Carbon Capture for Natural Gas-Fired Power Generation: An Opportunity for Hyperscalers

Key Takeaways

• Enhanced geothermal systems (EGS) overcome traditional geothermal energy limitations by engineering subsurface conditions rather than searching for them, enabling widespread deployment of clean firm renewable power.
• Induced seismicity from high-pressure injection has caused major EGS project cancellations, but advanced approaches like Sage Geosystems’ gravity-assisted fracturing mitigate this risk by avoiding overpressures and directing fractures downward away from fault zones.
• Sage’s $97 million Series B financing, co-led by Ormat Technologies and Carbon Direct Capital, will fund the first commercial EGS facility at an existing Ormat plant—accelerating the transition from innovation to grid-scale deployment.
• For hyperscalers racing to power AI infrastructure, EGS offers a credible path to firm, 24/7 low-carbon power at scale.

Geothermal Energy: The Heat (And Pressure) Is On

For decades, geothermal energy has occupied a compelling yet narrow place in the clean energy landscape. It offers what the grid increasingly needs— firm, renewable, low-carbon power—yet has remained constrained by limited siting flexibility, high upfront resource risk, and persistent concerns around induced seismicity. 

Enhanced geothermal systems (EGS) change that equation. Instead of searching for ideal subsurface conditions, EGS engineers them directly. In doing so, EGS rewrites the rules of where geothermal energy can be deployed and how far it can scale, with the potential to transform this historically niche resource into a widely deployable form of clean firm power.

One such solution, Sage Geosystems, uses a pressure-managed EGS approach to extract geothermal energy from engineered subsurface reservoirs, while explicitly addressing the seismicity risks that have constrained earlier projects. 

How EGS Scales Geothermal Energy

Conventional geothermal power relies on a narrow set of subsurface conditions: sufficiently high temperatures, naturally occurring fluid, and enough permeability to circulate fluid through hot rock. In practice, those conditions coexist in only a few places—nearly all US commercial geothermal power generation is concentrated in California, Nevada, and a handful of sites across Utah and Hawaii.

EGS reduces this constraint by engineering permeability and fluid access rather than relying on their natural presence. While fluid access and permeability are harder to find, heat is not: the Earth’s natural geothermal gradient ensures that hot rock exists almost everywhere at sufficient depth. 

By reducing the number of variables that must be discovered rather than designed, EGS expands siting flexibility and lowers the resource risk that has historically constrained geothermal development. The Department of Energy (DOE) estimates this approach could unlock more than 5,500 GW annually of US resource potential, which, when converted to electric power, is roughly comparable to the total installed power capacity of the US today.

One remaining challenge has been induced seismicity. When you inject pressurized water into rock and create fractures, you are adding lubrication to geological systems that have been static for millions of years. If those fractures propagate into existing fault zones, the faults can slip, producing earthquakes. Projects in Basel, Switzerland (2006) and Pohang, South Korea (2017) triggered magnitude 3.4 and 5.4 events, respectively, both leading to project cancellations and regulatory backlash that set the industry back years.

Sage's approach to EGS is designed to address this risk directly. Rather than relying on high-pressure hydraulic stimulation, Sage uses a gravity-assisted fracturing approach that helps avoid the high overpressures that can drive fault slip. Further, its approach biases fracture growth downward and away from shallow, critically stressed fault systems. By understanding causes and conditions, Sage aims to work with the subsurface, not against it. 

This is not a minor technical detail. It is the difference between a technology that can scale with community acceptance and one that faces opposition at every site. For a hyperscaler evaluating geothermal offtake agreements, seismicity risk translates directly into permitting risk, timeline risk, and reputational risk. 

The Clean Firm Power Gap Driving EGS Adoption

To understand why this matters, start with the problem hyperscalers are trying to solve. Solar and wind have scaled dramatically, but they face a structural limitation: they do not generate power when the sun is not shining or the wind is not blowing. Batteries help bridge short gaps, but current technology cannot economically cover multi-day periods of low renewable output. Nuclear provides firm generation, but faces permitting timelines that extend well past 2030.

This creates what might be called the 'clean firm power gap'—the difference between what hyperscalers need (24/7, low-carbon, scalable to gigawatts) and what current markets can supply. A single large AI training cluster can consume more than 100 MW continuously. Meta, Google, and Microsoft are planning data center campuses that will require gigawatts of capacity. The gap between demand and available clean firm power supply is widening, not narrowing.

Geothermal energy aligns closely with this need. Unlike solar or wind, geothermal power plants run continuously, with capacity factors that routinely exceed 90%. And unlike nuclear, geothermal projects can, in principle, be permitted and built on shorter timelines. The challenge has never been performance, rather availability: with the emergence of EGS, geothermal power is expanding where clean firm power can realistically be built, arriving at a moment when the grid’s need for dependable, low-carbon supply has never been greater.

Sage Raises $97 Million to Deploy Geothermal at Ormat Site

Sage Geosystems announced $97 million in Series B financing co-led by Ormat Technologies, the world's largest geothermal operator, and Carbon Direct Capital, a leading energy investing firm. Ormat will host Sage's first commercial facility at an existing Ormat plant.

The investment signals that EGS has become investable to the industry built to scale it. For Ormat, the logic is clear: conventional geothermal is constrained by resource availability. EGS expands the addressable market, but requires the subsurface capabilities that conventional operators don't typically possess by Sage does.

Why the Partnership Structure Works

EGS proposes that the fastest way to scalable power is to eliminate the resource risks that beset conventional geothermal projects. These risks do not simply disappear: they are transferred into subsurface and remain unproven at scale. Conventional operators locate naturally permeable reservoirs. EGS requires creating permeability in crystalline rock and managing induced seismicity risks that don't exist in hydrothermal systems. Sage is actively addressing the seismicity problem that ended projects in Basel and Pohang. Ormat brings everything else: turbines, plant operations, grid expertise, and six decades of operational knowledge.

Building at an existing Ormat site provides another advantage: established subsurface characterization, proven geological stability, and grid infrastructure already in place. For a first commercial deployment, this de-risks demonstration in ways greenfield sites cannot.

Both companies move faster together because the technical capabilities required to make EGS work don't naturally exist within a single organization.

What Hyperscaler Demand Means for the Power Sector

Meta's 150 MW power purchase agreement with Sage—announced in August 2024, with delivery planned for sites east of the Rocky Mountains—adds another dimension to this story. Hyperscalers have concluded that waiting for clean firm power technologies to mature before signing contracts means those technologies may not be available when needed. So they are becoming anchor customers, providing the revenue certainty that enables projects to secure financing.

For geothermal power specifically, this demand signal is transformative. Contracted offtake from creditworthy counterparties changes project economics fundamentally. It lowers the cost of capital, enables debt financing, and de-risks the investment case for additional capacity. The hyperscaler model has already accelerated deployment in solar, wind, and battery storage. Its application to geothermal power may prove similarly catalytic.

The Final Constraint

EGS is not a silver bullet, but it is beginning to look like a credible answer to a growing-problem: how to deliver clean firm power at scale, in more places, and on timelines that match accelerating demand. Advances in subsurface engineering are reducing the resource and seismicity risks that once confined geothermal to a narrow footprint, while partnerships with incumbent operators are showing how those advances can be integrated into existing energy infrastructure. 

At the same time, hyperscalers are reshaping the market by signaling demand early, underwriting first deployments, and pulling technologies forward rather than waiting for them to mature on their own. That combination of technical progress, industrial adoption, and committed buyers is what turns promising concepts into deployable systems. 

Whether EGS ultimately fulfills its potential will depend on repeatable and continued performance under real-world conditions. But the recent alignment of science, incumbents, and demand suggests EGS is moving beyond possibility and into a phase where the final constraint is no longer what the Earth can provide, but what the energy system is prepared to build. 

Frequently Asked Questions

What is an enhanced geothermal system?

An enhanced geothermal system, or EGS, produces geothermal energy by engineering underground conditions needed to circulate fluid through hot rock. Unlike conventional geothermal projects, which depend on naturally occurring heat, fluids, and permeability occurring together, EGS can create or enhance permeability and fluid circulation, greatly expanding the locations where geothermal power may be developed. 

Why is EGS important for data centers and AI infrastructure?
AI and data centers require large amounts of electricity around the clock, creating demand for power sources that are both low-carbon and firmly available. EGS could provide high capacity factor (greater than 90%), 24/7 clean electricity in more locations than conventional geothermal, making it a potentially valuable complement to intermittent renewable resources. 

What is induced seismicity, and how are new EGS technologies addressing it?

Induced seismicity refers to earthquakes caused by changes in underground pressures or stresses caused by human activities. Earlier EGS projects demonstrated that high-pressure fluid injection can activate existing faults and in some cases triggered noticeable earthquakes and intense public backlash. New EGS approaches are being designed to better control reservoir pressure, fracture development, and proximity to faults, reducing seismicity risk while maintaining the fluid circulation needed to extract geothermal energy.

Can EGS be deployed anywhere?

EGS substantially expands geothermal’s geographic potential, but it does not make every location equally suitable. Projects still depend on factors including underground temperature, how deep they need to drill to access that temperature, water availability, seismic risk, and whether the rocks are of type suitable to hold and sustain engineered fracture networks.

Key Takeaways

• Data centers are driving surging demand for new, firm electricity supply, accelerating natural gas-fired power generation.
• Carbon capture and storage (CCS) offers a practical way to balance long-term climate commitments with the need for new electricity generation in the near term.
• New natural gas-fired power plants must be capture-committed, not just capture-ready, potentially delivering power in 18 months and decarbonized power 18-24 months later.
• Capture-committed plants integrate planning and finance for the CO₂ capture, transport, and storage value chain from the start.
• Relae believes early investment in engineering, infrastructure, and community engagement is essential to meet capture commitments.

A New Era of Electricity Demand and Climate Pressure

The US and much of the developed world are experiencing profound growth in electricity demand. Two main forces are driving this trend: (1) the push to electrify existing uses, such as vehicles and heating, to improve energy security, enhance system efficiency, and reduce air pollution; and (2) the growth of energy-intensive sectors like manufacturing, telecommunications, and AI data centers.

Among these drivers, AI is creating unique demands that catalyze specific investments in electric power generation. Astonishing AI data center buildout, led by a handful of large technology firms (sometimes called “hyperscalers”) and their utility and construction partners, is accelerating energy consumption. These firms prioritize speed. When asked for their top five criteria for bringing new AI infrastructure online, one executive responded: “Speed, speed, speed, cost, and carbon emissions.”

Data centers require reliable, always-on power (referred to as “firm power”). This differs from other use cases, such as residential or commercial, which do not need the same amount of power across all hours. While hyperscalers and their partners are investing in renewables, nuclear, and geothermal energy at a remarkable pace, renewable resources alone do not yet meet the exploding demand for firm power. 

Natural Gas Provides Firm Power but Drives Emissions Higher

The mismatch between data center power needs and variable renewable generation is fueling a boom in natural gas-fired power generation. The pipeline of new natural gas-fired power plants is enormous. Plants under construction in 2025 would, by themselves, add roughly 25 million tonnes of greenhouse gases each year to the air and oceans. The full suite of plants in planning is at least 10 times larger. Existing gas plants are also being used more and staying online longer.

This rapid buildout is creating tension with corporate climate goals. Hyperscalers remain seriously committed to reducing emissions, but their ability to hit those targets is undermined by the need to procure new, large-scale electricity generation quickly.

Carbon Capture Aligns with Data Center Energy Demands

Carbon capture and storage is one way to bridge the gap. Data centers operate continuously and may have the ability to shift or curtail load. This demand profile suits the duty cycles of natural gas turbines and CCS facilities well. The potential to reduce direct emissions is profound: today’s CCS technology can capture 95% or more of CO₂ emissions at competitive costs in many markets.

This has led to a resurging interest in the concept of capture-ready gas power generation. New natural gas power plants can be built and brought online in 18 months. In a capture-ready plant, the developers integrate the necessary interfaces and reserve additional land, water, and energy to enable a carbon capture project to be built at a future date. In favorable locations, carbon capture can be added to a capture-ready plant in 18-24 months.  

However, past experience shows that capture-ready plants rarely deliver. The ambition and commitment of the developers were contingent on policy and market signals that were either too small or never materialized. While the base plant may have made economic sense in terms of energy value for investment, it does not appear anyone was willing to pay the climate premium for CCS.  

As David Hawkins of the Natural Resource Defense Council famously said, “If your plant is capture ready, my driveway is Ferrari ready.” To bring David’s humorous analogy back to the specifics here: don’t build a new driveway without at least a downpayment on the car.

How to Build Capture-Committed Power Plants for CCS

A better approach is building capture-committed plants, namely facilities that integrate CCS from the start. To be capture-committed, project developers must:

• Identify geologic storage for the many millions of metric tons of CO₂ that these plants will produce each year over the next 20-30 years.
• Plan reliable CO₂ transportation from power generation to geologic storage by pipeline, rail, barge, or truck.
• Engage credible vendors of carbon capture technology that serve their needs and fit their goals.
• Fund front-end engineering design (FEED) studies.
• Arrange, or help to arrange, financing for the construction, commissioning, and operation of all necessary components in the CO₂ capture, transportation, and storage value chain.
• Ensure natural gas supply has near-zero fugitive methane emissions.
• Partner with local and frontline stakeholders to incorporate community impact into project planning, design, and financing.

Capture-committed plants send strong market signals. They help build the permitting pathways and develop the workforce, infrastructure, and community acceptance needed to avoid extra expense and delays. Done well, early commitments and investments will likely create repeatable models that reduce build times and costs.  

A Path Toward Power That’s Clean Firm and Future-Ready

Eventually more carbon-free power in the form of renewables, nuclear, and geothermal energy will be deployed to serve national and international electric load growth for all types of electrification. Over time, these resources will likely displace natural gas. Until then, hundreds of millions of tons of CO₂ will be emitted each year unless commitments are made to take tangible action now. 

Capture-committed natural gas-fired plants offer a pragmatic solution. With the right planning, financing, and community engagement, they can provide reliable power without locking in emissions, and they can deliver enormous benefits compared to uncontrolled operation. Federal and state governments can accelerate this transition by honoring and increasing CCS grants, supporting shared infrastructure, and streamlining permitting for CCS plants as they have for other clean energy supplies. These investments will enable the construction of cleaner, more resilient power infrastructure for the industries driving demand, from AI data centers to heavy industry.

[cta]

Frequently Asked Questions

What is the difference between a capture-ready and a capture-committed power plant?

A capture-ready plant creates an option to add carbon capture in the future, whereas a capture-committed plant treats capture as part of the project from day one. In a capture-ready plant, developers install the right interfaces and reserve extra land, water, and energy, but nothing obligates them to build the capture project, ever. A capture-committed developer secures options for CO₂ transportation and geologic storage, relationships with capture equipment vendors, funding for engineering studies, and financing across the full value chain before the base plant comes online.

Why have capture-ready plants historically failed to add carbon capture?

Nobody was willing to pay the climate premium. Capture-ready developers built plants that made economic sense on energy value alone, then waited for policy and market signals to justify carbon capture. Those signals were either too weak or never arrived, so the option went unexercised and no capture project was ever designed. The base plant runs uncontrolled for decades while the reserved land sits empty. David Hawkins of the Natural Resources Defense Council captured the problem well: "If your plant is capture ready, my driveway is Ferrari ready." Preserving an option costs very little. Exercising it costs a great deal, and capture-ready facilities rarely came with the funding to do so.

If renewables, nuclear, and geothermal will eventually displace gas, why invest in CCS for gas plants now?

Because greenhouse gas emissions happen in the meantime. Gas plants being built today will operate for 20 to 30 years, long before carbon-free resources scale enough to displace them. Left uncontrolled, they will emit hundreds of millions of tons of CO₂ over that span. Capture on those plants avoids most of it. Today's technology can capture 95% or more of CO₂ emissions at competitive costs in many markets.

What can federal and state governments do to accelerate capture-committed projects?

Three kinds of support matter most: funding, infrastructure, and permitting. Governments should honor and extend existing CCS incentives. Developers make capture commitments years before any revenue arrives, so uncertainty in government funding undermines the confidence these projects require. Governments should also support shared CO₂ transport and storage infrastructure. Common pipelines, rail terminals, and storage hubs make it easier for developers to secure physical CO₂ offtake. Finally, permitting for CCS should be streamlined the way it has been for other clean energy supplies. Permitting delay is a leading cause of cost overruns, and a capture-committed plant should be able to pursue capture and storage with the same intensity and speed as electricity generation.

Key Takeaways

• Batteries have competed away the revenue that financed their first wave. In Texas, ancillary prices have fallen about 90% since 2023, and price spikes driving arbitrage revenue are becoming ever less frequent.
• The business has moved from trading energy to selling capacity. In Texas, tolls work as private capacity contracts, and resource adequacy payments make up roughly 70% of battery revenue in California. Project economics increasingly pencil on a fixed $/kW-month basis rather than price spikes.
• Long-term contracts help shift the risk, but they don’t fully close the gap. Under a 15-year toll, a new battery needs about 22% less revenue, but that is still nearly three times what Texas batteries earned in the market in 2025.
• Buildout now depends on who will pay for capacity. Merchant revenue alone no longer covers a new battery, so growth hinges on how many buyers will sign capacity contracts worth nearly three times what Texas batteries earn in the market. 

Battery Revenues Are Falling While Deployments Keep Rising

US developers are expected to install about 24 GW of utility-scale battery storage energy systems (BESS) in 2026, about 60% more than the record 15 GW added in 2025, according to the EIA. Yet in Texas, where batteries earn most of their money in the market, a typical battery earned roughly 85% less in 2025 than in 2023. 

Is this unjustified overbuilding by an industry that has grown too fast? Some battery executives argue this is not overbuilding, because merchant revenue earned in wholesale markets is no longer the business. In their view, falling merchant revenue matters less because longer-term offtake contracts, such as 15-year tolls with utilities and large power buyers, make lower revenue financeable. As a result, today’s installations bank on batteries being paid as capacity assets rather than as pure energy arbitrageurs. 

We tested those claims against market data from Texas (ERCOT) and California (CAISO) between 2022 and 2026. Our core finding is that project economics in both markets are increasingly reliant on fixed $/kW-month earnings rather than arbitrage revenue.

How Does a Grid Battery Make Money?

A battery has three ways to earn:

1. Energy arbitrage (merchant: paid per MWh traded, hour-by-hour): charging when power is cheap and discharging when it is expensive. Its value depends on the daily spread between the highest- and lowest- priced hours.
2. Ancillary services (merchant: paid per MW held in reserve, bought a day ahead): payments for standing ready to keep the grid stable, such as frequency regulation and reserves, whether or not the battery is called on to deliver energy.
3. Capacity contracts (contracted: fixed $/kW-month for many years): a utility or large buyer pays a fixed monthly fee, often called a toll, for the right to use the battery, and takes on the market risk itself. It is the most stable of the three.

Texas batteries earned almost entirely from the first two. California batteries have earned from all three, and as market revenue has fallen, resource adequacy (RA) contracts–the mechanism through which load-serving entities meet the state’s capacity requirements–have become their main source of income. 

That additional revenue stream is crucial in CAISO, making up roughly 70% of battery revenue by 2025, as net market revenue declined from $103/kW-year in 2022 to under $40/kW-year in 2025. Outcomes differ, too: California still curtails large amounts of midday solar despite its growing battery fleet, while ERCOT’s batteries have flattened price peaks. 

How Texas Batteries Lost Their Merchant Revenue

Ancillary services filled up first. In 2022-23, ancillary services supplied more than 80% of Texas batteries’ revenue. But these markets are small and fixed in size. ERCOT buys about 3 GW of the fast reserves batteries excel at, and 6–8 GW of ancillary services in total. 

The battery fleet passed the first mark in mid-2023 and the second in spring 2024; it is now about 19 GW of registered capacity. Average ancillary services prices fell 85-95% between 2023 and 2025, and ERCOT's day-ahead ancillary services spending fell from about $1.5 billion to $174 million. California’s ancillary markets saturated even earlier: batteries already won about 70% of its regulation-up procurement in 2022, and about 90% since 2024. Its ancillary services spending, much smaller than ERCOT’s to begin with, fell by about 60%, from $237 million in 2022 to $95 million in 2025.

Price spikes disappeared, too. ERCOT's average daily spread between its four highest- and four lowest-priced hours fell from $163/MWh in 2023 to $48/MWh in 2025, even as natural gas prices rose. Yet the typical day, adjusted for gas prices, looks much as it did in 2022. What vanished is scarcity: hours when real-time prices topped $1,000/MWh fell from about 60 in 2023 to 2 in 2025. 

Batteries discharge into exactly those hours, and there are now enough of them to flatten the peaks. Controlling for gas prices, renewables, and demand, each additional gigawatt of batteries is associated with a 9% narrower daily spread. 

Merchant revenue collapsed. Texas batteries earned about $193 per kW-year in 2023 and only about $29 in 2025, according to Modo Energy’s fleet index. California batteries’ market revenue fell from about $80 to under $40 per kW-year, before resource adequacy payments. Merchant revenue alone is no longer enough to finance a new battery in either market. 

Costs fell too, but not by as much: NLR’s benchmark installed cost for a four-hour system fell about a third from its 2022 peak to 2024, from roughly $517 to $334 per kWh in 2024 dollars. Costs then turned up again: Lazard’s levelized cost of storage rose in 2026, to near the top of its range since 2020, as tariffs cut access to cheaper Chinese cells and new foreign-entity-of-concern (FEOC) rules pushed buyers toward other suppliers.

‍

The Battery Business Has Moved From Energy to Capacity

As merchant revenue declines, storage developers are increasingly turning to capacity contracts to support revenue requirements. This has an important advantage: contracted cash flow can carry more debt at a lower cost. Under illustrative financing assumptions, moving a battery coming online in 2028 from merchant to a 15-year toll lowers its cost of capital from about 12.6% to 7.2%. 

Using NLR’s projected installed capital costs for 2028, that cuts the annual revenue a two-hour system needs to break even by about 22%, from $105 to $82 per kW-year ($64-101 across NLR’s low- and high- cost cases). However, that is still nearly three times what Texas batteries earned in the market in 2025. Given this gap, new BESS installations might appear to be uneconomic.

‍

Why Are Developers Still Building BESS? 

However, new BESS installations continue to be built. This is partly because many battery storage projects coming online now were financed on 2022-23 revenue (the average project can take over four years from entering the interconnection queue to operation), and partly because new contracts, such as tolls, are priced on capacity, not energy: the supplier is obligated to keep certain amounts of capacity available over a given time period, independent of energy delivery. 

In Texas (which has no capacity market), such tolls are, in effect, private capacity contracts. In California, resource adequacy already plays that role: batteries’ RA contracts averaged about $9 per kW-month in early 2025. 

An owner needs roughly $7 per kW-month for a two-hour battery (about $11 for four hours) to break even, in line with the roughly $7 per kW-month cap on a 20-year battery storage agreement Austin Energy approved in May 2026. By contrast, the energy and ancillary revenue an offtaker could capture by dispatching it is about $2.50–3.50 per kW-month. The difference is what buyers pay for firm capacity: insurance against scarcity events like Winter Storm Uri, capacity to meet their own load obligations, or a hedge against scarcity returning as demand grows. 

The Next Battery Wave Depends on Who Pays for Reliability

Batteries have done what they were built to do: absorb volatility and steady the grid. In doing so, they competed away the revenue that financed the first wave. At today’s prices, merchant revenue no longer covers a new battery. The build continues on projects financed in 2022–23 and on capacity contracts worth nearly three times what Texas batteries earn in the market. Whether the buildout continues depends on how many buyers will pay that premium. 

Deployment continues to grow in both markets. ERCOT registered about 4.2 GW of new batteries in January to July 2026, against 2.8 GW a year earlier. But the pace is slowing: Texas’s fleet grew about 50% in 2025 after nearly doubling in 2024, and interconnection queues are thinning: ERCOT battery applications fell by half in the second half of 2025. 

Wholesale markets financed the first boom. The next wave depends on buyers who value reliability, including utilities, retailers, and large loads such as data centers, and on how cheaply that risk can be priced and capitalized. Data centers in particular put a high value on firm power: research inferring their willingness to pay from cloud computing prices puts it at up to $13,600 per MWh for frontier AI chips on guaranteed service, far above ERCOT’s $5,000 per MWh price cap. 

For developers and investors, the question is no longer how much the market will pay, but who will sign, and for how long. For more on how interconnection queues shape how many storage projects get built and how queue wait times vary by technology, see Relae’s interconnection research.

Relae’s Power & Energy advisory team helps developers, investors, and large energy buyers understand and price the value of storage. This includes behind-the-meter modeling, resource portfolio optimization, and interconnection queue analysis. 

Frequently Asked Questions

Why are battery storage revenues falling if deployment keeps growing? 

Batteries have become victims of their own success. They earned most of their early revenue from two sources: ancillary services and high wholesale market prices. Texas now has more batteries than ancillary services the grid needs, so those prices have collapsed. And because batteries discharge during the most expensive hours, there are now enough of them to flatten the price spikes that used to pay the most.

What is a battery tolling agreement? 

A toll is a long-term contract in which a utility or large buyer pays a battery owner a fixed monthly fee for the right to direct how the battery is used. The buyer keeps whatever the battery earns in the market and takes on the risk if that falls short. The owner gets a steady income it can borrow against.

Who is signing these contracts? 

Mostly utilities, retailers and power traders, who need capacity to serve their customers or want protection against price spikes. In California, utilities sign capacity contracts with batteries to meet the state’s resource adequacy requirements. Large loads such as data centers place a high value on reliable power, but few have signed contracts with grid batteries so far; most of their battery deals are for on-site power.

Will battery storage buildout slow down? 

Not yet. Texas is still adding more batteries each year, largely projects that were planned and financed when revenues were higher. But the pipeline behind them is thinning: new battery applications in ERCOT’s interconnection queue halved in the second half of 2025. Whether buildout holds up depends on how many buyers will pay for capacity at a premium to what the market pays today.

Key Takeaways

• Accurate load forecasting is needed to distinguish and prioritize real demand, align capital deployment, and reduce delays in bringing new power capacity online. Traditional load forecasting was built for predictable, gradual demand growth, not for the scale, uncertainty, and dynamic behavior of data centers. 
• The system-level fix to data-center load forecasting requires probabilistic, more frequent, category-specific methods paired with mandatory data standards and policy alignment. Together, these give planners visibility into the range of possible futures and the likelihood of each. 
• Without that fix, today's forecasts conflate real demand with speculative submissions, reducing accuracy. Inaccurate forecasting in either direction is expensive: underbuild adds friction to economic development; overbuild risks raising retail rates. Both can erode public trust in planning.
• Behind-the-meter generation (BTM) and load flexibility can help achieve speed-to-power in the near term. Just 1% data-center flexibility could unlock 100 GW—more than the entire US nuclear fleet.

Load Growth Is Increasing, Uncertain, and Concentrated

For two decades, US electricity demand was flat. Utilities, transmission planners, and corporate buyers built their planning models around that reality. Then AI workloads changed it.

AI load growth is large, uncertain, and concentrated in major power markets. While load forecasting projections vary across studies, the trajectory is clear: electricity demand is scaling faster than the bulk power grid was designed to handle. Accurate load forecasting is needed to distinguish and prioritize real demand, align capital deployment, and reduce delays in bringing new power capacity online.

On April 30, 2026, Relae (formerly Carbon Direct) hosted a Trellis Group panel on load forecasting in the AI era. Panelists included Derya Eryilmaz, PhD, Vice President of Power Commercialization at Relae; John Miller, Director of Transmission Policy at the Corporate Energy Buyers Association (CEBA); Daniel Padilla, Strategy and Business Development Lead at Emerald AI; and Sam Hodas, Head of US Government Affairs at National Grid. Jake Mitchell, Director of Climate Tech Innovation at Trellis Group, moderated.

The conversation explored where load forecasts fail, what they cost, how to fix them, and near-term solutions to overcome grid constraints. Here is what the panel found.

What Is Load Forecasting?

Load forecasting is the practice of predicting how much electricity will be consumed across a region, at what times, and under what conditions. It informs the major capital and procurement decisions on the grid: where to build transmission, how much generation to procure, what capacity to bid into wholesale markets, and how corporate buyers secure clean, firm power.

Long-term forecasts inform multi-year decisions about transmission and generation. Short-term operational forecasts inform real-time grid operations and trading. The two often sit in separate workflows, but short-term operational forecasts should feed into long-term system planning to improve accuracy as demand patterns shift.

The Bulk Power Grid Is Under Strain

Large power users face constraints on clean, firm power, transmission capacity, multi-year interconnection queues, and aging infrastructure. The strain is most acute in PJM Interconnection (PJM) and the Electric Reliability Council of Texas (ERCOT), the two US markets expected to see the most significant load growth. Each constraint raises the cost of getting load forecasts wrong.

Hodas from National Grid describes the operational reality on the utility side: aging infrastructure inherited from a different demand era. “We’ve got transmission lines that are 70 to 100 years old in New York and Massachusetts, some of the oldest in the country, still in operation.” Replacing or upgrading that infrastructure requires investment, and ratepayers are already pressed. 

Why Today’s Load Forecasts Fail

Traditional load forecasting was built for predictable, gradual demand growth, not for the scale, uncertainty, and dynamic behavior of data centers. 

Most utilities and Independent System Operators (ISOs) produce load forecasts on annual or biannual cycles. They aggregate submissions from individual customers, run that data through a deterministic single-peak load estimate against a single capacity scenario, and pass the consolidated forecast up to regional planners. Regional Transmission Organizations (RTOs) roll those bottom-up utility forecasts into a regional view. 

This worked when demand was flat and predictable. It no longer works with nonlinear growth driven by data centers. Eryilmaz from Relae identifies key structural limitations. 

Four Structural Limitations to Traditional Forecasting Methods

• Over-stating and double-counting. Data centers bid into multiple regions while shopping for power, inflating regional forecasts and blurring the line between real and hypothetical demand—the speculative-load problem.
• Deterministic models (vs probabilistic models). Most planning runs a single peak load estimate against a single capacity scenario, missing the geographic concentration and uncertainty inherent in integrating large loads into the system.
• Aggregated submissions. Utilities report large loads as a single block of gigawatts, with no resolution into workload type, ramp schedule, or operational shape. Planners reverse-engineer peak-demand assumptions rather than measure them.
• Infrequent cadence. Annual or biannual forecasts cannot catch an 80% queue reduction or a multi-gigawatt addition between cycles.

The Speculative-Load Problem

The core challenge in load forecasting is distinguishing real versus hypothetical load. While data center electricity demand is projected to grow by 13-27% annually through 2028, the majority of the projects in the data center queue may not materialize, inflating regional load forecasts.

American Electric Power's Ohio utility (AEP Ohio) introduced a tariff requiring data centers to put up firm financial commitments before getting in line for grid connection. Its interconnection queue dropped from 30 gigawatts to 5.6 gigawatts. More than 80% of the submitted load was speculative: projects that disappeared once commitment became required.

ERCOT shows the same overstatement problem on a larger scale. Roughly 225 gigawatts of data center demand sits in the ERCOT queue against a historic system peak of 85 gigawatts. Texas Senate Bill 6 introduced similar financial obligations for new loads, but those rules apply only to interconnections after 2025, and the cleanup of speculative demand has not yet materialized.

The speculative-load problem shows up in interconnection times. An average new project in PJM can wait 4 to 5 years to become operational. Some of that delay is a real backlog. The rest comes from the inability to distinguish real submissions from speculative ones.

As Eryilmaz puts it, “Load forecasting is actually the center of all of these problems. It is a tool to help planners make the right investment decisions.”

The Cost of Inaccurate Forecasting

As Miller from CEBA notes, “A single misforecasted project can swing a transmission plan by hundreds of megawatts.” Significant inaccuracies can erode public trust in the planning process in two main ways. Underbuilding adds friction to economic development and can limit corporate access to clean power markets. Conversely, overbuilding risks raising retail rates if capacity remains underutilized. 

The goal is to achieve right-sized infrastructure investment. When planning aligns with actual large load growth, it can be net beneficial to retail rates. By spreading fixed costs across more usage, significant new demand can put downward pressure on the rates via the “denominator effect.” 

On the other hand, forecasting variability can distort capacity procurement and interconnection queue prioritization. When load forecasts spike upward, grid operators like PJM have to scramble to buy additional electricity capacity on short notice. These emergency procurements lock in major dollar commitments on the basis of unstable forecast numbers. 

PJM, Midcontinent Independent System Operator (MISO), and Southwest Power Pool (SPP) have also reshaped their interconnection queues to make room for new large loads, but those queue priorities depend on the same forecasts that are unreliable in the first place. 

“There is no substitute for good backbone regional transmission planning,” Miller says. “Full stop. That is the enabler of all of the load growth that we’re talking about.”

BTM Generation and Load Flexibility: A Near-Term Bridge

Hyperscalers’ need for power is way faster than that of utilities and RTOs. Generation alone cannot scale fast enough to meet this new demand, and hyperscalers need speed-to-power.

As Eryilmaz frames it, behind-the-meter generation and load flexibility are interim solutions to the timing mismatch between data center urgency and the grid's slower build cycles. BTM generation and flexibility work differently:

• BTM is power generated on the data center's side of the utility meter, bypassing grid interconnection entirely. The structure gives operators large, reliable blocks of power without waiting years for grid approval.
• Load flexibility is the demand-side approach. A data center modifies its grid draw in response to grid signals. In practice, that can mean curtailing compute workloads during stress events, pre-cooling facilities ahead of a heat wave, drawing from on-site batteries or generators, or shifting workloads to data centers in less-constrained regions.

The Value of Load Flexibility

Relae’s power system modeling quantifies the dollar value of load flexibility in ERCOT. Load flexibility can eliminate forced load shedding risk, even at 40 gigawatts of data center buildout, preventing $5.5 billion in annual consumer welfare losses by curtailing an average of 5% of demand for under 1% of operating hours.

Padilla from Emerald AI reinforces the scale and value of load flexibility: “With just 1% flexibility, we can unlock 100 gigawatts of data centers across the US. That’s more than the entire US nuclear fleet.”

Silicon Valley Power, a municipal utility, is the first US utility to tie flexibility to interconnection speed: flexible data centers get connected faster. NVIDIA, EPRI, Digital Realty, and PJM are partnering on the Aurora AI Factory, the first purpose-built reference design for flexible AI data centers. 

But standardized policy for load flexibility is lagging. Padilla highlights this challenge: “Today, if a data center wants to be flexible, they have nowhere to point. We need standardized tariffs, interconnection rules, and product definitions for large loads that reward them with upsizing interconnection in response to flexibility.” 

Flexibility Takes Many Forms, but it Isn't Universal

Flexibility means accepting brief, predictable downtime, and some workloads can't tolerate it. Hospital systems and mission-critical enterprise applications need 99.999% uptime, the "five nines" standard. As Padilla puts it: "99.9% uptime, with brief and predictable curtailments, is plenty" for most AI workloads. That distinction determines which data centers can participate in flexibility programs.

Miller points out that compute-level flexibility is not always feasible. BTM batteries and virtual power plants (VPPs) are among the alternatives that can offset what data centers withdraw when the grid is stressed, even at facilities whose compute workloads cannot pause directly.

Better Load Forecasting: The Longer-Term Fix

While BTM generation and load flexibility can help address near-term speed-to-power, the longer-term fix is improving load forecasting methods and the standardization of data provided by the data centers themselves.

Eryilmaz outlines three technical shifts for better load forecasting:

• Embed short-term operational forecasting into long-term planning. Short-term spikes, weather risk, and reserve considerations carry direct implications for multi-year capital decisions. The line between operations and planning breaks down when growth is nonlinear.
• Replace deterministic models with probabilistic methods. Risk metrics like loss of load hours (expected hours per year that demand exceeds supply) and expected unserved energy (total expected energy shortfall) measure both how much capacity the system has and the conditions under which it might fall short. The North American Electric Reliability Corporation (NERC) has suggested both metrics as part of its reliability framework.
• Forecast load by category. Treating all data center load as a single block hides the differences in load profiles, operational schedules, and ramp-up timing that drive system planning.

Policy Alignment

Technical forecasting improvements only scale with policy alignment, and Miller proposes a two-part fix:

On the top-down side, RTOs need authority to take an independent view of utility-submitted forecasts. They should require milestones, such as firm financial commitments and secured financing, before counting a submitted load against the regional forecast. 

On the bottom-up side, state regulators set the rules that govern how individual utilities prepare their forecasts. Large load tariffs play a big role in how utility-level forecasts come together. Federal and state authorities need to row in the same direction. Hodas frames the same alignment from the utility side: “Grid investment unlocks economic growth, but for us to make those investments, we need regulatory certainty.”

Standardizing Large-Load Data

The Federal Energy Regulatory Commission (FERC) has since moved: in June 2026 it issued show cause orders directing six RTOs and ISOs—CAISO, ISO-NE, MISO, NYISO, PJM, and SPP—to revise or justify their large-load interconnection rules, and in July 2026 it directed NERC to develop computational-load reliability standards and registration criteria by the end of the year. Both are useful first steps. But as Eryilmaz argues, voluntary disclosure has not closed the gap.

The industry cannot meaningfully compare ISO forecasts when each utility submits load data in different shapes (e.g., using different methods and data standards) on different schedules. Mandatory submission requirements and published methodologies, applied consistently across utilities, ISOs, and state regulators, are the only path to forecasts whose components are actually comparable.

Getting Load Forecasting Right Starts Now

The system-level fix to improving forecasting is through probabilistic, category-specific methods paired with data standardization and policy support. Together, these account for the scale, uncertainty, and dynamic behavior of data center loads, and give planners visibility into the range of possible futures and the likelihood of each.  

All forecasts will be wrong to some degree, but as Miller puts it, “It's ultimately not about having a perfect prediction. It's about baking in methods to account for uncertainty.” These system-level improvements won't eliminate errors entirely, but they will minimize them, leading to more confident investment decisions and a grid better prepared for what's ahead.

Frequently Asked Questions

What is load forecasting, and why is it harder now with AI data center loads?

Load forecasting predicts how much electricity a region will consume, when, and under what conditions—the basis for where to build transmission, how much generation to procure, and how corporate buyers secure clean, firm power. It was designed for two decades of flat, gradual demand growth. Data center load is none of those things: it is large, geographically concentrated, arrives in gigawatt blocks with no disclosed operating shape, and can be withdrawn as quickly as it appeared.

What is speculative load in an interconnection queue, and how do planners tell it apart from real demand?

Speculative load is capacity requested by projects that may never be built — often the same data center bidding into several regions at once while shopping for power, which counts the same gigawatts more than once. The tested filter is a financial commitment: when AEP Ohio required firm commitments before queue entry, the utility's reported data center pipeline fell from about 30 GW to roughly 5.7 GW. Milestone requirements, independent RTO review of utility submissions, and mandatory data standards are the tools planners have.

Is load flexibility proven and scalable today, or still emerging?

The modeling case for load flexibility is strong; the commercial case is still early. Duke's Nicholas Institute found the 22 largest US balancing authority areas could absorb roughly 76–126 GW of new load if it accepts modest curtailment, and Relae's ERCOT modeling shows demand response eliminating forced load shedding risk at 40 GW of data center buildout, avoiding $5.5 billion in annual consumer welfare losses. What is still missing is the market plumbing—standardized tariffs, interconnection rules, and product definitions—so a data center willing to be flexible has somewhere to sign up.

What should a company look for when evaluating a region's load forecast?

Ask whether the forecast is probabilistic or a single deterministic peak, how often it is refreshed, and whether large loads are broken out by category and operating shape rather than reported as one block of gigawatts. Then ask what milestone or financial commitment a project must clear before its megawatts count toward the forecast. A forecast that cannot answer those three questions cannot tell you how much of the queue ahead of you is real.

[cta]

Key Takeaways

• Power demand is outrunning buildout. Meeting large load growth requires more than new generation; it requires faster interconnection and congestion relief on existing transmission lines. 
• Dynamic line rating (DLR) is available today, deploys in months, and enables faster speed-to-power. On the right thermally congested lines, DLR can unlock more capacity at a fraction of new infrastructure cost. In one utility demonstration, 5% to 10% of additional capacity was enough to clear most of the congestion on the lines studied.
• DLR has been held back by weak incentives, but that is changing. Utilities earn a regulated return on capital they invest in new assets, which favors building new infrastructure over lower-cost solutions like DLR. Load growth and new Federal Energy Regulatory Commission (FERC) mandates are starting to shift the calculus.

The Grid Cannot Expand Fast Enough for AI Demand, But It Can Carry More

Power demand is booming as data centers scale across the US grid, and current grid infrastructure cannot supply it. This constraint is physical, not financial. Meeting this demand requires a significant amount of power generation and infrastructure upgrades. More than 2 terawatts of generation and storage sit in interconnection queues, roughly 1.5x the total installed generation capacity in the US. 

Regional markets are working to accelerate generation buildouts, but connecting that generation to the transmission network remains expensive and slow to match speed-to-power needs. New high-voltage lines take years to permit, cost between $2 million and $6 million per mile to build, and major projects routinely take five to ten years from identification to energization. For example, PJM Interconnection LLC (PJM) identified the Doubs–Goose Creek 500 kilovolt (kV) corridor as a bottleneck feeding Data Center Alley in 2023 and set June 2027 as the date a fix was needed. Dominion Energy's published schedule for its portion of that rebuild anticipates a completion date of 2031.

A number of studies^1,2 show there is headroom in the bulk transmission system. Grid-enhancing technologies, such as dynamic line rating (DLR), can convert part of that headroom into capacity today while new generation and transmission are being built. DLR lets suitable transmission lines increase their carrying capacity in real time, unlocking that headroom at a fraction of the cost of a buildout. Realizing that value is a targeting exercise with a key question: On which thermally limited lines can DLR actually relieve congestion? 

What Is Dynamic Line Rating?

Dynamic line rating is a method for calculating a transmission line's real-time carrying capacity using live weather and conductor-temperature data. It lets grid operators safely carry more power whenever weather conditions allow.

Most transmission lines operate under a static rating: a fixed, conservative limit on current, set for worst-case weather and held all year. The limit is based on temperature, because pushing too much current can overheat the conductor wire. Metal conductors expand as they heat, which can make them sag and touch trees or other obstacles, causing short circuits or fires. Real conditions almost always cool a conductor better than the worst-case assumption a static rating is built on. That means the line can carry more current while staying at the same maximum conductor temperature, and therefore within the same sag and clearance envelope. That headroom is exactly what DLR captures: instead of leaving it on the table, DLR recalculates the line's rating in real time so operators can use the extra capacity safely.

Beyond a static rating is the ambient-adjusted rating (AAR), which many utilities have begun adopting. An AAR recalculates the rating from forecast ambient air temperature, typically hourly and out to several days. DLR goes further, adding wind speed and direction, solar heating, and in some deployments the conductor's measured temperature.

DLR technologies rest on a heat-balance algorithm: how fast a line heats up (from electric current and sunshine) versus how fast it cools off (from wind and cold air). The calculations are standardized in IEEE 738 in North America and CIGRE 601 internationally. The data feeding those calculations can come from line-mounted sensors, weather models, or both, depending on a tradeoff between per-span accuracy and the cost of installing sensors along every span.

Even so, DLR remains limited in the US, and AAR has been slow to arrive. FERC's Order 881 required the transmission providers it regulates to adopt AAR by July 2025, but FERC has granted numerous extensions. PJM became the first to fully implement AAR in March 2026, while Midcontinent Independent System Operator (MISO) and New York Independent System Operator (NYISO) are not expected until 2028.

The Near-Term Value of DLR: Reducing Grid Congestion

DLR's value is immediate. It can be installed in months, not years, so a currently congested line can start carrying more power the moment conditions allow, reducing congestion right away. When cheaper generation is available upstream of that line, DLR cuts costs directly, because grid operators no longer need to dispatch pricier generation downstream of the congestion to supply load. That means DLR can reduce congestion costs in the current delivery year, compared to a transmission line rebuild that sits in a decade-long queue. 

Over a longer horizon, utility planners can build that headroom into long-term capacity models. This is important, because current capacity-expansion and integrated resource plan (IRP) models still run on static or seasonal ratings, and typically leave out the potential gains from grid-enhancing technologies like DLR. 

NERC's large loads white paper and FERC's RM26-4 rulemaking both raise the issue of how utilities can absorb multi-hundred-megawatt data center requests without a decade-long transmission build. Solutions like DLR are one of the few tools that can compress that timeline. 

The hardware itself is cheap: sensors and data management cost a small fraction of any physical upgrade. That means the economics comes down to identifying the lines that benefit most from DLR. This is particularly important because on most US grids, congestion concentrates on a small number of lines that repeatedly reach their limits. On those lines, DLR can cut congestion costs directly and defer costlier upgrades, while its potential on other lines may be far lower. As a result, identifying those high-potential, thermally congested lines is essential.

Proven DLR Examples in the Industry 

Real deployments show DLR can reduce a meaningful share of transmission congestion costs, with extra carrying capacity above the static rating running roughly 5% to 30%, depending on how often that capacity is available. In Oncor's ERCOT demonstration, 5% of additional capacity would have relieved up to 60% of congestion on the target lines, and 10% would have practically eliminated it. PPL Electric in Pennsylvania/PJM reports annual customer savings of $23 million after deploying DLR across its initial three lines. The DLR installation cost about $250,000, against a rebuild alternative that would have cost about $50 million and taken far longer. 

The contrast abroad is instructive. Austria's grid operator, APG, recorded about $13 million a year in congestion savings across roughly 15% of its network. While these savings are real, it's important to recognize that these results come from single, well-chosen, badly congested lines. 

The UK's National Grid began with a two-year DLR trial on a single 275 kV circuit in 2022, expanded to more than 275 kilometers of its network by 2025, with estimated consumer savings of about $26 million a year. In April 2026, National Grid signed a five-year contract covering 585 kilometers more, with most installations due by 2028 and potential savings of up to $66 million. Each expansion followed measured results from the stage before it.

Where the Headroom Is: Screening PJM's Data Center Alley

To illustrate the congestion savings from DLR, Relae screened PJM's five-minute real-time market record for every binding transmission constraint in 2025. For each one, we captured the shadow price, the marginal value of relaxing that constraint.³

Our analysis focused on thermal constraints, and then identified lines that bind frequently, in conditions milder than the worst case their static rating was set for, which is when a conductor's true rating sits above its static assumption. For the lines that we identified, congestion costs were added over the binding hours to set a bound on the savings that could result from DLR. That full amount would not necessarily be realized in practice, because the shadow price values only the next megawatt freed, and relieving one line can shift the constraint to the next. However, it serves as a useful estimate for the scale of savings that could be achieved.

We ran the analysis on the Dominion (DOM) zone in PJM, home to Data Center Alley in Loudoun County, Virginia. Figure 2 shows a high-level section of the grid. The 500 kV bulk grid steps down through transformers to the 230 kV substations feeding the data centers, with the lines that experience recurring congestion highlighted. A handful of those 230 kV lines showed up as binding thermal constraints again and again. 

The congestion in DOM isn't constant, and it concentrates in particular months and within the day in particular hours. Figure 3 shows three transmission lines within the DOM zone and the number of hours each was thermally congested in each hour-of-day slot over 2025. Binding concentrates in the warm months and, within the day, from late morning through early evening. 

At first glance, this period looks like the wrong window for DLR. The local weather record says otherwise. These periods turn out to be some of the windiest hours of the day, not the stillest. Median wind speed at Dulles ran about 3.5 m/s, above the 0.6 m/s crossflow a static rating conventionally assumes, with fewer than 5% of observations falling below that threshold. Median ambient temperature in those hours was about 26°C, against the 35–40°C a static summer rating is typically built for. Across all three lines, the large majority of congested hours coincided with weather that would have supported a materially higher rating. 

Valuing just one megawatt of DLR relief at each five-minute shadow price, the estimated savings are worth roughly $300,000 in this three-line example across about 263 line-hours.⁴

Because the value concentrates on a handful of thermally limited, heavily congested lines, and because the operational case has to be made line by line, capturing the opportunity is fundamentally an analytics problem: find the right lines, and prove the savings.

What Does It Take to Scale DLR?

DLR is cheap and effective, but two things stand between it and broader adoption: incentives and advanced grid analytics.

The utility cost-of-service model recovers investment in generation and transmission assets and earns its profit as a regulated return on the capital deployed. Because rates recover capital rather than power delivered, utilities have a stronger incentive to build or upgrade lines than to move more power across the ones they already own. That bias toward capital investment over optimization is why a mature technology has stayed niche in the US for years. Regulators have started to look more closely at this, but the main federal rule still mandates the milder AAR, not DLR, and leaves the return model untouched.

Contingency analysis compounds the problem. Current models are built around fixed line limits. A rating that changes hour to hour adds real modeling work, and more importantly, the system still has to hold under worst-case contingencies. So while operators already forecast weather daily for wind and solar, the harder step is trusting a forecast enough to commit a transmission limit against it. That takes significant predictive analytics built into system planning, not bolted on after.⁵

How Policy Is Starting to Shift the Calculus

Policy is starting to move the incentive problem. FERC's Order 881 made AAR the minimum for the transmission providers it regulates (effective July 2025, with several operators on extended timelines) and required markets to be capable of accepting dynamic ratings. PJM has started to implement this: PPL Electric has run sensor-based DLR on nine congested lines since 2022, feeding PJM's day-ahead markets. 

Order 1920, FERC's first long-term transmission-planning overhaul in more than a decade, now requires planners to formally evaluate grid-enhancing technologies like DLR against conventional builds. It stops short of mandating deployment, but it forces a comparison utilities used to skip. That comparison is now written into filed tariff processes (PJM filed its plan in December 2025). Those first cycles only began in 2026, and the order allows up to three years to reach a selection, so the results are still pending. 

A shared-savings incentive, letting a utility keep a slice of the congestion savings it creates, has been proposed to FERC and championed in the Advancing GETs Act, but it isn't yet a rule, so the core misalignment stands. DOE's GRIP program has funded grid-enhancing deployments, and by early 2026, 16 states had some form of advanced transmission technology requirement, with Colorado adding its Grid Optimization Act in April 2026.

The newest pressure is coming from the demand side. Through 2025–26, FERC began overhauling how large loads connect to the grid, and while none of it touches the utility's return on capital, it changes who sees the costs. FERC issued show-cause orders directing all six grid operators to justify or reform their large-load rules. This tees up consideration of alternative transmission technologies in study processes and greater transparency into costs. 

And the rules are moving toward making the large load pay for the upgrades its connection requires. Pennsylvania's model large-load tariff, for example, recommends utilities charge data centers for the upgrades their interconnection makes necessary. It also instructs utilities to let those customers self-construct certain upgrades, including some affecting the wider grid. That combination is what matters. The party paying the bill now has a reason to ask whether a cheaper fix exists and, in at least one state, a route to build it. We have not yet seen a DLR deployment selected this way, because these frameworks are only months old, but the cost gap between a DLR fix and a rebuild is becoming visible to the party who pays the difference.

How Relae Helps Find the Value of DLR  

Through our Power, Data, and Innovation practice, Relae combines transmission congestion data, line-level thermal constraints, and short-term weather forecasts into a single view of where dynamic ratings would actually pay. The output is a short list of candidate lines, each with a modeled capacity uplift and an estimated dollar value, turning a vague “DLR is promising” into a priced, line-by-line decision. It is the transmission-side complement to our work on the interconnection queue and demand-side flexibility. All three are ways of closing the gap between demand and delivered capacity faster than new construction allows.

Key Takeaways

• AI's carbon footprint has two distinct parts: embodied emissions from building data centers and operational emissions from running them, both accelerating as global data center electricity use is set to double by 2030, and AI-focused use to triple.
• Managing that footprint will require deliberately steering technology architecture, power sourcing, and materials choices, instead of leaving them to react to demand after the fact.
• Eight concrete strategies, from smarter chip design to firm clean power and carbon removal, can cut AI's footprint today, without waiting on new regulation.
• US data centers used 4% of the USA's total electricity in 2024, and are projected to use as much as 15% by 2030.

Introduction

The rapid growth of artificial intelligence (AI), particularly large-language models (LLM) and generative AI, has taken many by surprise. This surge has led to escalating electricity demands at data centers and raised concerns about the strain on the power grid. It has also sparked the construction of new, larger data centers, resulting in growing embodied emissions tied to building and maintaining AI physical infrastructure.

Managing the risks of increased greenhouse gas (GHG) emissions from AI requires investment, expertise, and new approaches to building and operating many aspects of AI operation and supply chains. The immediate task is to understand these risks, gather the necessary information, and to avoid poor outcomes by proactively managing construction, operation, and emissions associated with the growth in AI. In parallel to that work, it's important to recognize that AI can itself be a real force to reduce emissions incrementally and dramatically across a wide range of sectors.

What Is the Carbon Footprint of AI?

The carbon footprint of AI consists of two main parts: "embodied" emissions that come from manufacturing IT equipment and constructing data centers, and "operational" emissions that come from electricity consumed by servers, memory and networking equipment as they perform AI-related calculations. Both of these aspects of emissions are growing as more data centers are built and existing data centers increase their share of power-hungry AI applications like generative LLM searches, AI agents, and AI image generation.

Understanding Electricity Demand for Data Centers

Today, the electricity demand from AI-specific applications is estimated to be less than 1% of global electricity use. To understand this number, it helps to start with the electricity consumed by the 12,000+ data centers worldwide, which was about 1.5% of global electricity consumption in 2024. (This excludes another 0.4% from cryptocurrency mining.) However, most of the computation at these data centers is not AI; instead, it's more conventional applications like e-commerce, video streaming, social media, and online gaming.

The amount of AI-based computation at data centers is hard to determine, but AI-dedicated accelerated servers consumed about one third of overall data center electricity in 2025, or roughly 0.5% of global electricity. Notably, this is projected to grow at 30% annually, much faster than conventional (non-AI) data center electricity use. However, that electricity use results in a relatively small share of greenhouse gas emissions: about 0.5% of global fuel combustion emissions, with AI data centers representing only a small portion of that value.

Still, the demand for AI applications is rapidly growing, and this is likely to drive up the electricity used by data centers and the associated greenhouse gas emissions. The most important implications of this trend are in the US, which hosts about half the world's data centers. Currently, data centers use about 4% of US electricity, but projections for the future range from a low of 9.5% to a high of 15.3% in 2030.

[cta]

How Electricity Sources Impact Data Center Emissions

A large increase in electricity use doesn't necessarily result in a similarly large increase in greenhouse gas emissions. Currently, a significant portion of the electricity powering data centers comes from zero-carbon sources such as wind and solar. This is partly because of large, corporate power-purchase agreements (PPAs) signed by leading data center operators, particularly Amazon, Meta and Google. 

US technology companies have been buying renewable energy for years. Global corporate clean energy procurement hit a record 62 GW in 2024, then fell to 55.9 GW in 2025, the first annual decline in nearly a decade, as elevated power prices and policy uncertainty made even large buyers more selective. Meta, Amazon, Google, and Microsoft still accounted for roughly 49% of global clean energy procurement in 2025, with Meta and Amazon alone securing 20.4 GW combined, including 4.7 GW of nuclear power.

The use of low-carbon power means that the net emissions from these data centers is smaller than the electricity consumption numbers might suggest. Of course, a crucial consideration is whether this low-carbon power is truly "additional," meaning that it is being added to the grid and not simply taken away from other uses. Data center operators are also expanding beyond their traditional wind and solar PPAs by exploring novel approaches to try to meet this standard, including geothermal projects in the US and Taiwan.

However, the projected electricity demand from AI applications at data centers will be difficult to meet entirely with low-carbon power, at least in the near term. Despite installing over 43.2 GW of wind, solar and battery projects in the US in 2025, these generators face a long wait for interconnection approval in many parts of the country. Geothermal and hydro power, which offer steady ("baseload") low-carbon electricity, remain constrained in the near term. And the interest in scaling up nuclear power, from restarting full-scale reactors to novel small modular reactors (SMRs), faces significant regulatory, cost, and supply chain hurdles.

One important source of low-carbon electricity that has not received enough attention is natural gas-fired power equipped with carbon capture and storage (CCS). This technology has the potential to significantly reduce emissions from existing power plants and enable new projects to achieve near-zero emissions.

The Role of Embodied Emissions in Data Center Construction

Embodied emissions include all emissions associated with the extraction, production, transportation, construction, and disposal of materials used in construction.

The embodied emissions from constructing data centers are substantial, and include concrete, steel, and IT hardware. Scope 3 GHG emissions for data centers—which include embodied emissions—range from approximately one-third to two-thirds of overall lifetime emissions. At Microsoft, Scope 3 emissions made up about 86% of the company's total FY2025 footprint and grew roughly 12% year over year, with capital goods driving most of that increase. In FY2024, capital goods alone accounted for about 41% of Microsoft's Scope 3 emissions, and purchased goods and services (including IT hardware) accounted for another 34%. In response, Microsoft has started using wood in some data center construction to reduce this impact. While using wood offers a partial solution, it cannot fully offset the emissions of even a single facility, and wood supply chains remain limited.

Major data center operators are working hard to address this challenge, including emphasizing the need for standardized emissions measurements and disclosures for key building materials. Ultimately, achieving deeper decarbonization will require further action to address both operational and embodied emissions.

Eight Strategies to Reduce the Carbon Footprint of AI

1. Adapt Technology Architecture

Efficiency is the foundational strategy in any clean energy approach. As such, chipmakers are developing ways to cut energy use from the outset, such as incorporating more memory directly onto computer chips or hard-wiring basic calculations. These innovations have already reduced energy consumption in new computer chips substantially, in some cases a 96% improvement. Examples of this include NVIDIA's Blackwell platform and the company's newer Rubin platform, launched in 2026, continues that trajectory. Likewise, servers are being designed with new architectures that minimize internal data transfers, delivering additional efficiencies. Even more efficiency gains may be possible with emerging technologies like photonic computing.

2. Optimize Training Geography

There are also significant opportunities to manage AI's energy use through time and space optimization. For example, a large portion of the energy consumption for LLMs occurs during the training phase, prior to a model's deployment for inference. Because these training tasks are not location-dependent, they can be carried out in regions with abundant, low-cost, low-carbon electricity, as part of broader efforts to dynamically move computing tasks to reduce emissions, known as carbon-aware computing. Additionally, server requests for generative AI tasks, like ChatGPT searches, can potentially be routed through systems powered by low-carbon electricity. Although this may add only a few milliseconds of latency, it could substantially reduce emissions from computing operations.

3. Select Appropriately-Sized Models

Not all generative AI tasks, like ChatGPT queries, are equal in terms of energy demand. Leading AI companies are increasingly focusing on using smaller, more efficient AI models to perform these tasks, achieving nearly equivalent quality for far less energy consumption. A notable recent test of that idea came in January 2025, when China's DeepSeek released a model with competitive performance that was trained using less powerful chips and far fewer computing hours than its established rivals. Similarly, many AI applications, such as digital twinning and satellite-based pattern recognition, consume far less electricity than generative LLMs, because of their specialized, relatively efficient models. This can even save energy compared to non-AI approaches: for example, some of the most advanced AI-driven weather prediction models require far less energy than traditional weather simulations, running on a laptop rather than a supercomputer.

4. Address Fugitive Methane Emissions

As data center operators increasingly plan on using natural gas for new electricity supply, reducing upstream emissions from gas production and transmission will be crucial. In the U.S., the Environmental Protection Agency (EPA) 2024 Methane Rule was designed to cut these non-carbon dioxide greenhouse gas emissions by approximately 80%. However, Congress repealed the rule's methane fee in 2025 and barred the EPA from collecting it until 2034. The EPA has since extended compliance deadlines and loosened flare and vent-gas requirements, with litigation over those changes still ongoing. Meanwhile, tools from companies like Kayrros and organizations like Carbon Mapper help detect methane leaks and attribute them to specific operators. The best actors in the industry emit minimal methane, less than 0.5% of what is produced. This standard is achievable for nearly all gas producers.

5. Use Carbon Capture on Power Plants

For both new and existing natural gas-fired power plants, carbon capture and storage technology offers the potential for generating firm, low-carbon power. While many plants currently in operation continue to emit unchecked, this doesn't have to be the case: their emissions can be captured and securely stored geologically. Hyperscalers and project developers should pursue new investments and business models for CCS to reduce existing emissions by 95% or more. For new generation projects, options like NetPower, Arbor, and CES will soon enable emissions abatement of 100%, or even more if combined with biopower to deliver carbon dioxide removal as well. Achieving this will require the development of carbon dioxide pipelines, barges, and storage facilities, which face their own challenges, such as permitting and community approval, that must be addressed directly.

6. Add More Zero-Carbon Power to the Grid

Roughly 8,200 solar, wind, and battery projects in the U.S. are seeking grid interconnection. By the end of 2025, the interconnection queue held roughly 2,060 GW of proposed generation and storage across thousands of projects, and its composition shifted meaningfully. Solar, wind, and storage volumes in the queue all declined year over year (although remained at high absolute levels) while natural gas capacity in the queue grew by 86%. Our blog post, The $5.5 Billion-Dollar Case for Enabling Data Center Load Flexibility, covers one way hyperscalers are working around the wait rather than simply enduring it. These delays need to be addressed, and permitting reform remains an unresolved, live debate. The Manchin-Barrasso bill, which once looked likely to pass, was tabled in December 2024 and never became law. As of 2026, no comprehensive federal permitting law has replaced it. One potential innovation is to use AI to accelerate the development of power flow models and streamline the paperwork required to complete the regulatory process.

7. Invest in Low-Carbon Building Materials

While wood is a promising low-carbon building material, we'll also need glass, concrete, steel, aluminum, and computer chips with minimal embodied carbon emissions. Hyperscalers currently face significant challenges accessing low-carbon versions of these materials, which will eventually be produced using low-carbon hydrogen, carbon capture and storage, and low-carbon electricity. However, these systems require significant investment, workforce development, and permitting to be built. Without these advancements, the embodied emissions from data centers will increase rapidly and significantly in the US, Europe, and globally.

8. Increase Carbon Dioxide Removals

It's already clear that AI applications at data centers will generate emissions from electricity use and embodied carbon that cannot be avoided in the near term. Estimates of current greenhouse gas emissions exceed 300 million tons per year and are likely to grow this decade. These emissions should be measured using full life-cycle analysis and then offset through high-quality carbon removal projects, preferably those with high durability.

To effectively reduce the environmental impact of AI, all eight strategies discussed must prioritize the communities most affected: frontline communities near new infrastructure, consumers facing price increases, and tribal authorities with limited legal protections. Our own research on community opposition to AI data centers found that transparency, not cost or environmental impact alone, is the dominant driver of pushback across 46 stalled or blocked projects. We explore this concern further in our blog, Who Pays for the AI? The Hidden Costs of Rising Data Center Demand, including how ratepayers, not just data center operators, often absorb the cost of new grid infrastructure. Planning should begin by understanding the needs of these communities, ensuring that efforts focus on minimizing harm while maximizing benefits. Equity and justice must be embedded in every stage of planning, production, and permitting across all strategies.

AI's Power Demand Is Indicative of Broader Electricity Demand

AI is just one part of a broader trend of rapidly growing electricity demands, including from electric vehicles, heat pumps, industrial electrification, green hydrogen, and various e-fuels. The challenges AI presents to hyperscalers, communities, regulators, and investors serve as a preview of the complex, far-reaching impacts emerging in other sectors. The same questions keep recurring. Who secures reliable, affordable power fast enough? Who ends up carrying the cost and emissions burden of getting there the wrong way?

Managing AI's power demand will require building the technology architecture, clean firm power supply, and materials strategy to meet that demand deliberately, rather than reactively. AI's carbon footprint underscores the critical need for expertise in clean electricity, grid management, decarbonization, and carbon removal—expertise that will become increasingly vital as more companies realize the complexity and cost of the journey ahead.

Fortunately, AI itself can be part of the solution. With applications in grid management, material science, and advanced manufacturing, AI has the potential to play a powerful role in the climate response.

Read the full 2025 report: ICEF Sustainable Data Centers.

Frequently Asked Questions

How much electricity do AI data centers actually use? 

AI-specific computation likely accounts for around 0.04% of global electricity use today, but data centers overall (most of it non-AI computation) used about 1.5% of global electricity in 2024. In the US, which hosts roughly half the world's data centers, Lawrence Berkeley National Laboratory puts current usage at 4% of US electricity, projected to reach 9.5-15.3% by 2030 as AI-specific demand grows.

Will more efficient AI models like DeepSeek reduce data center energy demand? 

Not necessarily. DeepSeek's 2025 debut showed that competitive models can be trained with less powerful chips and fewer computing hours, but whether that translates into lower total energy demand is contested. Historically, efficiency gains in computing have tended to get absorbed by increased usage rather than reducing total consumption, so the honest answer is that it depends on whether demand growth outpaces the efficiency gained.

What is being done about the embodied emissions from building AI data centers?

‍Embodied emissions, from concrete, steel, and IT hardware, can account for one-third to two-thirds of a data center's lifetime emissions. Strategies include using lower-carbon materials like wood where feasible, developing low-carbon concrete, steel, and chips, and standardizing emissions disclosures for building materials so operators can compare and choose lower-footprint options.