Flywheels are a well-known energy storage technology, at least on the power side of the equation. They work by spinning up a heavy disk or rotor to high speeds, and then tapping that rotational energy to discharge high-power bursts of electricity. Companies like Vycon, Active Power and Beacon Power provide emergency ride-through power for buildings, or fast-responding frequency regulation services for grid operators, to name two typical use cases.

But it’s a lot harder to use flywheels to store energy for hours at a time. Mainly, that’s due to “coasting losses” -- the inevitable mechanical and electromagnetic forces that slow down a heavy spinning object. These challenges have pretty much relegated long-duration flywheels to research labs -- at least, until now.

Last week, Amber Kinetics unveiled a four-hour duration flywheel system, one it says combines the efficiency and flexibility of an electrochemical battery with the durability and lifespan of a simple mechanical device. That’s a twofer that could be very valuable to utilities and grid operators -- and with a 20-megawatt, 80-megawatt-hour contract with utility Pacific Gas & Electric, the stealthy startup has a venue to prove it works.

Amber Kinetics has been working for the past five years on two key flywheel advances, CEO Ed Chiao said. First, it’s using a steel rotor -- much cheaper and simpler to make than the carbon-fiber composites used for most flywheels today. Second, it’s applying a set of technologies, developed at UC-Berkeley and Lawrence Livermore National Laboratory, aimed at cutting those coasting losses down to a bare minimum.

“We have significantly mitigated the issues of coasting losses,” he said in an interview at the company’s Union City, Calif. headquarters and shop last week. “They can never be eliminated, but we’ve shown a pathway to bringing their economic costs to a de minimis level.”

Seth Sanders, the professor at the center of much of this research, met Chiao in 2010, when Amber Kinetics was working on an earlier flywheel design (PDF) for testing in a Department of Energy-funded pilot project. Chiao was struggling with coasting losses, and “Seth told me, basically, ‘I know exactly what to do,’” he said. 

Sanders merged his compay, Berkeley Energy Science Corp., with Amber Kinetics in 2013, with technology licensed from Lawrence Livermore and UC-Berkeley to apply to its next generation of flywheels. These are the units it plans to deploy with PG&E, and are now being tested with the DOE and San Diego Gas & Electric, as well as at a test site in nearby Alameda, Calif. 

Among the startup’s advances, Sanders noted last week, is its steel rotor, using a “high homogeneity of material, which can be challenging, and is something we’ve mastered.” It’s also reduced residual gas in the vacuum chamber, which is the main cause of coasting loss, and has a magnetically assisted ball-bearing system to reduce load and mechanical losses, he said.

The startup has also benefited from advances in electric motors and power electronics over the past five years, which have helped lower efficiency losses in the systems that convert electricity to rotational force and back again. All of these technical advances are available to others, he noted.

But so far, the short list of companies developing flywheels for commercial markets haven’t put them together the way Amber Kinetics has, Chiao said. “There’s no one silver bullet that allows us to unlock the potential of flywheels. It’s just a laundry list of things you have to engineer.”

The core system is a 25-kilowatt-hour flywheel, capable of charging and discharging for more than one duty cycle per day, Chiao said. That’s a big differentiator from the lithium-ion batteries used in most grid applications today, which tend to limit their charge-discharge cycles to avoid degrading battery performance and lifespan.

It’s also expected to last a lot longer than lithium-ion batteries, which tend to come with the expectation that they’ll lose a significant portion of their operating capacity over a 10-year performance period, he noted. Amber Kinetics, by contrast, promises “a 30-year life, with 30,000 full charge-discharge cycles, and no degradation.”

Ten flywheels in a storage container make up what the company is calling an “energy block,” he said -- the core module of its Energy Nuevo project with PG&E, to be built in California’s Central Valley.


But unlike many of the energy storage agreements that PG&E signed in its 75-megawatt storage procurement last week, Amber Kinetics’ 20-megawatt resource is coming without restrictions on how often PG&E can use it per day, he said.

“The key thing we think about is how...we give the utility maximum flexibility,” he said. “We don’t know exactly how we’re going to use it. But we’d like them to work it hard, and use it 24/7/365.”

Chiao wouldn’t provide any cost or price figures on a per-kilowatt or per-kilowatt-hour basis. But he did say they’re competitive with lithium-ion batteries today for multi-hour duration applications. The startup is targeting another 50 percent reduction by the end of the decade, he said.

Chiao also addressed the critical issue of safety -- fast-spinning steel disks can be dangerous if they spin out of control, as the June explosion at flywheel startup Quantum Energy showed. Amber Kinetics is building aerospace-standard safety features and insulation designs into its flywheels, and has tested its suspension system against the G-forces it could expect during a California earthquake, he said.

Amber Kinetics plans to sell its systems to others, but it's also looking at project development, as it’s doing with PG&E, as a significant source of new business, he said. “The strategy is to go out and develop our own projects, and be vertically integrated. We’re trying to mirror the successes of the First Solars and SunPowers of the world.”

As for why the startup has been so quiet until now, Chiao said, “The energy storage space has a lot of companies that make a lot of claims and then don’t deliver. [...] We wanted to have tangible things we could point to, whether it’s a technical report or a customer contract, before we made any announcements.”

Amber Kinetics certainly has a long list of proving points to meet its 20-megawatt commitment with PG&E, although it has some time to complete them -- the project isn’t slated for commissioning until 2020. In the meantime, the startup is pursuing a 10-megawatt deal with an independent power producer in Southeast Asia, among other new opportunities, he said.

The startup has raised about $25 million from early-stage investors, as well as a “very wealthy family office, to fund our current phase of growth,” he said. It also has some key former utility executives on its board of directors, including Tom Higgins of Southern California Edison and Dick Rosenblum of Hawaiian Electric Industries, to help navigate the business of selling new technologies to a tradition-bound industry.

“My experience is, utilities will tend to undervalue flywheels until they get some actual experience with them,” Rosenblum said before a board meeting at the startup’s office last week. “They tend to adopt things in a fashion that has high reliability, that they’re confident in. That’s what they’re doing with storage.” But that takes time -- “to find the value means, to have it on your system and utilize it.”

At the same time, “Their technology seemed to me to be exactly what the industry needed,” he said. “It doesn’t fade, it’s a mature technology, and it fulfills a much larger need than a classic battery would. From that standpoint, this is a solution to utilities’ problems as we go into high-renewables world.”