The ability to store electricity for later use is becoming significantly more important as the U. S. increases its use of renewable power. Wind and solar-electricity generation is intermittent and sometimes unreliable when the sun doesn’t shine or the wind doesn’t blow. A passing cloud, for example, can easily cut photovoltaic-power output by 80% within seconds.
Storage lets utility companies “stockpile” excess energy and deliver it when wind and solar power are unavailable or when electricity demand increases.
Today’s electrical grid has virtually no storage capacity. The few storage facilities that do exist rely on pumped hydropower: surplus electricity is used to pump water uphill to a reservoir, then water flows downhill through turbines to generate electricity when needed.
While pumped-hydropower storage is proven and often cost effective, it is only feasible in a few locations. The Dept. of Energy’s Advanced Research Projects Agency – Energy (ARPA-E), through its Grid-Scale Rampable Intermittent Dispatchable Storage (GRIDS) program, is funding research with the aim of developing new energy-storage technologies that match the reliability and cost of pumped hydropower, but are modular and can be installed most anywhere.
One such ARPA-E-funded project focuses on flywheels. A team led by Beacon Power Corp., based in Tyngsboro, Mass., is developing a next-generation flywheel energy-storage module that will hold four times the energy at one-eighth the cost of stored energy of current commercial flywheels. The proposed design will be capable of more than 40,000 full charge/discharge cycles and have a 20-year life, making it suited for grid-scale energy storage.
Flywheels — in essence, mechanical batteries — store energy by accelerating a cylindrical rotor to high speeds and maintaining the kinetic energy in the system as rotational energy. Slowing the flywheel releases energy when needed.
Performance is measured in units such as kilowatt-hours (kW-hr), indicating the amount of power available over a given period of time. Multiple flywheels can be used in tandem to provide megawatt-level storage capacities.
For example, Beacon’s current Smart Energy 25 flywheel can store and deliver 25 kW-hr of power for up to 15 min. The heart of the flywheel is a rotating rim fabricated from a graphite and fiberglass fiber-resin composite. Its high strength and light weight let it store more energy than comparable metal flywheels. A metal hub and shaft supports the composite rim, and a motor/generator mounts on the shaft. Together the rim, hub, shaft and motor/generator assembly comprise the rotor. When charging (or absorbing energy), the flywheel’s permanent-magnet motor draws power to accelerate the rotor. When discharging, the rotor’s inertial energy drives the motor, which acts as a generator, and produces electricity that flows back into the grid.
The flywheel spins at up to 16,000 rpm and the surface speed of the rim (tip speed) reaches approximately 1,500 mph. To minimize friction, drag, and energy losses, the rim levitates on a combination of permanent magnets and electromagnetic bearings and operates in a vacuum-sealed chamber. With extremely low parasitic losses, the flywheel can efficiently spin for extended periods with little power required to maintain operating speed.
Smart Energy 25 flywheels can handle hundreds of thousands of charge-discharge cycles over their 20-year life and operate reliably for many years with little or no maintenance. They are typically used for utility-grid frequency-regulation applications, and are being examined for smoothing intermittent power delivery from renewable-generation facilities.
Taking flywheel energy storage to the next level, however, will require a radical rethinking of the overall design, according to Beacon officials. Their proposed 100-kW, 100-kW-hr ARPA-E system will supply four times the energy of current 25-kW-hr units and deliver energy for an hour, versus 15 min in the smaller Smart Energy 25.
“It’s all about cost,” explains Richard Hockney, chief engineer at Beacon Power, and his team is working to bring the price of storage down from $4/kW-hr to $0.50/kW-hr. “This is not an incremental cost reduction effort, it’s a quantum leap, and that brings with it significant technical challenges,” cautions Hockney.
But that’s the point of the ARPA-E program, he stresses, backing projects so risky that no company would under take them on its own. If successful, however, such research promises extremely high payoff. “This program has those two elements: a factor of eight reduction in cost per kilowatt-hour and some extreme technical challenges to get there,” he says.
One way to reduce costs is to increase the amount of storage per unit. As you do that, says Hockney, there is a benefit in that the flywheel’s fixed costs become a smaller percentage of the overall cost of storing energy.
But building a bigger, higher-capacity unit entails considerable changes to current designs. First and foremost is eliminating the central shaft and hub to create what’s called a hubless flywheel, Hockney explains. “The hub in our 25-kW-hr unit connects the rim — the composite structure — to the shaft, on which we mount the bearings and motor.” Eliminating the shaft and hub yields significant costs savings, and it permits a much thinner rim that can attain higher tip speeds.
“Full speed will be about 9,000 rpm, versus 16,000 rpm in the 25-kW-hr unit,” he says. But the hubless design will let Beacon increase the diameter from 32 to 70 in. While rotational speed has dropped, more than doubling the diameter actually increases tip speed by about 25%. And tip speed, not rotational speed, is what counts in terms of a flywheel’s energy-storage capacity, according to Hockney.
The 100-kW-hr flywheel will have much the same carbon-fiber composite construction as in the smaller unit. But rotor weight will increase to about 4,000 lb compared to 2,500 lb in the 25-kW-hr design. “Raising energy storage by a factor of four with only a 60% increase in rotor mass gives a feel for how hard we’re pushing the technology,” he says.
The new flywheel is essentially just spinning a composite ring, notes Hockney. But without a shaft and hub, how do you suspend the ring and transfer power in and out?
After all, it still requires a means of support and a motor/generator, just like existing designs. Beacon’s answer is to mount magnets to the inside surface of the ring that will serve two purposes — create a magnet bearing as well as a permanent-magnet motor.
“Our current 25-kW-hr flywheel only uses an axial magnetic bearing, so it levitates magnetically but radially rides on ball bearings,” explains Hockney. “The new 100-kW-hr ARPA-E wheel will be completely magnetically levitated with no rotating elements, so there will be no contact between rotor and stator.
“Most parasitic losses in the current 25-kW-hr flywheel are in the motor-generator,” he adds. “But that’s because it is typically used for frequency regulation where the unit is constantly cycling.” For these applications, a high-efficiency motor-generator with moderate idle losses works best.
“The ARPA-E wheel is for a different application, supplying energy for up to an hour versus 15 minutes. With energy-storage time increasing by a factor of four, you become more concerned with standby losses,” says Hockney. Minimizing them requires a different motor-generator configuration.
The hubless design offers high-capacity storage with low losses, but the devil is in the details. “The hard part of the magnetic bearing and motor is the magnets we’re gluing to the ID of the rim,” he says. No material with the needed combination of mechanical and magnetic properties exists today that can do the job.
At full speed the rim strains about 1% and the ID increases about 0.70 in. “This means the magnets have to ‘grow’ with that strain and not crack,” says Hockney, which requires a powerful magnetic material that is also physically strong and extremely flexible.
Literature searches, manufacturing surveys, and discussions with more than a dozen magnet suppliers have turned up no suitable options. Again, stresses Hockney, that’s the nature of ARPA-E projects, and it’s forcing Beacon to develop a new material based on neodymium-boron-iron chemistry with a custom binder.
In addition to devising new magnets and a hubless design, Beacon also looks to reduce manufacturing costs. For example, the company is investigating a proprietary and, potentially, much faster method for making the carbon-fiber composite rim.
Finally, there’s safety. A 2-ton ring spinning at supersonic speeds poses obvious concerns should the rim contact the housing — say due to an earthquake or magnetic-bearing failure.
In either case, the rim must be contained and brought to a halt. So Beacon is developing a full-scale demonstrator of what’s called a touchdown bearing that can capture and stop the rotor. Engineers are currently analyzing several mechanical-bearing concepts before proceeding with prototypes and testing.
The program began last September and will run for two years. At that point, Beacon intends to successfully demonstrate the magnets, enhanced manufacturing processes, and the emergency shutdown bearings. “If this comes out well, it would represent a quantum leap in flywheel performance, and the payoff would just be enormous,” says Hockney. A prototype could be running by October of 2013, he adds.