Skip to Content
Climate change and energy

We’re going to need a lot more grid storage. New iron batteries could help.

Flow batteries made from iron, salt, and water promise a nontoxic way to store enough clean energy to use when the sun isn’t shining.

February 23, 2022
install of ESS storage on site

One of the first things you see when you visit the headquarters of ESS in Wilsonville, Oregon, is an experimental battery module about the size of a toaster. The company’s founders built it in their lab a decade ago to meet a challenge they knew grid operators around the world would soon face—storing electricity at massive scale.

Unlike today’s lithium-ion batteries, ESS’s design largely relies on materials that are cheap, abundant, and nontoxic: iron, salt, and water. Another difference: while makers of lithium-ion batteries aim to make them small enough to fit inside ever shrinking phones and laptops, each version of the iron battery is bigger than the last.

In fact, what ESS is building today hardly resembles a battery at all. At a loading dock on the back side of the ESS facility, employees are assembling devices that fill entire shipping containers. Each one has enough energy storage capacity to power about 34 US houses for 12 hours.

The company, which last year became the first long-duration energy storage company to go public and has ambitions to open factories around the world, will soon begin work on a battery that will dwarf even these truck-size versions. In partnership with the utility company Portland General Electric, ESS plans to construct one that will fill a half-acre building on land adjacent to its factory. It’s expected to have almost 150 times the capacity of the biggest batteries the company ships today.

ESS’s key innovation, though, is not the battery’s size—it’s the chemistry and engineering that allow utilities to bank a lot more energy than is economically feasible with grid-connected lithium-ion batteries, which are currently limited to about four hours of storage.

The iron “flow batteries” ESS is building are just one of several energy storage technologies that are suddenly in demand, thanks to the push to decarbonize the electricity sector and stabilize the climate. As the electric grid starts depending more on intermittent solar and wind power rather than fossil fuels, utilities that just a couple of years ago were looking for batteries to store two to four hours of electricity are now asking for systems that can deliver eight hours or more. Longer-lasting batteries will be required so that electricity is available when people need it, rather than when it’s generated—just as ESS’s founders anticipated.

Good chemistry

Craig Evans and Julia Song, the founders of ESS, began working on an iron flow battery in their garage in 2011. A married couple, they met while working for a company developing fuel cells. Song (now the chief technology officer of ESS) is a chemist, and Evans (ESS’s president) is an engineer and designer.

They saw the price of renewable energy systems dropping dramatically and predicted that this would drive demand for energy storage. An electric grid that is 80% powered by solar and wind, for example, would require an affordable way to store energy for at least 12 hours.

Currently, about 95% of the long-duration energy storage in the United States consists of pumped-storage hydropower: water is pumped from one reservoir to another at higher elevation, and when it’s released later, it runs through turbines to generate electricity on its way back down. This simple method works well but is limited by geography.

Batteries don’t have that limitation. However, most grid-scale batteries operating today are lithium-ion batteries. Relatively expensive, they also deteriorate within a few years and are made from difficult-to-recycle materials that can burst into flames or explode. Worse, if you want to double the storage capacity of your battery array, you have to buy twice as many batteries. That makes it too expensive to store energy for longer than a few hours, says Scott Litzelman, who manages a program that focuses on long-term energy storage at ARPA-E, the US agency that funds research and development of advanced energy technologies.

Flow batteries, like the one ESS developed, store energy in tanks of liquid electrolytes—chemically active solutions that are pumped through the battery’s electrochemical cell to extract electrons. To increase a flow battery’s storage capacity, you simply increase the size of its storage tank. When the battery grows to the size of a building, those tanks become silos.

Inside the flow battery’s electrochemical cells, two electrolytes are separated by a membrane. One electrolyte flows past a positive electrode as it’s pumped through the cell, and the other electrolyte flows past a negative electrode. In ESS’s battery, these two electrolytes are identical: iron salts dissolved in water.

As the electrolytes flow through the cell, chemical reactions take place on both sides of the membrane. When an electric current is charging the battery, the electrolyte at the battery’s negative electrode gains electrons, and dissolved iron salts are deposited onto the electrode’s surface as solid iron.

When the battery discharges, the process is reversed: the electrolyte loses electrons at its negative electrode, the plated iron returns to its dissolved form, and the chemical energy in the electrolyte is converted back to electricity. At the positive electrode, the opposite process occurs: the electrolyte loses electrons and “rusts” to a brownish fluid while the battery is charging, and this process reverses during discharge.

In a conventional lithium-ion battery like the one in a mobile phone or electric car, the cell and electrolyte are contained inside a single package. “What you have at the start is what you get,” says Evans.

But with a flow battery, keeping the electrolyte in an external tank means that the energy-storing part is separate from the power-producing part. This decoupling of energy and power enables a utility to add more energy storage without also adding more electrochemical battery cells.

The trade-off is that iron batteries have much lower energy density, which means they can’t store as much energy as a lithium-ion battery of the same weight. And flow batteries require more up-front investment and maintenance than lithium-ion batteries.

However, when it comes to safely storing large amounts of energy for long periods, they’re hard to beat. And that’s exactly what grid operators will need to do a lot more of in the coming years.


The batteries that utilities use today typically store power for four hours or less. That’s fine for tasks such as smoothing out short-lived frequency fluctuations and supply drops, but as the electricity sector moves toward 100% clean energy, “you absolutely can’t do it with four-hour batteries,” says Hugh McDermott, senior vice president for sales and business development at ESS.

To accommodate the ups and downs of solar and wind generation, most grid operators use natural-gas “peaker plants,” which can start up rapidly when electricity is in high demand. A battery that can provide 16 hours of storage would be cheaper to install than any peaking system, McDermott says.

Flow batteries are a small but growing part of the grid-storage market. By the end of 2019, they were used in only 1% of large-scale battery installations in the United States, according to an August 2021 update by the US Energy Information Administration on trends in the battery storage market. A few utilities began installing large-scale flow batteries in 2016 and 2017, but those batteries use a vanadium-based electrolyte rather than iron. Vanadium works well, but it’s expensive.

Evans and Song initially set out to design a vanadium flow battery but changed course when they stumbled across some iron-based chemistry done at Case Western Reserve University in 1981. Iron struck them as a low-cost alternative to vanadium, “but it had challenges,” says Evans.

One challenge was how to prevent roughly 1% of the electrons on the negative side of the battery from bonding with stray hydrogen ions in the water-based electrolyte instead of plating iron. Over time, this side reaction generates a buildup of hydrogen gas and causes the two sides of the battery to depart from a chemical balance in which both electrolytes return to their original, identical state when fully discharged.

“All batteries have side reactions,” says Evans. But because it’s easy to access the chemicals that circulate through a flow battery (unlike the chemicals closed inside a conventional battery), designers can include a mechanism to recover from these side reactions.

Evans and Song dealt with the problem by adding a “proton pump” to their battery. It’s a fuel-cell-like unit that converts hydrogen gas back to protons, which reduces the pH of the electrolyte and brings the two sides of the battery back to the same state of charge. With the pump, the battery is expected to be able to cycle an unlimited number of times, for at least 20 years.

At Case Western, researchers have tried another approach: plating dissolved iron onto the particles in an iron slurry rather than onto a fixed electrode, so that the plated metal is stored in the battery’s external tank. It worked well in smaller cells, but in bigger cells the slurry caused clogs.

Both Case Western and ESS have received ARPA-E funding to build and demonstrate iron flow batteries. The $2.8 million, five-year grant ESS received in 2012 enabled the company to develop the proton pump and move to commercial production.

Breakthrough Energy Ventures, a fund established by Bill Gates and other investors concerned about climate change, has also backed ESS. The company sold its first product in 2015: a battery that enabled a California vineyard to store solar energy during the day and power an irrigation system in the evening.

Today ESS has a backlog of orders for its shipping-container-size battery, which has a capacity of up to 500 kilowatt-hours. The company has begun delivering some to SB Energy, a clean-energy subsidiary of SoftBank, which agreed to buy a record two gigawatt-hours of battery storage systems from ESS over the next four years. The deal is valued at more than $300 million.

Buying time

ESS batteries can currently hold four to 12 hours of charge depending on how they’re configured, but eventually some energy-storage systems may need to work for days or even weeks to accommodate seasonal fluctuations in wind power. Massachusetts-based Form Energy is developing an iron-air battery technology, which uses oxygen from ambient air in a reversible reaction that converts iron to rust. The company claims its battery could store power for up to 100 hours. Its first installation will be a one-megawatt pilot plant in Minnesota, scheduled to be completed in 2023.

Utilities aren’t just thinking about how to store energy as they move toward renewables; they’re also thinking about how to make the grid more resilient to extreme weather and other effects of climate change. Long-duration batteries have a role to play there, too.

In a project with San Diego Gas & Electric, ESS’s iron flow batteries will be paired with a solar array in the wildfire-prone town of Cameron Corners, California. If the utility needs to shut down transmission lines to prevent or respond to a fire, the solar-battery microgrid can keep the town’s critical services functioning. The project is slated to come online later this year.

ESS’s Wilsonville facility has room to ramp up production, but the number of orders it gets will depend to a large extent on the fate of clean-energy tax credits that are part of the Build Back Better bill currently stalled in Congress. Proponents of energy storage argue that long-duration storage deserves the same incentives as renewable energy.

If lawmakers agree, those credits could help make energy storage technologies like the iron flow battery cheap enough for utilities to begin using them widely. Both the ARPA-E program and the US Energy Department’s Long Duration Storage Shot aim to have cost-competitive systems that can store 10-plus hours of energy on the market within a decade.

For ARPA-E, that means getting the levelized cost of energy storage—which takes into account all costs incurred and energy produced over a lifetime—down to less than five cents per kilowatt-hour, Litzelman says, which would be a 90% reduction from 2020. The initial cost of a battery is just part of that equation.

Flow batteries aren’t the only promising technology being developed for long-duration energy storage. Other companies and researchers are experimenting with different types of batteries, as well as with hydrogen storage and mechanical systems such as compressed air or “mobile masses” that are hoisted and lowered to convert electrical energy to kinetic energy. One experimental system funded by ARPA-E stores energy by pumping water into rocks, and extracts energy when the water gets squeezed back out.

All these systems have a shared goal, says Litzelman: “24/7 clean energy.” Getting there will very likely require multiple new storage technologies, and many more companies will have to reach the point where ESS is today. Unless, of course, a different kind of technology breaks through.

Deep Dive

Climate change and energy

How fish-safe hydropower technology could keep more renewables on the grid

Natel Energy is trying to design turbines that are safer for fish passing through.

AI is an energy hog. This is what it means for climate change.

How worried should we be about AI’s effects on the grid?

Here’s the problem with new plastic recycling methods

Technology is giving us more options for plastic waste, but new methods are still far from perfect.

Why bigger EVs aren’t always better

The world is moving toward larger vehicles, and EVs are following the trend.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.