The Carbon Harvest: Inside the World of Direct Air Capture Plants

In the Icelandic highlands, where volcanic landscapes stretch toward the horizon, a quiet industrial revolution is unfolding. Giant fans draw in the crisp Arctic air, filtering out carbon dioxide so it can be pumped deep underground, where it reacts with basalt and turns to stone. This is Mammoth, the world’s largest operational direct air capture (DAC) plant, and it represents a pivotal moment in humanity’s struggle against climate change.

Mammoth, operated by Swiss company Climeworks, is designed to capture up to 36,000 tons of CO₂ annually—roughly ten times larger than its predecessor, Orca. While this is a minuscule fraction of global emissions—which exceed 35 billion tons per year—Mammoth stands as a proof of concept for a technology many scientists consider essential for meeting climate goals. The Intergovernmental Panel on Climate Change (IPCC) estimates that the world must remove between 3 and 12 gigatons of CO₂ annually by 2050 to limit warming to 1.5°C. Direct air capture is expected to play an important role in that effort.

How Carbon-Removal Plants Work

Direct air capture plants use chemical processes to extract CO₂ from ambient air. The air is drawn through fans into contact with a sorbent material—either a liquid solvent or solid filter—that selectively binds with CO₂ molecules. When the sorbent becomes saturated, it is heated to approximately 100°C to release the captured CO₂, which is then purified and stored.

In Iceland, Climeworks partners with Carbfix, which injects the captured CO₂ dissolved in water into basalt formations about 1,000 meters underground. The solution reacts with the rock, forming solid carbonate minerals within two years. This mineralization process provides exceptionally secure, permanent storage that can be verified by independent third parties.

The plants are powered by renewable geothermal energy supplied by ON Power in Iceland, making the entire process net-carbon-negative when properly accounted for. Unlike point-source carbon capture, which captures CO₂ from industrial flue stacks, DAC removes legacy emissions that have been accumulating in the atmosphere since the Industrial Revolution.

The Pioneers and Their Plants

Climeworks has emerged as the leader in operational DAC deployment. After launching its Orca plant in September 2021—the world’s first commercial DAC and storage facility—the company broke ground on Mammoth in June 2022 and began operations in May 2024. The plant uses a modular design with 72 collector containers, allowing for phased construction and scaling. Climeworks processes close to 200 million data points daily from its plants, using operational experience to continuously improve performance.

Looking ahead, Climeworks has an ambitious roadmap: megaton capacity by 2030 and gigaton scale by 2050. The company is developing multiple megaton hubs in the United States, including Project Cypress in Louisiana, and is exploring projects in Norway, Kenya, and Canada. Its founders emphasize that reaching gigaton scale requires running the technology in the real world as fast as possible, learning through relentless deployment.

Meanwhile, the American company Carbon Engineering has been developing its Stratos facility in Ector County, Texas. Stratos is designed to capture more than 500,000 tons of CO₂ annually—more than an order of magnitude larger than Mammoth. The plant launched commercial operations in January 2026, with Microsoft as a key buyer of its carbon credits. Backed by Breakthrough Energy, Mitsubishi, and Siemens, Stratos signals that major capital is flowing into DAC at commercial scale. However, the project has faced delays: Occidental Petroleum, which is developing Stratos, had initially hoped to have the plant online by the end of 2024, but encountered issues with some components during testing.

A third wave of developers is entering the market with novel approaches. UK-based Airhive opened a 1,000-tonne-per-year system in Alberta, Canada, in late 2025. Using fluidised bed technology—a proven industrial process—Airhive claims its system will achieve costs below $500 per tonne, significantly below the approximately $1,000 per tonne that current DAC credits command. The company has a clear path to lower costs further as it scales. Airhive is also partnering with Coca-Cola Europacific Partners to install DAC units at bottling plants, using captured CO₂ to carbonate beverages.

The Cost Challenge and Learning Curve

Cost remains the most formidable barrier to DAC deployment. Current costs for capturing and storing CO₂ at Climeworks’ Mammoth plant are around $1,000 per ton. The company projects reductions to $250–$350 per ton by 2030 and $100 per ton by 2050 through scaling and efficiency gains. The Canadian government estimates that DAC costs could reach $100–$300 per ton at scale, while the World Economic Forum notes that DAC’s affordability will depend heavily on advancing the technological learning curve.

The economics of DAC are challenging because CO₂ is dilute in the atmosphere—at about 420 parts per million, processing trillions of cubic metres of air is required to capture millions of tons. The capital expenditures are substantial: Climeworks’ COO described Mammoth’s construction costs as being “in the order of low triple-digit millions” of US dollars. Energy demands are also significant, as heating sorbents and running fans requires substantial power.

However, the market for DAC credits is growing. Companies including Microsoft, JPMorgan, Shopify, Google, and Meta have prepurchased carbon removal credits, providing revenue certainty for developers. Frontier, a carbon removal buying consortium, has been particularly active in aggregating corporate demand. DAC credits trade at a premium because the removal is both verifiable and permanent—once sequestered, CO₂ stays underground for 10,000 years or more.

Technical and Scaling Hurdles

The path to gigaton-scale deployment is strewn with technical challenges. Sorbent materials—the chemicals that capture CO₂—face durability issues. Amine-based sorbents can degrade when exposed to atmospheric oxygen and heat during regeneration. Improving sorbent durability to 10,000–15,000 cycles would reduce replacement costs by 60–70%, but achieving this while maintaining capture performance remains an active research challenge.

Next-generation materials are being developed in laboratories worldwide. Metal-organic frameworks (MOFs) with tuneable pore sizes offer 30–50% improvements in CO₂ selectivity and capacity over current materials. A 2025 paper in Nature Energy demonstrated a MOF-based sorbent achieving more than 10,000 cycles with less than 5% capacity degradation, though it has not yet been tested at pilot scale. Humidity-swing sorbents that release CO₂ when exposed to moisture could eliminate the thermal energy requirement entirely.

Scaling poses additional challenges. The approximately 18 operational DAC plants currently capture only about 10,000 tons of CO₂ annually—a tiny fraction of what is needed. Achieving megaton and gigaton scale requires not just technological advances but also massive supply chain development, workforce training, and regulatory frameworks. Monitoring, reporting, and verification (MRV) frameworks for DAC are still maturing, though mineralization storage provides high confidence in permanence.

Alternative Carbon Removal Pathways

Direct air capture is not the only engineered carbon removal approach. Bioenergy with carbon capture and storage (BECCS) involves burning biomass for energy and capturing the resulting CO₂ emissions. This approach can theoretically generate energy while removing carbon, but has faced significant controversy. In the UK, analysis by energy think tank Ember warns that BECCS at the Drax power plant could require up to £30 billion in subsidies—exceeding the entire UK budget for carbon capture—while delivering uncertain climate benefits. The plant already receives approximately £744 million annually in public support for burning imported woody biomass, and Ember argues these removals are far from guaranteed given uncertainties around carbon reabsorption through replanting. Burning woody biomass actually releases more carbon than coal, and the plant has faced allegations of sourcing wood from primary forests in North America.

Other engineered approaches include biochar—heating biomass in the absence of oxygen to create a stable carbon-rich material—which costs $80–200 per ton and is the most cost-effective engineered CDR currently available. Enhanced rock weathering, which grinds CO₂-absorbent rocks and spreads them on agricultural fields, costs $50–200 per ton. Ocean-based methods such as direct ocean capture and ocean alkalinity enhancement are also being explored.

Policy, Investment, and the Future

The growth of carbon removal depends heavily on supportive policy frameworks. Canada has developed a competitive advantage with industrial carbon pricing, emissions regulations, investment tax credits, and abundant low-carbon electricity and geological storage. The country is home to 78 CDR companies and 48 active or planned projects. The US Department of Energy has awarded over $600 million in funding for megaton DAC hubs, with Project Cypress in Louisiana receiving initial funding. The UK, despite its ambitious net-zero targets and substantial North Sea storage capacity, has yet to see any commercial-scale DAC project break ground.

The debate over carbon removal is intensifying. Critics argue that DAC remains unproven at scale and risks diverting billions from proven solutions like renewables, heat pumps, and efficiency measures. Supporters counter that because some emissions—from aviation, cement, steel production, and agriculture—are extremely difficult to eliminate, permanent carbon removal is essential for net-zero targets. They point to the IPCC’s conclusion that all pathways to 1.5°C use CDR.

For the carbon-removal industry, the coming decade will be decisive. Climeworks aims for megaton capacity by 2030. Stratos, once fully operational, will capture half a million tons annually. New technologies like Airhive’s fluidised bed system and emerging MOF-based sorbents promise lower costs and faster deployment. Whether carbon-removal plants can scale from thousands to megatons—and eventually to gigatons—will depend on sustained investment, technological breakthroughs, policy support, and perhaps most importantly, whether the world views them as a complement to, rather than a substitute for, aggressive emissions reduction.

The carbon harvest has begun. The question is whether it can grow fast enough to matter.