H. P. Friedrichs (AC7ZL) Homepage
In Search of Perfect Power:
Thoughts on the DoD's "Wearable Power" Contest
In the November 2007 issue of the ARRL's QST Magazine, "Eclectic Technology" columnist Steve Ford, WB8IMY, reported on the Department of Defense's ongoing "Wearable Power" contest. The purpose of this contest is to drive the development of a low-mass, high-density energy source to power the ever-increasing number of electronic gadgets our soldiers are expected to carry onto the battlefield. The contest is open to contributions by the general public, which means that this is a golden opportunity for America's best basement engineers, garage inventors, and ham radio enthusiasts to compete in an arena usually dominated by large defense contractors. A first-prize winner takes home a million dollars. The runner-up will have to "make-do" with a mere half-million.
Unfortunately, a payoff like that won't come easily. To qualify for the prize, the winner must demonstrate an electrical energy source capable of delivering an average of 20 watts (with peaks of up to 200 watts) for periods of up to 96 hours. At the same time, this energy source must weigh less than 4 kilograms (about 9 pounds). Finally, the mechanical proportions and characteristics of the device must be suitable for attachment to a vest or similar garment.
The idea behind the contest is an interesting one and it caught my attention because ham radio operators, particularly those interested in QRP, portable, or emergency communications, are always on the lookout for new and better power sources. At the same time, I would be lying if I claimed that a million dollar prize wasn't equally inspiring. Either way, I decided that it was worthwhile to do some Web surfing and a little bit of number-crunching to get a sense of how difficult it might be to meet the contest's design criterions.
How Much Energy Are We Really Talking About?
To simplify things I decided set aside, for the moment, the weight and peak power specifications and consider the average power delivery requirement. If we multiply 20 watts times 96 hours we get 1920 watt-hours. Energy may also be expressed in the unit joules, and a unit conversion from watt-hours to joules yields 6.9 million joules.
How does this quantity of energy relate to the real world? If we divide 100 watts into 1920 watt-hours, we can see that a winning energy source would be able to light a 100 watt light bulb for more than 19 hours. That's a lot of juice.
Another way to look at this is to consider the energy required to lift something heavy. Lift energy is equal to an object's mass, times the gravitational constant, times height. Assuming a typical U.S. military Humvee weighs 5200 pounds, a winning power source would be capable of providing sufficient electrical energy to lift that vehicle to a height of almost 1000 feet. As you can see, the DoD is not looking for a toy-- this is a real meat-and-potatoes power source.
Bunches of Batteries
My next step was to see how this energy requirement meshes with the delivery capability of common power sources. Surfing the Web, I found an ELK-1280 lead-acid gel-cell battery. This is the kind of battery commonly seen in computer UPS's and alarm boxes. The battery is rated at 12 volts and 8 amp-hours, a 96-watt-hour capacity. Divide this into the 1920-watt-hour requirement, and we find that to build a contest-winning power pack from these batteries, you'd need at least 20 of them connected in tandem. These batteries weigh in at more than 6 pounds apiece. A full set of 20 weighs 123 pounds, or 56 kg. In short, a power pack based on these cells fails the DoD weight requirement by a factor of 14, and we haven't yet accounted for real-world factors that would drive the need for additional batteries.
For example, the amp-hour rating of batteries is based on a certain rate of discharge, and the relationship between load and battery life is not linear. You might actually need more cells (and therefore more weight) to assure sufficient power delivery throughout the power pack's discharge cycle. Let's not forget that battery capacity can be adversely affected by temperature extremes, and note that we haven't even considered how 200-watt peak demands might impact the design. No, lead-acid batteries are not the answer.
Granted, lead-acid technology is old-school, so I surfed the Web some more, looking for data on nickel-cadmium cells. I found a NiCad battery pack, part number PN-3600, with a terminal voltage of 36 volts, and a capacity of 3.6 amp-hours. Multiply these two numbers and divide the result into the 1920-watt-hour requirement, and you discover that you need at least 15 of these packs to meet the DoD specification. Unfortunately, the combined weight of all the cells is almost 90 pounds, or 41 kilograms. Again, we fail the weight requirement by a large factor.
Lithium-ion batteries are more "cutting-edge," but how do they stack up? On the web, I found a lithium-ion replacement battery for the Ipod, part number IPOD-3G37V500. The pack is rated at 3.6 volts, with a capacity of 0.5 amp-hours. The retail information states the weight at 0.2 ounces. If you run the math, you need at least 1037 batteries. The combined pack weighs in at nearly 13 pounds, or 5.9 kilograms. The pack is one and one-half times too heavy. We're getting closer to the goal, and yet, we are still so far away!
Very quickly, it becomes evident that any common battery technology is not going work in this application.
Not a Plane, Not a Bird... It's Super Capacitor!
Thirty years ago, when I began learning about electronics, I remember reading about capacitors and their role as energy storage devices. Universally, text from that era goes on to explain that while the measurement unit of electrical capacity is the farad, practical capacitors are always measured in microfarads (millionths of a farad) or picofarads (trillionths of a farad.) At the time, there was no such thing as a 1-farad capacitor.
How things have changed! So-called "ultra" or "super" capacitors are starting to show up everywhere. Now, it is not only possible to purchase a 1-farad capacitor, it's possible to purchase capacitors rated at many hundreds of farads. There is also a great deal of discussion as to whether these devices might eventually represent real competition to traditional electrochemical rechargeable batteries. I decided to investigate super-capacitors in the context of the DoD's contest, and returned to the Web to conduct some more research.
The results were at once intriguing and disappointing. A visit to Maxwell Technology's Web site provided datasheets for 650-farad capacitors, from which I could calculate energy capacity and weight. The model BCAP300-E270, for example, has an advertised energy density of 5.52 watt-hours per kilogram. You'd need at least 348 kilograms worth to provide the energy required by the contest-- way too heavy.
A Wikipedia entry on the phrase "supercapacitor" references a 2006 patent (number 7033406) filed by EEStor. According to the article, energy densities as high as 342 watt-hours per kilogram are claimed. This is most impressive, because it actually exceeds the energy density that I had calculated for the lithium-ion battery, and it shows that super-capacitor technology has a bright future in rechargeable electronic devices. That said, a power pack composed of EEStor capacitors still fails the contest's weigh limits by a large amount.
Energy Density and Liquid Fuels
A Google search on the phrase "energy density" brought me to an excellent Wikipedia entry that lists energy density values for a variety of fuels and energy sources. The energy density values in this table are expressed as watt-hours per kilogram, so direct comparisons are made easier if we translate the contest requirements to similar units. We know a winning power pack design must supply 1920 watt-hours and weigh no more than 4 kilograms. Divide the second figure into the first, and we get a target energy density of 480 watt-hours per kilogram. The point is this: In order for a contest entry to be viable player, its energy density must meet or exceed 480 watt-hours per kilogram.
Right off, we can dismiss certain "alternative" ideas like using compressed air as a power source (only 34 watt-hours per kilogram). Similarly, while significant research has gone into the idea of using flywheels for energy storage, with an energy density of only 120 watt-hours per kilogram, they don't make the cut.
What stands out in the Wikipedia table are the hydrocarbon fuels. All of them exhibit energy densities that are consistent with the demands imposed by the DoD's contest. The question is how to perform the necessary chemical-to-electricity conversion.
Burning fuel in an engine to drive a small permanent magnet generator is an obvious approach, but as I see it, it's not a very good one. Internal combustion engines are notoriously inefficient. Even if we engineer our way past their weight, there is a real problem with getting rid of the waste heat they produce. A soldier wearing armor and NBC (nuclear/biological/chemical) gear in the 112 degree heat of a place like Baghdad is not likely to tolerate a "wearable" power source that blows hot exhaust gas against his body. Engines also fall short in applications where stealth is essential. Even muffled engine exhaust can be heard at great distances, and through an infrared viewer, exhaust pipes may as well be flashing beacons proclaiming, "Here I am! Shoot in this direction!"
External combustion engines like low delta-T Stirling motors may provide the answer, particularly the newer piston-less, travelling-wave varieties. Pressure variations within these devices have been used to drive piezo disks or linear alternators, resulting in electricity production without rotating parts. However, as in the case of internal combustion engines, efficiency, weight, noise, and heat signature issues may render Stirling variants unsuitable for application to "wearable" battlefield power sources.
It is possible to "burn" fuel without a flame, and to harness that reaction to directly produce electricity using an apparatus called a fuel cell. The principle behind the fuel cell was recognized as early as the first half of the 19th century. Hydrogen fuel cells have found use in U.S. spacecraft, where they provide reliable power and even drinking water. In more recent years, research efforts, both public and private, have sought to reduce the cost of fuel cell technology, improve the efficiency and working-life of the equipment, and to broaden the range of fuels that can be consumed in the cell.
Revisiting Google and the Web, one of the first fuel cell search "hits" I stumbled upon was the Web site of UltraCell Products. Prominently featured is their model XX25 fuel cell. The cell is capable of providing 25 watts of power continuously, which meets the basic output specification as required by the contest. The cell is fueled by a methanol/water mixture that is introduced into the cell with preloaded cartridges. A single cartridge will supply the cell with fuel for up to 9 hours, and cartridges are "hot-swappable," so periodic replacement of empty cartridges with fresh ones would allow the cell to produce energy almost indefinitely.
According to the manufacturer, the design and packaging of the XX25 meets military standards for operation under various humidity, temperature, dust, vibration, and shock conditions, so practical implementation concerns have already been addressed.
The XX25 itself weighs only 1.24 kilograms (2.7 pounds), well below the target set by the contest. However, if we add the weight of 10 or 11 fuel cartridges, enough fuel for 96 hours of operation, the total system weight crosses the 5-kilogram mark. Another possible shortcoming of the XX25 is that there is no indication on the product's datasheet that the unit is capable of 200 watt surges. So, strictly speaking, the XX25 does not meet the contest criterions either.
Despite its shortcomings, the XX25 fuel cell comes tantalizingly close to meeting the DoD's specifications. It occurs to me that ethanol is similar to methanol, but contains 23% more energy per unit weight. If it was possible to reconfigure the cell to work with ethanol, it should be possible to produce the same amount of electricity using less fuel or a smaller fuel cell. This strategy might reduce the total system weight to the desired 4-kilogram mark.
With regard to the contest's peak power requirement, my thought would be to couple the fuel cell with a large super-capacitor to give the system the ability to provide for demand surges.
It will be very interesting to see who wins the contest and what technology is employed to capture the $1,000,000 prize. For what it's worth, I'd place my bet on fuel cell technology or a fuel cell/ super-capacitor hybrid.
The following is a brief collection of interesting URLs. Some of these provided the raw data used to prepare the text above.
Document Revision 1, 10/30/07
Document Revision 2, 10/31/07
Document Revision 3, 11/14/07