Most of New York’s fossil-fueled peak power plants would be vulnerable to competition if longer-duration storage technologies hit the market, a new study says.

The Empire State passed a law to eliminate carbon from the grid by 2045. Now it’s getting to work on replacing its peaker fleet. Since those plants are often the dirtiest and are typically located in disadvantaged or minority neighborhoods, it’s an environmental and racial justice imperative as well as a clean-energy policy.

But currently available lithium-ion batteries don’t run long enough to keep up. They can replace the peakers that run for just a few hours here and there, but they can’t crank for, say, 15 hours during a summer heat wave. Today’s battery options could only match the operations of 6 to 11 percent of New York peakers, according to a 2019 study by consultancy Energy + Environmental Economics.

2019-era batteries won’t be the only options forever. Indeed, Massachusetts-based startup Form Energy recently won a contract for a 150-hour duration battery plant in Minnesota, way beyond any battery duration offered previously. It’s hard to plan grid resources without understanding the technologies that are coming online, so Form decided to publish its own research to inform the debate about what will soon be possible.

“We’re trying to start a conversation and get people thinking about the potential of this emerging asset class,” said Scott Burger, analytics lead at Form, who wrote the report along with Isabella Ragazzi.

This isn’t a peer-reviewed journal article; it’s ultimately making the case for the category of product Form wants to sell. But the company hired enough researchers from nearby MIT to apply more rigorous modeling and analysis than is typical of a corporate white paper.

“This is MIT academic-grade, in terms of methodology,” said Marco Ferrara, SVP for analytics and business development.

When they ran the modeling, the researchers found that a mix of long- and short-duration storage could cost-effectively replace up to 83 percent of New York’s peaking units by the middle of the decade in the most favorable scenario. That translates to 4,886 megawatts out of 7,500 megawatts of peaking capacity and covers 89 percent of the units located in disadvantaged communities in New York City.

Direct replacement of fossil peakers

To get that result, the researchers looked at 10 years of hourly operating data for all the peaker plants in New York state, tracking their performance under a variety of weather patterns and stresses. Then they modeled different types of storage to see what could cost-effectively replicate the real peaker performance.

An environmental rule targeting nitrogen oxide emissions from peakers will force the oldest and dirtiest ones to upgrade or shut down. Given that context, the study considers energy storage cost-effective if it is cheaper than the cost of compliance by building a conventional peaker. The lithium-ion pricing comes from the National Renewable Energy Laboratory’s expectations for 2025 and 2027. The long-duration storage pricing represents “emerging aqueous air technologies” in that time range — they reflect Form’s understanding of where its technology is headed but capture a broader class of devices.

The study also takes a conservative approach to how the plants will operate.

“We said storage and storage alone has to match the operations of a peaker over a 10-year period,” Burger said.

In practice, it would be cheaper to meet peak demand with a portfolio including storage, renewables and flexible demand, but those resources invite debate over how to compare the reliability of distributed resources versus central plants. This study instead looks at one-for-one replacement, asking what it would take to swap out a fossil-fueled plant for a storage plant of the same nameplate capacity.

It also excludes other roles the storage could play outside of the few hours when needed as peakers. The storage found to be economic in this study is doing so with a single, very limited use, when in practice it could perform additional roles to make even more money.


Long-duration storage companies frequently pitch their technology with critical talking points on the limits or weaknesses of lithium-ion batteries. Form does not — co-founder Mateo Jaramillo built one of the most recognizable lithium-ion storage brands at Tesla. The company is “bullish” on the growth prospects for lithium-ion, Burger said, in the shorter ranges where it competes best.

Lithium-ion can replace the peakers that run most infrequently and for durations of six hours or less. Another cohort of peakers can be economically replaced by long-duration storage on its own; they go up to 200 hours of discharge duration. A third cohort calls for a hybrid approach, using both lithium-ion and long-duration storage at the same site.

In that model, the lithium-ion would tackle fast starts and stops, and the long-duration device would take over for usage that spans many hours. It bears some resemblance to projects that added batteries to gas plants or pumped hydro to tackle fast response duties, although those use cases were driven by physical limitations of the generating technology.

For this vision to materialize, plant owners need to figure out how to make money with the new long-duration asset class. Even conventional lithium-ion business models are still new for New York, with limited examples of large battery business models. But the report stakes out the potential for long-duration storage to serve the state’s energy goals, and that will only happen if its grid planning and contracting start to incorporate this technology.