Carbon capture—dream or nightmare—could be coming. Or not.

Editor’s note: This story was originally published by Yale e360. It appears here as part of the Climate Desk collaboration.

In early September, at an industrial facility located about 25 miles southeast of Reykjavik, Iceland, the Swiss company Climeworks will mark the opening of a new project named “Orca.”

At least in a conventional sense, Orca doesn’t actually make anything. It is comprised of eight elongated boxes that resemble wood-clad tanks. Each of these boxes—known as “collectors”—is roughly the size of a tractor trailer, and each is festooned with 12 whirring fans that draw a stream of air inside. Within the collectors, a chemical agent known as a sorbent will capture carbon dioxide contained in the air wafting through. Periodically, the surface of the sorbent will fill up. And at that point the carbon dioxide trapped within it will need to be released. At Orca, this task is accomplished with a blast of heat, which is sourced from a nearby hydrothermal vent. The extracted carbon dioxide will then be piped from the collector boxes to a nearby processing facility, where it will be mixed with water and diverted to a deep underground well.

And there it will rest. Underground. Forever, presumably. The carbon dioxide captured from the Icelandic air will react with basalt rocks and begin a process of mineralization that takes several years, but it will never function as a heat-trapping atmospheric gas again.

Climeworks maintains that Orca, once it’s running around the clock, will remove up to 4,000 metric tons of carbon dioxide from the atmosphere each year. And there isn’t much reason to doubt the facility can achieve this feat. For one thing, the technology for the plant, known as direct air capture, or DAC, is a variation on ideas that have been utilized over the course of half a century in submarines and spacecraft: Employ chemical agents to “scrub” the excess carbon dioxide out of the air; dispose of it; then repeat. More to the point, perhaps, is the fact that Climeworks has already built smaller, less sophisticated plants in mainland Europe, which have in turn pulled hundreds of tons of carbon dioxide per year from ambient air.

What seems most significant about Orca, then, is how it represents the possibility that direct air capture has moved closer to something resembling a commercial business. Climeworks now has dozens of customers—individual consumers who have purchased carbon removal services directly from the company, as well as corporations, like the insurance giant Swiss Re—who will pay for the permanent carbon offsets that will be buried underneath Icelandic soil. What’s more, the Orca facility will be the largest functioning direct air capture plant in the world to date—by the company’s estimation, it represents a “scale-up” of its carbon removal efforts by about eighty-fold over the course of four years.

And yet, Climeworks and Orca will likely soon be eclipsed. Plans for even larger DAC plants—one in the US Southwest, slated for completion at the end of 2024; another in Scotland, to be finished about a year after the American project—will be built by a competitor, Carbon Engineering, of British Columbia. Employing a somewhat different technology, Carbon Engineering’s facilities, as initially planned, will be powered by renewable energy and will eventually each remove, on net, about a million metric tons of carbon dioxide a year from the atmosphere.

“In our view, this will decisively answer the question: Is direct air capture feasible at large scale and affordable cost,” Steve Oldham, the CEO of Carbon Engineering, told me recently. “As I see it, we are out of academic research and feasibility and now into engineering reality.”

One way to consider the global value of these efforts is to place them within the humbling math of climate change. In the most recent report by the Intergovernmental Panel on Climate Change (IPCC), a number of models were used to chart possible future emissions scenarios, and to make sense of how we might experience a rise of, say, 1.8 degrees Celsius or 2.5 degrees C (3.2 degrees to 4.5 degrees Fahrenheit) by the year 2100. Last year, about 31 billion metric tons of carbon dioxide were released into the atmosphere. Probably that number will rise even higher this year, as the global economy begins to recover from the Covid-19 pandemic. But to have a chance at limiting warming to 2 degrees Celsius we would have to effectively bring our emissions near to zero by around the middle part of this century.

Without question, the best way to begin doing so is to drastically transform our electrical, transportation, and industrial systems to emissions-free energy sources and processes. However, we may need to actively compensate for economic sectors—air travel, for instance, or steel production— hat prove too hard to rapidly decarbonize. This would mean we would have to actively take carbon dioxide out of the atmosphere at the same time. Our carbon removal efforts could involve natural means, such as sequestering atmospheric carbon dioxide in soil or new forests. But we could also utilize more technological approaches, such as DAC or bioenergy with carbon capture and storage—known by the acronym BECCS—which involves growing plants, burning or fermenting them for energy, and then capturing the carbon dioxide emissions and burying them.

And here the numbers get daunting. Zeke Hausfather, a climate scientist at the nonprofit Berkeley Earth involved in charting possible emission pathways for the IPCC report, told me that in one of the most optimistic scenarios, which limits temperature rises to 1.5 degrees Celsius by 2100, the conditions require massive mitigation efforts as well as about 17 gigatons—that’s billions of tons—of carbon dioxide removal per year by the end of the century. And as much as planting trees might seem an ideal solution to help us reach such a goal, new forests are likely not a sufficient or durable carbon removal solution, especially in the wake of huge wildfires in Siberia and the American West. “Natural ways of removing carbon dioxide are generally less desirable than long-term geologic storage through BECCS or DAC,” Hausfather says, adding that this is because storing carbon above ground in biomass is most likely temporary.

Whether DAC can make a meaningful contribution to carbon removal goals remains a lingering question. But the new Climeworks and Carbon Engineering plants suggest significant progress, not just hype. “You’ve got these two companies that are ready to go today,” Jennifer Wilcox, an official at the US Department of Energy (DOE) and an expert in carbon capture technologies, told me. “But the question is, how do they get from thousands of tons to millions of tons?” After that, of course, comes an even bigger question: Could they actually get to billions?

In several instances over the past few years, in the course of reporting on climate solutions and carbon removal strategies, I’ve been told by venture capitalists—investors I respect for their grasp of technology, and who have a deep understanding of climate change—that direct air capture could turn out to be among the world’s biggest industries by midcentury. I would not be shocked if this turns out to be true. On the other hand, even modest growth of the DAC industry seems entirely conditional. Reducing emissions—a Herculean task for the world’s governments and industries as they begin to phase out fossil fuels — remains the primary challenge. And beyond that, whether DAC “scales up” and makes a significant climate impact depends mainly on its expense. Or to put it another way: How much will it ultimately cost to separate a metric of ton of carbon dioxide from the air and put it into the ground, or into a long-lasting product like concrete or carbon fiber? It seems a matter of consensus that if it can get nearer to $100 per ton, direct air capture might become an essential and useful technology

But we’re still not sure. A few years ago, Climeworks executives told me their cost of direct air capture was somewhere between $500 and $600 per ton. The company is not publicly estimating—and indeed may not yet know—how much the Orca plant will improve on that measurement. Still, there are physical and thermodynamic limits, according to the Energy Department’s Wilcox, that can help serve as a lower bound. Wilcox surmises that scientific constraints may make it difficult for Climeworks or Carbon Engineering plants to get much below $100 per ton to remove carbon from the air. And this may remain the case no matter how hard engineers work in the future to bring down expenses through cheaper materials and assembly-line industrialization. A recent analysis of the industry, authored by scientists at Carbon Engineering, predicts a similar outcome: An eventual cost range for DAC of between $94 and $232 per ton. That could be decades away—or it may never come to pass.

The point of new plants like Orca or Carbon Engineering’s mammoth project in the Southwest, however, isn’t to perfect the carbon removal process. The point is to take a large technological and commercial step forward. We can only speculate on what DAC technology may achieve next based on the future targets of executives at carbon removal companies. But if costs reach, say, $400 or $350 per ton in the next few years, it would suggest this remains a promising tool in need of further refinement. It may indicate this could prove to be a viable option for companies like airlines, say, or fertilizer manufacturers (or even for government entities) that may eventually be compelled to buy offsets so as to compensate for their carbon emissions.

It’s probably best to interpret the opening of the new plants as the start of a complex, multi-decade, global deployment process that follows years of research and development. “We’re confident our costs will continue to fall,” Oldham, the Carbon Engineering CEO said. “But only if we deploy. If you never deploy, your costs never go down.”

Oldham’s view, moreover, is that the world might, through epic efforts at mitigation, be able to eliminate 70 to 80 percent of emissions by 2050. “But that would leave about 20 to 30 percent of the carbon footprint we’re going to have to remove,” he says—probably the equivalent of about 10 to 12 billion tons per year of carbon dioxide. As a thought experiment, that would require 10,000 Carbon Engineering plants like the ones the company is now planning. “I think if the world sets its mind to it, we can produce many of these plants,” he said. “And we’ve done this in the past. Look at the way we scrambled for Covid vaccines. Or look at the way we scrambled for wars and got into the mass production of planes.”

One concern is that DAC might prove increasingly controversial if it erodes global efforts at mitigation. If carbon can effectively and affordably be removed from the air, in other words, it may slow the rush to eliminate fossil fuels. For now, at least, that remains a hypothetical risk. And Oldham and colleagues in his field told me they believe new state and government policies are moving his industry in the right direction. A US tax credit for companies called 45Q, for instance, is helping to subsidize some of the high costs of carbon capture and sequestration. The possible passage of a federal infrastructure bill in the coming months may likewise allocate as much as $3.5 billion to help construct large DAC plants. Meanwhile, a push from the private sector has been a boon to fledgling DAC firms. A slew of tech companies interested in becoming carbon neutral or carbon negative—Microsoft, Stripe, and Shopify are the most prominent — have invested substantial sums in Climeworks and Carbon Engineering. Their commitments have, in turn, helped the companies move forward with planning and construction.

At the same time, investment dollars are beginning to flow into “next generation” DAC ideas. The US Energy Department recently invested more than $12 million in a slew of early-stage approaches and component technologies. A number of venture firms, notably Breakthrough Energy and LowerCarbon Capital, have placed tens of millions more into startups. One new firm, San Francisco-based Noya, uses existing power plant cooling towers to create a “distributed” system of direct air capture stations that the company hopes will prove cheaper than building DAC plants from scratch; another, a Detroit-based startup known as Remora, fits carbon-capturing sorbent technology on trucking rigs to vacuum up carbon dioxide on long hauls. As a sweetener, a new $100 million X-prize, sponsored by Elon Musk, involves a four-year global competition that will reward the most promising young carbon removal firms for ideas that can be scaled up to gigatons per year.

So within the industry, there is plenty of money and plenty of enthusiasm. What is in short abundance, in light of the hottest month on record and near-term projections for future global temperatures, is plenty of time.

At this stage in DAC’s evolution, it’s worth recalling that it can be notoriously hard to predict how long it takes for technologies to mature. In the mid-1950s, just after the first practical photovoltaic solar cell was invented at Bell Labs in New Jersey, one of its inventors, Daryl Chapin, calculated it would cost around $1.5 million to deploy the devices as an electricity source on a typical American house. Nowadays, you can outfit a home with solar panels for around $20,000, according to the Solar Energy Industries Association—and that investment pays off over time through the benefit of cheaper electric bills. In some locations, solar PV is now the cheapest source of energy on the planet.

The future challenge for direct air capture technologies—the uncertain, downward arc of its cost—is therefore a familiar one. One possible future is that the DAC industry will continue to reduce expenses but may not get near enough to a price structure, such as $100 per ton, that makes it economically appealing as an offset. To Klaus Lackner, a pioneer in the direct air capture field who runs the Center for Negative Carbon Emissions at Arizona State University, an essential question is whether DAC, as it evolves, mimics the astonishing cost declines of solar photovoltaic panels and wind turbines, or whether it remains a boutique technology that crashes into inherent economic limits.

“My opinion is that if it behaves like many other mass-manufactured technologies, it is not unreasonable to assume that by growing roughly 300-fold, we should be skirting $100 per ton for carbon removal,” Lackner says. “Beyond that, my crystal ball is a little cloudy. But if this keeps growing a thousand-fold, we should be at $50, or maybe $70 or $80 per ton.”

Lackner believes that DAC may actually be at a better stage than solar photovoltaics were during, say, the 1970s, when prices were prohibitively expensive. For widespread adoption, solar technology needed to reduce costs by about 100 times, he says. DAC needs only to reduce costs by 10 times to make it desirable. He acknowledges there is no guarantee that DAC will succeed in the same way as it scales up. And he warns that even if direct air capture costs drop dramatically, the world will still need a regulatory framework for its application, so as to make a significant impact on the climate. As crucial as it may be to improve the technology, it will be equally important to compel industries and governments “to treat carbon dioxide as a waste product,” he says, and therefore pay to clean it up.

In light of this, the next few years should be telling. We may soon know whether DAC is a go—or whether the technology, often viewed by its critics as quixotic, will hit a wall of inefficiency. If it’s the former, we will have a useful tool in the climate toolbox. But if it’s the latter, it will almost certainly make the goal of achieving a livable world more complicated. The work ahead of us, already monumental in its political, technological, and economic challenges, would become even more difficult.