In May 2026, in the town of Big Stone City, South Dakota — 300 kilometers west of Minneapolis, amid cornfields — one of the largest battery installations in history went online. Capacity: 5 gigawatt-hours, roughly 1600 industrial Tesla Megapacks rolled into one. The domestic dream of a 13-kilowatt-hour Powerwall looks like a button cell beside it. And all of this is stored not in lithium, not in sodium, not in chemistry at all: more than 200 steel containers, packed with stacks of solid carbon, heated by electricity beyond what most metals can withstand in industrial furnaces. Put simply: a giant brick kiln fired by wind.
And fuel has an unfamiliar meaning here. The carbon blocks do not burn — they glow. Heated to incandescence, they store energy the way a cooling iron or cast-iron radiator does: as pure heat. The philosophy is simple: when prairie winds spin turbines for nothing and electricity on the market costs almost zero, the surplus goes into the furnace. During the day, when the POET plant — the world’s largest biofuel producer — requires a continuous supply of industrial steam, heat is drawn off in precisely the volume needed. The “battery” charges in waves, discharges in a trickle — around the clock, days ahead.
The most charming thing about this high-tech is its complete lack of novelty as such. Or rather, all the novelty lies in the scale.
In postwar Britain of the late 1940s, power stations ran into a problem: cities hummed and consumed by day, the country slept at night, but shutting down the boilers was more expensive. Thus the Economy 7 tariff was born: at night, a kilowatt cost several times less, and people installed a brick cabinet with a heating element in the hallway, where it roasted itself all night and released its warmth into the room in the morning. Dimplex released the first mass-produced “night storage heater” in 1961; at their peak, in the late ’80s, they sold a million a year. Grandma’s “little heater” beneath the windowsill is exactly the same technology now sitting at a biofuel plant in South Dakota: heating element + brick + cheap nighttime tariff. Only the scale has grown by six orders of magnitude, and the brick has been replaced with more energy-dense carbon.
The engineering moral: if an idea works for three generations in homes, it is ready to work in industry.
The next layer is what turned a “stove” into a battery capable of returning not just heat, but electricity. It is thermophotovoltaics, TPV — an idea so audacious that saying it aloud 20 years ago felt embarrassing. What if, instead of heating steam and spinning a turbine, you simply watched incandescent carbon? It glows, after all. So put a photovoltaic cell opposite it — like a solar panel, only facing a furnace.
Twenty years ago, TPVs had a laughable efficiency — single-digit percentages; then the figure got stuck at 30% for a full 40 years, since 1980. The breakthrough came in April 2022: Asegun Henry’s team at MIT, together with NREL, assembled a two-layer TPV from exotic III-V semiconductors with a gold mirror at the bottom. The mirror sent unabsorbed low-energy photons back into the incandescent tungsten emitter — where they heated it again instead of leaking away. Efficiency exceeded 40% — from a source at 2400 °C. For comparison: the average steam turbine has converted around 35% of heat into electricity for a century, while the best heat engine in history has only reached 60. TPV is solid-state: nothing to move, nothing to break. And Henry sees a clear path to 50%.
Henry’s idea was magnificently cynical: do not try to make turbine materials heat-resistant — get rid of the turbine. Heat graphite to 2400 °C if you like (carbon remains solid up to 3500), then place a photovoltaic cell in front of it.
Today, this is already a market, and it looks more interesting than any crypto project.
Antora Energy (Sunnyvale, California) is the flagship. Founded in 2016, its early money came from ARPA-E under the first Trump administration, then Breakthrough Energy Ventures, Shell Ventures, BHP Ventures; in February 2024, a $150 million Series B. Its record: a commercial module storing heat above 1800 °C — the hottest full-scale thermal storage unit on the planet. And in 2023, the company launched the world’s largest production line for... TPV cells. A photovoltaic-cell factory built to look at a furnace — now that is what I call strategic positioning.
Rondo Energy took another route: no exotic materials, refractory brick at just over 1000 °C, steam — and its first major contract... in oil. Its 100-megawatt-hour battery in Kern County, California (October 2025), heats steam for injection into oil wells instead of using a gas boiler — saving around 13,000 tons of CO2 a year, while extracting oil. The company calls it a necessary compromise: a paying customer now matters more than reputational purity. A Faustian bargain, but the arithmetic is convincing.
Australia’s MGA Thermal conceals a phase transition in an alloy with a miscibility gap — but that is already a neighboring episode, as is the Xcel Energy and Google contract for 30 GWh of storage using Form Energy’s iron-air batteries: rust as a battery.
The South Dakota project itself is interesting less for its physics than for its financial architecture. From the first stake in the ground to energy delivery — less than 12 months. The project’s sole outside investor was Grok Ventures — the private fund of Mike Cannon-Brookes, Atlassian’s co-founder, the man behind the Australia–Singapore solar megaproject Sun Cable and a combatant in the fight over AGL with another billionaire, Andrew Forrest. (This Grok has nothing to do with Elon and his Grok — it is a coincidence of names that trips up even attentive readers.)
The subtlety lies in the contract: Antora sells POET neither electricity nor “capacity,” but heat and steam themselves, under a long-term heat offtake agreement. That turns the battery into an industrial boiler with someone else’s meter. Add to that a special tariff that Antora, together with the local utility Otter Tail Power, negotiated from the state commission: charging during moments of local surplus, “without raising costs for other customers.” There it is, the dark art of regulation that strangled the whole idea for 10 years: technically, everything was already working in 2020; the obstacle was the tariff structure.
The closed-loop irony of the ending: the installation that makes a biofuel plant cheaper is itself assembled from carbon warmed by wind, and its economics pencil out only because surplus electricity is nearly free. Wind creates surplus → carbon stores it → steam drives the plant → the plant makes fuel. The cycle closes beautifully.
Skepticism, as usual, lives in the details. Thermophotovoltaics are still expensive when it comes to returning electricity: III-V semiconductors cost several times more than silicon, so at Big Stone Antora sells heat only — TPV will wait for future products. The industrial-heat market is conservative to the point of paranoia: a plant that goes down for a day because of a temperamental steam supplier loses more than an entire year’s savings. And above all: the main engineering risk of any thermal battery is not the block itself, but the interfaces — linings, radiation windows, heat-extraction systems. Brick holds heat beautifully; the weak point is always the joint.
But the trend can no longer be stopped: at a moment when data centers alone consume more than small countries, and Xcel and Google order 30 GWh of storage under a single contract, a “battery made of bricks” ceases to be an oxymoron and becomes a load-bearing structure of the new energy system.
Grandma’s nighttime little heater turned out not to be a relic. It was simply waiting for wind power and a customer with a factory.
Sources: POET and Antora Energy press release, “Landmark project rapidly deploys over 200 of Antora's thermal batteries...” (Big Stone City, May 19, 2026, antora.com/insights/big-stone-release); pv-magazine Australia — “Grok Ventures backs 5 GWh thermal energy storage project in US” (May 21, 2026); RenewEconomy — “Cannon-Brookes bankrolls one of biggest batteries in the world in huge bet on thermal storage” (May 2026); Science, Robert F. Service — “‘Thermal batteries’ could efficiently store wind and solar power in a renewable grid” (April 13, 2022); MIT News — “A new heat engine with no moving parts is as efficient as a steam turbine” (April 13, 2022; LaPotin et al., Nature 2022); Antora Energy press release, “System launch” (September 12, 2023); Energy-Storage.News — “Thermal energy storage startup Antora raises US$150 million for industrial decarbonisation” (February 27, 2024); Canary Media — “Rondo Energy turns on first major thermal battery — at an oil field” (October 22, 2025); storageheaters.com — “History of storage heaters and the economy 7 tariff”; antora.com/technology.