In a conversation about climate change, the phrase “carbon removal” used to elicit scepticism — a technofix fantasy that would let the world keep burning fossil fuels. That scepticism has not entirely disappeared, but the conversation has changed. Direct air capture (DAC) — technology that pulls carbon dioxide from ambient air and stores it permanently — has attracted billions of dollars of investment, been endorsed by the Intergovernmental Panel on Climate Change as necessary to meet the Paris Agreement goals, and produced its first operational facilities. The question is no longer whether it works. It is whether it can scale.
How DAC Works
Direct air capture is, in principle, simple. Air is drawn through a chemical medium that binds CO₂. When the medium is saturated, heat or steam releases the CO₂ in a concentrated stream, which is then compressed and either stored underground (sequestration) or used in industrial processes (utilisation — making fuels, chemicals, or building materials). The medium is then regenerated and the cycle repeats.
Two main chemical approaches are used. Solid sorbent systems, used by companies like Climeworks, use amine-based filters that capture CO₂ at ambient conditions and release it when heated to roughly 100°C. Liquid solvent systems, used by Carbon Engineering (now owned by Occidental), use an aqueous hydroxide solution that reacts with CO₂ to form carbonate, which is then heated in a kiln to release the CO₂.
The challenge is scale. CO₂ in the atmosphere is extremely dilute — about 420 parts per million, or 0.04%. To remove one million metric tons of CO₂ per year — a tiny fraction of global emissions of roughly 40 billion tons — a DAC plant must process an enormous volume of air. That takes energy, land, and capital.
Who Is Building What
Climeworks
The Swiss company Climeworks commissioned Orca, the world’s first commercial DAC facility, in Iceland in 2021. Orca captures roughly 4,000 metric tons of CO₂ per year — miniscule relative to emissions, but a proof of scaleup. Its CO₂ is mixed with water and injected into basalt rock formations by Carbfix (another company), where it mineralises into stone within a few years.
In 2024, Climeworks started up Mammoth, a larger facility that aims to capture roughly 36,000 tons per year when fully operational — nine times Orca’s capacity. It is still modular: each collector unit is a shipping-container-sized box with fans and filters, and multiple units are ganged together. The cost per ton is high — widely estimated at $600 to $1,000 per ton of CO₂ — and the company’s goal is to reduce costs by orders of magnitude as it scales, targeting $300-400 per ton by 2030 and below $200 per ton eventually.
Occidental Petroleum / 1PointFive
Occidental Petroleum (Oxy), through its subsidiary 1PointFive, is building a far larger DAC plant called Stratos in the Permian Basin of Texas. Based on Carbon Engineering’s liquid-solvent technology, Stratos is designed to capture up to 500,000 metric tons of CO₂ per year when complete — an order of magnitude larger than Mammoth. The captured CO₂ will be partially sequestered and partially used for enhanced oil recovery, a practice that environmental groups criticise as prolonging fossil-fuel use. Oxy argues that using CO₂ for oil production improves the economics and creates a bridge to full sequestration.
In 2023, Occidental acquired Carbon Engineering for $1.1 billion, consolidating the liquid-solvent technology under a single owner. Stratos broke ground in 2023 and is expected to begin operations in 2025-2026.
Other Players
Several other companies are pursuing DAC with different chemistries and business models. Heirloom Carbon, a startup, uses limestone — a low-cost, abundant material — as its sorbent, accelerating the natural weathering process. RepAir uses an electrochemical approach. Mission Zero Technologies uses an electro-swing process. These are earlier-stage but reflect a competitive landscape with multiple technical approaches.
The Economics
The dominant barrier to DAC is cost. Existing facilities capture carbon at well over $500 per ton. For context, a $500-per-ton carbon price is far above almost any existing carbon price in the world; the EU Emissions Trading System, the largest carbon market, saw prices in the range of €50-100 per ton in 2024. Without a much higher carbon price, substantial government subsidy, or a voluntary carbon-offset market willing to pay a premium, DAC is uneconomic.
Subsidies are beginning to arrive. The U.S. Inflation Reduction Act increased the 45Q tax credit for carbon capture, providing up to $180 per ton for direct air capture with sequestration. The European Union and several member states are funding DAC projects. Major technology companies — Microsoft, Stripe, Shopify — have committed hundreds of millions of dollars to advance purchase agreements for carbon removal, creating a small but real market. Microsoft alone committed to purchasing removal credits from multiple DAC companies, including Climeworks and Heirloom.
The Scale Challenge
To make a meaningful contribution to climate goals, DAC would need to remove billions of tons of CO₂ per year by mid-century. That would require thousands of plants, an enormous energy supply (renewable, or the net benefit is reduced), and hundreds of billions or trillions of dollars of investment. Optimists point to historical scaling of solar and batteries; pessimists point to the fact that DAC is capital-intensive manufacturing, not semiconductor-like miniaturisation, and that the scaling curve is likely to be slower.
The Moral Hazard Question
Critics worry that DAC provides — and will continue to provide — a rhetorical cover for continued fossil-fuel use: a promise that “we can clean it up later” that delays emissions reductions now. This concern has some basis, particularly in the positions of some oil companies. Others argue that, given the IPCC’s finding that limiting warming to 1.5°C or even 2°C requires both deep emissions cuts and substantial carbon removal, DAC is a necessary complement, not a substitute.
Both can be true: DAC is essential, and it must not delay decarbonisation. The risk is that policymakers and companies use prospective DAC as an excuse for insufficient action today.
The Energy Penalty in Detail
The energy cost of DAC is not a detail; it is the problem. Capturing a ton of CO₂ from the dilute atmosphere requires moving and processing an enormous amount of air, which demands substantial thermal and electrical energy. For liquid-solvent systems like Carbon Engineering’s, roughly half the energy is thermal (for the calcination step that releases captured CO₂) and half is electrical (for fans, pumps, and compression). The total is on the order of several thousand kilowatt-hours per ton, depending on the system. If that energy comes from fossil fuels, the net benefit shrinks, and at some point the process becomes counterproductive. Pairing DAC with dedicated, firm, clean power — geothermal, nuclear, or abundant solar with storage — is essential for credibility.
Enhanced Rock Weathering and Other Removal Pathways
DAC is not the only carbon-removal technology, and it is not even the largest. Enhanced rock weathering — spreading finely ground minerals on agricultural land to absorb CO₂ as they weather — uses natural processes and has lower energy costs but requires enormous volumes of rock and land. Afforestation and reforestation are cheaper and scalable but are reversible and compete for land. Ocean-based methods — alkalinisation, iron fertilisation — remain experimental. Each approach has its advocates, and a responsible portfolio would develop several in parallel, because no single one can reach the billions-of-tons scale alone.
The Business Model Dependency
DAC companies rely on voluntary carbon markets and advance-purchase agreements, which are inherently fragile — they depend on the willingness of large corporations to pay a premium that no regulation requires. The 45Q tax credit provides a policy floor in the U.S., but at $180 per ton for DAC, it does not fully cover costs. If the voluntary market falters — as previous carbon-offset markets have — the economics collapse. Government procurement, or a robust compliance market that values permanent removal at a high enough price, is the only long-term foundation for the industry.
Measuring Permanence
One of DAC’s advantages, in principle, is that sequestration in geological formations — deep saline aquifers, basalt formations — is effectively permanent, far more so than biological sequestration. But verifying that permanence requires extensive monitoring: the injection site, the cap rock, the absence of leakage over centuries. Standards for monitoring, reporting, and verification are being developed (by organisations like CarbonPlan and independent certifiers), but they are not yet standardised or universally trusted. The industry’s credibility depends on getting this right, because if “permanent” turns out to be “for a year,” the entire economic case collapses.
Conclusion
Direct air capture has moved from concept to operational reality, with commercial plants capturing thousands of tons and larger ones under construction. The technology works. The economics do not yet work without subsidy, and the scale required — billions of tons per year — is daunting. DAC will almost certainly be part of the climate toolkit, but it is a supplement to, not a replacement for, aggressive emissions reductions. The most important climate technology today remains the one that produces the most energy from zero-carbon sources and consumes the least from fossil ones. DAC is a backup plan, and the fact that it is being built at all reflects both the ingenuity of its engineers and the seriousness of the problem it is meant to solve.
