There’s a colorless, odorless, and largely benign gas that humanity just can’t get enough of. We produce 40 trillion kg of carbon dioxide each year, and we’re on track to cross a crucial emissions threshold that will cause global temperature rise to pass the dangerous 2°C limit set by the Paris climate agreement.
But, in hushed tones, climate scientists are already talking about a technology that could pull us back from the brink. It’s called direct-air capture, and it consists of machines that work like a tree does, sucking carbon dioxide (CO2) out from the air, but on steroids—capturing thousands of times more carbon in the same amount of time, and, hopefully, ensuring we don’t suffer climate catastrophe.
There are at least two reasons that, to date, conversations about direct air capture have been muted. First, climate scientists have hoped global carbon emissions would come under control, and we wouldn’t need direct air capture. But most experts believe that ship has sailed. That brings up the second issue: to date, all estimates suggest direct air capture would be exorbitantly expensive to deploy.
For the past decade, a group of entrepreneurs—partly funded by billionaires like Bill Gates of Microsoft, Edgar Bronfman Jr. of Warner Music, and Gary Comer of Land’s End—have been working to prove those estimates wrong. Three companies—Switzerland’s Climeworks, Canada’s Carbon Engineering, and the US’s Global Thermostat—have been working to build machines that, at reasonable costs, can capture CO2 directly from the air. (A fourth company, Kilimanjaro Energy, closed shop due to a lack of funding.)
Over the past year, I’ve been tracking the broader field of carbon capture and storage, which aims to capture emissions from sources such as power plants and chemical factories. Experts in the field look at these direct air capture entrepreneurs as the rebellious kids in the class. Instead of going after the low-hanging fruit, one expert told me, these companies are taking moonshots—and setting themselves up for failure.
Climeworks just proved the cynics wrong. On Oct. 11, at a geothermal power plant in Iceland, the startup inaugurated the first system that does direct air capture and verifiably achieves negative carbon emissions. Although it’s still at pilot scale—capturing only 50 metric tons CO2 from the air each year, about the same emitted by a single US household—it’s the first system to take CO2 in the air and convert the emissions into stone, thus ensuring they don’t escape back into the atmosphere for the next millions of years.
The impossibility of direct air capture can be illustrated by simple physics. Imagine if you were allowed to eat as many M&M’s as you wanted—as long as you only eat red ones. If in a bag of M&M’s there was one red M&M for every 10 pieces of candy, it would be easy to find them and eat them with glee. But imagine if the concentration fell to one in every 2,500. You might give up searching for even a single M&M.
At a coal-power plant, the exhaust flue gas contains about 10% carbon dioxide (i.e., about one in 10 gas molecules are CO2). Capturing the greenhouse gas at these relatively high concentrations requires less energy than capturing it from the air, where it is present at just 0.04% concentration (about one in 2,500 gas molecules). A 2011 report from the American Physical Society estimated that it may cost between $600 and $1,000 per metric ton of CO2 captured from the air. Capturing it at the source—at a coal-burning plant, for example—could cost less than one-tenth of that.
But air capture still matters. First, we currently don’t have any possible way to deal with CO2 released by cars, ships, and planes. Second, because we are on track to emit more CO2 than we need to keep under the 2°C limit, we likely need a means of sucking back up some of that extra greenhouse gas.
Consider the chart below. It shows our historical emissions and then charts emissions targets required to hit climate-change mitigation goals. Sadly, however, “approximate emission pledges” made under the Paris climate agreement show our emissions increasing into the 2020s and 2030s, when they should be decreasing (following the trajectory shown in red on the chart).
Most systems to capture CO2 depend on a process called “reversible absorption.” The idea is to run a mixture of gases (air being the prime example) over a material that selectively absorbs CO2. Then, in a separate process, that material is manipulated to pull the CO2 out of it. The separated CO2 can then be compressed and injected underground. Typically there’s a limited supply of the material, so it will get put through the cycle once again to capture more of the greenhouse gas.
Climeworks and Global Thermostat have piloted systems in which they coat plastics and ceramics, respectively, with an amine, a type of chemical that can absorb CO2. Carbon Engineering uses a liquid system, with calcium oxide and water. The companies say it’s too early in the development of these technologies to predict what costs will be at scale. “It’s like if you asked someone in 1940 what the cost of commercial rockets would be today,” says David Keith of Carbon Engineering. What they are willing to share are cost targets.
Jan Wurzbacher, Climeworks’s director, says it hopes to bring costs down to about $100 per metric ton of carbon dioxide. That’s close to the price Carbon Engineering is targeting, according to Geoffrey Holmes, the company’s business development manager. Peter Eisenberger, co-founder of Global Thermostat, says their technology will be even cheaper: when scaled up, he says, costs will drop to as low as $50 per metric ton. (Intriguingly, the startup’s projected capture cost seems to be inversely proportional to the money each says it has raised: about $15 million in private investment for Climeworks and Carbon Engineering, and $50 million for Global Thermostat.)
Each of the startups has built a functional pilot plant—each of which can capture hundreds of kg of CO2—to prove their technology. And all boast that their tech is modular, meaning they can build a direct air capture plant as small or large as somebody is ready to pay for. Even at $50 per metric ton of capturing emissions, if we have to capture as much as 10 billion metric tons by 2050, we are looking at spending $500 billion each year capturing carbon dioxide from the air. It seems outrageous, but it may not be if climate change’s other damages are put in perspective—and that’s what these startups are betting on.
Global Thermostat’s pilot plant in Stanford Research Institute in Palo Alto was sitting idle when I visited it. Carbon Engineering is waiting to build out the technology more—including the ability to convert captured CO2 into fuels—before it starts scaling up its business. Only Climeworks has been able to show success in commercial applications.
In May this year, Climeworks set up its first commercial unit near Zurich, Switzerland, capturing about 1,000 metric tons of CO2 from the air each year (equivalent to 20 US households’ annual emissions). The captured CO2 is supplied to a nearby greenhouse, where a high concentration of the gas boosts crop yield by 20%.
But the company’s newest installation in Iceland is even more impressive, because it’s the first true “negative emissions” plant.
Buried once and for all
To build it, Climeworks first needed a “carbon-neutral” power plant. They found one in Hellisheidi, Iceland, where the public utility company Reykjavik Energy runs a geothermal power plant. The Hellisheidi plant, about 15 miles (25 km) southeast from the capital Reykjavik, uses naturally occurring heat from a volcanically active region to produce electricity and heat, by pumping water through an underground network of pipes. As the water travels underground, it heats up and turns into steam which can be used to run turbines. The plant produces about 300 MW of electricity (enough to power 200,000 American households) and about 130 MW of heat. Though geothermal is a clean energy source, the process of recovering heat releases gases—a mixture of carbon dioxide, hydrogen sulfide, and hydrogen. It’s not a massive amount of CO2—for each unit of energy produced, a geothermal plant produces 3% of the carbon emissions of a coal-fired power plant—but it’s still something.
In 2014, Reykjavik Energy along with the help of academics in the US, Europe, and Iceland, formed a project called CarbFix to test technology to get rid of the small amount of gases it does produce. And since the Paris climate accords, the utility company has been ramping up its emissions-reduction efforts to help Iceland reach its country-wide goals.
Instead of using reversible absorption, CarbFix uses Iceland’s plentiful natural resources to capture carbon dioxide. Every time you open a can of soda, you are drinking dissolved CO2 in water. The Hellisheidi plant works on a similar principle. When the gases released from geothermal vents are mixed with water, they get absorbed ever so slightly. The mixture—27 kg of fresh water for each kg of CO2—is then injected 700 meters underground.
In a 2016 study, scientists found that the carbon dioxide in these water mixtures was reacting with Iceland’s vast basaltic rock—dark igneous rock usually found under ocean floors—to form minerals. This process usually takes hundreds or thousands of years, but what was surprising in Iceland was that the mineralization occurred in less than two years. The speed probably has something to do with the unique local geology. Sandstone aquifers—which have been the most well-studied type of rock system when it comes to carbon dioxide-injection systems—react very slowly with CO2. Basalt rock, on the other hand, seems to react much more quickly, because of the presence of metals like iron and aluminum.
The study showed, for the first time, that storing carbon dioxide underground is easier and safer than it has been made out to be. Once locked into the minerals, the carbon dioxide cannot enter the atmosphere for millions of years. Better still, this sort of basalt rock is present in large deposits around the world—enough to take in many decades of fossil fuel emissions, direct air capture, and more.
Over the past three years, more than 18,000 metric tons of CO2 have been injected into the ground as part of the project, according to Edda Aradóttir, a CarbFix geologist. Moreover, they’ve been able to do it for less than $30 per metric ton of CO2.
This month, Climeworks installed a unit that captures carbon dioxide directly from the air and transfers it to CarbFix to be injected underground. Because CarbFix has been monitoring the injection sites for the last three years, they can be sure there will be no leakage: once mineralized, the CO2 will remain trapped for thousands or millions of years. This makes the Climeworks-CarbFix system the world’s first verified “negative emissions” plant.
(A chemical plant in Chicago, where Archer Daniels Millard ferments corn to produce ethanol, claims to be a negative-emissions plant. The company says that because corn consumes CO2 to grow in the first place, even though fermentation releases CO2, it is carbon neutral—simply returning what it took from nature. If you then capture and inject those emissions underground, you could technically consider it negative emissions. But experts haven’t yet run the numbers to verify Archer Daniels Millard’s claim.)
Climeworks says it is now looking to customers who want to buy their way into programs that cut their emissions. The delivery company DHL, for example, has committed to reaching zero emissions by 2050. But even if they move their entire road vehicle fleet to run on all electric cars, there is currently no technology to cut emissions from the airplanes DHL relies on. The hope is that DHL will pay money to Climeworks to bury those excess emissions into the ground.
Academics used to think that direct air capture would be too expensive for any practical purposes. They still tend to think that for carbon capture and storage more broadly. But what Climeworks and its competitors are showing is that, if direct air capture can be made cheap enough for there to be commercial interest, then carbon capture at point-sources will likely work, too. And if nothing else, the existence of direct air capture gives humanity a high-premium insurance policy against what would surely be a much more expensive disaster.
Read next: The revolutionary technology pushing Sweden toward the seemingly impossible goal of zero emissions
The reporting was supported by a fellowship from the McGraw Center for Business Journalism at the City University of New York Graduate School of Journalism.