American Scientist

Cruising along on the highway, you spot a traffic jam ahead and hit the brakes. In seconds, most of the kinetic energy of your car is transformed to heat on the brake pads and rotors, slowing you (thankfully), but at a cost: An enormous amount of energy is wasted. If you're driving a hybrid, your car recaptures a portion of this kinetic energy by using it to generate electricity, which in turn goes into a large battery for later use. But that approach isn't very efficient, because chemical batteries are not particularly good at taking a charge in large bursts.

So imagine how difficult the situation becomes when one tries to perform this kind of regenerative braking on something as big and energetic as a diesel locomotive. The solution, at least according to some, is to adopt an entirely different kind of energy-storage technology, one that appears poised to serve for this and other novel applications: the flywheel.

Using a flywheel to store energy is nothing new—potters have been doing it for millennia. Modern high-speed flywheels are, however, much more sophisticated than the massive stone wheels used in ancient times to turn clay vessels. Instead of using rock or metal, today's designs use comparatively light composite materials, which are stronger and can be spun at high speeds without coming apart. This approach is advantageous because the amount of energy stored in a flywheel scales linearly with mass but increases with the square of the rotational speed.

But getting a large object to turn freely at many thousands of rotations per minute is not without its difficulties. The standard solution is to evacuate the chamber holding the flywheel, so there is no air resistance, and to support the spinning mass using magnetic bearings, again to eliminate friction. The exchange between electrical and mechanical energy takes place by electromagnetic induction, as it does in an ordinary electric motor or generator. Such flywheel energy-storage systems have been built for many years, but being considerably more expensive than conventional batteries, they have had limited application.

One place that flywheels might eventually find a niche is space. NASA has contemplated using flywheel energy storage for the International Space Station and has  funded considerable research in this area (although so far this technology has not been adopted). The impetus was to find a way to hold the electrical energy generated by the station's solar panels, in darkness a good fraction of each orbit, without having to suffer the vagaries of chemical batteries, which tend to wear out after many charge-discharge cycles. In space, flywheels could serve double duty, replacing both the batteries that would otherwise have to be carried and the "reaction wheels" that are often used to adjust attitude by taking up or giving back angular momentum. For such control, one would install several flywheels at different orientations and then move energy among them to obtain the desired angular momentum for the set.

Although they have worked on space flywheels, engineers at the University of Texas at Austin are also employing this technology in what they call the Advanced Locomotive Propulsion System. The hardware they have been building in some ways resembles what one finds today in hybrid cars. But instead of an internal combustion engine, their system uses a gas turbine, and in place of a chemical battery, it uses a flywheel—perhaps the largest high-speed flywheel in the world in terms of the energy it can store: 133 kilowatt-hours, when it operates at its maximum design speed of 15,000 rotations per minute. At that rate, the perimeter of the rotor moves at approximately 1,000 meters per second, which is faster than a round from an AK-47 assault rifle. So far this flywheel has been run only with a down-sized rotor. The full-size one has just been assembled, and spin tests will begin shortly.

Why build one huge flywheel instead of many small ones? "On a cost per mega-joule basis, a single, large flywheel was the most efficient," explains John D. Herbst, who is co-principal investigator on the project. Robert E. Hebner, who shares the lead with Herbst, says, "We're not aware of anything larger than ours," but quickly adds, "and right now, we don't know a market for one as big as ours," referring to the fact that their cutting-edge work might or might not ultimately lead to commercialization.

But there is another large flywheel now undergoing tests that is being readied for commercialization—and soon, perhaps during 2007. Although smaller than the University of Texas flywheel, it will be able to hold 25 kilowatt-hours of energy when operating at its maximum speed of around 16,000 rotations per minute. What is more, its designers plan to link multiple flywheels together in what they term an "energy matrix," which will be able to store greater amounts of energy, absorbing and releasing it much faster than can normal batteries. Its developer, Beacon Power Corporation of Wilmington, Massachusetts, plans to use its flywheels for a rather novel application: stabilizing the frequency of electric-power grids.

The connection between energy storage and frequency regulation requires some explanation. The alternating current carried by electric utilities in the United States nominally oscillates at 60 hertz, but when demand exceeds supply, the many whirling armatures generating electricity for the grid slow, and the frequency drops slightly. Conversely, when demand is less than supply, the extra energy goes into spinning those generators slightly faster than normal, raising the AC frequency. Grid operators try to maintain a stable frequency, which is to say that they seek to balance supply and demand at all times. That's tricky, of course, because they are not in control of demand, which changes every time someone flicks on a light.

Having many people connected to the grid, some switching things on while others turn things off, evens out demand to a great extent. But there are still unpredictable changes that occur minute by minute. To cope with such short-term variations, which might amount to something like 1 percent of the total, grid operators arrange with certain power generators to reserve a small portion of their capacity, so that they can adjust the supply of electricity on the fly. Such regulation service commands a hefty price tag, because it typically comes from turbines burning expensive fuel (natural gas) and because it requires these generators to operate at something other than their most efficient power levels.

Here Beacon Power believes it can compete profitably using an array of flywheels to absorb the excess energy when demand falls short of supply and to return that energy to the grid when the reverse happens. Equipment for doing just that with an array of 6-kilowatt-hour flywheels is now operating on a small scale in both California and New York in an effort, supported by the Department of Energy, to iron out technical wrinkles and to obtain certification from the grid operators.

The management of Beacon Power believes they are tapping into a marketthat will only grow—not because people are becoming more prone toturning their washing machines on or off at the same time but becausewind power is likely to increase in significance as utilities embracethis and other sources of renewable energy. Even modest changes in thespeed of the air flowing over the blades of a wind turbine causesubstantial swings in energy output. So grid operators will increasinglyhave to struggle with unpredictable variations in supply as well asthose that have always existed in demand. Bill Capp, chief executiveofficer, says, "We think this is going to be a large issue that wecan help solve." If Beacon or other flywheel makers can eventuallydo that with devices that are both faster than a speeding bullet andmore powerful than a locomotive, that would, of course, be just super.