD control board was used to record the data. Data
were acquired at 1 Hz.
In both cases, the bus, fuel cell, and battery voltages, and the
motor inverter, fuel cell, and battery currents were acquired. On
the other hand, the motor, fuel cell, and battery power values
were calculated. The motor calculated power values correspond
to the power at the input of the inverter.
A. Unregulated Architecture
Fig. 10 shows the motor, fuel cell, and battery power values,
with a natural power balance between the battery and the fuel
cell. The motor power is approximately 40 kW. At the beginning,
as the battery is fully charged, it supplies a little bit more than
half of the power and the fuel cell supplies the rest of the power.
The power supplied by the fuel cell increases with time due to
the battery discharge. Fig. 11 shows the motor, fuel cell and
battery currents. Fig. 12 shows the bus voltage, which coincides
with the fuel cell and battery voltages. The fuel cell current and
2862
Fig. 12. Measured voltage with unregulated hybrid source during take off.
Fig. 13. Comparison of simulation and test results for power values for the
unregulated architecture during take off and climb.
Fig. 14. Comparison of simulation and test results for battery discharge for
the unregulated architecture during take off and climb.
power increase, and the bus voltage decreases as the battery
discharges.
The motor torque request was abruptly reduced at nearly
8 min, thus the motor power and current decreased conse-
quently. Large ripples observed in the measurements shown in
Figs. 10–12 are due to the operation of the controllable hydraulic
brake that replaced the airplane’s propeller during the test bench
campaign.
Figs. 13 and 14 show the comparison between simulated and
measured power values and battery discharge, respectively. The
simulation results are in close agreement with the test results.
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO. 12, DECEMBER 2009
Fig. 15. Measured power values with regulated hybrid source during take off.
Fig. 16. Measured currents with regulated hybrid source during take off.
B. Regulated Architecture
Fig. 15 shows the motor, fuel cell, and battery power values.
The SBC control successfully increases the power supplied by
the fuel cell, making it work near or at full power. The fuel cell
was forced to operate at 22 kW gross power. Fig. 16 shows the
motor, fuel cell, and battery currents. The SBC control made
the fuel cell work at high current, reaching 100 A during 3 and
a half minutes of the take off and climb. Fig. 17 shows the bus,
fuel cell, and battery voltages on the left ordinate, and the SBC
output voltage on the right ordinate. The difference between
the fuel cell voltage and the bus voltage is the voltage boosted
by the SBC. The battery and bus voltages coincide. The output
voltage of the SBC is maximum during the first 3 and a half
minutes and decreases afterwards. At the beginning, the SBC
attempts to make the fuel cell operate at the programmed current
(100 A), but it is limited by its maximum output voltage (26 V).
Therefore, by boosting its maximum voltage it makes to operate
the fuel cell 26 V below the bus voltage. However, it is not
enough to set the fuel cell voltage to the voltage corresponding
to the programmed current (216 V). As the battery discharges
with time, the difference between the fuel cell desired voltage
and the bus voltage decreases, and therefore, the output voltage
of the SBC decreases. Once the battery is discharged enough, the
SBC is able to force the fuel cell to operate at 100 A. After that,
the output voltage of the SBC decreases with time, following
the further battery discharge.
BATALLER-PLANES et al.: POWER BALANCE OF A HYBRID POWER SOURCE IN A POWER PLANT FOR A SMALL PROPULSION AIRCRAFT
Fig. 17. Measured voltages with regulated hybrid source during take off.
Fig. 18. Comparison of simulation and test results for power values for the
regulated architecture during take off and climb.
Fig. 19. Comparison of simulation and test results for battery discharge for
the regulated architecture during take off and climb.
The motor torque request was abruptly reduced at nearly 8
min, thus the motor power and current decreased consequently.
The ripples observed in measurements of Figs. 15–17, on the
contrary to those of Figs. 10–12, are low because, in this case, the
test was done using the propeller of the aircraft as the mechanical
load instead of the controllable hydraulic brake.
Figs. 18 and 19 show the comparison between simulated and
measured power values and battery discharge, respectively. The
simulation results are in close agreement with the test results.
The models can compare the battery savings for the regulated
versus the unregulated architecture. For a typical take off and
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
2863
climb (40 kW motor power for 7 min) a regulation forcing the
fuel cell to operate at 22 kW of gross power could allow a battery
capacity saving of 440 W·h or 2 A·h; and a regulation forcing
the same fuel cell to operate at 23 kW of gross power could
allow a battery capacity saving of 540 W·h or 2.5 A·h.
Taking a typical energy density of 70 W·h/kg for Li ion bat-
teries, the capacity saving would correspond to a weight saving
of 6.3 kg when forcing the fuel cell at 22 kW and 7.7 kg when
forcing the fuel cell at 23 kW. On the other hand, taking a typical
handling power density of converters of 0.5 kW/kg, including
heat sinks, the additional weight due to the converters would
correspond to 6 kg. The additional hydrogen mass consumed by
the fuel cell is no higher than 50 g in both cases. Therefore, for
this particular system and this particular application, the total
weight saving would correspond to approximately 0.3 kg when
forcing the fuel cell at 22 kW and 1.7 kg when forcing the
fuel cell at 23 kW. The weight savings would increase with the
duration of the operation.
The models can also compare the time the hybrid power
source can supply a constant power while avoiding the fuel cell
power exceeding a programmed limit in the throttle conditioning
control.
In the unregulated architecture, the natural distribution makes
the fuel cell increase its power contribution within time, fol-
lowing the battery discharge. Once the fuel cell reaches the
programmed power limit, the fuel cell power would carry on
increasing while the battery power would carry on decreasing.
However, at this point, the throttle control starts decreasing the
torque command sent to the inverter motor, thus succeeding to
decrease the power demanded by the motor.
In the regulated architecture, the fuel cell is forced to operate
at its power limit, and, once the battery gets discharged enough,
the fuel cell power would increase over its limit while the bat-
tery power would continue decreasing. Once again, the throttle
control avoids the fuel cell going beyond the programmed limit.
However, due to the battery savings, this time is reached later
than in the unregulated architecture.
Therefore, by using a regulated architecture, the hybrid power
source will supply a constant power demand for longer.
The simulations results show that for a typical takeoff and
climb power rate (40 kW motor power) and when the fuel cell
power limit is programmed to be 22 kW gross power, the reg-
ulated hybrid power source can supply the constant power for
one more minute (from 8 and a half minutes to 9 and a half
minutes). For the same typical take off and climb, when the fuel
cell power limit is programmed to be 23 kW gross power, the
regulated hybrid power source can supply the constant power
for 1 and a half more minutes (from 9 and a half minutes to 11
minutes).
VI. QUALITATIVE EVALUATION OF BOTH ARCHITECTURES
In this section, the advantages and disadvantages of both
architectures are discussed according to different criteria that
might be crucial depending on the particular application.
1) Complexity: The unregulated architecture is simpler. It
does not require any additional wide input voltage-wide
2864
output voltage dc–dc modules and the implementation of a
current loop to control them under all different conditions.
2) Reliability: Due to its simplicity, the unregulated archi-
tecture is more reliable in terms of components failure.
Although protections are necessary to avoid damaging the
sources with heavy load conditions, that could be caused
by any fault, those are also necessary in the case of the
regulated architecture.
3) Efficiency: The efficiency of the unregulated architecture
is higher than that of the regulated architecture since the
power of the fuel cell is not processed by any dc–dc con-
verter.
4) Cooling: Related to the efficiency, the cooling require-
ments for the unregulated architecture are lower than those
of the regulated architecture because of the lower losses.
5) Impact on battery size: The regulated architecture allows
minimizing the size of the battery that supports the fuel
cell during the take off and climb because it forces the fuel
cell to supply all the load current up to its programmed
limit. In this way, the battery capacity is minimized.
6) Weight: Although the weight of the regulated architecture
is penalized by the dc–dc converters, the benefit of having
a smaller battery can overcome this weight penalty. As
a consequence, the regulated architecture system can be
lighter.
7) Duration: Since the regulated architecture saves some bat-
tery energy, the hybrid power source is able to supply a
certain power level for longer without exceeding the fuel
cell power capabilities. The total energy of the hybrid
power source is then enhanced.
8) Robustness: Since in the unregulated architecture the load
sharing between the fuel cell and the battery is strongly
dependent on their parameters (state of charge, inter-
nal impedance, capacity, OCV, temperature, etc.), its be-
havior is dependent on the nominal tolerances, variation
with temperature and age, etc. In the case of the regu-
lated architecture, the SBC and the control loop make
the behavior of the system almost insensitive to those
variations.
VII. CONCLUSION
Two different architectures can be used in a hybrid power
source that comprises a PEM fuel cell and a Li ion bat-
tery. The unregulated architecture implies a natural balance
between the sources. The regulated architecture controls the
power supplied by each source by means of a serial boost
converter connected to the distribution bus in series with
the fuel cell. The output voltage of the converter is regu-
lated through a control loop that maximizes the fuel cell
power.
Both proposed architectures offer advantages and disadvan-
tages. The unregulated architecture is simpler, more reliable,
more efficient, and needs a simpler cooling. The regulated ar-
chitecture discharges less the battery, allowing the use of smaller
batteries, and, thus, it offers a weight reduction possibility; an-
other consequence of the battery energy savings is that the hybrid
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO. 12, DECEMBER 2009
power source can supply a certain power level for longer with-
out exceeding the fuel cell power capabilities, maximizing, this
way, the total energy of the hybrid power source; moreover, the
regulated hybrid power source has a more reproducible behavior
because it has a lower dependency of the external conditions,
such as temperature, ageing, or operation.
Both concepts have been simulated, implemented, and thor-
oughly tested in the Fuel Cell Demonstrator Airplane, the first
manned prototype to prove straight level flight with fuel cell
power. The unregulated architecture was more appropriate for
the demonstrator mainly because in this particular application
the battery energy saving, and so the possible weight reduction,
was not significant enough. The increase of the takeoff and climb
time offered by the regulated architecture was neither required
because 5 min were sufficient to reach the cruise altitude. On the
other hand, the safety and reliability of the system were crucial.
The unregulated configuration was inherently more reliable due
to its simplicity and did not force the fuel cell to operate at its
maximum power output, thus reducing the risk of a fuel cell
trip off during take off and climb, and decreasing the amount of
water and hydrogen to be carried onboard during the test flights.
In addition, a bigger battery could provide the full power for
a shorter take off and climb in the case of a fuel cell trip off,
allowing the aircraft to climb up to a safe altitude.
REFERENCES
[1] S. Williamson, M. Lukic, and A. Emadi, “Comprehensive drive train
efficiency analysis of hybrid electric and fuel cell vehicles based on motor-
controller efficiency modeling,” IEEE Trans. Power Electron., vol. 21,
no. 3, pp. 730–740, May 2006.
[2] A. Emadi, S. S. Williamson, and A. Khaligh, “Power electronics intensive
solutions for advanced electric, hybrid electric, and fuel cell vehicular
power systems,” IEEE Trans. Power Electron., vol. 21, no. 3, pp. 567–
577, May 2006.
[3] T. Markel and A. Pesaran. (2006, Dec.) Energy storage for fuel
cell hybrid vehicles. Energy Storage Requirement Analysis for
Advanced Vehicles (Fuel Cell, Mild Hybrid, and Plug-in Hy-
brid), NREL/CD-540-40658 [Online]. Available/: http://www.nrel.gov/
vehiclesandfuels/energystorage/pdfs/40658.pdf
[4] D. Daggett, S. Eelman, and G. Kristiansson, “Fuel cell APU for commer-
cial aircraft,” presented at the ’03 Fuel Cell Semin. Abstracts, Miami, FL,
2003.
[5] K. Rajashekara, “Hybrid fuel-cell strategies for clean power generation,”
IEEE Trans. Ind. Appl., vol. 41, no. 3, pp. 682–689, May/Jun. 2005.
[6] N. Lapena-Rey, J. Mosquera, E. Bataller, and F. Orti, “The boeing fuel
cell demonstrator airplane,” presented at the SAE ’07 Aerotech Congr.
Exhib., Los Angeles, CA, 2007 (publication number: 2007-01-3906).
[7] N. Lapena-Rey, J. Mosquera, E. Bataller, F. Orti, C. Dudfield, and A. Or-
sillo, “Environmentally friendly power sources for aerospace applica-
tions,” in Proc. 10th Grove Fuel Cell Symp., 2007, pp. 353–362 (publica-
tion number: POWER-D-07-02171).
[8] L. Gao, Z. Jiang, and R. A. Dougal, “Evaluation of active hybrid fuel
cell/battery power sources,” IEEE Trans. Aerosp. Electron. Syst., vol. 41,
no. 1, pp. 346–355, Jan. 2005.
[9] M. Marchesoni and C. Vacca, “New DC–DC converter for energy storage
system interfacing in fuel cell hybrid electric vehicles,” IEEE Trans.
Power Electron., vol. 22, no. 1, pp. 301–308, Jan. 2007.
[10] S. K. Mazumder, R. K. Burra, and K. Acharya, “A ripple-mitigating and
energy-efficient fuel cell power-conditioning system,” IEEE Trans. Power
Electron., vol. 22, no. 4, pp. 1437–1452, Jul. 2007.
[11] L. Changrong and L. Jih-Sheng, “Low frequency current ripple reduction
technique with active control in a fuel cell power system with inverter
load,” IEEE Trans. Power Electron., vol. 22, no. 4, pp. 1429–1436, Jul.
2007.
[12] X. Haiping, L. Kong, and X. Wen, “Fuel cell power system and high
power DC–DC converter,” IEEE Trans. Power Electron., vol. 19, no. 5,
pp. 1250–1255, Sep. 2004.
BATALLER-PLANES et al.: POWER BALANCE OF A HYBRID POWER SOURCE IN A POWER PLANT FOR A SMALL PROPULSION AIRCRAFT
[13] S. C. Hageman, “Simple PSpice models let you simulate common battery
types,” Electron. Design News (EDN), vol. 38, pp. 117–132, Oct. 2009.
[14] A. A. Ferreira, J. A. Pomilio, G. Spiazzi, and L. de Araujo Silva, “Energy
management fuzzy logic supervisory for electric vehicle power supplies
system,” IEEE Trans. Power Electron., vol. 23, no. 1, pp. 107–115, Jan.
2008.
[15] Z. Jiang, L. Gao, and R. A. Dougal, “Flexible multiobjective control of
power converter in active hybrid fuel cell/battery power sources,” IEEE
Trans. Power Electron., vol. 20, no. 1, pp. 244–253, Jan. 2005.
[16] W. Friede, S. Rael, and B. Davat, “Mathematical model and characteri-
zation of the transient behavior of a PEM fuel cell,” IEEE Trans. Power
Electron., vol. 19, no. 5, pp. 1234–1241, Sep. 2004.
[17] G. Fontes, C. Turpin, S. Astier, and T. A. Meynard, “Interactions between
fuel cells and power converters: Influence of current harmonics on a fuel
cell stack,” IEEE Trans. Power Electron., vol. 22, no. 2, pp. 670–678,
Mar. 2007.
[18] S. K. Mazumder, K. Acharya, C. L. Haynes, R. Williams Jr., M. R. von
Spakovsky, D. J. Nelson, D. F. Rancruel, J. Hartvigsen, and R. S. Gemmen,
“Solid-oxide-fuel-cell performance and durability: resolution of the effects
of power-conditioning systems and application loads,” IEEE Trans. Power
Electron., vol. 19, no. 5, pp. 1263–1278, Sep. 2004.
[19] P. P. J. Wingelaar, J. L. Duarte, and M. A. M. Hendrix, “Computer con-
trolled linear regulator for characterization of polymer electrolyte mem-
brane fuel cells (PEMFC),” in Proc. IEEE Int. Symp. Ind. Electron. Conf.,
2004, pp. 821–826.
Elena Bataller-Planes received the M.Sc. degree
in industrial electronics from the École Superieure
d’Électricité (SUPELEC), Gif-sur-Yvette, France,
in 2001, the M.Sc. degree in industrial engineer-
ing from the Universidad Politécnica de Valencia,
Valencia, Spain, in 2001, and the M.Sc. degree in in-
dustrial electronics from the Universidad Politécnica
de Madrid, Madrid, Spain, in 2009.
At the Boeing Company, she was responsible for
the electrical system of the Boeing Fuel Cell Demon-
strator Airplane, which was the first manned fuel cell
aircraft in aviation history and proved a level manned flight with a fuel cell
system as the only source of power. Her research interests include power man-
agement, system design and integration, and electrical vehicles.
Nieves Lapeña-Rey received the M.Sc. degree in
chemistry from the Universidad Complutense de
Madrid, Spain, in 1993, the M.Sc. degree in chem-
istry of materials from the University of Kent, Eng-
land, U.K., in 1994, and the Ph.D. degree in chemical
engineering from the Imperial College of Science,
Technology and Medicine, London, U.K.
She was the Leader of the Fuel Cell Demonstrator
Airplane project. Since January 2007, she has been
the Team Leader and the Technical Manager for the
Environmentally Progressive Air Transport Depart-
ment, Boeing Research and Technology Europe, Boeing Company, Madrid,
Spain. Her current research interests include novel applications of fuel cell
technology to aviation and reduction of hazardous heavy metals in aeronautical
parts.
Jonay Mosquera received the M.S. degree in aero-
nautical engineering from the Technical University
Munich (TUM), Munich, Germany, in 2003.
He worked at he Boeing Company as the Aero-
nautical Engineering of the Boeing Fuel Cell Demon-
strator Airplane.
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
2865
Fortunato Ortı́ received the M.S. degree in aero-
nautical engineering from Polytechnic University of
Madrid, Madrid, Spain, in 1968, and the Ph.D. de-
gree in aeronautics and astronautics from Stanford
University, Stanford, CA, in 1971.
He was with IBM from 1974 to 1994. From 1994
to 1998, he was on the faculty of the University Al-
fonso X El Sabio, Madrid, Spain, and was engaged
in teaching of fluid mechanics. He is currently a Se-
nior Scientific Advisor with the Boeing Research and
Technology Europe, Madrid, Spain. His current re-
search interest includes environment.
Jesús Á. Oliver (M’05) was born in Toledo, Spain, in
1972. He received the M.S. and Ph.D. degrees in elec-
trical engineering from the Universidad Politécnica
de Madrid, Madrid, Spain, in 1996 and 2007
respectively.
Since 2001, he has been an Assistant Professor of
electrical engineering at the Universidad Politécnica
de Madrid. He has authored or coauthored over 40
technical papers and holds one patent. He has been
actively involved in over 25 R&D projects for com-
panies in Europe, the U.S., and Australia. His current
research interests include modeling and control of power electronics systems,
fuel-cell-powered systems, and energy-efficient design.
Óscar Garcı́a (M’98) was born in Madrid, Spain,
in 1968. He received the Master’s and Doctoral de-
grees in electronic engineering from the Universidad
Politécnica de Madrid (UPM), Madrid, in 1992 and
1999, respectively.
He is an Associate Professor at the UPM. He has
been involved in more than 40 research projects, holds
five patents, and he has authored or coauthored more
than 100 papers in IEEE conferences and journals.
Félix Moreno received the B.S., M.S., and Ph.D.
degrees in telecommunication engineering from the
Universidad Politécnica de Madrid (UPM), Madrid,
Spain, in 1980, 1986, and 1993, respectively.
He is currently an Assistant Professor in elec-
tronics and microelectronics at the Escuela Técnica
Superior de Ingenieros Industriales (ETSII), UPM.
His current research interests include embedded sys-
tems (and/or intelligence) design, systems on chip,
and evolvable hardware applied to digital signal pro-
cessing and digital control systems. He participated
in different research projects related to those fields of interest.
Jorge Portilla, photograph and biography not available at the time of
publication.
Yago Torroja was born in Madrid, Spain, in 1963. He received the Electronics
Eng. and Ph.D. degrees from the Universidad Politécnica de Madrid (UPM),
Madrid, in 1990 and 2000, respectively.
He is an Associate Professor of Electronics at the Automatic Control, Elec-
tronics, and Informatics Department of UPM since 2002 and part of the Research
Centre of Industrial Electronics (CEI-UPM). His research interests are focused
to wireless sensor networks, embedded systems, electronic interfaces for musi-
cal expression, and low power design and estimation.
2866 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 24, NO. 12, DECEMBER 2009
Miroslav Vasić received the M.S. degree from the
School of Electrical Engineering, University of Bel-
grade, Belgrade, Serbia, in 2005, and the Master’s de-
gree in industrial electronics in 2008 from the Univer-
sidad Politécnica de Madrid (UPM), Madrid, Spain,
where he is currently working toward the Ph.D. de-
gree. His current research interests include switching-
mode power supplies, power architectures, and digital
control applied to power electronics.
Santa Concepcion Huerta received the B.Eng. de-
gree in electronic engineering from the Instituto Tec-
nologico de Celaya, Mexico City, Mexico, in 2003.
She is currently working toward the Ph.D. degree
from the Universidad Politécnica de Madrid (UPM),
Madrid, Spain.
Her current research interests include digital con-
trol of switching-mode power supplies, power elec-
tronics, and dynamic voltage scaling techniques.
Matı́as Trocki received the M.S. degree in industrial engineering from the
University of Madrid, Madrid, Spain, in 2008.
He has been associated with the Centro de Electrónica Industrial, Universidad
Politécnica de Madrid (UPM), Madrid.
Authorized licensed use limited to: Ericsson. Downloaded on February 14,2020 at 00:23:12 UTC from IEEE Xplore. Restrictions apply.
Pablo Zumel (M’06) received the M.S. and Ph.D.
degrees both from the Universidad Politécnica de
Madrid (PUM), Madrid, Spain, in 1999 and 2005,
respectively, and the M.S. degree from the Ecole Cen-
trale Paris, Paris, France, in 2000.
In 2003, he joined the Departamento de Tecnologı́a
Electrónica, Universidad Carlos III de Madrid,
Madrid, where he is currently an Assistant Professor.
His current research interests include power electron-
ics for renewable energy sources and digital control.
José A. Cobos (M’92) received the Master’s and
Doctoral degrees in electrical engineering from the
Universidad Politécnica de Madrid (UPM), Madrid,
Spain, in 1989 and 1994, respectively.
Since 2001, he has been a Professor at UPM. He
was engaged in the field of power supply systems
for telecom, aerospace, automotive and medical ap-
plications. His current research interests include low
output voltage, magnetic components, piezoelectric
transformers, transcutaneous energy transfer, and dy-
namic power management. He has authored or coau-
thored over 150 technical papers and holds three patents. He has been actively
involved in over 40 R&D projects for companies in Europe, the U.S., and
Australia.
Dr. Cobos is an AdCom Member of the IEEE Power Electronics Society
(PELS), and the Chair of the Technical Committee on DC Power Systems. He is
an Associate Editor of the IEEE-PELS letters and the IEEE TRANSACTIONS ON
POWER ELECTRONICS. He received several awards, including the UPM Research
and Development Award for faculty less than 35 years of age, and the Richard
Bass Outstanding Young Power Electronics Award of the IEEE (year 2000).
He is the Vice Dean for Research and Doctoral studies of the Escuela Técnica
Superior de (ETS) Ingenieros Industriales of the UPM.