The Orca and the Mammoth
In September 2021, a modest industrial facility in Hellisheiði, Iceland, began operations. It looked like a collection of shipping containers fitted with fans, and its promise was audacious: pull carbon dioxide directly from the ambient air, concentrate it, and inject it underground, where it would mineralize into solid rock. The facility, called Orca, was built by Climeworks, a Swiss company, and had a nameplate capacity of 4,000 tonnes of CO₂ per year — roughly the annual emissions of 870 cars. It was, by any industrial standard, tiny. But it was the largest direct air capture (DAC) facility in the world, and it proved that the technology worked at commercial scale.
In May 2024, Climeworks opened Mammoth, a facility ten times the size of Orca with a capacity of 36,000 tonnes per year. A few months later, the company broke ground on Project Cypress in Louisiana, a US Department of Energy-supported facility designed to capture up to 1 million tonnes per year — roughly the emissions of 200,000 cars. A million tonnes sounds enormous until you consider that humanity emits approximately 37 billion tonnes of CO₂ annually. To capture just 1% of that — 370 million tonnes — would require 370 facilities the size of Project Cypress, at a cost of hundreds of billions of dollars. The technology works. The question is whether it can scale to the point where it matters.
How Direct Air Capture Actually Works
Direct air capture comes in two main flavors. Solid sorbent DAC, used by Climeworks, passes air over a filter material (usually amines bonded to a solid substrate) that selectively binds COâ‚‚ molecules. When the filter is saturated, it’s heated to about 100°C, releasing a concentrated stream of COâ‚‚. Liquid solvent DAC, used by Carbon Engineering (acquired by Occidental Petroleum in 2023), passes air through a liquid solution (typically potassium hydroxide) that reacts with COâ‚‚ to form a carbonate. The carbonate is heated to release pure COâ‚‚. In both cases, the energy input is substantial — DAC is fundamentally fighting entropy, unmixing COâ‚‚ from air where it’s present at just 420 parts per million.
The captured COâ‚‚ can be permanently stored (injected into geological formations where it mineralizes over time) or used as a feedstock for synthetic fuels, chemicals, or building materials. The “utilization” path is smaller in scale — the total market for COâ‚‚-derived products is a tiny fraction of what needs to be captured — but it can generate revenue that partially offsets capture costs. The “storage” path is the one that matters for climate: permanent geological sequestration that removes COâ‚‚ from the carbon cycle.
Iceland is the ideal location for DAC because it has abundant, cheap geothermal energy (solving the energy problem) and basalt rock formations that react with COâ‚‚ to form stable carbonate minerals (solving the storage problem). In most of the world, these conditions don’t coincide. The energy grid that powers DAC facilities needs to be clean — capturing COâ‚‚ using fossil-generated electricity defeats the purpose — and suitable geological storage isn’t universally available. These constraints limit the addressable geography, though not as severely as critics sometimes suggest; saline aquifers, the most common storage formation, are widespread, and clean energy is getting cheaper everywhere.
The Cost Curve That Must Bend
The critical variable for DAC’s future is cost. Climeworks has stated that its near-term cost target is $400-600 per tonne of COâ‚‚ captured. The long-term target — the number that would make DAC economically viable at climate-relevant scale — is $100-150 per tonne. For comparison, the current price of carbon in the European Union’s Emissions Trading System is roughly €70-80 per tonne. For voluntary carbon markets, prices for high-quality DAC credits range from $500 to $1,000 per tonne, purchased by companies like Microsoft, Shopify, and Stripe (through its Frontier fund) that are willing to pay a premium for permanent, verifiable carbon removal.
The cost reduction pathway is well-understood: larger facilities, standardized manufacturing, learning-curve improvements in sorbent materials, and cheaper clean energy. The analogy that the industry uses is solar power, which saw costs fall 90% from 2010 to 2020 through a combination of manufacturing scale, technological improvement, and policy support. But solar is an energy technology; DAC is a chemical engineering technology. The learning curves aren’t identical, and the optimism of solar’s cost trajectory may not translate directly to carbon capture.
An independent analysis published in Joule in 2023, led by researchers at ETH Zurich, estimated that DAC costs could reach $226-385 per tonne by 2030 under optimistic assumptions — still above the $100 target that would make DAC broadly economic without subsidy. The study concluded that DAC “could become a billion-tonne industry by 2050” but that achieving that scale would require “sustained policy support on the order of $100-200 billion over the next decade.” That’s real money, but it’s in the same ballpark as the subsidies that built the solar and wind industries — and dramatically less than the cost of unchecked climate change.
The Moral Hazard Debate
DAC’s most controversial dimension isn’t technical or economic. It’s moral. Critics argue — with some justification — that the promise of future carbon removal reduces the urgency of emissions reduction in the present. Why stop burning fossil fuels if we can capture the emissions later? The fossil fuel industry, notably, has been an enthusiastic supporter of carbon capture research, which critics see as an attempt to preserve the industry’s business model rather than genuinely address climate change.
The advocates respond that this is a false choice. The IPCC’s Sixth Assessment Report makes clear that keeping warming below 1.5°C requires both rapid emissions reduction and large-scale carbon removal. Even if the world reaches net-zero emissions by 2050 — an extraordinarily ambitious target — the cumulative emissions in the atmosphere will continue to warm the planet. Carbon removal isn’t an alternative to emissions reduction. It’s a complement that addresses the emissions we’ve already released and the hard-to-abate sectors (aviation, agriculture, heavy industry) that won’t reach zero for decades.
The moral hazard argument has merit, but it’s also a form of delay in its own right. Every year spent debating whether carbon removal distracts from emissions reduction is a year not building the removal capacity we’ll need. The responsible position is “both, urgently”: drive emissions to zero as fast as possible while simultaneously building the carbon removal infrastructure that physics demands. DAC isn’t a substitute for mitigation. It’s insurance against the mitigation we’ve already failed to do — and against the mitigation we may continue to fail at in the decades ahead.
The Verdict: Not Yet, But Getting Closer
Direct air capture in 2025 is where solar power was in roughly 2005: technically proven, economically challenged, and dependent on policy support to reach the scale where costs become competitive. The facilities exist and they work. The cost trajectory is downward. The policy support — the US 45Q tax credit ($180/tonne for DAC), the EU Innovation Fund, the Frontier advanced market commitment — is real. But the scale gap is enormous, and closing it will require sustained investment and political commitment over decades, not years.
The honest assessment is that DAC won’t solve the climate problem. Nothing can — the problem is too large for any single technology. But DAC can be part of the solution, alongside electrification, renewable energy, energy efficiency, and ecosystem restoration. Whether it becomes a meaningful part depends on whether the cost curve bends — and whether the political will to fund the transition outlasts the fossil fuel industry’s resistance to it. The technology works. The economics are improving. The politics are the variable that will determine whether direct air capture becomes a climate solution or a climate footnote.
The Alternatives That Don’t Get Enough Attention
Direct air capture dominates the carbon removal conversation, but it’s not the only approach — and it may not be the most cost-effective one. Enhanced weathering — spreading crushed silicate rocks like basalt on agricultural land, where they react with atmospheric COâ‚‚ to form carbonates — is estimated to cost $50-200 per tonne and has co-benefits for soil health. Ocean alkalinity enhancement — adding alkaline materials to seawater to increase its COâ‚‚ absorption — could potentially operate at enormous scale. Afforestation and reforestation remain the cheapest carbon removal methods per tonne, though they’re limited by available land and face permanence concerns (forests can burn).
The emerging consensus in climate policy is that a portfolio approach is necessary: no single carbon removal technology can scale to the billions of tonnes needed by mid-century. DAC, enhanced weathering, ocean-based approaches, and nature-based solutions will each contribute, and the optimal mix will vary by geography, cost, and political context. The carbon removal industry is still in its infancy — the Sector’s annual capture capacity is measured in thousands of tonnes, against a need measured in billions — but the infrastructure buildout is beginning. The question, as with so much of climate policy, is whether it’s beginning fast enough.
