This article is part of a series of publications exploring how next-generation technologies could accelerate the energy transition if they reach their full potential.
Throughout the past decade, massive technological breakthroughs, along with dramatic declines in the cost of renewables and batteries, have accelerated the energy transition globally. But despite all this progress, cost-competitive green molecules and solutions for many areas of industrial decarbonization remain out of reach. Quantum computing could change that.
We estimate that quantum computing could eventually unlock 3 to 7 gigatons (Gt) of annual emissions savings. At the midpoint of that range—roughly 5 Gt—the savings would be equivalent to nearly a tenth of global emissions. Most of that potential is concentrated in hard-to-abate sectors, such as steel, cement, chemicals, trucking, aviation, and shipping, where few economic solutions exist today.
But computational breakthroughs do not translate immediately into emissions reductions. Companies must still commercialize and deploy quantum-enabled solutions across industrial assets, some of which can operate for 20 to 40 years. Assuming normal asset-replacement cycles, we estimate that only about 10% to 15% of quantum computing’s full emissions-savings potential could be realized by 2040. For business leaders and investors, however, long deployment cycles make quantum relevant long before the technology reaches maturity. Capital plans and investment decisions made today could determine how swiftly companies can adopt quantum-enabled technologies as they become commercially viable.
The rapid growth of AI has demonstrated how quickly a computing breakthrough can create new demands on the energy system. Our analysis suggests that quantum could follow a very different trajectory, generating a relatively small carbon footprint even as its applications scale.
If quantum-enabled advances in materials, chemistry, and industrial processes reach commercial scale, the implications could extend far beyond the emissions savings. They could fundamentally change parts of the energy system and broader economy—electricity generation and storage, food production, and carbon removal. We have identified three ways that quantum could accelerate the energy transition over the long term, and we have noted several key decisions that businesses and investors can make today to prepare for those opportunities.
Quantum Computing Could Scale Without an AI-Sized Carbon Footprint
The energy demands of AI are a central concern for businesses and policymakers, and it’s important to know whether quantum computing could create a similar challenge as it scales. Our analysis suggests that it is unlikely to do so. (See “Our Methodology.”)
Our Methodology
Emissions Generated. We estimated both manufacturing emissions and operational emissions for the quantum computing fleet through 2040. Our calculations for manufacturing emissions are based on the number and size of new machines built, including the physical qubits required for error correction and expected chip yields. We gauged operational emissions on the basis of the number of active machines, their power requirements and runtimes, and the carbon intensity of the electricity used to operate them. We assumed that there would be four generations of machines through 2040, with full-scale machines drawing up to approximately 1 MW of power and a global fleet of roughly 230 to 1,400 machines. By our calculations, manufacturing and operations together will produce an estimated 0.001 to 0.15 Gt of CO2e in 2040.
Full Emissions-Savings Potential. We assessed approximately 50 known quantum applications and identified nine—primarily molecular and materials simulations—in which quantum could uniquely enable emissions reductions by solving computational problems that classical methods cannot practically solve today. For each use case, we started with the sector’s projected emissions and estimated both the share that could be decarbonized and the share for which quantum could provide a distinctive enabling technology. Across the nine applications, this yields a full-deployment emissions-savings potential of approximately 3 Gt to 7 Gt of CO2e annually, with an average of approximately 5 Gt.
Potential Realized by 2040. Our full savings estimate assumes complete deployment of quantum-enabled solutions. To estimate how much could actually be realized by 2040, we accounted for when each solution might become available, how long it would take to build or retrofit supporting infrastructure, and what the replacement cycle of the relevant physical assets is likely to be. Given normal asset-turnover cycles, we estimate that approximately 10% to 15% of the full emissions-savings potential could be realized by 2040. This assumption is deliberately conservative: sufficiently attractive quantum-enabled solutions could encourage companies to retrofit or replace assets earlier, which would accelerate adoption and emissions savings.
Quantum computing itself consumes significant amounts of energy, but its overall emissions footprint is likely to remain relatively small. Our modeling indicates that quantum computing will generate approximately 0.09 Gt of carbon dioxide equivalent (CO2e) in 2040, equivalent to less than 0.2% of the expected 50 Gt of global emissions that year. This would equate to approximately 7% of potential 2040 data center emissions at the midpoint. These emissions come primarily from one-time manufacturing processes associated with building each machine, in addition to the ongoing energy required to operate the fleet. (See Exhibit 1.)
Compared with roughly 5 Gt of potential annual emissions savings, that level of emissions implies a climate benefit of approximately 60 to 1. The potential savings would be equivalent to eliminating nearly 10% of global emissions today and roughly 90% of current US emissions.
The relatively small footprint is not an indication that quantum computers themselves are energy efficient. A full-scale quantum computer could draw roughly 1 MW of power, comparable to the energy demands of a small data center. But quantum is unlikely to require the extensive infrastructure built to support AI. More than 11,000 data centers, ranging in operating requirements from roughly 1 MW to 100 MW, already power AI and other types of high-performance computing, with some hyperscale facilities approaching 1 GW. We estimate that approximately 230 to 1,400 machines could serve global quantum computing demand by 2040, even after accounting for spare capacity and machines operated by governments, laboratories, and universities. That relatively small fleet reflects how quantum computing is likely to be used: not as a wholesale replacement for classical computing, but for a specialized set of problems where it can provide a computational advantage.
Additional applications could emerge as the technology matures, increasing demand beyond what we model today. But currently identifiable use cases suggest that the relatively small number of machines required will limit quantum’s direct emissions, despite the high power consumption of each machine.
Quantum’s indirect emissions from enabling advances in oil and gas, AI, or reencryption are likely to be limited, too. In oil and gas, companies are exploring early applications in areas such as discovery and optimization, but the emissions impact is likely to remain limited. Nor do we expect quantum to materially increase generative AI emissions, because its advantage lies in solving different types of problems rather than in making general-purpose computing cheaper. And while the shift to quantum-safe cryptography could require upgrades to infrastructure such as satellites and fiber networks, the associated emissions would largely be one-time rather than recurring.
Where Quantum Could Unlock New Paths to Decarbonization
Of the roughly 50 known quantum computing applications that we assessed, 16 could plausibly reduce emissions. In seven of those cases—such as solar photovoltaic cells and electric cars—the same savings could be achieved through existing technologies, or the savings were caused by factors that greater computational power would do little to address. That leaves nine use cases where quantum could enable emissions reductions that classical computing and AI cannot achieve today. (See Exhibit 2.)
These nine use cases share an important characteristic: they depend on simulating how molecules and materials behave and interact. Designing a better catalyst, battery electrode, or carbon-capture material requires accurately modeling interactions among electrons. The process of accurately modeling a small molecule would take thousands of years for even the largest supercomputers.
Quantum computers are much better at modeling complex interactions at the atomic and molecular scale. This could enable researchers to discover a much wider set of catalysts, sorbents, and compounds, accelerating the search for solutions that possess the properties needed to decarbonize some of the hardest-to-abate industrial sectors.
We have highlighted three applications to illustrate the range of climate problems that quantum could help address: carbon capture, green hydrogen and ammonia, and batteries. (See Exhibit 3.)
Carbon Capture
Carbon capture can alleviate emissions that remain difficult or expensive to eliminate at their source, but cost remains a major barrier. Direct air capture (DAC), for example, costs approximately $600 to $1,000 per ton today, well above the roughly $100-per-ton level that experts often cite as necessary for adoption at scale.
Part of that cost involves the materials used to capture CO₂. An effective sorbent must bind strongly enough to capture carbon but weakly enough to release it without requiring large amounts of energy. Classical computers can screen known candidate materials, but accurately evaluating millions of novel and complex structures becomes increasingly difficult.
Quantum computing could expand the set of materials that researchers can realistically evaluate, enabling more accurate simulation of how novel sorbents interact with CO₂ without relying as heavily on approximations derived from past examples. Finding materials that capture and release carbon more efficiently could reduce the energy—and therefore the cost—required for carbon capture, helping make it viable at much greater scale.
Green Hydrogen and Ammonia
Ammonia forms the basis of synthetic fertilizers that feed roughly half of the world’s population, but producing it accounts for approximately 1% to 2% of global emissions. Today’s process for creating ammonia relies on hydrogen derived largely from natural gas, and it requires extreme temperatures and pressure to make the chemical conversion.
A better catalyst could reduce the energy necessary for that reaction and help make low-emission ammonia more economical. The challenge is finding one. Predicting how a novel catalyst will perform requires modeling complex interactions between its electrons and nitrogen (N2) molecules, precisely the type of problem that becomes difficult for classical computers to solve as molecular complexity increases.
Sufficiently capable quantum computers could simulate a broader range of novel catalysts more accurately, helping researchers identify promising candidates before synthesis and testing. The result could be a faster path toward lower-cost, lower-emission ammonia production, supporting fertilizer that depends less on fossil fuels and strengthening the case for lower-carbon shipping fuel. These same advances would also make the catalysts used to produce green hydrogen more efficient, lowering its cost, improving the business case for hydrogen-based steelmaking, and thereby opening another route to cutting hard-to-abate emissions in steel production. The largest gains from unlocking green hydrogen and ammonia come not from lower carbon production itself, but from the downstream industries that they help decarbonize.
Batteries
Better batteries are critical to several parts of the energy transition, from electrifying long-haul trucks to storing renewable electricity for times when supply exceeds demand. But today’s batteries face significant challenges in energy storage, lifespan, and cost, which restricts their use in some applications where they could have the greatest climate impact.
The core challenge is to find the right combination of materials for the electrode and electrolytes—the components that determine how much energy a battery holds, how fast it charges, and how long it lasts. Classical computers can simulate simple, well-understood chemistries, but as materials become more complex, the millions of potential configurations become impractical to simulate or test physically, which severely limits the number of options that researchers can explore.
Quantum computers can accurately simulate how novel electrodes and electrolytes behave without having to physically build them. By directing experimentation toward the most promising chemistries, quantum could accelerate the development of batteries that store more energy, cost less, or perform better, expanding electrification into applications that today's batteries cannot serve practically or economically.
Will Quantum’s Benefits Actually Arrive in Time?
Across the nine applications that we identified, quantum computing could eventually enable approximately 3 Gt to 7 Gt of annual emissions savings if the resulting solutions were fully deployed. But even if quantum technologies unlock the full emissions-reduction opportunity, realizing that value will depend largely on factors outside quantum itself. If commercially viable quantum solutions emerge around 2035, companies will still need to translate those breakthroughs into industrial-scale processes, build or retrofit the infrastructure needed to produce them, and deploy them across existing assets.
That process could take decades, particularly in hard-to-abate sectors where much of quantum’s potential is concentrated. Steel, cement, chemicals, aluminum, oil and gas, aviation, shipping, and trucking are likely to account for roughly 30% of global emissions in 2040, and our analysis suggests that quantum could eventually abate 20% to 40% of their emissions. (See Exhibit 4.)
But many of these sectors rely on long-lived, capital-intensive assets such as steel mills, cement kilns, chemical plants, and carbon-capture facilities, which may be refurbished or replaced only every 20 to 40 years. As a result, even when quantum enables a better material or process, adoption may have to wait until a company replaces or substantially upgrades the underlying asset. A breakthrough in cement chemistry, for example, may have little near-term impact on a kiln that was rebuilt the previous year and is expected to operate for another 30 years. Shorter-lived assets such as trucks can adopt new technologies much faster, but the same principle applies: the pace of emissions reduction depends heavily on how quickly the industry leverages the new technology and how swiftly the underlying asset base turns over.
We estimate that industries may realize only approximately 10% to 15% of the full emissions-savings potential by 2040 under normal asset-replacement cycles. (See Exhibit 5.) Quantum’s climate impact is therefore likely to be heavily back-weighted, with physical deployment rather than technological potential determining how rapidly companies can capture the opportunity. But our 10% to 15% estimate is not a ceiling. A combination of sufficiently compelling economics, government mandates, and coordinated global industry standards could push that figure higher by 2040. Subsidies for industrial retrofits and asset replacement could accelerate deployment beyond normal turnover cycles, and greater prioritization of solutions such as DAC could create additional upside beyond our modeled scenario.
The investment opportunity is similarly primed for change. Some of quantum’s largest potential beneficiaries could be steelmakers, cement producers, chemical companies, and other industrial incumbents that can deploy these solutions at scale. These companies already own the physical assets, engineering capabilities, and operating infrastructure required to turn computational breakthroughs into real-world emissions reductions. Investors could therefore back startups engaged in developing new technologies while also directing capital toward the established industrial companies that will ultimately deploy them. Funding both innovators and deployers could help accelerate retrofitting and asset replacement, and advance emissions savings that might otherwise take decades to materialize.
Business leaders, investors, and policymakers should prepare before quantum reaches commercial maturity. Devising capital plans, developing asset replacement strategies, making infrastructure investments, and forming technology partnerships today could determine how quickly companies can put quantum-enabled breakthroughs to work once they become commercially viable. Well-conceived mandates, standards, incentives, and targeted investment could determine whether quantum-enabled solutions help accelerate existing asset-replacement cycles.
Three Ways Quantum Could Transform the Energy Transition
The emissions savings that our analysis quantifies capture only the direct impact of quantum-enabled technologies. If these breakthroughs become commercially viable at scale, their effects could extend much further, changing the economics of the energy transition and creating ripple effects across industries, economies, and societies.
These outcomes are not inevitable, and some are farther from today’s reality than others. But three examples illustrate how breakthroughs that begin with a quantum computation could ultimately transform energy systems—and why decisions about capital, infrastructure, and technology made today could shape who captures that value.
Better Storage Could Accelerate Global Electrification
IEA STEPS projects that renewable energy will grow from roughly 30% of global electricity generation in 2024 to approximately 60% by 2040. But wind and solar do not always generate electricity when needed. When generation exceeds demand, power is curtailed; and when it falls short, grids often rely on fossil fuels as backup. In 2024, renewable energy curtailment increased by approximately 55%, reaching 4% of wind and 3% of solar PV generation.
Quantum-enabled advances in battery chemistry could alleviate this problem. Batteries capable of storing energy for longer periods could capture electricity that would otherwise be curtailed and make it available when the grid needs it, reducing reliance on fossil-fuel backup and strengthening the case for renewables to supply a greater share of the global energy mix. Improvements in battery cost, energy density, and transportability could expand access to electricity in regions hampered by the difficulty of transporting today’s batteries, by suboptimal weather conditions for renewables, or by underdeveloped renewable infrastructure.
Businesses and investors have an opportunity to invest in regions best positioned for renewable generation while also backing storage technologies, infrastructure, and other key enablers.
Green Molecules Could Make Food and Energy More Abundant
Food security and access to electricity are fundamental to human development, yet both can be scarce in developing economies. Synthetic fertilizer supports food production for roughly half of the world’s population and has helped increase crop yields by 30% to 50%. But its production depends heavily on natural gas, which exposes fertilizer-importing countries to fluctuating gas prices and supply shocks. Studies show that crop yields could be profitably doubled if fertilizers weren't so expensive. In addition, hundreds of millions of people still lack access to electricity.
Quantum computing could help address both challenges. Better catalysts could make green ammonia more economical, reducing fertilizer production’s dependence on natural gas and supporting greater agricultural productivity and food security. Better batteries could make electricity cheaper and more accessible in regions where reliable power is currently too costly or impractical to provide.
The potential result is a development benefit that extends well beyond emissions: greater agricultural productivity and food security, alongside wider access to reliable power. Businesses, investors, and governments can identify developing regions where cheaper energy and agricultural inputs would have the greatest impact and invest in the energy and agricultural infrastructure needed to deploy them at scale.
Affordable Carbon Removal Could Create a Carbon Thermostat
DAC remains expensive, at roughly $600 to $1,000 per ton, and IEA STEPS projects that DAC will capture only about 0.2 Gt of CO2 by 2040. By comparison, the US Environmental Protection Agency’s estimate of the social cost of carbon—the estimated societal damage resulting from each additional ton of emissions—was roughly $200 per ton in 2023, and the EPA expected it to rise to approximately $300 per ton by 2050. Bringing DAC costs closer to the $100-per-ton level often cited for adoption at scale could make removing a ton of carbon cheaper than bearing the estimated societal cost of leaving it in the atmosphere.
Quantum-enabled advances in sorbents could help meet this challenge. If DAC becomes sufficiently economical, businesses and governments will have another option for addressing hard-to-abate emissions besides relying exclusively on every sector, company, and country to eliminate emissions at the source. In a more ambitious scenario, DAC might function as a kind of carbon thermostat: policymakers could set a desired climate trajectory and use carbon removal to narrow the gap between remaining emissions and the reductions required to achieve it.
Storage capacity and deployment speed would still determine how quickly these changes could occur. But lower-cost DAC could shift part of decarbonization from being a coordinating challenge across thousands of emitters to being a matter of scaling a technology that possesses increasingly attractive economics.
For businesses and governments, the priority today is to invest in the most promising early carbon-removal technologies and use government grants, partnerships, and other forms of support to accelerate their development and deployment.
Quantum computing is still years away from realizing its full potential, and there is no guarantee that the scenarios modeled in our research will materialize. But our analysis suggests that its significance for the energy transition could be substantial, thanks to the new solutions that it could make possible in sectors where emissions have been particularly difficult to reduce.
Realizing quantum computing’s full potential in the energy transition will require progress on two fronts. Quantum technology must mature enough to solve the complex chemistry and materials problems at the heart of these applications. Meanwhile, businesses and governments must be prepared to commercialize those discoveries and deploy them across the industrial assets, infrastructure, and supply chains where emissions reductions actually occur.
Quantum may provide computational breakthroughs, but the pace of climate impact will depend on how quickly companies can put those advances to work. Investments made today—in quantum capabilities, infrastructure, and the companies positioned to deploy new solutions—could determine how much of quantum’s potential comes to fruition and who captures the value that it creates.
