The climate math behind dozens of national net-zero pledges depends on a technology that barely exists at commercial scale, and new research suggests the assumptions propping it up are dangerously optimistic. Researchers at the MIT Energy Initiative have found that widely used climate-stabilization models assume direct air capture will become cheap and widespread on timelines that do not align with engineering realities. Meanwhile, the International Energy Agency has warned that even lower-cost approaches like bioenergy with carbon capture and storage carry price tags that cannot be verified without large-scale operations and rigorous monitoring. Together, these findings expose a widening gap between what policymakers are planning for and what the technology can actually deliver.
What the research actually shows
The MIT Energy Initiative’s analysis targets a specific problem: the integrated assessment models that governments and the Intergovernmental Panel on Climate Change use to chart pathways to 1.5°C and 2°C warming limits. Those models frequently assume DAC costs will fall to $100 to $200 per ton of CO2 removed within the next two to three decades. Current demonstrated costs, however, sit between roughly $400 and $1,000 or more per ton, depending on the technology and energy source. The MIT researchers argue that the learning curves baked into these models borrow too heavily from the experience of solar panels and lithium-ion batteries, technologies that benefited from decades of incremental manufacturing gains, standardized designs, and enormous production volumes. DAC faces a fundamentally different thermodynamic challenge: pulling a gas that makes up roughly 0.04% of the atmosphere out of open air requires significant energy inputs no matter how the hardware improves.
On the bioenergy side, the IEA’s November 2023 commentary on driving down the cost of carbon removal makes a related point. The agency notes that advertised low costs for BECCS have not been corroborated by operational-scale projects with strong monitoring, reporting, and verification (MRV) systems. Early demonstration plants tend to sit at the high end of published cost ranges, and meaningful reductions depend on learning-by-doing across many project cycles. Building out the MRV and permanence frameworks that buyers and regulators require adds costs that promotional estimates routinely omit.
Peer-reviewed work reinforces these cautions. A 2023 analysis published in Energy & Environmental Science laid out thermodynamic limits, realistic deployment trajectories, and infrastructure requirements for scalable DAC pathways, offering a level of technical detail that corporate announcements and press releases typically lack. The broader body of research in that journal, which has become a key venue for carbon capture and storage analysis, consistently highlights materials science constraints and system-level trade-offs that temper optimistic projections.
Where real-world projects stand
Despite the cost uncertainties, concrete steps are moving forward on the ground. In April 2024, the U.S. Environmental Protection Agency issued final Class VI permits to Oxy Low Carbon Ventures LLC for geologic sequestration wells in Ector County, Texas. Class VI permits, authorized under the Safe Drinking Water Act, require detailed technical review of subsurface formations, well construction plans, and long-term monitoring protocols. The permits followed a full public process that included a formal comment period and hearing. They represent one of the first final Class VI authorizations the EPA has granted directly, rather than delegating to state agencies.
The permits are tied to 1PointFive, Occidental’s carbon management subsidiary, which is building what it calls the Stratos facility in the same region. Stratos is designed as a large-scale DAC plant and, if completed as planned, would be among the first to pair direct air capture with dedicated geologic storage under federal permits. Separately, in Iceland, the Swiss company Climeworks began operating its Mammoth plant in 2024, the largest DAC facility in the world at the time of its launch, though its capacity of roughly 36,000 tons of CO2 per year remains a fraction of what climate models envision.
Occidental’s broader bet on carbon removal includes its August 2023 acquisition of Carbon Engineering, a Canadian DAC technology developer, for approximately $1.1 billion. That deal gave Occidental ownership of core DAC intellectual property and engineering capacity. Tracked through the Department of Energy’s clean energy demonstrations office, the acquisition signaled that at least one major oil and gas company views carbon management as a long-term business line, not a side project. Whether that corporate confidence translates into commercially viable removal at competitive prices is a question that will take years of construction and operation to answer.
The U.S. Department of Energy is also placing large public bets. Its Regional Direct Air Capture Hubs program, funded with $3.5 billion from the Bipartisan Infrastructure Law, selected two initial hub projects in August 2023: one in Texas led by 1PointFive and one in Louisiana led by Battelle. These hubs are intended to demonstrate DAC at a scale of at least one million tons of CO2 removed per year each, but construction timelines stretch into the late 2020s, and final costs per ton will not be known until the facilities operate.
What remains uncertain
The biggest open question is whether any first-of-a-kind DAC project will produce verified cost data that matches the optimistic figures embedded in climate models. No operational DAC facility has yet published independently audited per-ton removal costs at commercial scale. Until that data exists, the debate over realistic pricing will rely on estimates and projections rather than field results.
MRV systems present a related challenge. The IEA has stated that permanence and verification frameworks add real costs, but the exact magnitude varies across geologies, capture methods, and regulatory regimes. A ton of CO2 injected into a Texas well formation faces different monitoring requirements than carbon stored through enhanced weathering or ocean-based approaches. Some methods require decades of subsurface pressure and plume tracking; others depend on periodic sampling of soils, rocks, or seawater chemistry. Without a common measurement framework, buyers in voluntary carbon markets and governments designing compliance programs cannot easily compare removal claims across technologies.
Permitting presents its own bottleneck. The EPA’s Class VI process for the Ector County wells took years from application to final authorization. Scaling that model to dozens or hundreds of sites, potentially across state lines and into jurisdictions with different geological and political conditions, introduces legal and logistical questions that have not been resolved. Cross-border transport of CO2, long-term liability for storage integrity, and mutual recognition of removal credits across trading schemes all remain unsettled.
There is also the question of demand. Voluntary carbon markets have faced credibility crises over offset quality, and corporate buyers are increasingly wary of purchasing removal credits that lack robust verification. Compliance markets, where governments mandate emissions reductions, have been slow to integrate engineered removal. Until policy frameworks create durable revenue streams for verified carbon removal, even fully permitted and technically sound projects may struggle to attract the financing needed to operate at scale.
Why the gap between models and reality matters now
The practical stakes are not abstract. The MIT researchers’ central finding, that models assume costs and deployment speeds that do not match current engineering realities, has direct implications for how governments allocate climate budgets and how corporations set decarbonization timelines. If DAC cannot scale at the speeds baked into net-zero roadmaps, those plans need revision. Delayed or insufficient carbon removal capacity would mean either steeper near-term emissions cuts, which carry their own economic and political costs, or higher peak warming temperatures.
For companies, investors, and policymakers tracking this space, the most defensible approach as of May 2026 is straightforward: treat any carbon removal cost figure below current demonstrated ranges as aspirational until an operating facility confirms it with third-party verification. The IEA, MIT, and the growing body of peer-reviewed literature all point in the same direction. Engineered carbon removal is a valuable and likely necessary tool, but it is not yet a proven one at the scale or price that climate plans require. The most robust path to net zero still prioritizes rapid emissions cuts while treating removal as a critical complement under development, not a guaranteed backstop that lets the world delay harder decisions.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
