By all measures that usually matter, Aquion Energy should have succeeded.
The startup, which sold big batteries for renewable projects and power grids, was founded by Jay Whitacre, a materials science professor at Carnegie Mellon who previously developed batteries for Mars rovers at NASA. It raised nearly $200 million from prominent investors, including Bill Gates, and venture capitalists at Kleiner Perkins and Shell. Perhaps most important of all, the company entered the market with a clear-eyed awareness of earlier missteps by battery startups. It took pains to avoid the use of rare materials. It relied on repurposed manufacturing equipment. And it identified niche markets where it might gain a foothold.
But on March 8, after failing to raise additional funding, Aquion filed for bankruptcy protection, cut 80 percent of its staff, and halted manufacturing. It was the latest of several stumbles for venture-backed storage startups. EnerVault, which was developing what are known as flow batteries, put itself up for sale after failing to find additional investors in 2015. Later that year, liquid-metal battery startup Ambri laid off a quarter of its staff. Around that same time, LightSail Energy, which was struggling to develop technology to store energy as compressed air in carbon-fiber tanks, pivoted to selling its containers to natural-gas suppliers. Taken together, these struggles have deflated hopes for the emergence of affordable and practical grid storage anytime soon.
And that’s a problem. Without cheap ways of storing excess energy generated from intermittent sources like wind and sun, there are limitations on how much these renewable sources of power can contribute to the grid’s overall electricity generation—and, by extension, how much we can cut the greenhouse-gas emissions driving climate change. There are already days when California solar farms have to shut down because they’re generating more power than the grid can use at a particular time. And yet the system still needs to run enough fossil-fuel backup plants to meet total demand whenever the sun dips behind clouds.
A year ago, MIT Technology Review ranked Aquion number five on its list of 50 Smartest Companies. It’s hard to decipher the particulars of what went wrong, as the bankruptcy filing offers few additional details. Whitacre declined requests to speak until an auction is complete, but he made clear that he hopes the company or technology can proceed in some form after that point.
What is clear is that despite the compelling need for better grid storage technology, any startup today faces several daunting realities. First, the slowly developing market for advanced grid storage still isn’t large, in part because the technologies are immature and expensive. Second, and more important in the immediate term, the price of existing technology in the form of lithium-ion batteries has dropped far faster than expected, narrowing the promised benefits of new approaches like Aquion’s.
“Don’t hold your breath for the things that come after lithium-ion,” says Ilan Gur, founding director of the Cyclotron Road program for energy entrepreneurs, who previously cofounded a battery company that was acquired by Bosch. “We’re much more likely to ride the lithium-ion cost curve for another few decades.”
By now it’s a familiar phenomenon in clean energy. Companies must make a massive up-front investment to develop new hardware and scale up manufacturing, all while chasing moving price and performance targets as incumbent technologies improve. Faced with such challenges, vanishingly few succeed.
The danger, in this case, is that many observers believe lithium-ion isn’t the right technology for full-scale baseload grid storage, because there seem to be limits on how cheap and long-lasting the technology can ever become. Yet its success frosts over an already chilly investment market for technologies that may be only minimally better today, but could have far greater potential to transform the energy system in the long run.
How can we support such technologies long enough to get to that point?
When MIT Technology Review first wrote about Aquion, in 2012, the company said it hoped to manufacture its batteries for less than $300 per kilowatt-hour of capacity. The technology paired a saltwater electrolyte with a manganese oxide cathode and a carbon-based anode. The price positioned the product between low-end lead-acid batteries and costlier lithium-ion technology. But a big part of the company’s pitch was that the batteries lasted far longer and performed better in the heavy charging and discharging cycles inherent to the electric grid. That promised to lower the lifetime costs of the system.
It’s unclear where Aquion’s pricing stood at the time of its bankruptcy filing, but lithium-ion batteries themselves fell below the $300-per-kilowatt-hour threshold last year, according to Bloomberg New Energy Finance. Specifically, the price dropped by nearly half in just two years, reaching $273 in 2016, as worldwide production ramped up to meet growing demand for phones, electric cars, and solar-power backup systems. And prices for lithium-ion batteries are likely to continue declining. Bloomberg projects they’ll fall to $109 per kilowatt-hour by 2025, and $73 by 2030.
Donald Sadoway, a materials scientist at MIT and cofounder of Ambri, is skeptical that such estimates reflect total costs for lithium-ion storage systems at grid scale. He also argues that future projections from Bloomberg New Energy Finance are unrealistic, noting that the 2030 estimate seems to dip below the cost of raw materials. But either way, the sudden drop in price has already had a “huge ripple effect” on the storage market, he says. “When lithium-ion comes out with these audacious claims, it makes investors pause,” Sadoway says. “They look and say, ‘Wow, you guys are toast.’”
Meanwhile, customers considering a multimillion-dollar storage system have little incentive to bet on an emerging and riskier technology. Lithium-ion batteries already meet many of the specific needs of large-scale utility customers, offering a reliable product from stable providers, says John Zahurancik, president of the energy storage division of AES, a power project developer increasingly focused on battery systems. When California utilities were forced to quickly replace a natural-gas storage site after a major leak in 2015, they opted for three separate lithium-ion systems, assembled by AES and Tesla. “We see it as a technology that’s here, now, and ready,” Zahurancik says.
But there are good reasons to hedge our battery bets, says Jessika Trancik, an associate professor in energy studies at MIT. As much as prices have dropped, lithium-ion batteries still aren’t cheap enough to serve bigger roles in the grid. And it’s not clear whether they’ll get there, she says.
Pricing is also more complicated than what a single kilowatt-hour captures. The more telling metric is dollars per kilowatt-hour per cycle, a figure that would reflect lifetime costs of the system. And that gets us to a big problem for lithium-ion: these batteries can fade fast, as every smartphone owner knows. That makes them less than ideal for the strains of the baseload storage required to balance out the growing portion of intermittent sources like wind and solar in the grid.
It’s still anyone’s bet what grid storage technology, or combination of them, could ultimately displace lithium-ion batteries. In addition to advanced battery possibilities, there are flywheels, compressed air, hydrogen fuel cells, electric-vehicle-to-grid systems, and even energy-storing air conditioners. (Pumping water up a hill and letting it run down later remains by far the biggest storage technology.) But all these technologies suffer from significant shortcomings, and all are likely to require deep sustained investments to further develop them, test whether they are truly viable, and make them cost competitive.
The question is where such investments will come from. The Trump administration has taken steps to dismantle the federal funding program for clean-energy startups, and venture capital investments for such technologies have dropped nearly 30 percent since 2011, from $7.5 billion to $5.2 billion, according to a recent report from the Brookings Institution.
The good news is that the market for large storage systems is widening as more wind and solar energy projects are built, and as factories look to shave costs during peak usage times. And the economics of grid storage are becoming more favorable as additional renewable generation comes online and more aging fossil-fuel plants shut down, say Severin Borenstein, an energy economist at the University of California, Berkeley. “As we roll out more intermittent renewables, storage becomes more valuable,” he says.
The fear, however, is that market mechanisms alone aren’t working fast enough given the pace at which the world needs to cut emissions to avoid the greatest threats of climate change. One tool that could accelerate the shift is smart public policy. Carbon taxes, or cap-and-trade programs like the ones in place in California, New York, and other states, raise the cost of using fossil-fuel plants and increase the incentives to add more renewables and storage to the grid. There are more direct ways the government can support the sector as well, such as California’s requirement that state utilities add more than 1.3 gigawatts of storage to the grid by the end of this decade, or federal grant and loan programs for promising startups.
But ultimately, nascent storage technologies are fighting forces embedded deep in the core of capitalism. Markets consolidate around dominant technologies and companies. It often takes a radical advance to shake up the old order, and in the energy sector those don’t come along often.