It’s hard to imagine a decarbonized grid without batteries that can last longer — far longer — than the four hours today’s grid-scale, lithium-ion batteries can pump power onto the grid. But who’s going to pay for it?
That’s the question developers and researchers are puzzling over as the U.S. electricity grid struggles to replace aging generation and transmission infrastructure. At the same time, forecast demand for electricity is surging thanks to electrification of transportation and home heating, factory construction, and, of course, data centers. With solar (still) coming online, there’s a need to spread out the plentiful power generated in the middle of the day — or even year — across other hours and seasons.
In much of the country, electricity markets are set up to optimize the delivery of energy on very short time frames at the lowest cost, and to ensure ancillary services that can keep the grid stable from second to second. Then there are capacity markets, where electricity generators receive payments in exchange for their future availability in order to maintain long-term reliability.
Molly Robertson, an associate fellow studying electricity market design at Resources for the Future, a nonprofit research institution, is skeptical about how long-duration energy storage can fit into this market. “If we think about the market as compensating for those three things, there’s two questions,” she told me. “One is, is the market covering all of the things that the grid needs? And are there enough products that are being purchased that actually cover all of the needs of the grid?”
Long-duration batteries fit awkwardly into that equation. “Right now, I think you don’t see long duration storage because there are resources that are more cost competitive” for what existing wholesale markets reward, Robertson told me.
But the grid today may not be the grid of tomorrow — or at least that’s the argument of the long-duration energy storage industry.
“This energy transition was always going to be necessary around this time frame, regardless of the decarbonization agenda or anything like that,” Jon Norman, the president of Hydrostor, a Canadian company developing large-scale, compressed air batteries, told me. “Most of the infrastructure was built in the 80s and 90s and it’s hitting its natural end-of-life cycle. So these traditional coal-fired power plants, gas-fired power plants would either need to be rebuilt or new infrastructure built.”
“There’s no way of avoiding that,” he added.
Norman, of course, thinks that long-duration storage is a “good replacement for a lot of those assets.” Large-scale batteries like Hydrostor’s can store surplus electricity from when renewables are producing more than the grid needs, and then discharge that energy when needed — and for far longer than today’s batteries.
Lithium-ion is the dominant chemistry for battery energy storage systems today, thanks to its high energy density and ability to withstand many charging and discharging cycles, the same factors that have made it the default choice for electric cars. Because of both lithium-ion’s physical limits and the specific needs of the grid, however, the vast majority of grid-scale systems top out at four hours of discharge.
From a grid planning perspective, the difference between those batteries and long-duration storage, which can discharge for 10 or more hours at a time, means that the latter “can reliably replace” existing fossil fuel generation, Norman said. That makes Hydrostor’s batteries less like an “energy” product and more like capacity — a role typically filled by coal and natural gas, which get paid handsomely for doing so.
Restructured electricity markets work fine at wholesale electricity pricing for infrastructure that already exists, Norman argued. In the late 1990s and early 2000s, when electricity markets were deregulated, “you didn’t need a lot of buildout,” he said. Instead, the question was, “How can we most efficiently dispatch this stuff? How do we send the right signals to the generators?”
But sudden demand growth and the ravages of time have brought a new set of challenges. “The issue that we’ve seen over the past 10 years — and it’s coming to a head now — is, how do you build new capacity? Nobody’s really investing in these markets because there’s a real disconnect between those power market signals that are in real time and short term and the long-run cost of building infrastructure,” Norman told me.
Relying on market forces to come up with new capacity has not worked, he said. “This experiment has failed.”
Management of the PJM Interconnection, the country’s largest electricity market, has practically had to beg developers to bring more firm power onto the grid. It’s also overhauling its internal processes to get projects approved for interconnection more quickly.
In the meantime, as capacity payments and reliability worries continue to spiral, the market’s managers have introduced a pair of proposals that would subject new large sources of electricity demand (i.e. data centers) to mandatory shutoffs and allow utilities to get back into building generation. The former would essentially undo the foundational “duty to serve” model that’s been at the heart of electricity policy for over a century, and the other would reverse decades of electricity market deregulation and restructuring.
Suppliers and customers alike revolted against the idea of mandatory curtailment, and both proposals are now on hold. Whether or not either is ever realized, the fact that they’re even being discussed shows how dire the capacity crisis is.
Even in Texas, the most deregulated market in the country, a plan to offer cheap financing to natural gas-fired power plants to shore up the reliability following the 2021 Winter Storm Elliott disaster has found few takers and few viable projects. You have to get outside restructured electricity markets in states like Tennessee or Georgia, where utilities also control the generation of electricity, to find any appetite for large-scale generation projects like nuclear power plants. These markets are able — for better or worse — to pass along the cost of new power plants to ratepayers. It’s no coincidence that all the new nuclear power — a large source of firm power on the grid that takes a notoriously long time to develop — built this century has come in vertically integrated markets.
Everywhere else, building long-lasting infrastructure assets requires planning to lead the market, Norman told me. “Run really sophisticated competitive procurements — competitive mechanisms that allow you to hit a particular objective instead of the objective supposedly being decided by the market in real time,” he explained.
He pointed to California, where regulators tell utilities to procure clean firm generation like geothermal and long-term energy storage (or the state does it itself). Virginia, which is a vertically integrated market within PJM, has targets for energy storage procurement by its utilities.
Norman’s critique of restructured power markets rhymes with those of former Federal Energy Regulatory Commission Chairman Mark Christie, who said that there’s “missing money” in the electricity markets that exposes consumers to financial and reliability risks. He also asked whether restructured electricity markets, “especially the multi-state capacity markets, have been successful in ensuring a sufficient supply of the power necessary to sustain reliability,” as he wrote in widely noted in a 2023 law review paper.
For her part, Robertson cautioned that there are real technological and logistical questions for how long-duration storage would work in an electricity market, even if you can figure out a way to get them on the grid.
“When we think about longer-duration storage, we have to think about, how would those generators operate, and what timelines are they operating on? If you have a multi-day storage opportunity, how are you going to determine the best time to charge and discharge over that long of an opportunity window?” she asked.
In a RFF paper, Robertson and her co-authors argue that long-duration batteries “likely will not be sufficiently incentivized by price fluctuations within a 24-hour period,” as four-hour batteries are, and will instead have to “take greater advantage of long-term revenue opportunities like capacity markets.” But even then, she cautioned, markets would need to see big swings in prices over potentially multi-day periods to make the charging and discharging cycles of long-duration batteries economical.
Norman, however, had harsh words for critics who say this kind of procurement and planning will lead to inflated costs for infrastructure that may or may not be useful in the future. “What bugs me about keeping our head in the sand is that then results in us saying, Well, we just don’t want to pay for that, so we’re not going to set this target, and we’re going to let the markets decide,” he told me. “All we’re doing is deferring the problem and causing it to cost way more. And so I think we need a bit of a wakeup call.”
