Carbon Capture's Scale Gap Leaves 1.5°C Goal in Doubt

What happened

A fresh round of climate reporting has converged on an uncomfortable arithmetic problem: the world's climate models increasingly assume massive amounts of carbon dioxide removal, but the actual carbon capture and storage industry is nowhere near that scale. The trigger is a new United Nations Environment Programme report, Limiting Overshoot, released alongside independent research examining how real-world carbon storage projects have performed against their own projections 18910. The UNEP report concludes that crossing 1.5°C of warming is now effectively unavoidable and lays out an "overshoot, peak and decline" pathway as the least-bad option: cut emissions as fast as possible, hold the temperature peak down, and then use carbon dioxide removal to bring warming back below 1.5°C later in the century 910. Critically, UNEP insists removal must supplement, not replace, deep emissions cuts, and estimates that even ambitious deployment would only reverse a few tenths of a degree of warming this century 910.

Against that backdrop, a New Republic analysis drawing on research by Hugh Daigle and Arvind Ravikumar examined every operational or formerly operational permanent geological CO2-storage project worldwide and found a consistent pattern of underperformance 111. Projects hit only 59 percent of their projected injection rates on average, and commercial-scale facilities captured just 0.19 megatonnes of CO2 per year on average 11. Shell's Gorgon plant in Australia, the largest project studied, currently sequesters about 0.4 megatonnes annually versus the 3.5 megatonnes projected when it launched in 2019 — a shortfall corroborated by a separate peer-reviewed review of the Gorgon project's technical and modeling challenges 1113. Reaching even modest U.S. net-zero targets, the researchers calculate, would require building roughly 250 Gorgon-scale facilities every year through 2050, and drilling thousands of wells annually — a tall order given the U.S. had only 581 active drilling rigs as of mid-2026 11.

The scale problem, in numbers

UNEP estimates that stabilizing temperatures back below 1.5°C after an overshoot could require removing 15 to 24 gigatonnes of CO2 annually by 2100 111. Current global carbon removal totals roughly 2.2 gigatonnes a year, and the vast majority of that comes from conventional land-based methods like reforestation, not engineered capture 11. The International Energy Agency's own database shows just over 50 megatonnes of operational capture-and-storage capacity worldwide as of early 2025, with the existing project pipeline projected to reach around 430 megatonnes of capture and 670 megatonnes of storage capacity by 2030 12. That growth is real, but it still leaves the industry at roughly 2 to 3 percent of the annual removal volume UNEP's pathway implies will eventually be needed.

A separate 2025 study published in Nature adds another constraint: prudent, safely usable geological storage capacity may be about ten times smaller than earlier estimates — roughly 1,460 gigatonnes of CO2 rather than the far larger figures previously assumed — meaning full use of that storage could cut warming by only about 0.7°C rather than the 6°C once cited 15. Carbon Brief's analysis of the newest CMIP7 climate scenarios reinforces the tension: even scenarios that keep the world close to Paris Agreement goals require pumping 800 to 1,750 gigatonnes of CO2 underground by 2150, pushing the storage industry toward or beyond the upper bound of what geologists consider prudent 16.

Real projects, real progress — and a persistent scale gap

Not all the coverage is downbeat. The European Commission and trade press covered the September 2026 inauguration of Yara's Sluiskil plant in the Netherlands, described as Europe's largest industrial carbon capture facility 1718. The plant is designed to capture and liquefy up to 800,000 tonnes of CO2 annually from ammonia production, ship it to Norway's Northern Lights terminal, and inject it roughly 2,600 meters beneath the North Sea seabed, for a projected total of 12 million tonnes over 15 years 1718. Northern Lights, a joint venture of Equinor, Shell and TotalEnergies, already has its first phase of 1.5 million tonnes annual capacity fully booked, with a second phase approved to expand to at least 5 million tonnes a year, backed by EU funding 18. Equinor says it is targeting 30 to 50 million tonnes of annual transport-and-storage capacity by 2035 18. Elsewhere, an Illinois sequestration project has returned after a state moratorium lapsed 4, a $3.7 billion low-carbon ammonia plant is under construction in Louisiana 7, a Louisiana columnist argues CCS will boost the state's economy 5, and FuelCell Energy touted its first carbon capture demonstration and a $3.6 billion order backlog 6. The IEA notes 2024 brought several firsts — the first natural-gas power plant with CCS to reach a final investment decision, in the UK, and a Swedish combined heat-and-power CO2-removal project — alongside growing momentum in China and the Middle East 12.

Where the reporting agrees

Across UNEP's own documents, the IEA database, the Gorgon-focused academic reviews, and the skeptical advocacy reports, there is broad convergence on several points. First, carbon dioxide removal is treated as necessary in virtually every credible pathway that limits warming, but insufficient on its own 191016. Second, current global deployment — whether measured by IEA's 50 megatonnes of operational capacity or the 2.2 gigatonnes of total removal cited by UNEP — is a small fraction of what climate models eventually assume will be needed 1112. Third, geological storage is technically harder to scale than capture equipment alone; the Gorgon shortfall, the FluidFlower modeling benchmark study, and the ACS review of global storage projects all point to heterogeneous rock formations, injection-rate limits, and monitoring challenges as persistent bottlenecks rather than solved problems 111314. Fourth, no outlet in this set argues that carbon capture should substitute for cutting emissions — even the most industry-friendly coverage of the Yara and Northern Lights projects frames CCS as a complement to decarbonizing hard-to-abate industry, not a replacement for renewable energy or electrification 1718.

Where it doesn't

The clearest divergence is tonal and interpretive rather than factual. UNEP's own materials 910 frame overshoot as a manageable, if urgent, engineering and policy challenge — a curve to be minimized rather than a catastrophe. The European Commission and gCaptain's coverage of Sluiskil 1718 adopt a similarly forward-looking frame, presenting the project as proof that a full commercial CCS supply chain — capture, shipping, offshore injection — now works at scale. By contrast, The New Republic, the Environmental Working Group, and Environment America 1111920 argue that isolated technical successes do not establish scalability, and that CCS has historically been used to justify continued fossil fuel operation rather than replace it. EWG goes further than any other source here, asserting flatly that existing CCS facilities capture less than 1 percent of global emissions and citing Petra Nova's collapse as evidence the technology has repeatedly failed to meet its own promises 19 — a more categorical dismissal than UNEP or the IEA offer.

There is also a numerical tension worth flagging: the Nature-based estimate that safe geological storage capacity is roughly ten times smaller than previously thought, capping potential temperature reduction near 0.7°C 15, sits uneasily against CMIP7 scenarios that still model 800 to 1,750 gigatonnes of underground storage as achievable 16. Carbon Brief acknowledges this tension directly rather than resolving it, noting that even the less aggressive scenarios would require the storage industry to handle more CO2 than the fossil fuel industry currently moves in oil. Finally, the Stanford-based comparison cited by Environment America — that a fossil-fuel-plus-capture pathway costs 9 to 12 times more than a renewables-based pathway 20 — is a claim unique to that study; none of the other sources attempt a comparable economy-wide cost comparison, though EWG's finding that CCS could raise a new gas plant's levelized cost of electricity by up to 61 percent points in the same direction 19.

The reading the evidence supports

Taken together, the record does not support either extreme. It is not accurate to say carbon capture is worthless — Sluiskil and Northern Lights demonstrate that a functioning capture-to-storage chain can be built and operated commercially 1718. But it is also not accurate to treat that success as evidence the technology can be multiplied fast enough to matter at the gigatonne scale UNEP's own pathway requires. The Gorgon shortfall, corroborated independently by both journalistic and peer-reviewed sources 1113, is the most telling data point in this set: it shows that even a well-funded, first-of-its-kind project run by a major oil company fell dramatically short of its own five-year-old projections. If the industry's flagship project undershoots by nearly 90 percent, the assumption embedded in climate models — that capture capacity will simply scale as costs fall — looks like the weakest link in the mitigation strategy, not the strongest. The most defensible synthesis is UNEP's own framing, minus the optimism gap: carbon removal will likely be needed for residual emissions and limited overshoot, but nothing in current deployment data justifies treating it as a substitute for cutting emissions now.