Driving down the cost of carbon removal: Why innovation matters

The full version of this analysis can be found in Chapter 9 of The State of Energy Innovation report published by the IEA earlier this year. This excerpted version – updated with the latest data – will frame a dedicated discussion on the topic at the IEA Energy Innovation Forum 2025, which will take place in Toronto, Canada, on 29 October 2025 ahead of the G7 Energy and Environment Ministers’ Meeting.

Carbon dioxide removal (CDR) covers a growing set of technologies that pull carbon dioxide (CO2) out of the air and store it permanently. As such, CDR has the potential to enable deep emissions cuts and reduce the legacy carbon already in the atmosphere.

The CDR sector has expanded rapidly in recent years, in large part thanks to corporate demand, policy support and venture capital. However, costs remain high, and large-scale projects are still rare. Scaling up an industry that can remove millions of tonnes of CO2 in the coming decades will demand advancing a portfolio of approaches; some of the lower-cost options face limits to deployment, while others with much larger potential are currently more expensive and still at an early stage of development. Sustained innovation across this spectrum will be essential to reduce costs, improve performance and strengthen tools for monitoring, reporting and verification (MRV).

To unlock this potential, innovators need opportunities to test and scale their designs while policy makers should continue to focus resources on the most promising pathways. As with all new technologies, governments can play a uniquely catalytic role, and countries that provide stable and sustained support are most likely to receive the economic benefits, such as investment flows and job creation, that would be associated with the sector’s growth.

Carbon dioxide removal is no longer a fringe concept. In just five years, the number of start-ups in the sector has grown fivefold and venture capital investment has increased sevenfold. Some of the newest approaches – such as enhanced rock weathering, biomass storage and ocean-based capture – have emerged rapidly, together attracting almost one-third of investment in the sector in 2024. Innovators are now commissioning sizeable facilities that will be able to capture 15-40 thousand tonnes (kt) of CO2 per year – the equivalent of taking between 7 000 and 17 000 conventional cars off the road. Meanwhile, utilities and oil and gas majors are moving ahead with even larger projects, targeting 200-800 kt of annual removals. Strategic interest in CDR projects among energy sector companies has been growing: one oil company acquired a start-up for USD 1.1 billion in 2023, while an oil industry joint venture is considering a USD 500 million investment in a CDR project, for example.

Even as investment flows towards a range of CDR technologies, operational capacity today primarily uses just a few approaches that have been tested around the world over the past decade. The same holds for announced projects that could be running by 2030. The most popular method is bioenergy with CO2 capture and storage, or BECCS, whereby CO2 is withdrawn from the atmosphere via photosynthesis in plants, then captured in a more concentrated form when the carbon is oxidised again for industrial purposes, after which it is stored permanently. The other main approach today is direct air capture and storage, or DAC, whereby a machine pulls CO2 from the air; it is then permanently stored in deep geological formations or minerals.

A major step change in deployment for these two approaches is anticipated in the coming years, with several first-of-a-kind projects expected to be commissioned soon in Denmark, Iceland, Norway, Sweden, and the United States. Projects already under construction alone could more than double current BECCS capacity and expand DACS capacity by a factor of 50 worldwide. If all announced projects proceed, total global CO₂ removal capacity could grow close to 80-fold by 2035 from today’s level of around 1 million tonnes (Mt) of CO₂ per year.

Alternative approaches that could allow CDR to be deployed in more locations are also gaining traction, though at much smaller scale so far. Underground biomass storage companies have commissioned their first 10-15 kt facilities to grow plants, treat them and then store them in ways that prevent their decay for millennia, thereby preventing the captured CO2 from being re-released. Ocean-based carbon removal is moving forward with kt-scale pilots that separate and store CO2 from the sea, where CO2 concentration is higher than in the air, in ways that ensure the ocean then reabsorbs more CO2 from the air. Mineralisation-based approaches, which bring reactive minerals into contact with ambient CO2 and trap it into rock, are advancing through field trials. These efforts remain small in absolute terms but, with further research and industrialisation, could become the most cost-effective options in many locations. They are therefore vital to diversifying the portfolio of future CDR options.

Cost remains one of the most important barriers to scaling industrial carbon dioxide removal. Technologies today span a wide cost range, with large uncertainties over how quickly costs can fall.

DAC is expensive because the concentration of CO2 in the atmosphere is very low – only about 420 to 430 parts per million. This means a lot of energy is needed to process large volumes of air to capture each tonne of CO2. Current projects cost an estimated USD 500 to 1 900 per tonne of CO₂. Advances in capture materials and scale effects could bring costs down to around USD 300 per tonne by mid-century, while some next-generation designs are targeting the USD 100 per tonne mark.

BECCS has the potential to be significantly cheaper. Removal costs are estimated at USD 40-50 per tonne in biorefineries where CO2 concentrations are higher, and USD 95-120 per tonne of CO2 for more diluted sources such as heat and power plants or pulp and paper mills. However, achieving these lower costs will depend on optimising plant designs, benefiting from economies of scale, and minimising the costs of CO2 transport and storage for existing plants that are not located close to storage sites. First-of-a-kind BECCS projects currently range from USD 75-300 per tonne of CO2. Removal costs also depend heavily on factors like the emissions from producing and transporting biomass, as well as how consistently plants can operate. Improving feedstock options and refining operations will be key to bringing costs down.

Other approaches advertise lower costs today – for example, underground biomass storage firms report less than USD 100 per tonne – but the absence of operational facilities at scale make it difficult to corroborate these values. Robust monitoring, reporting and verification frameworks that provide confidence in the permanency of storage methods are yet to be developed for some of these approaches, and this process can add non-trivial costs.

Importantly, cost reductions largely depend on deployment. Only by building, testing and iterating can CDR technologies climb their learning curves. Early projects have been largely backed by public funding, with over USD 5 billion announced over the past five years, the majority of which set to support first-of-a-kind BECCS and DAC projects. Yet such funding is not set in stone. Voluntary carbon markets have also played a critical role by providing stronger demand signals and revenue certainty. These markets, however, remain highly concentrated: about 65% of purchases of carbon credits from CDR projects in 2024 came from a single buyer, Microsoft. Without more stable and diverse demand signals – and without sustained public support – investment could stall before CDR can reach commercial maturity.

Unlocking the sector’s full innovation potential would require far more than current levels of private capital and voluntary market demand. Governments are uniquely positioned to mitigate risks of early projects, fund knowledge-sharing and stimulate markets in ways that accelerate cost reductions. Building on best practices for promoting technological innovation, and considering the specific needs of CDR, several important levers stand out:

Governments can co-fund a global portfolio of pilot and demonstration projects. Not every CDR option will scale to the million-tonne level within a decade, but testing a wide set of approaches is essential to identify which are the most promising, while also filling critical RD&D gaps.

Open-access CDR testbeds can dramatically lower the risks innovators face. Facilities where new capture materials or system designs can be independently evaluated cut both time and costs, enabling better access to finance.

Governments can help build demand and boost investors’ confidence. Advance purchase commitments for CDR credits are one way to provide the long-term demand certainty that voluntary markets alone cannot deliver, and they can help attract private capital. Public procurement can also generate valuable data – on costs, performance, and monitoring, reporting and verification – that governments can share to help the entire field advance. The integration of carbon removals in compliance and international carbon markets, along with supporting policies such as carbon contracts for difference (CCfD), offer additional pathways to build sustainable CDR markets.

Robust frameworks for monitoring, reporting and verification should evolve alongside deployment. These processes are crucial to ensure the integrity of removals, yet significant knowledge gaps remain in assessing and certifying storage permanence, particularly for open-system approaches. Governments are uniquely placed to coordinate two-way data flows between projects and researchers, and to improve models, guide field trials and inform certification frameworks. They can also ensure the adoption of common standards and definitions across jurisdictions. International initiatives such as Mission Innovation’s CDR Mission already show how cross-border collaboration can accelerate learning and minimise duplication.