Gary Gill wants a shot at the biggest uranium deposit in the world. So do dozens of other researchers who one day may open up a mine they see as virtually inexhaustible: the ocean.

There is enough uranium dissolved in seawater to power nuclear power plants for centuries. Scientists have known this for decades, but the problem has been how to collect highly dispersed uranium molecules.

Gill, a researcher at the United States Department of Energy’s Pacific Northwest National Laboratory (PNNL) has a vision of a vast underwater field of fibers waving in the currents, gradually pulling uranium molecules out of the water. Others envision chains of wiffle-ball-like spheres doing the same, while dangling off the base of offshore wind turbines.

These ideas are likely decades or more away from reality, at least on any large scale, but researchers are refining the technology needed to extract uranium efficiently. Their aim is to close the gap between what it costs to mine the metal on land and what it might cost to take it out of the ocean.

A recent breakthrough recently came from Gill’s marine lab in Sequim, Washington, in partnership with Idaho-based energy firm LCW Supercritical Technologies. In the lab they were able to create 5g (0.2oz) of yellowcake – the powdered form of uranium used in nuclear power plants – collected from seawater.

If this extraction process can be scaled up, it would open up what Gill calls an “essentially renewable” uranium reserve, potentially eliminating the need for dangerous and environmentally damaging uranium mining.

Economics, however, dictate that this won’t be happening anytime soon, and if it eventually does, there could be as-yet-unknown environmental impacts on ocean ecosystems. Nuclear energy, too, is a controversial low-carbon power source for many reasons that go beyond uranium mining, so it’s long-term future is also unclear. But there could also be environmental benefits from the technology, even beyond the reduction in mining, such as using the same methods to clean up toxic metals from polluted waterways.

One certainty is that there is plenty of uranium floating around the sea. It could be as much as 4 billion metric tons, or 500 times more than estimated terrestrial uranium reserves, according to the United States Nuclear Energy Agency and the International Atomic Energy Agency. That’s about a 80,000-year supply based on current global consumption rates, said Gill.

“That borders on a renewable supply,” he said, adding that it is also gradually replenished by rivers. But the uranium is so dilute in the ocean – just three to four parts per billion – that collecting enough of it to be of any consequence is difficult and expensive. Salt, by comparison, is about 35 parts per thousand in seawater.

Still, like gold panners hoping to sift through enough river water to strike it rich, governments have dreamed for a long time of extracting that marine uranium lode. A number of companies undertook the initial research in the 1950s, before Japan made a major push, even operating an “experimental marine uranium adsorption plant” in the 1980s.

The U.S. Department of Energy (DOE) became interested around 2012, first reproducing the Japanese technologies, then trying to surpass them. They’ve done that. Other U.S. researchers are also working on similar variations of uranium extraction technologies, including at Stanford University and the University of Maryland. The majority of research efforts rely on fibers coated with a substance called amidoxine that, as the Japanese discovered in the 1990s, can attract the uranium molecules in the seawater. What’s key is that the absorbent properties of the fibers can be reversed, allowing the uranium sticking to them to be released, collected and processed into yellowcake.

Even though researchers’ collection materials differ, the science of what they are working on is very similar, said Travis Dietz, a University of Maryland materials science researcher working on one project. The Stanford researchers send electric pulses down a conductive fiber to collect more uranium. The University of Maryland researchers use a nylon fiber. PNNL rely on an acrylic one.

The PNNL initiative has recently been able to cut out a step in the process to reduce costs and improve the efficacy of the fibers by about a factor of three beyond what the Japanese accomplished, said Gill. The next step is to test their uranium-gathering technology outside the lab. They hope to do that in the Gulf of Mexico – they’ve found they can collect more uranium in warmer waters than in colder. They hope to do that next year and are currently seeking the funding and permits.

Right now, Gill said, one of the early analyses of the DOE program’s work was that it would cost a little over $1,000 to get a 1kg (2.2lb) of uranium out of seawater. Now, he says, he estimates it as a little over $200 per kilogram ($440 per pound).

But even if costs have fallen, to extract uranium from seawater “on a commercial scale you would have to demonstrate that the cost would be competitive with land-based sources,” said Edwin Lyman, a senior scientist at the Union of Concerned Scientists who works for the nonprofit’s global security program. A 2017 study published in the journal Progress in Nuclear Energy found that extraction of uranium from seawater would reach an economical “tipping point” when uranium prices are consistently $175–$250 per pound.

But in the wake of the 2011 Fukushima nuclear power plant disaster in Japan, many nuclear projects were canceled. With a glut of uranium on the market, prices plunged from $65 per pound before the nuclear meltdown and now hover between $18 and $24 per pound.

Lyman said he supports the research, but “right now uranium prices are pretty near historic lows. It could be many decades or even centuries for prices to change and make it economical … Industry today is not at all interested in this sort of research because they’re focused on the bottom line of the next couple years.”

He sees seawater extraction technology as “a sort of backstop … So if we start to deplete uranium sources on land, we’ll know there’s still this virtually inexhaustible resource out there.”

Dietz said, “The timeline very much depends on the cost of the technology. Right now, uranium is fairly cheap coming from terrestrial sources,” and current stockpiles of uranium should last at least 120 years. In the meantime, however, Gill noted that the technology works on more than just uranium.

“It picks up other things, as well,” he said, including vanadium, nickel, copper, cadmium and zinc. In large enough quantities, those elements – sometimes, ironically, the by-products of mining – can be toxic in waterways, and the fibers could in the shorter term be deployed to clean up polluted waters, he said.

Still, the long timeline is just fine for those worried about how a deployment of the uranium-absorbing plastic fibers on a vast commercial scale might affect the ocean and coastal ecosystems.

Gill imagines a deployment resembling a field of kelp, tethered to the seafloor and rising up to within 15m (50ft) of the surface. But that field would have to be vast – roughly the size of three Manhattan islands to power a 1,000MW nuclear reactor for a year. Installed in coastal waters, where it would be easier to harvest the uranium and replace materials, finding space could get tricky. And how that would affect the countless marine species that live in those ecosystems – as well as the impact on commerce and and recreation – remains unknown.

“A thorough examination of likely environmental effects and legal interactions would be essential before any significant deployment of this technology,” the 2017 study said.

Dietz says the prevailing concept now involves something that looks like hollow balls stuffed with clumps of the uranium-absorbing fibers, linked up in a chain, and moved through the water below offshore wind turbines on a sort of slow conveyor belt.

But even though such a setup might take up less space and could be further out at sea, it would still require extremely large amounts of plastic fiber. Plastic microfibers are already a major problem in the oceans, sloughing off fleece jackets and other products. “The biggest concern right now is having just a lot of plastic,” Dietz said, though some groups have explored using organic polymer materials derived from shrimp shells. “Material accountability is a big part of the research that is going on,” he said.

Changing the ocean’s biogeochemistry, at least, wouldn’t appear to be too problematic. “Uranium is one of the elements that aren’t required for biological life,” said Gill.

The biggest, most obvious barrier to such technology is simply that anything to do with nuclear technology is controversial and poses other safety, security and environmental risks. “Some have described [seawater extraction] as making nuclear renewable,” said Matthew McKinzie, director of the Natural Resources Defense Council’s nuclear program. “That’s not my perspective.”

While extraction from seawater could one day reduce the many problems of mining uranium on land, McKinzie said, it doesn’t address the other “unsolved problems” nuclear poses, most notably what to do with radioactive waste, how to prevent accidents like Fukushima and how to keep uranium out of the hands of terrorists. Although the NRDC supports energy research and development, including nuclear, he would like to see research in renewables, energy efficiency and upgrading the electric grid given a higher priority.

“There’s no reason not to continue research in nuclear energy with the hope of a breakthrough,” McKinzie said. “But if you switched out the very front end of what’s called the nuclear fuel cycle, you still haven’t changed everything to follow.”

This article was corrected to reflect that current uranium stockpiles are expected to last at least 120 years (not 50-100) and that Dietz’s research at the University of Maryland does not use amidoxine.