UCD- ITS- RR- 04- 16
Design Considerations for a PEM Fuel Cell Powered Truck
APU
David Grupp, Matthew Forrest, Pippin Mader, CJ Brodrick, Marshall Miller, Harry Dwyer
Institute of Transportation Studies, University of California - Davis
ABSTRACT
In recent years interest has been growing in using fuel cell
powered auxiliary power units ( APUs) to reduce idling in
line- haul trucks. Demonstrations of this technology have
been constructed at universities and within industry, each
with its own advantages and disadvantages. Invariably, in
every design, tradeoffs need to be made and this has
resulted in a multitude of different APU solutions that
address different aspects of the problem.
This paper reviews some of the recent work related to fuel
cell APUs for large trucks. The paper also examines what
characteristics are important to consider in the design and
integration of a fuel cell APU and outlines the strategy and
methodology taken by the University of California Institute
of Transportation Studies in designing and building a
viable demonstration fuel cell APU.
INTRODUCTION
Currently large numbers of transport trucks idle during
times when they are not actively transporting a load.
Power demanded during these times has been called
“ hotel load” because it is the power necessary for the
comfort of the driver while at rest. Earlier studies have
indicated that trucks may idle anywhere from 6 to 16
hours out of the day and consume around a gallon of
diesel fuel every hour that the engine is idling. The
amount of fuel used and emissions produced during this
time of inactivity is not insignificant. [ 1][ 2][ 3][ 4]
Policy making bodies, understanding the adverse effects
of diesel emissions, noise pollution, and fuel consumption
have started to pass legislation in an effort to curtail truck
idling. At the local level many states as well as some
municipalities have instituted idling bans, with many more
considering anti- idling legislation. [ 5]
With idling reduction gaining more attention, the UCD ITS
has designed and built a fuel cell demonstration APU for a
Class 8 truck. The purpose of which was to demonstrate
fuel cell APU technology and compare it to existing
technology options.
RECENT FC APU DEMONSTRATIONS
Fuel cells are experiencing a period of renewed attention.
In the last ten years federal and private funding for
research into fuel cell technology and uses has been
steadily increasing. As a result many fuel cell APU
projects have been demonstrated.
Freightliner successfully demonstrated a PEM fuel cell
APU on a Century Class Freightliner truck. [ 8] The APU
incorporated two fuel cells that provided a total output of
1.4 kW of power. The system utilized a 52 gallon ( 197 L)
compressed hydrogen storage tank at 2,500 psi. The
system was able to provide 120 VAC power or 12 VDC
power, however it was not able to supply both at the same
time. Climate control was provided by a diesel fuel fired
heater and a 120 VAC air conditioner described as a
typical home, in- window air conditioning unit.
The authors examined the possible fuel and emissions
savings that would be possible if engine idling were to be
eliminated. The analysis assumed idling times of
between 1,818 and 2,424 hours per year. In their
calculations for economic payback period for a FC APU
they also assumed 1 gallon of diesel fuel was consumed
per hour on average for an idling engine running typical
accessories. Using these assumptions it was calculated
that between $ 3,127 and $ 4,169 was spent on engine
idling alone per year. The paper also made some
estimates as to the reduction in emissions that could be
realized by using a FC APU.
Delphi Automotive Systems in cooperation with BMW has
demonstrated a proof of concept SOFC APU. [ 9] This
demonstration was built to power an electrical air
conditioner and electrical loads in passenger vehicles.
While not specifically designed for a Class 8 truck
application the fuel cell APU shares many design
characteristics that would make it easily adaptable to the
truck market.
SOFC technology was chosen for the Delphi APU. Some
of the reasons cited include the ability of the SOFC to
utilize liquid hydrocarbon fuels, less expensive non- noble
catalyst, a simpler reformer design, and elimination of the
need for humidification or complex water management.
Diesel fuel is readily available and would not require a new
fueling infrastructure or a secondary fuel to be dispensed
at the station.
The Sacramento Municipal Utility District ( SMUD) built
and tested a fuel cell APU. [ 10] Their demonstration
project integrated a supercharged, water cooled, 5kW fuel
cell with a Class 8 truck. The system used 5.3 kg of
compressed hydrogen storage as the fuel source.
Special attention was given to durability and the ability to
operate in temperature extremes from - 40C to 50C. The
completed FC APU underwent track testing for
environmental tolerance including both high and low
temperatures and road dirt.
FC APU DESIGN CONSIDERATIONS
In order to better understand the market for FC APUs and
the features that would be required for a desirable FC APU
many sources of information where considered. The UCD
ITS conducted a survey of 365 truck drivers at locations
nationwide soliciting responses to questions involving
trucker idling behavior, usage of idling reduction
technologies, and preference for APU features. The
results from previ ous studies, measurements taken from a
stock test vehicle, and survey results were used to guide
development of the demonstration fuel cell APU.
CUSTOMER REQUIREMENTS
The purpose of a truck APU is to provide the user with
climate control and electrical power necessary in order to
eliminate the need to idle the main engine that is normally
used to provide these services. Beyond simply replacing
the services provided by the main engine, the APU must
provide a superior user experience. In the absence of
regulation, the adoption of APU technology will proceed
only if the APU design provides a better user experience
at a lower cost than what is currently available by idling
the main engine. It is very important to understand what
compromises the user is willing to make in terms of
performance and price in order to design an APU that is
useful and marketable.
Venturi and Martin [ 11] examined the APU market in three
regions; North America, Europe, and South America
( using Brazil as a representative country). They
concluded that the North American market was driven by
high truck comfort and estimated its size at some 10,000
units. The European and South American market, due to
shorter distances traveled and higher fuel prices, placed a
higher value on fuel efficiency. They also published
estimates for expected auxiliary loads from various truck
appliances that might be used on a long haul truck.
Lutsey and Broderick also did work to estimate the power
requirements of various appliances, and conducted
multiple surveys to understand the distribution of these
appliances and the use profiles among a representative
set of line- haul truck drivers. [ 14][ 15] For line- haul trucks
they estimated the size of the market at 100,000 trucks
annually plus retrofits. They also estimated target
parameters for a fuel cell powered APU that would be
acceptable for this market. [ 16]
Truck Auxiliary Power Unit
Peak Power 3 - 6 kW
Approximate APU Cost 4,000 - 8,000 $
Target System Cost 500 - 1,000 $ / kW
Target Weight 50 – 100 W / kg
Target Volume 30 – 50 W / l
Table 1 – Truck APU Market Parameter Targets [ 16]
Climate Control
In a stock truck, heating is provided by the idling engines
coolant system. Cooling is provided by an engine driven
refrigerant compressor. In each instance there is a
thermal / mechanical load associated with the heating or
cooling plus an electrical load associated with the fans to
circulate the temperature controlled air through the cab.
In comparison to anticipated electrical loads, climate
control loads are much larger and are an important
consideration when specifying the peak power capacity of
the APU.
An APU climate control system can be sized smaller than
the peak capacity for the stock system because the APU
system is designed to provide climate control over a
smaller set of operation conditions than the stock system.
The stock HVAC system is designed to provide climate
control for the entire cab and sleeper. The APU system
needs to be sized only for the sleeper compartment which
is much smaller. The stock system must be sized to
bring the entire cab area to a desired temperature in a
short amount of time, a process called “ pull- down”. The
APU will primarily be operated after the cab has already
been brought to the desired temperature, usually after the
operating truck has reached it’s destination for the night.
If needed the truck could be idled for a short period of time
to provide pull- down without negating the advantage
provided by the APU. Finally, during APU operation the
truck will be parked, putting less of a thermal load on the
system due to reduced convective losses to the outside
air stream and reduced thermal losses due to drafts and
air leakage.
Cooling
The American Trucking Association publishes
recommended engineering practices for the OEM truck
manufacturing industry. They have developed
recommended practices and testing criteria for APU
climate control systems as part of RP 432. Table 2
shows the recommended performance and testing criteria
for an APU cooling system.
ATA RP432 Cooling Recommendations
Performance
Sleeper Temperature ( max) 78 ° F
Duration 10 hrs
Test Criteria
Factory Curtains Closed
Initial Cab / Sleeper Temp 73 ° F
Ambient Temperature 100 ° F
Relative Humidity 50 %
Solar Load ( Overhead Source) 600 W / m2
Table 2 – ATA RP432 APU Cooling [ 17]
In order to better estimate appropriate APU system
sizing, stock HVAC system measurements were taken on
the Freightliner test truck. Tests consisted of measuring
the difference between ambient air temperature and the air
temperature within the truck. A/ C testing was completed
under maximum HVAC settings with the engine idling at
1000 RPM.
Air conditioning systems in stock trucks are designed
with capacities of up to 24,000 BTU/ hr. This capacity is
sufficient for cooling the entire cab, driver area and
sleeping compartment under all operating conditions. The
Freightliner Century class test truck came standard with a
factory installed Sanden SD7H15 compressor. This
compressor is driven mechanically by a belt off the
engine, and may require from 1.5 – 6 kW of engine power
depending on idling speed and ambient temperatures. At
a condenser temperature of 52° C this compressor is
capable of providing more than 10,000 BTU/ hr cooling
capacity at idle speeds, with greater compressor speeds
the capacity approaches 24,000 BTU/ hr. Results from
stock cooling system tests are shown in Table 3.
Stock Cooling Performance
Max Rated Vent Airflow
425
( 250)
m3/ hr
( CFM)
Max Rated Capacity
7.0
( 24,000)
kW
( BTU/ hr)
Vent Temp Delta
[ Vent – Sleeper]
13.0
( 23.4)
° C
(° F)
Sleeper Temp Delta
[ Exterior – Sleeper]
7.8
( 4.3)
° C
(° F)
ATA RP432 Delta
15
( 27)
° C
(° F)
Table 3 – Stock Cooling Performance
Heating
The ATA also publishes recommended practice for
engine- off HVAC heating performance shown in Table 4.
As a rough comparison of the different duty requirements
for heating and cooling the temperature differences that
must be maintained can be compared for each task. For
the cooling system the testing criteria calls for a 27 ° F
temperature difference ( plus additional solar load) to be
maintained for a period of 10 hours, in the winter a 68 ° F
temperature difference must be maintained.
ATA RP432 Heating Recommendations
Performance
Sleeper Temperature ( min) 68 ° F
Duration 10 hrs
Test Criteria
Factory Curtains Closed
Initial Cab/ Sleeper Temp 73 ° F
Ambient Temperature 0 ° F
Solar Load 0 W / m2
Table 4 – ATA RP432 APU Heating [ 17]
Heating is provided by waste heat from the engine cooling
system. The engine on a large diesel truck has more
than enough waste heat to supply all that is needed to the
heating system. Unlike the cooling system, the heating
system requires no additional mechanical input energy,
except that needed in order to run low power auxiliary
pumps and fans. In order to better estimate appropriate
APU system sizing, stock HVAC system measurements
were taken on the Freightliner test truck.
Benchmark testing was also performed on the stock
heating system. Tests consisted of measuring the
difference between ambient air temperature and the air
temperature within the truck. Heating system testing was
completed under maximum HVAC settings with the
engine idling at 800 RPM. These tests are for the sleeper
cab only, the main cab climate control system was not
used. Measurements where taken after the sleeper
compartment temperature reached steady state. Table 5
shows the results for the heating performance on the
stock test truck.
Stock Heating Performance
Max Rated Vent Airflow
425
( 250)
m3/ hr
( CFM)
Max Rated Capacity
8.8
( 30,000)
kW
( BTU/ hr)
Vent Temp Delta 49.7 ° C
[ Vent – Sleeper] ( 89.4) (° F)
Sleeper Temp Delta
[ Exterior – Sleeper]
41.6
( 74.8)
° C
(° F)
ATA RP432 Delta
37.8
( 68)
° C
(° F)
Table 5 – Stock Heating Performance
Electrical Power
The electrical power needed for an APU system can be
broken down into different needs. Low voltage DC power
to run onboard truck integrated accessories such as
dome lights, dash radios, CB radios, and other DC
accessories. The other need, made possible by the APU,
is for high voltage 120V AC to power electrical air
conditioners, refrigerators, televisions, VCR’s, computers,
and other household appliances.
12V DC Power
Measurements were taken on devices powered by the 12
VDC system on board a representative line- haul test
truck. The loads from accessory devices, such as lights
and the dash radio, were found to be very small. The
power consumption by the climate control fans was found
to be 100W– 200W; however these fans are not used
during APU operation because these functions are
performed by other devices. The numbers in parenthesis
for the blowers in Table 6 indicate fan speed setting.
12V Truck Loads
Truck State Ave. Power [ W]
Off 7
Run 72
Climate Control
Cabin Air Blower ( 1) 153
Cabin Air Blower ( 2) 185
Cabin Air Blower ( 3) 215
Cabin Air Blower ( 4) 229
Sleeper Blower ( 1) 83
Sleeper Blower ( 2) 118
Sleeper Blower ( 3) 135
Accessory
Dome Light 29
Reading Light 14
Radio 12
CB Radio 15
Table 6 – 12 VDC Accessory Loads
120V AC
Unlike the stock 12 VDC accessories that are present on
nearly all trucks by the manufacturer ( ie. lights, fans,
radio), the number and type of 120 VAC accessories in
use varies greatly from truck to truck. The results of a
nationwide survey performed by the UCD ITS found that a
significant number of truck drivers use 120 VAC
appliances in their trucks, presumably by utilizing a small
inverter run off the truck’s 12 VDC power system. Even
though most trucks lack APUs, surveys show that these
accessories are being used in a large majority of trucks.
Some trucks don’t have any AC accessories, while other
trucks have several.
120V Truck Appliances
Entertainment Present Pk. Power [ W]
TV 74% 100
VCR 53% 30
Stereo 66% 50
DVD Player * 30
Game System * 20
Communication
Cell Phone 62% 10
Laptop Computer 23% 35
Comfort
Air Conditioner - 1200
Refrigerator 59% 160
120V Lamp 46% 100
Microwave 19% 1200
Coffee Maker 15% 1200
Hot Plate/ Crock Pot/ Grill * 750
Other * 11% -
* Survey lumped enumeration as part of the “ Other” category
Table 7 – 120 VAC Electrical Loads
The survey, for the most part, did not distinguish between
120 VAC appliances and those that directly operated on
12 VDC. However, all the appliances that are listed in
Table 7 could be operated on a 120 VAC power supply if it
were available. This would most likely be preferable to the
truck driver because 120 VAC appliances cost less and
usually run more efficiently than their low voltage DC
counterparts.
The majority of the electrical load on the APU comes from
120 VAC appliances and climate control devices. The
most significant load is from the electrical air conditioning
system, which demands a large amount of power that
must be maintained for long periods of time. The
remaining 120 VAC loads are much smaller and more
intermittent.
Packaging and Integration
Size and weight are important to truck drivers because the
drivers are usually limited in how much they can haul by
one of these two constraints. A truck that is loaded to its
full weight capacity is said to be “ grossed out”, a truck
loaded to it maximum storage capacity is said to be
“ cubed out”.
The UCD ITS survey asked a sample of line- haul truck
drivers how heavy and how large of an APU they would be
willing to purchase. The survey found that about 40% of
the respondents would find an APU that is less than 249
pounds acceptable.
APU Allowable Weight
0
5
10
15
20
25
30
35
0- 100 100- 249 250- 499 500- 999 > 1000
Weight [ lbs]
Frequency
0%
20%
40%
60%
80%
100%
120%
Frequency
Cumulative %
Figure 1 - Allowable Weight Response
It also found that about 80% of truckers are willing to
accept an APU less than 16 cu. ft. in volume.
Allowable APU Volume
0
2
4
6
8
10
12
14
0- 2 3- 4 5- 8 9- 16 17- 32 > 32
Volume [ cu. ft.]
Frequency
0%
20%
40%
60%
80%
100%
120%
Frequency
Cumulative %
Figure 2 – Allowable Size Response
The opinion of the truck driver is useful in understanding
what is acceptable in terms of APU size and weight, but it
is not necessarily the best gauge of what is acceptable.
It may be hard for a person to visualize what a complete
system would look like based on numerical descriptions
of size and weight. A well engineered and integrated
system may be acceptable even if it is heavier and bulkier
than the target audience reveals in a survey.
Cost and Economics
As part of the UCD ITS survey truck drivers were asked
what factors are important in their decision to purchase an
APU. The truck driver was given choices of different
factors that might be important to them. Among these
choices was the price of the APU. If the driver felt this
was an important factor, he or she was then asked how
much they would be willing to spend in order to purchase
an APU. The results indicated a wide range of prices that
the drivers were willing to pay ranging from $ 100 to
$ 10,000. This indicates that some drivers value an idling
solution highly while others are satisfied with the current
solution. The drivers that put a low value on an idling
solution may not idle that much, may not understand the
benefits of using an alternative idling solution, or may not
have been required to stop idling by regulations. When
drivers where asked what payback period would be
necessary for them to consider the purchase an APU the
average answer was 27 months.
ENGINEERING CONSIDERATIONS
Based on the design guidance gathered from previous
studies and directly from truck drivers, general engineering
requirements for a fuel cell APU were established. The
needs of the truck drivers were translated into engineering
design requirements. These requirements were used to
bound the design within practical limits and prioritize the
importance of APU features and capabilities.
Power Output
Climate control power requirements have been found to far
exceed the power requirements for auxiliary electrical
appliances currently being use. The power requirements
of the APU system will be largely dictated by the power
requirements of the climate control system. Auxiliary
electrical loads from small electrical appliances that a
truck driver is currently likely to use will make up a small
percentage of the total APU power used.
A few options are available for cooling a truck cab without
using the main engine. There are auxiliary compressors
powered by small diesel engines which consume
considerably less fuel than the main engine. Pony Pack
and Power Pak both make small, frame mountable
auxiliary diesel engine APUs that have an A/ C refrigerant
compressor and auxiliary alternator. Webasto
Thermosystems has developed a phase change chilled
storage system in which a fluid is frozen while the truck is
operating and later used to cool air when the engine is
turned off. They have integrated this with a diesel fuel
fired heater to supply both heating and cooling needs with
a single climate control unit. There are also several
companies that supply electrically powered air
conditioning systems that are suitable for running off of
120 VAC power supplied from grid electric plug- in sources
or an inverter.
Of these options the electrically powered vapor
compressor cooling system makes the most sense for
use with a fuel cell APU system. The system is very
similar to the stock cooling system with the only
difference being that the compressor is powered by an
electric motor instead of being powered mechanically by
the main engine.
Several electrically powered air conditioning systems
exist in the 10,000 – 14,000 BTU/ hr range that are
purpose- built for mobile applications such as boats, RVs,
and trucks. An air conditioning in this size range should
be sufficient to cool the cab to the level specified by
ATA’s RP432 recommended practice. A unit with a
10,000 BTU/ hr capacity consumes in the range of 1,200W
of electricity at rated conditions but could require up to 3
times this amount for very short periods ( about 1 second)
at startup. This is a significant amount of power and must
be supplied in addition to any power needed by other
powered appliances.
Heating options for fuel cell powered APUs include
electrically powered resistance heaters, fuel cell waste
heat utilization, and fuel fired heating systems. After
careful analysis the only system that made sense from an
energy utilization standpoint was the fuel fired heater.
Heating requirements may be anywhere from 2 - 3.5 kW
depending on the weather conditions and the quality of the
truck’s insulation. If a 2 kW heating requirement where to
be provided by an electrical resistance heater the input
fuel energy into the fuel cell would be 4 kW ( assuming a
50% conversion efficiency). The total thermal efficiency of
this configuration would be 50%.
The waste heat could in theory be recovered from the fuel
cell and used to heat the cab. However, even if 100% of
the waste heat from a fuel cell operating truck appliances
of 500W was recovered, the total heat input into the cab
would be only 1 kW. This would require extensive
engineering, complicate safety considerations by
introducing a direct path for hydrogen to enter the cab,
increase complication by requiring extra heat exchangers,
and shorten the usable life of the fuel cell by using high
quality electrical power when only low quality heat is
required. In any case, the amount of heat input to the cab
is only half of the 2 kW required and would still need to be
augmented by electrical resistance heating. Any
conceivable heating system utilizing resistance heaters
and recovered fuel cell waste heat results is an inferior
solution to that of simply combusting the available fuel
directly for heat.
Direct diesel fuel fired heaters are inexpensive, highly
efficient, and currently exist. The units are over 90%
efficient and can use diesel fuel directly from the trucks
main tanks. They also use a small amount of electrical
energy to power pumps and fans. End point emissions
are very low, but if desired could be further reduced by
combusting natural gas or hydrogen instead of diesel fuel.
The direct fuel fired heater is the best option for supplying
auxiliary heating needs for a truck.
Electrical power needs fall into two categories. 12 VDC
power for onboard truck integrated accessories such as
dome lights, fans, and dash radio and 120 VAC power for
small household appliances that are operated on board
the truck.
Having 120 VAC power available frees the on board 12
VDC system from many loads it might otherwise need to
supply. Almost any appliance that a truck driver might
want to use can be powered by 120 VAC current. Many
appliances such as television, lamps, and refrigerators
have been converted to accept 12 VDC power, however
these appliances are more expensive and less efficient
than their 120 VAC counterparts. It is expected that most
of the electrical power needed will be in the form of 120
VAC. The remaining 12 VDC load that is not replaced by
120 VAC loads is estimated to be small in comparison to
the amount of energy the batteries can supply.
The test truck has three batteries with about 130 Ah of
capacity; this is typical for most line- haul trucks. This
amount of battery capacity can supply a 100 W load for
more than 39 hours. Based on this estimate, the UCD
APU has made no provision for charging the on board
battery bank. It is expected that the truck will be in use
and the batteries recharged before they are depleted in
real world driver applications.
APU Loads
Ave. Power [ W] Pk. Power [ W]
12 VDC
Auxiliary 100 200
Heating 25 50
120 VAC
Auxiliary 300 1,000
Cooling 1,200 3,600
Table 8 - Power Requirement
There are many difficulties involved in estimating the
amount of power that must be supplied by an APU. The
average and peak power demanded will vary from
application to application depending on the weather, the
number and type of appliances in use, and the preference
of the truck driver. Based on information about the type of
appliances drivers tend to use, the power consumption of
these appliances, and the revealed preference of the truck
driver, Table 8 was compiled.
This table shows the electrical power requirement in four
categories that is expected to be sufficient for the majority
of truck driver needs. The estimated average power is
used to size the fuel cell and fuel storage capacity. The
peak power, in many cases, must only be supplied for
very short periods of time during transients and power
surges. The electric air conditioning compressor
illustrates that the difference between average and peak
power requirements can be quite different. These short
periods of increased power demand can be handled by
oversizing the fuel cell or by using power storage
components such as batteries or capacitors. Because
fuel cell efficiency tends to increase as power demand
drops, using a fuel cell that is larger than necessary for
average power usage has the advantage of increasing
overall efficiency. However, the extra efficiency comes at
the cost of having a bigger, heavier, and more expensive
fuel cell. As fuel cell costs drop and power densities
increase this will be of less concern.
Packaging and Integration
The packaging and integration of the APU is important in
terms of customer acceptability as well as durability and
safety. Consideration should be given to placing major
components as close together as possible to minimize
the transmission distance of high current electricity and
fuel. The number of connections between the enclosure
and the truck should also be kept to a minimum. This
has the effect of increasing efficiency, and minimizing the
size and weight of the APU. It also increases safety
because there is less chance of fuel leakage and less of a
chance of severing either high current wiring or fuel lines in
the event of an accident.
In keeping with this design philosophy the number of
joints and unions in the fuel lines should also be kept to a
minimum. Longer unbroken lengths of tubing are
preferable to many lengths of tubing teed or spliced
together. Joints that are necessary should be in well
ventilated areas or to the exterior of any enclosure if
possible.
The size and weight of the fuel cell APU will dictate where
the system can be integrated on the truck. The most
obvious place is on the frame rails of the truck. Line- haul
trucks are quite large and there is usually some unused
space along the frame rails. As APU systems get
smaller and lighter other possible mounting locations may
become feasible. Many trucks have a wind deflector on
top of the cab. This space has the advantage of being at
the highest point of the vehicle and is well ventilated.
Both of these characteristics help to make the APU
installation safer. Another area to consider would be
behind the cab near the step area between the truck and
the trailer.
Durability
Fuel cells are not considered as robust as internal
combustion engines. Fuel cells require clean air, clean
fuel, and have not proven themselves in high vibration
environments. Fuel cells also need to be protected from
temperature extremes. A complete APU system will
include an inverter, batteries, and other sensitive
components that have their own environmental
requirements. Protecting the fuel cell and other
components from an adverse operating environment can
have a profound effect on durability.
The APU enclosure should protect its components from
extremes in temperatures and weather. The competing
needs of protecting the components from rain, snow, dirt,
and grime and the need to provide adequate ventilation
and cooling during operation must be balanced. PEM fuel
cells generally need to be kept between 0° C and 100° C.
PEM fuel cells generally operate best at temperatures
around 80° C. Excursions above boiling or below freezing
can shorten the life or otherwise cause damage to the fuel
cell, even if it’s not operating during these excursions.
This requirement usually means that the enclosure needs
to be insulated and some method of temperature
monitoring and control must be provided. The need to
provide adequate ventilation and access to the APU
components tend to make the temperature control
requirements more difficult to meet.
Extremely high temperatures damaging to the fuel cell are
only likely to be experienced while the APU is operating
and producing high levels of power. High temperatures
extremes can usually be prevented by providing adequate
ventilation to the fuel cells. Should temperatures rise too
high, the APU can always be shut down or power output
reduced before damage is done.
Preventing the fuel cell from freezing is a more challenging
problem. Freezing can be a danger when the fuel cell is
not in operation and not producing large amounts of
power. Unlike extremely high temperatures, temperatures
below freezing are common and likely to be encountered
in real world situations. Protecting the fuel cell from
freezing requires active monitoring of the environment and
the ability to heat the enclosure should the temperature
fall too low.
Extremely cold conditions are usually avoided by
insulating the APU enclosure and providing some
mechanism for temperature control. Should temperatures
fall too low, the fuel cell can be turned on and the
generated waste heat can be used to maintain enclosure
temperatures at an acceptable level.
Isolation from vibration is also an important consideration.
Trucks are subject to vibration and shock inputs from
many sources; the road, the main diesel engine,
engagement and disengagement of the trailer, and
backing up against loading docks. Design actions must
be taken to isolate the APU and the fuel cell from these
damaging sources of shock and vibration while still
providing for robust attachment that can restrain the APU
in the event of an accident. Extensive work was performed
by Mathuria, et. al. on a vibration mount system for a fuel
cell APU system. [ 20]
Fueling
PEM fuel cells have been shown to operate using high
purity gaseous hydrogen and methanol as fuels. Because
trucks do not use either of these fuels additional fuel
storage will need to be devised. Methanol has many
advantages as a fuel. Methanol is a liquid, it has a high
energy density, and it is currently used in many chemical
and consumer applications. However, methanol is toxic,
even in small amounts, and it is soluble in water. This
combination of properties makes its safe distribution and
storage a concern. It is not uncommon for underground
gasoline fuel tanks to occasionally leak. A methanol leak
could potentially cause much more environmental damage
than a gasoline fuel leak. There are many examples of
safe use and storage of methanol if care is taken.
Consumer products such as stove fuel, racing fuel, and
windshield washer fluid all contain methanol. Windshield
washer fluid is interesting because it usually contains
35%- 45% methanol and is currently safely distributed and
used. A fluid very similar to windshield washer fluid could
conceivably be used to power a fuel cell APU.
Hydrogen is the most commonly used fuel for PEM fuel
cells. Hydrogen can be directly used by fuel cells but
requires more complex methods of storage. Cryogenic
liquid hydrogen storage is being considered by some
automobile fuel cell manufacturers. Gaseous compressed
hydrogen storage is a less complex method and its
availability is much greater. Compressed hydrogen is
available at all of the hydrogen fueling installations in the
US and is also readily available in standard industrial
cylinders.
The amount of fuel that must be stored depends on the
expected fueling interval, the efficiency of the fuel cell
APU and the expected average fuel cell APU load. Table
9 shows the estimated amount of fuel need for varying
levels of average power required from the APU. The
estimates where made for a fuel cell APU system with an
overall efficiency of 45% powering the estimated load
continuously over an 8 hour period. Note that the units for
hydrogen are given in kg of gas, for gaseous storage the
weight and volume of the system are usually much greater
than the weight of the gas. Advanced compressed
hydrogen storage systems have a weight storage
efficiency of up to 8.5%, meaning that the complete
storage system might weigh 12 times as much as the
stored fuel. For methanol the units are given in liters; this
does not include the likely additional volume that is
needed for dilution water as direct methanol fuel cells
generally use a methanol/ water mixture for fuel.
Average Power Hydrogen Methanol
500 W 0.23 kg 1.12 L
1000 W 0.45 kg 2.23 L
1500 W 0.68 kg 3.35 L
2000 W 0.90 kg 4.47 L
Table 9 - Storage Estimates
Many times truck drivers do not stop at truck stops or
other fueling stations to spend the night. In fact, one of
benefits of the fuel cell APU is the ability for the truck
driver to stop anywhere in order to sleep. The near silent
and point pollution free operation of the fuel cell APU will
open up new areas previously not available to the truck
driver. Idling ordinances or noise ordinances are no longer
a concern to the truck driver using a fuel cell APU. To
maintain this freedom and to ameliorate the scarcity of
fueling points, the on board storage system should be
sized to allow operation for multiple days without the need
to refuel. The exact number of days will depend on the
number of fueling points the driver expects to be near in a
given period of time. Initially fueling points may be located
at major warehousing and distribution points. If one were
to assume that the APU would be fueled once every three
days, then the APU storage system would need to be
capable of storing about 2 kg of hydrogen fuel or 10 L of
methanol assuming an average nightly load of 1500W.
The industry is in the early stages of standardizing
hydrogen fueling stations, fueling connectors, and storage
systems. Some hydrogen connectors similar to those
used for natural gas vehicles are being developed and
standardized. Hydrogen fuel filling stations generally
dispense fuel at pressures of 3,600 psi and 5,000 psi
some stations are capable of 10,000 psi. The tank must
be matched with the appropriate fuelling dispenser. Apart
from industry standardization and self regulation the US
DOT approves hydrogen storage systems for vehicle use.
Any storage system intended for on- highway use must
meet US DOT standards.
Safety
Safety precautions taken in the design of a fuel cell APU
should take many forms. The APU should have an easily
accessible emergency shutoff switch that can remotely
stop fuel flow to the APU and electrical current from it.
Any sources of high current electrical storage should be
fused and sustained overload conditions to the truck
should be protected for via circuit breakers.
In addition to precautions taken to avoid fuel leakage, a
robust design should incorporate features that make the
APU tolerant to leakage should a failure occur. As a first
step the APU should employ passive ventilation strategies
to assure that any fuel that does leak is vented to the
surrounding atmosphere even when there is no power
available to the system or monitoring is disabled. In
terms of gaseous hydrogen this means incorporating
venting at the high points of enclosures, placing tubing
joints outside of any enclosure, and eliminating direct
paths for fuel encroachment into the cab of the truck.
Apart from passive design features, active monitoring and
forced ventilation using fans should also be used when
possible. The lower flammability limit of hydrogen is 4%
by volume. Sensors should monitor for hydrogen leaks
and take action to dilute leaks through active ventilation
strategies while simultaneously interrupting fuel supply
and alerting the driver.
DEMONSTRATION FUEL CELL APU
The UCD ITS goal was to construct a functional FC APU
suitable for public outreach and energy flow data
collection and analysis. The capabilities and features of
the APU were chosen with this goal in mind. Design for
packaging and durability where considered secondary and
compromises where made with the understanding that
future advances in technology are expected to make
these compromises unnecessary.
Figure 3 - UCD ITS Fuel Cell APU
SYSTEM DESIGN
Figure 4 shows a block diagram of the UCD ITS
demonstration fuel cell APU system. The APU system
consisted of two fuel cells connected in parallel, two lead
acid batteries, a power inverter, a battery charger, and
associated power distribution and safety components.
Figure 4 – APU Architecture
Power Output
The sizing of the APU in terms of power output is arguably
the most important consideration in the design of a truck
APU. If the APU is undersized it will not be able to
provide sufficient climate control and electrical power
needs, resulting in customer dissatisfaction. If the APU is
oversized it will cost more than necessary and will be
prohibitively expensive. The goal is to size the APU no
bigger than what is necessary to provide satisfactory
performance to a large percentage of users.
Challenges to optimum sizing stem from the mismatch in
the amount of power needed on average with what is
needed for short periods. The difference between peak
power demand and the average power demand can be
significant. Some components may require large
amounts of power during startup with a much lower
average power consumption level. Electric motors, such
as those used in electric air conditioners are an example
of this class of load. Startup power for an electric motor
may be as much as 3 times the average power required
and is usually needed for a second or so. Many loads are
operated in an on- off manner with an associated duty
cycle. If the duty cycle is 25% than the peak load might
be 4 times the average load. Examples of appliances that
operate in this manner would be hot pots, refrigerators,
microwave ovens, and coffee makers. The cycle times of
these loads can be anywhere from seconds to minutes.
Along with peak and average power demands from the
components themselves the customer use profile of these
components also introduce peak load demands. At some
time the trucker may want to use the refrigerator, coffee
maker, television, and the air conditioner all at the same
time. At a later time, for instance as the truck driver
retires to bed, the only appliance in operation may be the
air conditioner. Peaks in demand due to the number of
appliances in use may be on the order of minutes or
hours.
A fuel cell APU designed with a very large continuous
power output large enough to handle the maximum
possible load that is ever expected would result in a very
large APU with excess capacity during most of its
operational period. A better strategy is to size the
average power output of the APU to match the expected
average power requirement of the appliances likely to be
used, and to provide sufficient power for peak intermittent
loads.
Examining the expected appliances and usage profiles it
was found that the requirements of the air conditioning
system would dictate the sizing of the APU in terms of
both average power output and peak power output. Based
on estimated cooling loads an air conditioning system
with a capacity of 10,000 BTU/ hr was chosen. This air
conditioning system had an average power requirement of
1.2 kW. Based on appliances that were expected to be
used and the usage profile of these appliances another
600 W of power was expected to be used on average, for
a total expected average load of 1.8 kW. This
requirement dictated that the fuel cells powering the APU
must be capable of continuously supplying this average
load of 1.8 kW.
Parallel FC Stack Design
The fuel cell industry is small and currently few choices
exist for complete fuel cell systems. To provide the
primary power for the APU the Nexa series fuel cell built
by Ballard Power Systems was chosen. The Nexa
system is rated at 1.2 kW and has output characteristics
similar to that of a battery, although over a much
increased range of voltage. Under open circuit conditions
the system voltage is 42V. The output voltage steadily
drops as it approaches its 1.2 kW full load output of 50A
at 25VDC.
In order to supply the 1.8 kW average power requirement
it was necessary to use two Nexa fuel cells. The fuel
cells were connected in a parallel configuration to
maintain an operating voltage range of 42 - 25 VDC with
100A available at full load. A series configuration was
considered because of its inherently more efficient higher
voltage and lower current characteristics. This
configuration was ultimately rejected because the higher
84 – 50 VDC voltage range of operation would have been
more difficult to integrate with common off the shelf
( COTS) components and more stringent higher voltage
safety considerations.
Hybrid System Design
The peak power required by the air conditioner was found
to be about 3 times its average power consumption during
startup. This higher transient power requirement is known
as “ locked rotor” load. This load was expected to be the
most extreme peak load that the APU would be required
to power. The 10,000 BTU/ hr air conditioner system
contained an electric motor that required around 30 amps
for about 1 second during startup. The APU therefore
needed to be able to provide 3.6 kW for at least 1 second.
From these estimates it was found that average APU
power demand would be about one half the peak power
demanded.
The large difference in peak power demand to average
power demand suggested that a hybrid APU design might
be advantageous. By using load leveling energy storage
components in the APU design the peak capacity of the
fuel cell could be reduced from the full 3.6 kW to 1.8 kW.
The decision to build a hybrid APU allowed the APU to
use two fuel cells instead of the three that would have
been required to provide 3.6 kW of power. If custom built
fuel cells optimized for these requirements were used, a
single 1.8 kW fuel cell would have been sufficient.
In a hybrid APU design, the storage element provides
extra power during times of peak demand and recharges
during times of lower power requirements. Battery and
ultracapacitor load leveling elements were considered. It
was determined that ultracapacitors could have easily
provided the power necessary for all expected peak loads
and would have been smaller and lighter then their lead
acid battery counterparts. The final design, however,
utilized a pair of small lead acid batteries connected in
parallel with the fuel cells. Batteries were used because
the inverter that was chosen required a constant DC input
to maintain its parameter memory. Capacitors have a
much smaller energy storage capacity than batteries and
also have the tendency to “ leak down”, or lose their
charge over a period of hours if they are not recharged,
making them less desirable in this configuration.
Passive Control Strategy
The APU incorporates a passive control strategy. The
strategy used ensures that fuel cell capacity is fully
utilized before the energy storage components are called
upon. This has the effect of minimizing the number of
charge discharge cycles on the batteries and increases
efficiency. It also reduced the cost of the system by
using simple components and eliminating the need for
expensive computer control.
The key to the passive control system is careful matching
of the fuel cells, batteries, and inverter. Because each of
these components operate at different voltages, provisions
had to be made to integrate them into the final design.
The fuel cells operate in the voltage range of 42 – 25 VDC.
If the fuel cell output voltage drops below roughly 24 VDC
they open an internal contactor and shutdown. The two
fuel cells alone have a rated output of 50A at 25VDC for a
total of 2.4 kW. During times of peak power demand,
such as when the air conditioner starts, the voltage
required would be in excess of this amount causing
system shutdown. To solve this problem, two 13Ah lead
acid batteries were also integrated into the system to
provide power during these peak demand periods.
Under charging conditions the two lead acid batteries,
connected in series have a nominal voltage in the range of
around 28 VDC. This is somewhat higher than the
standard 12.4 VDC of a lead acid battery under steady
state conditions and is due to the charge system
maintaining a float voltage in the range of 14 volts per
battery. Under sustained load the battery voltage would
quickly fall to 12.4 VDC per battery. As implemented the
batteries start to augment the power output by the fuels
cells when the bus voltage falls below 28 VDC. As power
demand rises and the bus voltages falls even lower the
batteries start to increase their output. As the bus voltage
approaches the fuel cells supply their maximum rated
power, more than 100A at 25 VDC. To this the batteries
are also supplying more than 100A at 23VDC. The total
output available to the load is greater than 4.4 kW which
is more than enough to supply the transient power needs
for most components. This system works very well
because power is instantly drawn from the batteries as it
is needed without any active control.
Lead acid batteries tend to charge best at slow rates and
are usually sustained at a voltage above their open circuit
voltage after they have reached full charge. This voltage is
called the batteries “ float voltage” and for lead acid
batteries is usually 14.7 VDC per battery. In this
installation two batteries are connected in series resulting
in an optimum float voltage of 29.4VDC. Recharging the
batteries after a load has been drawn from them required
some extra design effort. The batteries cannot simply be
connected to the main bus because the fuel cells operate
at voltages up to 42 VDC; this is much too high a voltage
for the batteries to graciously handle. To solve this
mismatch problem, a power diode was placed between
the batteries and the main bus. This allowed the batteries
to discharge power to the bus when the buses voltage fell
below the battery voltage, but it prevented power from the
bus from reaching the batteries when the bus voltage was
above the battery voltage.
Charging of the batteries was accomplished by using a
battery charge controller. The charge controller worked by
modulating a current switch at high frequency to limit the
current that the batteries could absorb. This device is
sometimes termed a chopper because of the way it
modulates the current from the power supply to the load.
This arrangement allowed the battery to charge at
voltages up to the 29.4 VDC float voltage even when the
bus voltage was at higher voltages. The workings of the
charge controller also prevented the batteries from
charging when the bus voltage fell and both the batteries
and fuel cells were providing power to the load, because
the batteries would be at a higher potential than the bus
due to the power diode.
Instrumentation
Instrumentation and data collection was accomplished
using current shunts placed at various locations in the
power distribution network. The shunt voltages were read
and recorded by a 16 channel data acquisition system
connected to a laptop computer by way of a USB bus.
The information acquired with the system was used to
monitor power flows and bus voltages.
Packaging and Integration
The APU enclosure was built using modular aluminum
extrusions. The enclosure was mounted to the frame rails
behind the cab and isolated from the truck by rubber
mounts. The fuel cells where further isolated from the
enclosure by a second set of rubber ring isolators. The
truck was isolated from the road by its suspension
system.
Weatherproof conduit was used to connect the APU to
the truck cab. 120 VAC power was fed to the cab through
a metal sheathed set of wires to a junction box located
below the sleeper bunk. This junction box served to
distribute power to a set of outlets located on the lower
part of the bunk and to the air conditioning system.
Controls for the air conditioning and for the fuel fired heater
where also located on the lower bunk panel.
Fueling
The tank used was an earlier generation 150 liter
composite wrapped hydrogen vessel rated at 3,000 psi.
Because the rated tank pressure did not exceed the 3,600
psi fill pressure available at local hydrogen fill stations it
was decided that filling would be performed from industrial
hydrogen cylinders. The hydrogen cylinders were
delivered at a pressure of 2,000 psi, well below the rated
pressure of the hydrogen tank. After connecting the
cylinder to the tank, the hydrogen pressure in the cylinder
was allowed to equilibrate with the tank pressure.
Because of the difference in volume between the cylinder
and the tank, much of the hydrogen gas is transferred
using this method provided that the tank pressure is kept
low. The partially depleted hydrogen cylinders were then
used elsewhere until fully depleted. This method of filling
is not the most efficient, but it did allow for multiple
demonstrations without the need to use a gas
compressor. It was also very inexpensive.
Safety
The enclosure incorporated many passive and active
ventilation and safety features to minimize the possibility
of a fuel leak and mitigate a leaks impact should one
happen. The bottom of the enclosure had large vents on
either side and the top panel had spacers between the lid
and the frame to allow any light hydrogen gas to escape.
The fuel cells were positioned in the box so that the
cooling fans would draw air from the bottom and expel it at
the top, ensuring that during operation a continuous flow
of fresh air was circulated through the enclosure. The fuel
cell also had hydrogen sensors integrated into them that
would shutdown the system should a leak be detected.
Each of the two fuel cells had a hydrogen sensor thus
giving the system an element of redundancy.
The hydrogen storage tank had a built in pressure relief
device ( PRD) that would vent excess hydrogen to a vent
mast connected to the rear of the cab in the event of an
overfill or if excessive heat, as from a fire, caused
pressures to rise above the 3,000 psi rating of the tank.
An emergency stop was located on the exterior of the
enclosure within reach of the fill port. This stop would
interrupt both gas supply from the tank and DC power to
the inverter. The fuel cells had many safety features built
into them also. Safety startup self checks made sure that
the fuel cells were operating properly and checked for fuel
leaks and other abnormalities. The fuel cells also
shutdown should the temperature get too high or the bus
voltage fall too low. A high current fuse was placed
between the batteries and the main bus to protect against
a high current short circuit. The high voltage electrical
system safety was primarily handled by the individual
components. The inverter had an integrated circuit
breaker that would trip if the output line voltage dropped
too low or current draw became too high.
COMPONENTS
Freightliner Century Class Test Vehicle
The fuel cell APU was integrated into a 1994 Century
Class test tractor. This tractor is representative of most
line- haul class 8 trucks.
Figure 5 - Freightliner Century Class Truck
Ballard Nexa Fuel Cell
The fuel cells chosen for this project were Ballard Nexa
fuel cells. As of this writing the Nexa fuel cell is the PEM
fuel cell that is closest to being commercially available.
The Nexa system has extensive control and monitoring
systems integrated into the stack allowing for push button
operation. This stack was chosen for its availability, ease
of use, compact size, and consistent operation.
Figure 6 - Ballard Nexa Fuel Cell
Ballard Nexa PEM Fuel Cell
Outputs
Rated Power 1200 W
Rated Voltage 26 V
Mass 13 kg
Operating Life 1500 hrs
Inputs
Fuel 99.99% Hydrogen
Consumption < 18.5 SLPM
DC Startup Power 18 - 30 V
Environment
Operating Temp. Range 3 - 40 ° C
Storage Temp. Range - 29 to 70 ° C
Storage Freeze Cycles 50 Cycles
Table 10 - Fuel Cell Specifications
Testing was performed at the UCD ITS Fuel Cell Lab to
determine the performance of these fuel cells.
Polarization curves and efficiencies curves were generated
for single fuel cells.
Fuel Cell Performance
0
5
10
15
20
25
30
35
40
45
50
0 10 20 30 40 50 60 70
Current [ A]
Voltage [ V]
0
200
400
600
800
1000
1200
1400
1600
Power [ W]
V ( Gross)
V ( Net)
Power ( Gross)
Figure 7 - Fuel Cell Polarization Curve
The fuel cell’s power is greater than the rated output,
approaching 1.4 kW at 65 amps. The fuel cells are rated
for 50 amp output, so the 65 amp outputs represents
output that is 30% in excess of rated output. The effects
of operated the fuel cell in excess of rated output is not
know, therefore the APU design was implemented so that
the fuel cells would not exceed maximum ratings.
Fuel Cell Performance
0
10
20
30
40
50
60
70
80
0 200 400 600 800 1000 1200 1400 1600
Power [ W]
Efficiency [%, LHV]
? ( Gross)
? ( Net)
Figure 8 - Fuel Cell Efficiency Curve
The system efficiency of the fuel cells reaches a peak net
of around 48%. At very low loads the net efficiency is
poor because of the balance of plant components that
must be powered in order to maintain operation.
Trace Xantrex Inverter
The inverter chosen was the Trace Xantrex SW4024. The
SW4024 is primarily manufactured for home use, most
often for homeowners who wish to power household
120VAC electrical needs without being connected to the
grid. DC electrical power is supplied to the inverter unit
generated by solar, wind or hydro sources. The generated
power is stored in lead- acid batteries at 24VDC nominal
and inverted to 120VAC as demanded.
The same characteristics that make this inverter a good
choice for home power needs also make it a good choice
for a fuel cell APU. The fuel cells have voltage and power
output characteristics similar to a solar array or a turbine,
and the unit is able to handle the fuel cell source with only
minor modification. Modification to the inverter consisted
of changing its allowable input voltage range to handle
input voltages of 19 – 38 volts.
The inverter unit also has features that allow it to be tied
to the grid power in order to supply 120VAC output
electricity even when the primary power source is not
available. It also produces a clean 120VAC sinewave that
can be fed to the national electrical grid. For a
homeowner these features allow power to be drawn from
or sold to the grid based on the availability of the
renewable energy source. For the truck APU these
features allow the truck also to be easily connected to the
grid, a feature generally know as “ shore power”. It also
allows power to be resold to the grid from the fuel cell
generated energy. This feature might be beneficial for
distributed power generation.
Figure 9 - Trace Xantrex SW4024 Inverter
Trace Xantrex SW4024 Inverter
General
DC Input Voltage 22 - 33 VDC
Output Voltage 120 VAC
Continuous Power 4,000 VA
Continous Output 33 A
Max Output 78 A
Efficiency ( Peak) 94 %
DC Input
Full Rated Power 200 A
Short Circuited Output 360 A
Input Range ( Modified) 19 - 38 V
AC Output
AC Output Wavform Sinewave
Voltage Regulation +/- 3 %
Physical
Size 15" x22.5" x9"
Weight 105 lbs
Table 11 - Inverter Specifications
Taylor Made Air Conditioning Unit
The Taylor Made A/ C is a self contained electrical air
conditioner purpose- built for small area climate control.
The unit can be installed into boats, RVs, campers, and
trucks. Some trucks already use this A/ C in combination
with a diesel powered APU or when connected to shore
power, as such no modification was needed to integrate
the unit with the fuel cell APU.
Figure 10 - Cruisair ASC Air Conditioning
Cruisair ASC Air Conditioner
Cooling Output 10,000 BTU/ hr
2.93 kW
Rated Voltage 120 VAC
Running Amps 10 A
Starting Amps 30 A
Power Consumption 1.2 kW
Table 12 - A/ C Specifications
Webasto Fuel Fired Heater
The Webasto Air Top 2000 diesel fuel fired heater is also
purpose built for small area climate control and truck use.
The Air Top is powered by 12VDC from the trucks
batteries and is integrated totally separate from the fuel
cell APU. The unit draws very little power and analysis
showed that the stock battery capacity of the truck could
power the heater without danger of depletion.
Figure 11 - Webasto AirTop 2000
Airtop 2000 Specifications
Heat Output .9 - 2.0 kW
Fuel Consumption .12 - .24 l/ hr
Power Consumption 9 - 22 W
Table 13 - Heater Specifications
Dynetek Composite Fuel Tank
The Dynetek hydrogen fuel tank used is an earlier
generation composite design. Current tank offerings
achieve operating pressures of 10,000 psi. The tank is
constructed out of an aluminum cylinder wrapped with
carbon fiber and resin. This design makes the tank
lightweight and strong.
Figure 12 - Dynetek Composite H2 Tank
Dynetek Tank
Volume 150.0 L
Rated Pressure 3,000 psi
Capacity 2.6 kg
Table 14 - Tank Specifications
Hawker Genesis Batteries
The Genesis battery is a high performance lead acid
battery. This battery is capable of high current discharges
and is very robust to deep discharge and harsh charging
cycles. These characteristics made it a very good choice
for the fuel cell APU application.
Figure 13 - Hawker Genesis Batteries
Hawker Genesis G13EP
Mass 4.9 ( 10.8) kg ( lbs)
Capacity 13 Ah
Discharge ( 5 min rate) 70.8 A
Table 15 - Battery Specifications
Instrumentation
The NI DAQPad data acquisition system was used to
gather energy flow data on the FC APU. The system has
16 single ended analog inputs and transmits collected
data to a laptop computer running LabView where the data
was analyzed and stored.
Figure 14 - National Instruments DAQPad- 6020E
DAQPad- 6020E
Bus USB
Analog Inputs 16SE / 8 DI
Input Resolution 12 bits
Sampling Rate 100 kS / s
Input Range +/- .05 to 10 V
Table 16 – Data Acquisition Specifications
Current shunts were used to measure energy flows from
the fuel cells, batteries, and to the inverter. Typical
voltage drops across the shunts were 100mV at rated
amperage, which was 100A for the fuel cells and batteries
and 200A to the inverter.
Figure 15 – Empro Shunt
Electrical Distribution
An International Rectifier HFA180NH40 rectifier diode was
used to protect the batteries from over voltage. The diode
has a high forward current capability allowing it to handle
power transients from the batteries during peak demand.
Figure 16 – Rectifier Diode
International Rectifier HFA180NH40 Diode
Cathode to Anode Voltage 400 V
Forward Voltage 1.1 V
Forward Current ( 100° C) 160 A
Max Power Dissipation 625 W
Table 17 – Rectifier Diode Specifications
To provide safety shutdown protection to the fuel cell APU
system a Kilovac EV200 contactor was used. Typically
this component is used as a motor switch and is able to
connect and disconnect high current loads reliably and
robustly. This contactor can accept coil voltages from 9
to 36VDC making it compatible with the 24VDC battery
backup power included in the design.
Figure 17 - Kilovac EV200
Kilovac EV200 Contactor
Contactor
Carry Current 250 A
Break Current ( 1 Cycle) 2,000 A
Contact Resistance ( max) 0.4 mOhms
Close Time 15 ms
Mechanical Life 1,000,000 cycles
Coil
Coil Voltage 9 - 36 VDC
Hold ( min) 7.5 VDC
Hold ( avg @ 24VDC) 0.07 A
Table 18 – Contactor Specifications
SYSTEM PERFORMANCE
Climate Control
Climate control tests were performed on the Freightliner
test tractor. The Freightliner tractor was parked and
allowed to come to temperature equilibrium. The climate
settings on the APU powered HVAC systems were set to
maximum ( either heating or cooling) and the temperature
was recorded after the interior sleeper temperature
reached equilibrium, usually around 3 hours. The
temperatures are expressed as temperature deltas. For
the vent temperature delta this is the difference in
temperature between the vent outlet and the sleeper cab.
For the sleeper temperature delta this is the difference
between the exterior temperature and the sleeper
temperature. Tests procedures were not conducted in the
exact manner as dictated by the ATA recommended
practices due to resource limitations, however results
obtained here are expected to be close to those that
would be obtained with the stricter test procedures.
Heating
APU Heating Performance
Rated Vent Airflow 110 m3/ hr
Rated Capacity 2.0 kW
Vent Temp Delta
( Vent – Sleeper)
> 82.9
( 149.2)
° C
(° F)
Sleeper Temp Delta
( Exterior – Sleeper)
36.2
( 65.2)
° C
(° F)
ATA RP432 Delta
37.8
( 68)
° C
(° F)
Table 19 - FC APU Heating Performance
Tests performed on the APU heating system indicated
that the system was very nearly able to achieve
performance as recommended by the ATA for APU
heating systems, if the temperature difference was still
achievable at ATA test conditions. In order to improve
performance increased sleeper cab insulation or a slightly
larger heating system could be used.
Cooling
APU Cooling Performance
Rated Vent Airflow 552 m3/ hr
Rated Capacity 2.93 kW
Vent Temp Delta
( Vent – Sleeper)
8.6
( 15.4)
° C
(° F)
Sleeper Temp Delta
( Exterior – Sleeper)
10.9
( 19.6)
° C
(° F)
ATA RP432 Delta
15
( 27)
° C
(° F)
Table 20 - FC APU Cooling Performance
Tests performed on the APU cooling system showed that
it also was not able to achieve the recommendations as
stated by the ATA recommended practices. It should be
noted however that strict testing conditions were not kept
in regards to ambient temperature and solar load, indeed
it is suspected that the solar load was greater than what
was specified. To improve performance better insulation
or a slightly larger air conditioning unit could be utilized.
The fuel cell APU as implemented would be able to power
an air conditioning unit that is 50% larger. This increase
would most likely be enough to meet recommendations.
System APU Power
Testing was performed on the DC portion of the APU
system. These tests were used to better understand the
interaction of the fuel cells, batteries, and the passive
control system and what effects the design had on power
output and system efficiency.
APU Performance
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140 160 180 200
Current ( A)
Voltage ( V)
Figure 18 - APU Polarization Curve
The polarization curve of the system was similar to the
behavior seen with a single fuel cell stack. The voltage
initially drops rather quickly as current is drawn from the
system. As more current is demanded the curve seems
to flatten out and the voltage drop per amp demanded
decreases. A look at the current being supplied by each
individual component shows why this is happening.
APU Components
0
20
40
60
80
100
120
140
160
180
200
40 37.5 35 32.5 30 27.5 25 22.5
APU Voltage ( V)
Current ( A)
FC 1
FC2
Batt
APU
Figure 19 - APU Component Power Output
Figure 19 shows that the fuel cells supply all the power
demanded until the bus voltage drops to around 27.5 volts.
As the bus voltage falls from this point power is drawn
from the battery pack. It is apparent that the batteries are
a much more rigid source of current as compared to the
fuel cells. That is, while a voltage drop of 15 VDC from 42
to 27 VDC will cause about 40 amps ( 2.66 A/ V) to be
drawn from the fuel cells, a 5 VDC drop will cause around
100 amps ( 20 A/ V) to be drawn from the battery pack.
Transient Performance
The APU system was designed to handle high transient
power needs without shutting down. This characteristic
was especially important for the A/ C compressor startup.
During startup the electric motor that runs the compressor
draws around 3 times the amount of current needed for
steady state operation. To supply this large amount of
current, batteries were integrated into the system.
Air Conditioner Startup Transient
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
- 1 - 0.5 0 0.5 1 1.5 2
Time [ s]
Power [ W]
Figure 20 - A/ C Power vs. Time
During A/ C compressor startup high levels of power are
needed in excess of 4kW. These high levels are only
needed for a fraction of a second, after startup 1 – 1.2 kW
are needed depending on temperature conditions and fan
settings.
Air Conditioner Startup Transient
0
10
20
30
40
50
60
- 1 - 0.5 0 0.5 1 1.5 2
Time [ s]
Voltage [ V]
FC V
Batt V
Figure 21 - A/ C Startup Voltage vs. Time
Figure 21 shows the relationship between fuel cell and
battery voltage levels. Due to the way that these
components where implemented the fuel cell voltage is
allowed to be higher than the battery voltage, however the
battery voltage is not allowed to exceed the fuel cell
voltage.
Air Conditioner Startup Transient
0
20
40
60
80
100
120
- 1 - 0.5 0 0.5 1 1.5 2
Time [ s]
Current [ A]
FC1 I
FC2 I
Batt I
Figure 22 - A/ C Current vs. Time
The final figure shows the current draw from the batteries
and the fuel cells during the air conditioner startup
transient. As is expected, the batteries provide much of
the power needed during the transient.
CONCLUSION
System Performance
Stock APU Units
Heating
Max Rated Vent
Airflow
425
( 250)
110
( 65)
m3/ hr
( CFM)
Max Rated Capacity
8.8
( 30,000)
2.0
( 6,830)
kW
( BTU/ hr)
Vent Temp Delta
( Vent – Sleeper)
49.7
( 89.4)
82.9
( 149.2)
° C
(° F)
Sleeper Temp Delta
( Exterior – Sleeper)
41.6
( 74.8)
36.2
( 65.2)
° C
(° F)
Cooling
Max Rated Vent
Airflow
425
( 250)
570
( 335)
m3/ hr
( CFM)
Max Rated Capacity
7.0
( 24,000)
2.9
( 9,900)
kW
( BTU/ hr)
Vent Temp Delta
( Vent – Sleeper)
7.8
( 4.3)
8.6
( 15.4)
° C
(° F)
Sleeper Temp Delta
( Exterior – Sleeper)
13.0
( 23.4)
10.9
( 19.6)
° C
(° F)
Electrical 12 VDC
Average Power 1.2 .25 kW
Peak Power 12 12 kW
Electrical 120 VAC
Average Power n/ a 2.4 kW
Peak Power n/ a 4.0 kW
Table 21 – System Performance Comparison
The University of California Institute of Transportation
Studies has performed a study on PEM fuel cell APUs.
Based upon previous studies and truck driver input, a set
of performance targets were established that would satisfy
a majority of truck drivers needs. These targets were
used to guide the design of the fuel cell APU system.
The system was built at the UCD ITS fuel cell lab and
integrated with a class 8 tractor. Testing was performed
on the tractor with the integrated APU to determine the
performance of the APU as compared with the stock
system.
Testing has revealed that satisfactory APU performance
can be achieved by using off the shelf components,
however many challenges remain in the effort to
commercialize FC APUs. As research continues,
technical barriers are being removed and regulations are
starting to be adopted that may encourage the adoption of
this technology in the coming years.
ACKNOWLEDGMENTS
The University of California Institute of Transportation
Studies would like to thank the people and companies
that made this research possible.
National Science Foundation IGERT
Freightliner, LLC
California Air Resources Board
South Coast Air Quality Management District
Chevron Texaco
Carrier Transicold
TIAX
American Trucking Association
Neil C. Otto
Xantrex
Taylor Made Environmental
Webasto Thermosystems
UCD National Truck Survey Team
REFERENCES
1. Storey, John ME; Thomas, John F; Lewis, Samuel A
Sr.; Dam, Thang Q; Edwards, Dean K; DeVualt,
Gerald L; Retrossa, Dominic J; 2003. Particulate
Matter and Aldehyde Emissions from Idling Heavy -
Duty Diesel Trucks. Society of Automotive
Engineers. Paper 2003- 01- 0289.
2. Brodrick, Christie- Joy; Farshchi, Mohammad;
Dwyer, Harry A; Harris, Bruce D; King, Foy G Jr.;
Effects of Engine Speed and Accessory Load on
Idling Emissions from Heavy- Duty Diesel Truck
Engines. Journal of Air and Waste Management
Association. 52: 174- 185.
3. Lim, Han; 2003. Study of Exhaust Emissions from
Idling Heavy - Duty Diesel Trucks and Commercially
Available Idle- Reducing Devices. Society of
Automitive Engineers, Paper 2003- 01- 0288.
4. Pekula, N; Kuitz, B; Hearne, J; Marchese, AJ;
Hesketh, RP; 2003. The Effect of Ambient
Temperature, Humidity, and Engine Speed on Idling
Emissions from Heavy- Duty Diesel Trucks. Society
of Automotive Engineers. Paper 2003- 01- 0290.
5. Levinson, T; 2001. Anti- Idling Laws and Regulations.
The Maintenance Council Annual Meeting. Nashville,
TN. March 15.
http:// www. transportation. anl. gov/ ttrdc/ pdfs/ TA/ 73. pdf.
Accessed October 29, 2003.
6. Lutsey, N., J. Wallace, C. J. Brodrick, H. A. Dwyer,
and D. Sperling, 2004. “ Modeling Stationary Power
for Heavy- Duty Trucks: Engine Idling Versus Fuel Cell
APUs.” Society of Automotive Engineers Technical
Paper Series 2004- 01- 1479.
7. Brodrick, C- J; Lipman, TE; Farshchi, M; Lutsey, NP;
Dwyer, HA; Sperling, D; Gouse, SW, III; Harris, DB;
King, FG, Jr; 2003. Evaluation of Fuel Cell Auxiliary
Power Units for Heavy- Duty Diesel Trucks.
Transportation Research. Part D: Transport and
Environment.
8. Brodrick, C. J.; Farshchi, M.; Dwyer, H. A.; Gouse, S.
W.; Mayenburg, M.; and Martin, J., 2000.
Demonstration of a Proton Exchange Membrane Fuel
Cell as an Auxiliary Power Source for Heavy Trucks.
Society of Automotive Engineers, Paper 2000- 01-
3488.
9. Zizelman, James; Shaffer, Steven; Mukerjee,
Subhasish; 2002. Solid Oxide Fuel Cell Auxiliary
Power Unit – A Development Update. Society of
Automotive Engineers. Paper 2002- 01- 0411.
10. Warf, Bill; Mazaika, David M.; Scott, Paul B.;
Bergstrand, Richard; 2003. Integration of a Fuel Cell
as a Class 8 Truck APU. National Hydrogen
Association.
11. Venturi, Massimo; Martin, Andre; 2001. Liquid Fuelled
APU Fuel Cell System for Truck Application. Society
of Automotive Engineers, Paper 2001- 01- 2716.
12. Krumpelt, M; September 14, 2002. Auxiliary Power
Units: What, Why, How, When,… EU/ US
Symposium. Brussels.
13. Brodrick, C- J; Lutsey, NP; Sperling, D; Dwyer, HA;
Gouse, SW, III; 2001. The Market for Fuel Cell
Auxiliary Power Units for Heavy - Duty Diesel Vehicles:
First Widespread Application of Fuel Cells in
Transportation. Transportation, Energy, and
Environmental Policy: Managing Transitions.
14. Brodrick, Christie- Joy; Lutsey, Nicholas P; Keen,
Quentin A; Rubins, Daniel I; Wallace, John P;
Dwyer, Harry A; 2001. Truck Idling Trends: Results
of a Pilot Survey in Northern California. Paper 2001-
01- 2828.
15. Lutsey, N., C. J. Brodrick, D. Sperling, and C.
Oglesby, 2004. “ Heavy- Duty Truck Idling
Characteristics – Results from a Nationwide Survey.”
Proceedings from the 83rd Annual Meeting of the
Transportation Research Board. Washington, D. C.
January 11- 15.
16. Lutsey, N., C. J. Brodrick, D. Sperling, H. A. Dwyer,
2003. “ Markets for Fuel Cell Auxiliary Power Units in
Vehicles: A Preliminary Assessment.” Transportation
Research Record: Energy, Air Quality, and Fuels
2003 1842: 118- 126.
17. Technology & Maintenance Council ( TMC), American
Trucking Associations, Inc., 2200 Mill Road,
Alexandria, VA 22314 ( 703) 838- 1763
http:// tmc. truckline. com. Reprinted from TMC's
Recommended Practices Manual.
18. Surampudi, Bapiraju; Walls, Mark; Redfield, Joe;
Montemayor, Alan; Ingold, Chips; Abela, Jim; 2004.
42- Volt Electric Air Conditioning System
Commissioning and Control For a Class- 8 Tractor.
Society of Automotive Engineers, Paper 2004- 01-
1478.
19. Brodrick, C. J.; Gouse, S. W. III; 2001. Fuel Cell
Auxiliary Power Units for Heavy - Duty Diesel Trucks:
The First Commercial Application of Fuel Cells in
Transportation? Proceedings of the University of
California, Davis Institute of Transportation Studies
Biennial Asilomar Conference, Managing Transitions
in the Transportation Sector: How Fast and How Far?.
Sept 11- 14, 2001, Monterey, CA.
20. Mathuria, P. H.; Lim, T. C.; Shepard, W. S.; Bell, S. R.;
Venturi, M.; Dobbs, H. H.; Kallio, E.; 2002. Vibration
and Shock Considerations in the Design of a Truck-mounted
Fuel Cell APU System. Society of
Automotive Engineers, Paper 2002- 01- 3050.