The Carbon Capture Plant That Is Finally Working — At Real Scale

Direct Air Capture facilities are now capturing meaningful carbon at significantly reduced costs. Here is what breakthrough shifted the economics and what needs to happen to make it climate-relevant.

Climeworks' Orca plant in Iceland — the first commercial direct air capture (DAC) facility — had the capacity to capture 4,000 tonnes of CO2 per year when it opened in 2021. To put this in context: global CO2 emissions are approximately 37 billion tonnes per year. The Orca plant could capture enough CO2 to offset roughly one second of global annual emissions. This is not a criticism of the technology but a calibration of where it stood three years ago.

The 2026 situation is meaningfully different, though still nowhere near climate scale. Climeworks' Mammoth plant, which opened in Iceland in 2024, captures 36,000 tonnes per year — a 9-fold increase from Orca. Multiple additional facilities from Climeworks and competing companies including Carbon Engineering (now 1PointFive), Heirloom Carbon, and several Chinese entrants have cumulative capacity approaching 150,000 tonnes per year. This is still approximately 0.0004 percent of annual global emissions — but the trajectory matters as much as the current level.

The cost reduction trajectory is the specific metric that determines whether DAC becomes climate-relevant. Climeworks' Orca operated at approximately $1,000 per tonne of CO2 captured. Mammoth operates at approximately $600 per tonne. Company projections, which the industry track record suggests should be treated with appropriate scepticism, target $300 per tonne by 2030 and $150 per tonne by 2035. At $150 per tonne, DAC begins to compete economically with some nature-based carbon removal and becomes tractable within carbon credit market economics.

For the climate relevance question: IPCC models that hold global warming below 1.5°C almost universally require some deployment of carbon dioxide removal — the atmosphere's CO2 concentration is high enough that emissions reduction alone is insufficient. The question is which CDR methods scale fastest at lowest cost. Nature-based approaches (reforestation, soil carbon) are cheaper but land-constrained. DAC is expensive but can in principle scale anywhere with clean energy. The 2030s will determine which methods contribute meaningfully.