A Utah copper mine pulls about 20 tons of tellurium a year out of the sludge left over from refining, and that metal goes into solar panels. More than 71,000 of them now cover 210 acres of the mine's own land, powering the operation the sludge came from.

Tellurium is one of the scarcer things you can pull out of the ground. It sits at about 3 parts per billion in most rock, which the same source puts at roughly eight times rarer than gold. You don’t go looking for it directly. You get it as a leftover, mostly from copper, and if nobody bothers to recover it, it simply washes away with the rest of the refining waste.

For years, a lot of it did exactly that.

What makes the setup at Rio Tinto’s Kennecott mine near Salt Lake City worth a look is that it closes a loop most industrial processes never bother to close. The sludge left over from refining copper becomes tellurium. That tellurium becomes solar panels, which sit on the mine’s own land and feed power back into the operation that produced the sludge.

It’s neat enough that it almost sounds like marketing. But the mechanism is real, and it points at something genuine about where critical minerals actually come from.

The loop, plainly

Let’s start with the waste. When copper is refined using electricity, it leaves behind a residue at the bottom of the tank called anode slime. Kennecott began pulling tellurium out of that slime in 2022. Rio Tinto’s copper chief operating officer, Clayton Walker, said at the time, “We are proud to deliver a new domestic supply of tellurium to support the manufacturing of solar panels and other critical equipment here in the United States.” The recovery setup cost about $2.9 million and yields about 20 tonnes a year. 

From there the metal travels. It’s refined by 5N Plus in Canada and supplied mostly to First Solar, the biggest maker of a type of thin-film solar panel that uses cadmium and tellurium.

Then some of it comes home. In December 2025, Rio Tinto switched on a new 25MW solar array, which contains over 71,000 panels made with Kennecott’s own tellurium.

Rio Tinto’s Kennecott managing director, Nate Foster, described how the company sees it: “This new solar plant is more than a source of renewable power for our operations; it’s a demonstration of circularity and supply chain resilience.” That’s the operator’s own language, so take the promotional edge off it. What’s not in dispute is the physical chain: copper waste in, tellurium out, panels on the roof of the same enterprise.

Why 20 tons matters more than it sounds

Twenty tons is  what they poduce and it’s a rounding error next to most commodity figures. The strategic position is what gives it weight. China refines the great majority of the world’s tellurium, somewhere around 75% by the USGS count. As recently as 2021, the United States relied on imports for 95% of its tellurium. 

The reason so little tellurium gets recovered is that most producers don’t build the equipment for it. Walker’s point in 2022 was that the resource is already sitting in the copper stream. He put it as “Approximately 90 percent of the world’s tellurium resource is contained in copper ore.” How much goes uncaptured is striking. A USGS-led study, by Nassar and colleagues, estimated that in 2018 the tellurium sitting in copper sludge worldwide was close to four times the amount actually reported as produced. Most of it is thrown away.

That gap is the interesting part. Kennecott didn’t find a new deposit. It stopped discarding something it was already handling. The tellurium was always in the sludge. Building the equipment to catch it is a small decision with an outsized effect on where a strategically sensitive metal comes from.

What “circular” actually buys, and what it doesn’t

The loop is real, but 30MW of solar does not power a copper mine. Kennecott is one of the largest industrial operations in Utah, and 30MW is a slice of its demand, not the whole thing. Rio Tinto’s own figures put the two plants’ combined output at roughly what would power 1,026 homes, cutting site emissions by about 6%, some 20,000 tonnes of CO2 equivalent. Meaningful, but a fraction.

The panel-supply side is similar. The Department of Energy notes tellurium makes up only 10 to 15 percent of the cost of one of these panels, so 20 tons a year is a modest input to a large industry, not a supply that reshapes the panel market. The loop isn’t fully domestic either. The tellurium detours through Canada for refining before returning to the United States as finished panels. North American, yes. Wholly domestic, no.

A template, not a triumph

Strip the marketing away and what remains is a genuinely instructive model. Recovering a metal as a byproduct is one of the few ways to add critical-mineral supply without opening a new mine. 

A single mine catching 20 tons of a scarce metal it used to throw away, and using some of it to power itself, won’t rewire a global supply chain. What it does show is that a fair amount of critical-mineral capacity is hiding in waste streams that already exist. Whether the Kennecott model spreads depends on how many other copper refineries decide the sludge is worth keeping.