Here's the problem that decides whether the clean-energy transition actually works: the sun sets, the wind drops, and you still want the lights on at 9 p.m. Batteries are the obvious fix — but the lithium cells in your phone and your car get expensive fast when you ask them to hold a whole day's worth of electricity. So one of the more interesting answers doesn't use lithium at all. It uses carbon dioxide — the gas everyone's spent a decade trying to bury — squeezed into a liquid and let back out through a turbine. It sounds like a stunt. A US utility and Google just put money behind it.
What happened
On 15 June 2026, the Italian startup Energy Dome and Arizona's Salt River Project announced a 19-megawatt, 10-hour storage plant that banks energy in CO₂, co-funded by Google, targeted to come online around 2029. It builds on Energy Dome's Wisconsin project with Alliant Energy — the first commercial-scale CO₂ store in the US — due to start construction this year and finish in 2027.
The idea. Store energy not in chemistry but in pressure and heat, using CO₂ as the working fluid in a sealed loop.
The backers. A regulated utility (SRP) and a hyperscaler (Google) — exactly the buyers who need cheap, all-day power for the grid and for data centres.
The scale. 19 MW / 10 hours in Arizona; about 20 MW / 200 MWh in Wisconsin. Modest today — these are first-of-a-kind commercial plants, not a fleet.
The pitch. No lithium, cheap and abundant materials, and a duration — roughly 8 to 24 hours — where lithium's economics start to hurt.
One honest flag: this isn't proven grid infrastructure yet. The announcement is about two weeks old, the plants are first-of-a-kind, and the Arizona one isn't due until 2029. Treat it as a serious commercial trial, not a finished technology.
How a "CO₂ battery" actually works
Forget chemistry for a second — this is closer to a steam engine than to the AA in your remote. It's a sealed loop with one trick: CO₂ is easy to flip between gas and liquid.
Charging (storing power). When electricity is cheap or plentiful — think midday solar — the system uses it to compress CO₂ gas into a liquid, and carefully stores the heat that compression throws off. The gas lives in a big dome; the liquid waits under pressure in tanks.
Discharging (making power). When you need electricity back, you warm that liquid CO₂ with the stored heat and let it expand through a turbine, spinning a generator. The gas flows back into the dome, the loop reseals, and the same CO₂ gets used again and again. Nothing is burned; nothing is released.
The number that matters is round-trip efficiency: about 70% of the electricity you put in comes back out. That's lower than lithium's ~85–90% — but the parts are far cheaper, and for long-duration storage, cost beats efficiency. That trade is the whole thesis.
Why this is different
Versus lithium. Lithium is brilliant in short bursts — 2 to 4 hours — for your phone, your car, and quick grid balancing. But to store 10 or 20 hours, you just keep stacking expensive cells, and the bill climbs with every hour. A CO₂ battery's pricey bits — the compressor, the turbine, the dome — are mostly fixed; adding hours is closer to "a bit more steel and gas." The longer the storage, the better it looks.
Versus pumped hydro. The old champion of long-duration storage is pumping water uphill and letting it fall — but that needs a mountain and a reservoir in the right place. A CO₂ battery can be built on flat ground, almost anywhere the grid needs it.
The materials story. No lithium, no cobalt, no rare minerals with tangled supply chains — just steel, water and carbon dioxide. Abundant and boring, which for infrastructure is a compliment.
What it could mean
You won't install one of these. But the second-order effects are the point:
Cheaper all-day clean power. If a wind-and-solar grid can bank a full day of energy affordably, "renewables only work when the sun shines" stops being true — and that eventually shows up in your bill.
A hedge against the lithium crunch. Every EV and phone competes for the same battery minerals. Long-duration storage that sidesteps lithium takes pressure off that squeeze.
Why Google cares. Data centres — the things running your AI — need firm, around-the-clock clean power. A hyperscaler funding this is the tell: the big buyers want storage measured in days, not minutes.
What to watch next
The one number. Cost per kilowatt-hour at these first commercial plants. Undercut lithium for all-day storage and a CO₂ battery goes from curiosity to category.
Wisconsin first. Construction starts this year, completion targeted 2027 — the first real-world US scorecard.
Efficiency in the wild. 70% on paper; watch whether the built plants hold it — and whether cheaper materials make up the gap with lithium.
The rivals. CO₂ batteries are one of several non-lithium long-duration bets (iron-air, flow batteries, thermal). Watch which ones utilities actually re-order. (All of this is early — first-of-a-kind plants, years from a fleet.)
EDITOR'S TAKE
The unglamorous truth about clean energy is that the hard part isn't making the power — it's storing it long enough to matter. That's why a battery made of CO₂ is more interesting than it sounds: it attacks the exact gap — cheap, all-day storage — that decides whether a renewable grid holds up after dark. It isn't proven, and 70% efficiency is a real tradeoff. But when a utility and a hyperscaler co-fund your "commercial trial," the market is telling you it's hunting hard for a lithium-free answer. This is one of the front-runners.
Cut through the noise. Get the frontier in plain English — every weekday.
Frequently asked questions
What is a CO₂ battery?
A CO₂ battery, like Energy Dome's, stores electricity by compressing carbon dioxide into a liquid when power is cheap, then expanding it back through a turbine to generate power when it's needed. The CO₂ stays in a sealed loop, so nothing is burned or emitted. It's a form of long-duration energy storage that uses no lithium.
Is a CO₂ battery better than a lithium battery?
Not for everything. Lithium is more efficient (~85–90% vs ~70%) and better for short bursts of 2–4 hours. A CO₂ battery is aimed at longer durations (8–24 hours), where its cheaper, lithium-free materials can win on total cost — and it avoids battery-mineral supply crunches.
Who is backing Energy Dome's CO₂ storage?
Arizona utility Salt River Project and Google are co-funding a 19 MW, 10-hour plant in Arizona (online around 2029), and Alliant Energy is building the first commercial-scale US CO₂ battery in Wisconsin (around 20 MW / 200 MWh), due to finish in 2027.
Sources
This article is general information about science and technology, not investment advice. The projects described are first-of-a-kind and pilot/early-commercial scale; timelines (Wisconsin ~2027, Arizona ~2029) and the ~70% round-trip efficiency figure are as stated by the companies and had not been independently verified at scale at the time of writing.
