A geothermal plant pulls scalding salt water out of the ground, takes the heat, and pushes the water back down. It turns out that water is also one of the largest lithium deposits in North America.
Lithium is normally obtained in one of two ways, and neither is pretty.
In South America, brine is pumped into evaporation ponds the size of small towns and left for a year or more while the sun removes the water. In Australia, hard rock is dug out of open pits and crushed. One consumes enormous quantities of water in places that have very little. The other leaves a hole you can see from orbit. Both are why almost every proposed lithium project spends longer in front of a planning committee than in construction.
At the southern end of California's Imperial Valley, beneath the Salton Sea, sits salt water at around 300 degrees Celsius that is unusual for two separate reasons. It is hot enough to run a power station. And it carries a heavy load of dissolved lithium.
Geothermal plants have been pumping that brine to the surface for decades to make electricity, taking the heat, and returning the water underground. The lithium went back down with it.
Here is what happened
Construction began on the first large-scale plant of its kind in the United States. Ground was broken in Imperial County in January 2026 on the first commercial-scale direct lithium extraction facility in what is now routinely called Lithium Valley.
The US Department of Energy backed it with a loan commitment of up to 1.36 billion dollars. The conditional commitment from the Loan Programs Office is for Project ATLiS, a subsidiary of Energy Source Minerals.
ATLiS is designed to make about 20,000 tonnes of lithium hydroxide a year. That is roughly enough for 52 gigawatt hours of lithium-ion batteries annually.
A second company is building separately in the same field. Controlled Thermal Resources is developing Hell's Kitchen, an integrated lithium and power project targeting around 25,000 tonnes a year of battery-grade lithium hydroxide monohydrate.
The resource is unusually large. Department of Energy analysis has estimated the wider Salton Sea geothermal area may hold more than four million tonnes of lithium, enough in principle for over 10,000 gigawatt hours of battery capacity. That is a resource estimate, not a proven reserve.
How it works
The brine is already coming up. A geothermal plant lifts hot water from around two kilometers down, flashes it to steam to spin a turbine, and returns the cooled fluid to the reservoir. That cycle exists whether anyone wants the lithium or not.
Direct lithium extraction sits in the middle of it. After the heat has been taken, the brine passes through a material that selectively grabs lithium ions and lets everything else through. The lithium is then washed off and processed into battery-grade lithium hydroxide.
The water goes back down the hole it came from. No evaporation ponds, no year of waiting, and no net consumption of water, which in a drought-stressed valley is not a detail.
Selectivity is the whole engineering problem. This brine is dense with competing ions, and a filter that grabs sodium, potassium and calcium along with lithium produces an expensive mess. Every company in the field is really competing on the chemistry of that one step.
The economics inherit an existing asset. Conventional mining pays for the whole operation out of the metal. Here the pumping, the wells and the plant are already paid for by selling electricity, so the lithium only has to cover the extraction step. That is the argument, and it is the part that has never been tested at scale.
Why it matters
The waste stream turns out to be the product. Spent geothermal brine has always been something to dispose of. Nothing about the geology changed. What changed is what we noticed was dissolved in it.
Two industries collapse into one site. Power generation and mining have never shared a facility, let alone a working fluid. If it works, the same well is a power station and a mine at the same time.
It answers the objections people actually raise. The complaints about lithium are water use and open pits. This route has neither, which changes the politics of permitting as much as it changes the cost.
Battery supply is the binding constraint. Electric vehicles and grid storage both queue behind the same material, and we covered the search for alternatives in the battery built to break lithium's grip. Domestic lithium from a resource already being pumped attacks the same problem from the other end.
Geothermal gets a second revenue line. We wrote in how to feed a data center by drilling into 460 degree rock about geothermal finding a customer in AI. Lithium is a second one, and a plant with two products is a different investment case from a plant with one.
The honest catch
Nothing here has been proven at commercial scale, and the technology has a habit of being harder than it looks.
No commercial lithium has been produced from this brine yet. Both projects are building. Direct lithium extraction has been demonstrated, which is a very different thing from operating profitably for years.
The two projects are separate and should not be conflated. ATLiS belongs to Energy Source Minerals. Hell's Kitchen belongs to Controlled Thermal Resources. Different companies, different plants, different timelines, one shared resource.
Four million tonnes is an estimate of what is there, not of what can be got out. Recovery rates, brine chemistry and reinjection behavior will determine the real figure, and none of that is settled.
Reinjection is not risk free. Putting altered brine back into a reservoir at volume raises questions about scaling, corrosion, induced seismicity and long-term reservoir behavior that only years of operation will answer.
The lithium price does not sit still. A project financed against one price and commissioned into another is the ordinary way mining ventures fail, and government backing changes who bears that risk rather than removing it.
EDITOR'S TAKE
The part worth sitting with is that nothing was discovered. The lithium has been circulating through those wells for as long as the plants have run, and the brine has been pumped, cooled and reinjected past it thousands of times. What changed is that somebody stopped treating the fluid as a heat carrier with an inconvenient chemistry and started treating it as an ore with a useful temperature. That is a reframing, not an invention, and reframings tend to be worth more. The number to watch is not the four million tonne resource estimate, which is geology, but the operating cost per tonne once the first plant has run a full year. If a geothermal operator can produce lithium as a side effect of selling electricity and still make money, every hot brine field in the world gets re-examined. If it cannot, this becomes another demonstration plant with a good story attached.
Quick questions
What is direct lithium extraction, and why is it different?
It is a way of pulling lithium straight out of brine using a material that selectively grabs lithium ions while letting the other dissolved salts pass. The conventional alternative is to pump brine into vast evaporation ponds and wait a year or more for the sun to remove the water, concentrating the lithium as it goes. Direct extraction takes hours rather than months, uses a fraction of the land, and returns the water rather than losing it to the air. The difficulty is selectivity: geothermal brine is crowded with sodium, potassium and calcium, and a material that captures those alongside lithium produces something expensive to clean up. That single step is where the competition in this industry actually sits.
Does taking the lithium out affect the power generation?
In principle it should not, because the extraction happens after the heat has already been harvested. The plant lifts the brine, converts its heat to electricity, and only then does the fluid pass through the lithium step before being reinjected. Electricity and lithium are therefore sequential rather than competing uses of the same resource. The open questions are about the return leg: whether removing lithium and adding processing chemicals changes how the brine behaves when it goes back underground, whether it accelerates scaling or corrosion in the wells, and whether reservoir pressure and temperature hold up over decades of operation. Those are engineering questions that only sustained commercial operation will answer.
How much lithium could the Salton Sea actually supply?
Department of Energy analysis has suggested the resource may exceed four million tonnes, which would correspond to more than 10,000 gigawatt hours of battery capacity. That figure should be read carefully. It is an estimate of lithium present in the reservoir, not of lithium that can be economically recovered, and the gap between those two numbers is where mining projects usually live or die. The plants under construction are far smaller: around 20,000 tonnes a year at Project ATLiS and about 25,000 tonnes at Hell's Kitchen. Those are meaningful volumes for domestic supply, and they are a small fraction of the theoretical resource.
Sources
US Department of Energy, Loan Programs Office: the conditional commitment for Project ATLiS, with production capacity and project detail.
ThinkGeoEnergy: reporting on the 1.36 billion dollar loan and the geothermal context.
Controlled Thermal Resources: the Hell's Kitchen integrated lithium and power project, a separate development in the same field.
Frontier Signal explains frontier technology in plain English. Company and government figures should be independently verified. This is general information, not investment advice.