In a remote field in Saskatchewan, not far from the border with North Dakota, the Boundary Dam coal power plant sends its exhaust into a massive system of pipes and tanks that absorb CO2 and then compress the gas so it can be shipped off to a nearby oil producer for injection underground. The carbon capture system, which cost more than $1 billion to build, first started running nine years ago. It’s currently the only large-scale facility of its kind in use at a power plant anywhere in the world.

The EPA’s new proposed rules for power plant pollution—the first regulation of CO2 emissions at existing power plants, under the Clean Air Act—could change that. Under the rules, coal plants and some large gas plants in the U.S. will have to steeply limit pollution. The rules don’t say that fossil power plants have to use carbon capture; some utilities might choose to switch to renewables. (Thanks to the EPA v. West Virginia Supreme Court case last year, the agency can’t require utilities to use renewable energy.) But the new limits are based in part on what’s theoretically possible to do with carbon capture technology. Affected power plants will have to cut emissions by 90% by the 2030s.

The big question: Can it actually work?

Some environmental groups are skeptical. “Billions of dollars have been wasted trying to prove that this technology is real—and all we have to show for it are a series of spectacular failures,” Wenonah Hunter, executive director of Food & Water Watch, said in a statement when the EPA rules were unveiled last week. “Throwing good money after bad is not a climate solution—it’s an industry bailout.” On the other side, a coal power association, America’s Power, also argues that the technology isn’t economically or technically viable. In the past, multiple projects have flopped.

More than a decade ago, a carbon capture project at a large coal plant in West Virginia was canceled partly because of the cost, even though the federal government planned to foot half the bill. Other planned projects also failed in North Dakota, Indiana, Wyoming, Texas, and elsewhere. In Mississippi, a “clean coal” project that planned to capture emissions was canceled after costs more than tripled to $7.5 billion (the project also secured nearly $1 billion in federal grants and tax credits). Nearly 90% of carbon capture projects proposed in the power sector have failed, largely for financial reasons, according to a report from the Institute for Energy Economics and Financial Analysis.

At the Boundary Dam Power Station operating in Canada, the technology is designed to capture up to 90% of the plant’s CO2 emissions. In practice, the numbers have sometimes been much lower. When the plant first started running, the equipment got clogged with fly ash. In 2021, issues with the motor meant that the system stopped working for months, and it captured only 44% of the plant’s annual emissions.

Still, there’s proof that the basic technology works, says Jay Duffy, litigation director at the nonprofit Clean Air Task Force. There have been startup issues with early projects, he says, “but they were first-of-a-kind plants and significant learnings and advances have occurred since then.” In the first quarter of this year, Boundary Dam’s carbon capture system was available to run 93% of the time, capturing more than a quarter million tons of carbon.

“It’s still a very young industry,” says Scott Sklar, sustainable energy director at George Washington University’s Energy and Energy Management Institute. “It takes time to build a learning curve. Any conclusions that it won’t work—it’s still really too early to tell.” Sklar used to run the Solar Energy Industry Association, and says people initially had the same doubts about solar.

There are reasons why carbon capture isn’t an ideal solution. Boundary Dam, like some other proposed projects, sells the CO2 that it captures to an oil producer, which uses it to squeeze more oil out of old wells. The CO2 is sequestered underground (in theory, permanently), but it’s supporting another polluting industry. (Petra Nova, another large-scale, billion-dollar carbon capture project at a coal plant in Texas, shut down in 2020 as the price of oil plunged, though it may soon reopen.) And even if it’s possible to capture nearly all of the emissions at a power plant, coal mining and gas extraction add even more emissions earlier in the process.

From a climate perspective, if billions of dollars are going to be invested now, it makes more sense to invest in new renewable energy and battery storage. But because so much money has already been sunk into polluting infrastructure, it’s unlikely to immediately disappear, says Sklar. “I see [carbon capture] as an interim approach to deal with the issue,” he says.

The technology isn’t widespread now because the regulation hasn’t been in place, say experts from the Clean Air Task Force, and utilities don’t want to add additional costs if they aren’t forced. There was a similar situation in the 1970s when the EPA required power plants to add technology to capture sulfur pollution. At the time those standards were put in place, only three plants had been using the technology. The Clean Air Act is “forward looking,” Duffy said at a press conference about the proposed rules. “A pollution-control technology need not be on every street corner in order to be the basis of standards.”

The Bipartisan Infrastructure Law and Inflation Reduction Act both include incentives to help reduce the cost of carbon capture; 45Q, a tax credit for carbon storage, has been bumped up from $50 to $85 per ton of CO2 that’s permanently sequestered. The tax credits could mean that some power plants could actually make money by capturing carbon, according to an EPA official quoted on background in Politico. The cost of carbon capture is also likely to drop as the technology gets more use and startups work on new iterations of it.

Twenty large-scale projects are in development now, according to the International Energy Agency, and 11 of those are in the U.S. If the proposed EPA rules take effect in 2024—and survive legal challenges and a potential reversal if the administration changes—it’s likely that more projects will begin. Others may use green hydrogen, though that also has issues. Some utilities are also likely to move more quickly to renewable energy and battery storage as prices continue to drop. “I think it really depends on the specific utility,” says Mona Dajani, global head of renewables, energy, and infrastructure at the law firm Shearman and Sterling. “I’m representing a number of utilities that are very forward-thinking . . . they’re really embracing clean energy.”

Correction: This article was updated to correct the date that the EPA adopted standards for sulfur pollution.