To design a practical fuel cell system, a specific application was chosen to set target power,
energy, weight, and size requirements. The chosen application where a fuel cell device would
provide several unique advantages is a mid-sized UAV (total weight < 55 lbs (25 kg), 6-10 ft
(1.8-3 m) wingspan); these are mainly used in surveillance applications that require long
endurance, but do not necessarily require high power densities for maneuverability. To further
refine the target system requirements, the practical power, energy, and weight limitations for existing mid-sized UAVs were researched in literature and through discussions with experts in the industry to determine the practical design constraints for the fuel cell system focused on here.
Presently, UAVs are either powered using batteries or internal combustion engines (ICEs),
which offer unique advantages but come with different drawbacks. ICEs are used to power most
UAVs with long-term surveillance missions because of the extended durability afforded when
fueled with a hydrocarbon fuel. Some of these UAVs can fly for over 24 hours between refueling
operations. However, UAVs powered with ICE engines are often extremely noisy and therefore
have to fly at high altitudes to ensure that those being surveilled cannot hear it approaching.
These high altitude flights require expensive and complex optical equipment to properly view
their targets from such far distances, and it is highly desirable to reduce the noise output of the
propulsion system so that the flight altitude can be reduced and more cost-effective optics can be
used. Additionally, most ICE engines require significant maintenance after each flight, typically
with a major overhaul at 250 hours and a complete engine rebuild or replacement after 500
hours. ICEs additionally suffer from premature shut downs that may occur mid-flight, which are
142
Batteries are the other primary propulsion technology for UAVs. As batteries are much
quieter than ICE engines, they allow for lower altitude flights and cheaper, less complicated
optics. However, due to their low specific energy, they suffer from very short flight times that
are typically no longer than 1.5 hours, such as with the RQ-11B Raven developed by
AeroVironment [164]. As a result, UAVs powered with batteries can only be used for very short
range reconnaissance missions. To watch a target for extended periods of time using a battery
powered UAV, multiple UAVs are used for a single mission. At least three UAVs are in constant
use for a long endurance mission, where one UAV is over the target, one is returning, and the
other has just been launched towards the target. This greatly complicates and adds to the expense
of each mission as several operators are required for each UAV used.
A fuel cell system has the potential to provide the advantages both batteries and ICEs offer,
but with very few drawbacks. Because fuel cells generate power without any moving parts and
can be fueled with energy-dense hydrocarbon fuels, they offer quiet operation, long flight times,
and less maintenance than current battery and ICE technology. A fuel cell propulsion system that
reliably meets the target energy, power, weight, and durability requirements would provide
substantial advantages for military UAV missions. To date, however, few systems have been
developed that meet the target requirements that would enable widespread adoption of fuel cells
as the primary propulsion system for UAVs.
To determine the target requirements, discussions with engineers and program managers
within the U.S. military that have first-hand experience designing and operating UAVs were
conducted to determine the energy and power densities for existing mid-sized UAV power
plants. While the particular sources cannot be listed for confidentiality reasons, those engaged
143
determine the primary power, energy, and weight requirements for existing UAVs, as well as
what is desirable for future designs. It was quickly realized that the primary problems facing
small military UAVs used for surveillance is that those powered with ICEs are too noisy and
require significant maintenance, and those powered with batteries have unacceptably short flight
times. There is significant interest in technologies that can resolve both of these issues and fuel
cells are a very viable candidate.
Many types of fuel cell systems have been proposed and demonstrated for use in UAVs
where a particularly in-depth discussion for fuel cell UAV developments through 2010 can be
found in [165]. A few of those systems are highlighted here. AeroVironment has demonstrated
several fuel cell powered UAV systems intended to be used by the military, such as the Hornet
[166] and Puma [167]. These systems utilize LT-PEM fuel cells that are capable of flights up to
seven hours for the Puma, and are fueled with hydrogen stored in a sodium borohydride (NaBH4)
pellets that, when mixed with water, generate on-board hydrogen gas. Similarly, Kim et al. also
demonstrated a UAV powered with a PEM fuel cell that utilized sodium borohydride for on-
board hydrogen fuel storage [168]. Demonstrated flight times in their research have not quite
reached two hours, substantially less than the Puma; however, the work done by Kim et al. is in
the public domain, which provides significant insight into the fuel cell systems development. The
Stalker XE UAV [169], constructed by Lockheed Martin, is powered with a SOFC developed by
Ultra Electronics AMI [170]. The Stalker XE is fueled with propane and is capable of 8+ hours
of flight time. Hydrogen Energy Systems (HES) develops commercial fuel cells for integration
into UAVs [171]. They have developed methanol, NaBH4, and pure hydrogen on-board fuel
144
With the exception of the SOFC-based power plant developed by Lockheed Martin, the fuel
cell systems developed for UAVs all suffer from the main drawback in that they do not operate
on a commonly available fuel that is used by the U.S. military. Fueling power plants with a
commonly available fuel is an essential requirement for the U.S. military, as using uncommon
fuels creates significant logistics problems when trying to bring different types of fuels to the
military theatre. For equipment that requires refueling, the U.S. military has a single fuel policy,
and strongly desires for all machines to run on the logistics fuel called JP-8, which is similar to
diesel. Use of propane is one of the few exceptions allowed, particularly for special operations
that involve UAVs. Because propane can be purchased almost anywhere in the world, it is often
favored over JP-8 for some missions, as it can be purchased locally without raising suspicion.
From discussions with military personnel, it is highly desirable to have a UAV propulsion plant
that is fueled with JP-8 or propane, has a long run time, and is extremely quiet.
Beyond the existing problems with mid-sized UAVs currently faced by the U.S. military, the
general technical requirements for UAV power plants were also determined. From the
discussions, military UAV operators desire a minimum of ten hour runtimes for extended flight
missions, and they require at least twenty missions before the power plant fails or needs to be
replaced, equating to a total lifetime of 200 hours. It was also learned that a UAV will often
experience a catastrophic crash during flight or while landing before its expected twenty mission
lifetime, making longer power plant lifetime requirements somewhat irrelevant. Nevertheless,
premature crashes are not always the case, and while military operators require a minimum of
200 hours, they ultimately desire lifetimes of 1000 hours.
Regarding the power, energy, and weight requirements, it was also learned that existing
145
a third of the total take-off weight, while another third of the weight is taken up by the fuel and
fuel tank, and the rest includes the airframe, motor, and payload. Additionally, power plants are
generally designed to produce at least 35 W/kg (power per mass of the entire aircraft) for fixed
wing aircraft, and 140 W/kg for helicopters and multirotor aircraft. Current power plants used for
mid-sized UAVs generate approximately 250 to 3500 W. Some of the military personnel also
provided specific power output, weight, and volumetric data for the power plants currently used.
For example, a 50 kg fixed wing aircraft presently uses a 3000 W engine with a volume of
approximately 30 L. Although the engine in this example produces 3000 W, a 50 kg fixed wing
plane could fly with only 1750 W of power assuming the described requirement of only 35
W/kg. Thus, the minimal power density of the engine and BoP equipment can be calculated as
58.3 W/liter. However, it was described that volumetric constraints are not as important as
weight constraints, providing some flexibility in power density metrics.
For selecting a target fuel cell system size, many of the discussions with UAV operators
indicated there is present need for power plants that provide approximately 250 W of net power
for propulsion. In setting the target weight and volume requirements, it was assumed that the
BoP equipment (electronic controller, air compressor, cooling fan, flow controllers, valves, and
electronic control unit) consume approximately 20% of the total power output. Additionally, to
be conservative, it was assumed that a full-sized fuel cell stack will generate a power density that
is about 10% less than at the single cell level. Therefore, the gross power of the fuel cell stack
should be designed assuming 30% will be lost, indicating that 360 W is needed to produce 250
W for the primary propulsion. With 250 W put towards the primary propulsion power plant, the
UAV should weigh no more than 7.2 kg, and the fuel cell system including BoP equipment
146
total volume of the fuel cell stack and BoP equipment should be no more than 4.3 L. The fuel
cell system discussed in the remainder of this Chapter was designed to produce 360 W of power
weighing no more than 2.4 kg within a 4.3 L volume.