Back
Energy storage is having a moment — but the batteries that are taking off today only have enough juice to provide a few hours of grid power. Developers technically could stack up more batteries for longer-term storage, but that gets prohibitively expensive. For a renewables-dominated grid to ride out days of poor solar production or even just an entire night, a breakthrough in cost-effective, longer-term storage is needed.
Over the last couple decades, venture capitalists have recognized this transformative possibility and heaped billions of dollars into the sector known as long-duration energy storage, or LDES. They have little to show for their efforts. The startups that haven’t gone bankrupt have built some factories and early installations, but have not built any particularly large-scale projects, at least in the U.S.
A few weeks ago, I saw something in the desert outside Reno, Nevada, that got me thinking the investors and startups may have been barking up the wrong tree all along.
Former Tesla Chief Technology Officer JB Straubel unveiled a surprising new project in June at the Tahoe campus of his lithium-ion recycling company, Redwood Materials. Instead of ripping apart old electric vehicle battery packs, his engineers arranged them across a patch of desert and hooked them up to an adjacent solar field. This assemblage now stores so much clean power that it can run a small on-site data center, rain or shine, night or day.
In other words, instead of inventing a brand-new technology tailored for long-term storage, Redwood made it way cheaper to stack enough time-tested lithium-ion batteries to accomplish that goal.
Unveiling this new business line, Straubel wasn’t just diversifying his revenue streams. He was staking claim to the long-duration storage market writ large.
“We’re confident this is the lowest-cost storage solution out there,” Straubel said. “Not only just lower than new lithium-ion batteries, but lower than compressed-air energy storage, lower than iron-air, lower than a number of these other ones that carry a little more technology risk.”
As he spoke, Straubel pointed at a bar graph depicting the costs of those types of LDES technology, as well as thermal storage, pumped-hydro storage, and flow batteries. Naturally, the chart showed his used batteries clocking in cheaper than all of them.
It’s a big claim. Second-life battery development is even newer than the LDES field; prior to Redwood, only a handful of companies, like B2U Storage Solutions and Element Energy, had built large-scale second-life storage plants, and those were just in the last few years. The sector has a lot of work to do to convince customers and financiers that the gently used battery packs can be trusted to hold up over years of service. And with new lithium-ion packs getting ever cheaper, the discount offered by used batteries may prove tenuous.
Still, Straubel’s first operating project, which holds 63 megawatt-hours of energy storage, is already bigger than any novel battery installation in the U.S. If Straubel takes this concept mainstream, it could revolutionize the arms race for long-duration storage — and radically improve the odds of running the economy on a largely renewable grid.
At the June event, Straubel essentially asserted that his band of desert engineers, in just a few months of tinkering, has outmaneuvered the researchers and companies working on long-duration for decades.
That deserves some scrutiny — but even pinpointing the costs of the competition is challenging.
“There are a lot of flavors of long-duration storage. What all of them have in common is that actual deployments have been very limited up until now,” said Pavel Molchanov, who analyzes cleantech companies for financial services firm Raymond James. “To make any clear-cut statements about which particular flavor is cheaper than any other would be quite premature.”
Redwood says its second-life battery installations cost less than $150 per kilowatt-hour today, for systems that can deliver power over 24 to 48 hours. The company’s datapoints on the prices of other battery types were drawn from BloombergNEF’s 2024 analysis of the LDES field, augmented with Redwood’s internal estimates for what a complete iron-air system would cost today, since that technology isn’t yet commercially available.
Iron-air is under development, most famously, by Straubel’s former Tesla Energy compatriot Mateo Jaramillo at Form Energy, a VC darling that’s raised more than $1.2 billion to date. Redwood calculated iron-air costs at higher than $150 per kilowatt-hour, but Form has stated its intentions to sell batteries below $20 per kilowatt-hour when its factory reaches full production scale.
It’s worth noting that not all these technologies are directly comparable, because companies design and market them at different durations based on their technical sweet spots. If a technology works especially well at, say, 12 hours duration, the company might not even sell it for 48-hour configurations.
“Part of the issue with comparing long-duration storage systems and prices is that every company will give you their price point for a different duration,” said James Frith, a longtime battery analyst now at VC firm Volta Energy Technologies. “Then you’re thinking, how do I normalize this? How do we get to a base point that is comparable amongst the technologies?”
Epistemological issues aside, Redwood accurately diagnoses that the LDES sector’s struggle to deliver real installations at super-low cost leaves an opening for new competitors.
Brand-new lithium-ion batteries aren’t economically viable at longer durations, though their limits keep expanding as battery prices fall.
“Lithium-ion storage systems with longer durations require more battery cells, making the system capital-intensive and less economically competitive compared to emerging long-duration storage alternatives,” said Evelina Stoikou, head of battery technology and supply chain research at BloombergNEF.
Pumped-hydro and compressed-air energy storage work for longer durations, but they are huge, billion-dollar infrastructure projects of the sort that don’t get built anymore in the U.S. (Canadian company Hydrostor is attempting to break that curse with a $1.5 billion, 500 MW/4,000 Mwh compressed air project in California; if it gets permits to build, it might be online by 2030.)
Flow batteries — which store energy in tanks of liquid electrolytes — have been kicking around for decades with some success in China, where they benefit from government favor. In the U.S., they’ve not gained much traction.
Meanwhile, many LDES startups have made the strategic error of designing exotic storage solutions to eke out a few more hours, under the incorrect assumption that lithium-ion would never be able to compete at four, then six, and then eight hours.
Take ESS, which has developed an iron-based flow battery since 2011: Despite leaning into “long-duration” branding, the company was selling an Energy Warehouse with a bit over six hours duration, and only this year announced a “strategic shift to the 10+ hour product.” (Its board members had to throw in more cash last month to sustain the company through that shift, and gamely agreed to forgo personal compensation for the year.)
The LDES companies most vulnerable to competition from Redwood are the ones that aren’t actually very long-duration, and which haven’t gotten big enough to make their products cheaper.
That’s not to say the other LDES contenders are left quaking in their boots.
“We’re a long way away from proof that second-life batteries are a proper utility-grade asset, capable of 20 years of daily cycling,” said Ben Kaun, who for years analyzed LDES technologies for the Electric Power Research Institute and now works for battery startup Inlyte Energy. “I don’t see an existential threat to LDES.”
The sector has even been showing signs of life, at least compared to its dismal track record from the preceding decade. Form completed its factory in Weirton, West Virginia, and broke ground on its first commercial deployment, in Minnesota, last summer. The company plans to deliver its first batteries to the project in the coming weeks, for commissioning this fall. Over in the Netherlands, a Dutch startup called Ore Energy recently installed a small 100-hour system of its own iron-air battery, based on research at the Delft University of Technology.
Flow batteries have built up considerable installed capacity in China, but that trend hasn’t gotten much coverage in English-language press, said Eugene Beh, CEO and cofounder of California-based flow-battery startup Quino Energy. His strategy is to leverage the now-mature supply chain for flow-battery equipment but to drop in an electrolyte based on quinones, commonly used in clothing dyes, instead of the more expensive vanadium that’s popular in China.
Italian startup Energy Dome has moved swiftly from demo to commercial operations with an iconoclastic design: It stores energy by compressing carbon dioxide in a controlled environment; decompressing it turns a turbine and generates electricity. After building a pilot and a commercial project in Sardinia, Energy Dome just announced an equity investment from Google for an undisclosed amount and a commitment to build its systems to power Google’s data center expansion around the world.
These more out-of-the-box LDES companies might take solace in a few limitations that second-life battery developers must overcome to mount a serious challenge.
Second-life companies take hundreds of batteries from different manufacturers, with different patterns of wear and tear, then operate them all in concert. If that was easy, more people would be doing it by now. Firms that get this wrong could start fires, and fire safety is one of the key arguments used against lithium-ion installations, both by rival technologists and the general public.
Then again, Straubel has as much experience as anyone with the inner workings of lithium-ion batteries. At Tesla, he built the nation’s leading electric-car company and a wildly successful stationary-storage business with the Powerwall and Megapack.
Then there’s the question of longevity. The batteries were pulled out of vehicles for a reason: usually due to their capacity degrading, though other problems develop with age, like higher internal resistance, which makes batteries heat up during discharge. If second-life packs need to be swapped out too frequently, it undercuts the ease and cheapness of the model.
That leaves the matter of supply. Success in second-life depends on a steady and cheap source of gently used EV packs. Here Redwood has a unique advantage, in that the company was constituted to collect the nation’s battery waste and recycle it. Straubel said Redwood was receiving less than 1 gigawatt-hour of used EV packs two years ago, and now is pulling in more than 5 GWh per year.
The available supply of used EV packs is “going to follow roughly the same curve as electric vehicle adoption, but lagging by, let’s say, 10 years,” Frith said. “So we are going to start seeing the volume of packs growing, and I think the real volumes start to kick in closer to 2030.”
Indeed, he added, the growth in volume of used EV packs could parallel the growth of demand for long-duration storage: Few customers buy it now, but many analysts expect demand to grow by the end of the decade as renewables saturate the grid.
It’s too soon to know if used EV batteries will actually wipe the floor with the more unconventional long-duration battery technologies. But the scale and price point of Redwood’s first project announces them as a force to be reckoned with in this arena.
In doing so, Redwood puts a new spin on an energy-storage maxim that venture capitalists keep forgetting, or simply ignoring: Lithium-ion always wins.
Challengers that rely on different chemistries have to build up from negligible production scale and convince customers to take a chance on a design that few people have seen before. It’s a clear uphill battle.
Lithium-ion batteries, in contrast, command an unmatched and ever-expanding scale of industrial production, mostly in China but increasingly in the U.S. too. That manufacturing juggernaut unlocks incremental gains from economies of scale and continual innovation. It also confers consumer confidence, because the technology has such a clear track record of performance.
“Compared to most experts’ predictions, the costs have gone down faster and the performance has improved faster for lithium-ion than people predicted 10 years ago,” said Jeff Chamberlain, who helped the Department of Energy license battery technology to General Motors and LG Chem back in the late 2000s, and now invests in storage technologies as CEO of Volta Energy Technologies.
Nonetheless, investors continued to bet that the streak would end, and they could own a piece of the transformational tech that would triumph for longer-term storage.
“What a lot of startups and investors are doing is assuming the LDES market will exist and it will be enormous, and they’re assuming lithium-ion won’t solve the problem,” Chamberlain said. “I believe that is a very, very bad assumption.”
Over the last decade, lithium-ion has steadily chipped away at use cases where new battery inventions were supposed to win out. New lithium-ion is starting to push into six-hour configurations and beyond, said Stoikou, from BloombergNEF. Global average pricing for turnkey grid storage averaged $165 per kilowatt-hour in 2024, per the data firm’s 2024 survey.
Now, the cost savings from reusing lithium-ion packs accelerate the chemistry’s push into the long-duration market — something that would be a big win for grid-decarbonization efforts, while delivering the LDES hopefuls yet another stinging loss.
Clarifications were made on August 6 and August 7, 2025: This story has been updated to reflect Eugene Beh’s full title and to note that Hydrostor is attempting to build a large-scale compressed air project in California.
.svg)
