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

Key Takeaways

• On May 4, 2026, the North American Electric Reliability Corporation (NERC) issued a rare Level 3 “Essential Actions” Alert in response to repeated events in which 1,000+ megawatts (MW) of computation load dropped off the bulk power system in seconds, leading to major grid stability issues.
• The pattern has since escalated: on July 22, 2026, a transmission fault in Ashburn, Virginia took more than 3 GW of data center load offline in seconds—roughly 3% of PJM demand at the time.
• NERC also published Reliability Guidelines that push the same concerns into long-term planning, explicitly recommending resource adequacy models that capture firm vs. flexible load, behind-the-meter resources, and AI training operating windows.
• For transmission operators and balancing authorities, the releases compel new scrutiny of how computational loads affect stability and resource adequacy. For hyperscalers and other large loads, those assessments now sit on the critical path: if operators cannot show through advanced modeling that they can integrate the new loads, interconnection and buildout plans stall.
• Meeting the bar takes advanced grid modeling at multiple time and spatial scales, from sub-second stability through long-horizon capacity and resource adequacy, to evaluate the role of large load portfolios considering demand response, storage, and co-located generation.

Why Grid Frequency Matters for Large Loads

When we turn on the lights or charge our phones, it’s easy to forget that electricity travels through the power grid as alternating current. Sixty times a second—far faster than our eyes can see—the flow of electricity alternates back and forth along the wires making up both the transmission and distribution parts of the North American grid. 

Power generation equipment and most large industrial loads are designed to work with this 60 Hertz (Hz) alternating flow and must be synchronized precisely to this rhythm to function. Grid synchronization is so important that it can even have geopolitical implications.

For some electrical equipment, getting out of sync with the grid’s frequency can lead to malfunctions or even physical damage and destruction. That’s why grid-connected equipment is protected by circuits that automatically disconnect from the grid (“trip offline”) if the grid frequency begins to deviate by even one percent. For minor equipment, this is easily managed. However, when large amounts of generation or load trip offline quickly, it can lead to rapidly cascading grid blackouts affecting tens of millions of people with costs in the billions.

Grid operators pay extremely careful attention to factors that could cause grid frequency to deviate. The grid’s frequency stays near 60 Hz only when total power generation and consumption (load) are closely balanced. If load suddenly drops below generation, physical rotating generators like gas turbines can begin to speed up, making grid frequency rise. 

This becomes particularly dangerous when large grid-connected loads all trip offline simultaneously because of minor frequency deviations or other factors. If these loads are large enough, they can trigger a cascading sequence of rising frequency and further equipment and generator trips, potentially causing a complete “grid collapse” blackout. The North American grid may be getting closer to this scenario. 

What Triggered NERC’s Highest-Urgency Alert

Data center load drops are now a documented grid stability threat. On May 4, 2026, NERC issued a rare Level 3 “Essential Actions” Alert—its highest-urgency notification—in response to a pattern of customer-initiated load reductions in which 1,000+ MW of computational load (data centers) dropped off the bulk power system (tripped offline) in seconds. These were “customer-initiated” because protection circuits at data centers detected problems with grid-supplied power and automatically disconnected to protect their sensitive computing equipment from electrical damage. 

Paired with a new Reliability Guideline on emerging large loads, the alert highlights the urgent need to better understand the potential for these events to cause grid instability or even blackouts. Together, these two documents reset the bar for the detailed grid modeling and planning needed for any utility, independent system operator (ISO), or hyperscaler with material data-center growth in its footprint.

How the Alert Reshapes Grid Interconnection

For utilities and ISOs, the alert and guideline raise the standard of evidence required to connect computational loads safely to the grid. Modeling assessments now sit on the critical path for large load interconnection decisions, and the same studies will increasingly inform reserve margin, transmission, and dispatch program designs.

For hyperscalers and other large loads, the consequence is direct. Plans that assume firm service without supporting analysis will face longer queues and tougher interconnection conditions. Buildout timelines now depend on whether utilities and ISOs can show, through stability and resource adequacy modeling, that the system can absorb the load and respond safely to its disturbances.

For storage developers, particularly long-duration and fast-responding assets, these events elevate the reliability value of rapid response and load-shifting resources. The same grid modeling improvements that capture flexible load behavior also surface storage's full reliability contribution.

For flexibility platforms, the same modeling work that satisfies NERC's expectations unlocks faster, cheaper interconnection. Demand response, large-load shifting, and co-located dispatch coordination are now both technical and commercial enablers.

A Higher Bar for Power Analysis

These pressures point to a higher bar for power analysis at multiple time and spatial scales, for utilities and the large loads they serve.

At sub-second to second timescales, electromagnetic transient (EMT) models capture fast electrical switching and the uninterruptible power supply behavior that determines whether a data center stays connected during a disturbance (“rides through”). The alert asks for these models to be more detailed, validated against actual equipment, and shared between large loads, transmission owners, and planners.

At seconds-to-minutes, dynamic stability simulation covers system frequency response, voltage recovery, and oscillation behavior after disturbances. NERC now expects annual stability studies and explicit load drop contingencies in planning files.

At hours-to-years, capacity expansion and production cost modeling determine whether the system has enough resources, in the right places, with the right flexibility, to keep up with computational load growth. NERC’s May 2026 Large Loads Reliability Guideline is most explicit at this scale, calling for resource adequacy studies that represent firm and flexible load components, behind-the-meter resources, AI training operating windows, and probabilistic scenarios across many weather, load, and outage combinations on a network-aware footprint. 

Rising to the Challenge

Since the alert was issued, its expectations have begun hardening into rules. Registered entities were required to report to NERC on their progress against the seven Essential Actions by August 3, 2026, and on July 16, 2026, FERC directed NERC to go further: to develop mandatory reliability standards for computational loads and revise its registration criteria, with the first standards and Rules of Procedure changes due December 31, 2026 and a second-phase work plan due March 1, 2027. NERC's Large Loads Action Plan anticipates new "Computational Load Owner" and "Computational Load Operator" registered entity types alongside the first three computational load standards. 

The practical consequence is that the modeling described above is no longer only good planning practice: utilities, ISOs, hyperscalers, and other large loads should expect the data-sharing, study, and commissioning expectations in the alert to return as auditable requirements, and should build the capability before the compliance deadline rather than after it. 

Frequently Asked Questions

What is a NERC Level 3 Alert, and what does it require?

A Level 3 “Essential Actions” Alert is the most urgent of NERC's three alert levels, reserved for risks that need immediate, documented industry response. The May 4, 2026 alert directed registered entities to take seven essential actions on computational load—covering modeling, system studies, commissioning, protection, fault recording, and direct operational communication with large load operators. Written responses were due to NERC by August 3, 2026.

Why do data centers disconnect from the grid during minor disturbances?

Data centers run voltage- and frequency-sensitive computing equipment protected by automatic transfer systems that switch to on-site UPS or backup generation the moment grid power looks abnormal. Those protection settings trip faster, and at smaller deviations, than conventional industrial loads, so a fault lasting milliseconds can move a gigawatt of demand off the system in seconds. Because the shift is customer-initiated, grid operators get no warning and no time to rebalance.

How does the alert change interconnection for hyperscalers and other large loads?

Modeling assessments now sit on the critical path for large load interconnection. A plan that assumes firm service without stability and resource adequacy analysis behind it will face longer queues and tougher interconnection conditions, because the utility or ISO has to be able to show the system can absorb the load and respond safely to its disturbances. In practice, buildout timelines are now tied to someone else's study queue.

What modeling do utilities and large loads need to meet NERC's expectations?

Electromagnetic transient (EMT) models validated against actual equipment for sub-second ride-through behavior; dynamic stability simulation with explicit load-drop contingencies for seconds-to-minutes frequency and voltage response; and capacity expansion and probabilistic resource adequacy modeling that separates firm from flexible load, represents behind-the-meter resources, and reflects AI training operating windows. The paired Reliability Guideline is most explicit about the last of these.

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

• Geothermal shares the clean firm category with only three other technologies: nuclear, natural gas with carbon capture and storage (CCS), and solar or wind paired with long-duration energy storage (LDES).
• Buyers comparing geothermal to solar or unabated natural gas power contracts are comparing different products: one isn’t firm, the other isn’t clean, and both remain cheaper options because they satisfy less of what buyers currently need. 
• Geothermal’s cost premium is real, but so is the value it delivers. Geothermal can reach 90%+ capacity factor, delivering dispatchable clean power around the clock, and it sidesteps the gas turbine manufacturing backlogs and pipeline infrastructure delays slowing other firm power technologies. 
• Geothermal is closer to commercial readiness than most buyers realize. As demand for firm, low-carbon power intensifies, geothermal warrants a serious look alongside nuclear and gas with CCS.

Geothermal is Gaining Traction Despite Its Cost Premium

Hyperscalers, corporate energy buyers, and data center developers face unprecedented pressure to procure clean power. Wind and solar have played a major role in meeting this need, but they are intermittent, and load growth is accelerating and changing. This market pressure is creating bottlenecks that cannot be solved without new grid infrastructure and/or substantial additions of energy storage. Geothermal, which does not suffer from intermittency, has gained tremendous momentum as a challenger to wind and solar, yet it carries one seemingly glaring disadvantage: its cost. 

Yet, major utilities and hyperscalers alike are lining up for geothermal offtake anyway: Southern California Edison and Google have signed separate contracts with Fervo, and Meta has two 150 MW agreements in place with Sage and XGS. These contracts are signaling a growing shift in power markets: not all megawatt-hours are created equal, and the cheapest contracts may no longer contain what buyers truly need. 

Fast, Clean Firm Power: What Buyers Need Today 

In the clean firm power evolution, clean came first. Buyers signed solar and wind power purchase agreements (PPAs) and matched their annual consumption with renewable energy contracts, or offset them with renewable energy certificates (RECs). With lax carbon accounting, solar and wind remain unbeatable on price. But annual emissions tallies conceal something important—a buyer can procure as much renewable electricity as it consumes over a year while still depending on the rest of the grid during the hours when those resources are unavailable. To close that gap, buyers need a source that delivers clean power around the clock, firmly in every hour, on demand. 

Two unforgiving demands have followed since: speed and location. With today’s rapid buildout of new load, a power source needs to come online in time to deliver power when the customer needs it. It is also important to connect near a customer’s location because existing transmission assets are often fully subscribed. Failure on either axis can disqualify a project. However, a contract that can meet all attributes: clean in every hour, firm on demand, quick to deliver, and able to reach the load, might be the best (or only) option available. It is also likely to be considerably more expensive than those annual solar benchmarks that only met the clean condition. 

Four Technologies Make the Clean Firm Power Shortlist

Given this strict set of constraints, what power options should buyers consider? The right set of technologies should aim to solve as many of the aforementioned conditions as possible, and that leaves a fairly tight shortlist: 

1. Solar or Wind paired with Long-Duration Energy Storage (LDES): To achieve the level of reliability that many buyers, including data centers, require, solar or wind would have to be paired with LDES, beyond 4–8 hours, to be able to cover evening hours and remain online during increasingly common extreme weather events. While there are promising LDES technologies on the horizon, most remain expensive and not yet widely commercial. Costs rise significantly compared to a standard solar or wind PPA, and viability varies by location. This will be increasingly competitive as storage costs fall, but not yet a mature firm-power solution at scale.
2. Natural Gas with CCS: With large amounts of unabated natural gas generation coming online, CCS is an attractive option to mitigate a large portion of these emissions while retaining time to power (by building gas first and adding CCS later) and the dispatchability of gas-fired power. On the flip side, CCS is operationally complex, geographically restricted, and - unlike other options here - failure at any point in the CCS chain directly impacts the carbon intensity of power delivered. Gas-fired generation is also reliant on the gas delivery pipeline infrastructure, which may face capacity constraints. Finally, natural gas with CCS is 50% more costly per MWh than unabated firm gas.
3. Nuclear: Nuclear is seeing renewed interest from a wide range of corporate buyers and investors. Conventional nuclear and small modular reactors (SMRs) are attracting funding and offtake deals. While nuclear power holds high potential, long development timelines, high capital costs, and technology risk remain barriers to commercializing the next wave of nuclear technologies. 
4. Geothermal: Geothermal can provide firm, dispatchable, and low-carbon power with capacity factors typically above 90%. Development timelines can be faster than nuclear in regions where resources are already proven, with some projects approaching commercial readiness now. Geothermal technology is still geographically constrained, but enhanced geothermal systems (EGS) are poised to  expand their geographical flexibility by reducing the number of siting variables. 

How to Evaluate the Right Clean Firm Power Technology for Your Project

How do these various options stack up? We compared them across six dimensions:

1. Firmness: how reliably a resource can deliver power on the hour, on demand
2. Carbon: the lifecycle emissions intensity when considering the operational emissions of the generating facility and the broader system carbon impacts required to meet any reliability gaps
3. Time to power: a measure of how long it takes to get from “we need power” to delivery, including interconnection and schedule risk
4. Cost (current): the likely cost of a PPA if signed today
5. Execution risk: a measure of how likely a project’s claims regarding price, timing, and performance will hold
6. Siting flexibility: a measure of how constrained a resource is to a specific geography

This frame is ultimately something the buyer should score themselves; below you will find our current read. Red (unfavorable) represents a weakness that can end an evaluation on its own, green (favorable) means the dimension won’t constrain the decision, and yellow (moderate) sits somewhere in between. 

Comparing the clean firm power options is, perhaps ironically, challenging along the “clean” and “firm” lines. Every serious contender delivers low-carbon power around the clock, because that is the price of admission. Not one technology earns an unfavorable rating in those categories. And that is the catch: every weakness in the comparison table sits somewhere else. The dimensions that ultimately win (or end) these contracts may be the ones that the “clean firm” label can obscure or oversimplify.

A resource that cannot be energized for a decade, or sited near load, or delivered at a price a buyer cannot defend internally, may be infeasible regardless of what else it offers. None of the four clean firm generation technology groupings clear every hurdle today, but they can satisfy different buyer priorities.  

Solar or wind with LDES offers more buyer flexibility, at a price. Enough generation, storage, and other firming can deliver the reliability required, and the buyer sets that target. But the increments get expensive quickly, especially when reliability gaps remain long or unpredictable. Covering more hours means more storage and more generation to fill it, so each step toward firmness parity costs more than the one before it. Firming shortfalls also carry carbon consequences, depending on what gets called to compensate. Those emissions may be absent from GHG accounting ledgers, but that doesn’t stop them from hitting the atmosphere. 

Gas with CCS is challenged by execution and siting because the product depends on a chain: gas generation, capture, transport, and a permitted storage site capable of permanent CO₂ disposal at scale. Each piece can be developed and tested, but low-carbon electricity requires all of them to work together, which can only be demonstrated once all are built. A buyer cannot substitute a link, and cannot observe the whole chain before committing. Clean, in this sense, is conditional, and the delivered power could end up being far more carbon-intensive than advertised.

Nuclear ranks poorly on time, cost, and execution, but full-scale reactors and small modular reactors carry those weaknesses differently. A large reactor arrives in one piece, and its schedule and final cost come into focus as construction advances. By then, years and billions are committed; a buyer cannot learn from a fraction of a reactor before deciding whether to finish it. Small modular reactors are designed to break that capacity into increments a buyer could add as demand grows, which would ease the one-piece problem. But none are operating commercially in the US yet, and the signed deals target first power in the 2030s, so the modular promise is still years out. 

Geothermal’s weaknesses work somewhat differently. Conventional hydrothermal is tightly constrained by location, which buyers can screen before choosing where to procure, or co-locate where possible. Next-generation geothermal, including EGS, could widen that geography considerably. 

For EGS, the largest remaining uncertainty is execution: whether the developers can replicate successful drilling and reservoir performance at commercial scale. Part of that risk belongs to the rock, and no contract can make a reservoir perform. But drilling also produces information before the entire project has been built, and geothermal capacity can often be developed incrementally as needed. A buyer can require resource evidence before committing, tie later commitments to demonstrated performance, or spread procurement across projects rather than depend on a single field. 

Geothermal’s flexibility has limits. A buyer who needs full capacity by a fixed date may have little room to wait for wells to perform before committing to the next phase. And successful wells do not eliminate risk in project expansion or guarantee performance in other locations. Execution risk stays with the resource. What geothermal offers is more opportunity to learn about that risk while a buyer can still do something about it. 

Revisiting the Geothermal Premium

Geothermal contracts, especially EGS, remain expensive. But that premium looks different on this side of the comparison. Clean has a cost, but so does firmness, speed,  reach, and execution. Every clean firm option pays for those attributes somewhere. Solar and wind contracts will remain cheap by comparison, but it’s the wrong comparison: the market signing geothermal contracts is pricing a much larger product. 

That still leaves a hard decision. Geothermal and gas with CCS carry execution risk and siting constraints. Nuclear asks more of the buyer on cost and time. Firmed solar asks more as reliability requirements rise. The best option will change with each buyer, location, and project. Geothermal’s place in that choice is changing quickly. EGS is expanding where projects can be developed, drilling costs are falling, and commercial experience is growing. Those advances are bringing geothermal into consideration for buyers who previously had little reason to look its way. 

And that may be the more revealing aspect of the geothermal market today: buyers are signing those contracts with the alternatives in full view. Their decisions offer an early indication of what clean firm power delivered in the right place and at the right time is worth, and which risks buyers are willing to carry to get it. 

Frequently Asked Questions

Is Geothermal Cheaper Than a Solar PPA? 

No, a geothermal power contract can be significantly more expensive than a solar PPA. But they're not the same product: solar meets annual clean-energy targets, while geothermal delivers firm, dispatchable power around the clock, which is why buyers are increasingly comparing it to nuclear and gas with CCS instead.

Is Geothermal Cheaper Than Solar with LDES? 

Not necessarily on a per-contract basis, but the cost structures behave differently. Solar with LDES gets more expensive with every additional hour of reliability a buyer needs, since covering more hours means adding both more storage and more generation to fill it. Geothermal's premium, by contrast, is priced into the contract upfront as a firm, dispatchable product from day one.

How Does Geothermal's Risk Compare to Gas with CCS? 

Differently, in a way that matters to buyers. Gas with CCS depends on a chain (generation, capture, transport, and storage) that can only be verified once the whole system is built, so a buyer can't test it piece by piece. Geothermal generates performance data incrementally through drilling, letting buyers require resource evidence or tie commitments to demonstrated results before committing further, though this data does not come free.

Which Deploys Faster, Geothermal or Nuclear? 

Geothermal, in regions with already-proven resources. Fervo's Cape Station, for example, broke ground in 2023 and is on track for first power in late 2026 and 500 MW by 2028: a fraction of a typical new-nuclear large reactor timeline.

Key Takeaways

• Agentic inference has changed the economics of AI. Tokens are becoming units of work and the economic driver is now the work produced, not token generation. Per-token costs are falling and the willingness to pay for work produced is rising; these two trends compound. This tailwind enhances AI economics and has spillover impacts on all layers of the AI stack.
• The AI infrastructure question has shifted from whether demand will show up to whether the physical stack can scale quickly enough. That stack includes power generation, grid capacity, interconnection, compute, memory, networking, cooling, siting, and community acceptance.
• Carbon Direct Capital and Relae (formerly Carbon Direct Inc.) have a differentiated view because the two entities work across both sides of the constraint: Relae advises hyperscalers and energy buyers on power and grid bottlenecks, while Carbon Direct Capital invests in the technologies that relieve those bottlenecks.
• Carbon Direct Capital sees better risk-adjusted returns investing in the physical foundations of AI, including clean firm power, energy system efficiency, data center efficiency, and inference-optimized compute, rather than chasing late-stage AI application valuations.

A Better Question Than "Is AI a Bubble?"

The most important development in AI economics is agentic AI turning tokens into work, a shift that reframes the bubble debate which dominated investor conversations, sell-side notes, and Chief Information Officer surveys through early 2026. Hyperscalers spent approximately US$380 billion on capital expenditure (capex) in 2025 and have guided to approximately US$720 billion of capex in 2026.¹ Carbon Direct Capital and Relae have worked together to build project-level models for both training and inference facilities to demystify the numbers and understand financial and technical sensitivities. The core finding was that the assets could be bankable using standard assumptions and that the binding constraints were physical, not financial. That conclusion has been reinforced in recent months by new developments.

Concretely, AI is moving from single prompts and answers to multi-step workflows that plan, reason, call tools, verify outputs, and keep state. This shift to inference is the structural successor to training in the initial AI capex cycle; it changes power requirements, time to power, and compute architectures all at once. Goldman Sachs estimates that agentic AI could drive a 24-fold increase, relative to a 2026 baseline, to roughly 120 quadrillion tokens per month globally by 2030 as per-token costs continue to fall. SemiAnalysis makes the same point from another angle: the value of frontier tokens has risen as agentic workflows become useful, while hardware and software improvements have reduced the cost of producing each token.

This does not mean every AI company is attractive, every data center project works, or every valuation is justified. It means the easy bubble framing is missing the more investable question. If token demand is compounding and the unit value of work produced is rising, the scarce resource is not abstract enthusiasm. It is the physical infrastructure required to turn that demand into work produced.

The Data Center Model Still Matters, But the Box is not a Black Box

Our internal modeling for an illustrative 167-megawatt inference data center using Nvidia Blackwell graphics processing unit (GPU) servers suggests the potential for high-teens percent equity returns under a defined set of assumptions.² We built a bottom-up underwriting, beginning with the number of users served per inference data center, assuming approximately how many tokens they will demand daily, and translating that token demand into compute needed based on industry-standard quantization and utilization rates. We then inferred the number of GPUs and servers needed to achieve the desired compute, which ultimately drove the total invested capital and power demand based on assumed thermal power designs and power usage effectiveness (PUE). On the revenue side, we used GPU-as-a-service rental rates as one proxy for the market value of compute capacity. A hyperscaler would not rent scarce compute externally if it had higher-value internal demand for that same capacity. As we will detail below, GPU rental prices have been steadily increasing on the back of inflecting inference demand.

This model is not the entire argument; it is the starting point. An important lesson is that power cost alone does not break data center economics. Electricity is slightly over 10% of total costs in our model: a 50% increase in power price reduces equity-level returns by less than 2%. Access to power, speed of interconnection, and equipment availability matter more. In other words, the economics of the model facility are workable, but only if the facility can be built and powered on the timeline customers need.

That is where most AI commentary remains too superficial. It treats the data center as a black box: capex goes in, tokens come out. That misses the bottlenecks inside and around the box. AI racks are moving far beyond traditional cloud power density. Cooling is shifting from air to liquid and two-phase systems. Networking and high-bandwidth memory become binding constraints in inference architectures. Grid interconnection queue wait times stretch to years. Communities can and do block projects. The technical, physical, and political constraints are increasingly the drivers of potential returns.

Inference Makes the Constraint Structural

While training is episodic, inference is recurring. A training run can be delayed, accelerated, or redesigned. Inference happens every time a user asks a question, a developer runs an agent, a business automates a workflow, or an application calls a model in the background. Agentic inference multiplies that load because one user action can become many model calls, validation loops, and memory reads; industry benchmarks show that agentic systems consume 5–30 times more tokens than a standard chat interaction.

Inference demand is also resilient in both directions. If efficiency gains lower the cost per token, more workflows become economic and total token consumption rises - the classic Jevons Paradox. However, token prices do not necessarily need to fall for inference spend to grow. As the economic unit shifts from tokens generated to work produced, customers may pay more per token when an agent delivers work produced that is worth more than the inference cost. Regardless of token price, tokens must all route through the same physical bottlenecks and we are seeing an increase in inference demand.

The architecture of inference is also changing. Some workloads will prioritize low-latency answers. Others, especially agentic work without a human waiting on every token, will prioritize memory, state, context, and cost per completed task. That means the AI infrastructure stack will become more heterogeneous, not less: XPUs (specialized AI accelerator chips), custom silicon, photonics, memory hierarchies, and edge or regional deployment models will all matter. The pricing data shows demand for more AI infrastructure overall: on-demand GPU rental capacity is effectively sold out across all chip generations in early 2026, with one-year Hopper H100 contract pricing rising 15–20% month-on-month through March 2026 and Blackwell B200 rental rates up 23% in March alone. When rental rates rise into a wave of new chip supply, supply is not catching up to demand.

Power Is Not One Constraint, It Is Several

Saying "AI is power constrained" is true, but not specific enough. The real problem has several layers. First, data centers need more electricity than many local grids can deliver on hyperscalers' timelines. Crucially, some grids can supply sufficient power but not continuously for 8,760 hours per year, conflicting with traditional assumptions about service reliability and leading to novel strategies around flexibility and intermittent self-supply. Second, the grid must be able to absorb large, fast-moving computational loads without creating reliability risks. Third, customers need energy procurement strategies that satisfy cost, reliability, climate, and public-acceptance requirements. Fourth, projects must get built in real communities, through real interconnection processes and real permitting fights. Power is not simply a commodity to purchase. It is an infrastructure development problem.

This is where Relae is directly relevant. Relae has assembled a team of scientific, engineering, and market experts to support a paying power and energy advisory practice serving hyperscalers, energy buyers, and power producers. Its work answers the questions customers are asking before the market prices them: how to get more capacity out of existing physical grid infrastructure; how to assess the costs and value of load flexibility through advanced modeling capabilities; how to make clean firm generation bankable; how to reduce data center energy intensity; how to validate "bring your own power" and "bring your own compute" structures; and how to build projects that communities will accept. 

In the last twelve months alone, Relae has supported hyperscalers on bankability assessments for next-generation geothermal, scoped load-flexibility programs for multi-hundred-megawatt, single-customer sites, and modeled the carbon and reliability profile of "bring your own power" configurations against grid-tied baselines. Carbon Direct Capital leverages our network of technical experts at Relae, including power engineers, geologists, and electrochemists, to conduct credible technical diligence and to gain insights into early stage market trends and emerging preferences.

What Carbon Direct Capital Is Investing Behind

Our investment focus follows the bottlenecks. On the power side, we are investing in technologies that can deliver reliable power on AI timelines. Sage Geosystems is a next-generation geothermal platform with hyperscaler buy-in; Carbon Direct Capital co-led its US$97 million Series B with Ormat Technologies. We could not have made this investment without the deep expertise of the Relae research team which analyzed Sage's technical results to date to help underwrite future project feasibility. ION Clean Energy is a company that retrofits carbon capture technology onto natural gas combined cycle plants to create "blue electrons"; Relae is in active dialogue with multiple large power users on this topic. Carbon Direct Capital is also actively evaluating the enabling picks and shovels around geothermal, nuclear, fuel cells, and more.

On the data center efficiency side, we are investing in technologies that reduce the amount of power required for a unit of AI work. While it is encouraging to see incremental annual gains in chip efficiency, these are scaling far more slowly than compute demand, driving the need for more innovative technological solutions. As one example, a team at Relae helped us understand the fundamental energy consumption requirements of a standard complementary metal-oxide-semiconductor (CMOS) chip, and the potential of all-optical computing as an alternative. This led to Carbon Direct Capital investing in Neurophos, a photonic compute company targeting step-function gains in energy efficiency per chip that are beyond those achievable by existing GPUs. Carbon Direct Capital joined the company's US$110 million Series A alongside Gates Frontier, Microsoft's M12, Aramco Ventures, Bosch Ventures, and others. More broadly, we are studying other layers of the data center technology stack including networking, memory, cooling, and inference-optimized architectures because the next phase of AI infrastructure will not be solved by simply buying more of yesterday's hardware.

The Bear Case Deserves to Be Taken Seriously

There are real risks to the AI boom: Hyperscaler free cash flow can compress if capex grows faster than revenue. Model efficiency gains can reduce the amount of compute required for a given task. Training demand may be more episodic than the market assumes. Local opposition can slow or cancel data center and power projects. Some new data center capacity could become expensive cloud infrastructure competing on price if AI revenue disappoints.

Those risks are why Carbon Direct Capital frames this as an investment in constraints, not in AI enthusiasm. If efficiency improves, inference use cases expand and the bottleneck shifts to deployment, memory, power, and cost per unit of work produced. If training demand slows, inference and enterprise agents still require recurring capacity. If local grids cannot absorb load, technologies that unlock power, reduce energy intensity, or improve flexibility become more valuable. If some AI applications or model developers fail, the upstream physical bottlenecks remain for the rest.

The Investment Conclusion

The AI infrastructure opportunity sits at the intersection of frontier technology risk, project-finance economics, and energy-system engineering. Underwriting this opportunity well requires addressing all three at once; Carbon Direct Capital is built to do just that. The technical team at Relae has a pulse on emerging stakeholder preferences and scientific breakthroughs, understands novel technologies deeply, and is highly experienced in conducting detailed technical diligence to ensure that projects are viable and scalable. Carbon Direct Capital combines these market and technical insights with our commercial underwriting to facilitate new investments. We are not picking AI winners. We are not picking pure energy assets. We are investing in the companies and technologies that have to exist for AI to sustainably scale.

Frequently Asked Questions

Is the AI capex boom a bubble? While valuations vary, token demand and physical infrastructure needs are real and compounding. The correct question to ask is not whether AI is a bubble, but what the binding constraints are. Our modeling shows that constraints are physical, not financial. 

What are the real constraints on AI infrastructure growth right now? AI infrastructure growth is constrained by power availability, grid capacity, and interconnection speed, not capital availability.

Why does inference matter more than training for long-term AI power demand? Inference is recurring and grows with AI usage, it is not episodic like training runs.

What is Carbon Direct Capital investing in, and why? We are investing in clean firm power, energy system efficiency, data center efficiency, and inference-optimized compute—the physical bottlenecks rather than application-layer valuations.

Disclaimer

Carbon Direct Capital Management LLC is an investment adviser registered with the US Securities and Exchange Commission (SEC). Registration as an investment adviser does not imply any particular level of skill or training. Additional information about Carbon Direct Capital Management LLC, including our Form ADV Part 2A Brochure, is available on the SEC's website at adviserinfo.sec.gov.

This content is provided for informational purposes only and should not be construed as or relied upon as investment, legal, tax, or other advice. You should consult your own advisers regarding legal, business, tax, and other matters related to any investment. Any projections, estimates, forecasts, targets, prospects, or opinions expressed are subject to change without notice and may differ from opinions expressed by other employees of Carbon Direct Capital Management LLC, its affiliates, investors, portfolio companies and other individuals, groups or entities. Certain information contained herein may have been obtained from third-party sources believed to be reliable; however, Carbon Direct Capital Management LLC makes no representations about the accuracy or completeness of any such information or its appropriateness for any given situation. Any investments or portfolio companies mentioned are not representative of all investments made by funds managed by Carbon Direct Capital Management LLC, and there can be no assurance that any investment will be profitable or that future investments will have similar characteristics or results. Past performance is not indicative of future results. The content speaks only as of the date indicated. This content does not constitute an offer to sell or a solicitation of an offer to buy any security. Any such offering will be made only pursuant to formal offering documents.

Key Takeaways

• The electricity demand surge is real and accelerating. Just last year, data center load in the US was projected to increase from 25 GW to 120 GW by 2030. Today, Texas’ preliminary long-term load forecast projects over 187 GW of data center load by 2030, more than double today’s total peak demand of approximately 91 GW.
• Flexible loads that respond dynamically to policy signals are becoming a regulatory requirement. Texas Senate Bill 6 (SB6), signed into law in June 2025, is the clearest signal yet. It makes remote curtailment equipment a condition of interconnection for new loads of 75 MW or more, so utilities can disconnect them during declared firm load shed events, and separately creates a voluntary demand response program that those loads can elect to join. 
• Relae’s power system modeling puts a dollar value on what data center load flexibility is worth. Our ERCOT analysis shows that data center demand response can eliminate forced load shedding risk, even at 40 GW 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. 
• Flexible load curtailment and compute uptime do not need to be in conflict. In our modeling, demand response operates as a "ghost battery" at the data center's grid node, absorbing grid stress as a physical battery would. That construct has a direct real-world analog: on-site battery storage lets data centers draw from stored energy during grid stress events rather than curtailing workloads. Technologies such as those demonstrated by Emerald AI have shown 25% load curtailment at cluster scale while preserving compute service quality.

Data Center Electricity Demand is Testing Grid Limits

Electricity demand in the United States is growing at its fastest pace in decades. Leading the surge is a rapid buildout of data centers, driven by the expansion of artificial intelligence. ERCOT, the grid serving most of Texas, is projecting up to 187 GW of new data center load by 2030, against a total peak demand today of approximately 91 GW.

The scale of this shift extends across the country. According to NERC's 2025 Long-Term Reliability Assessment (LTRA), summer peak demand across the US bulk power system is forecast to grow by 224 GW over the next 10 years, 69% above the prior year’s 10-year projection of 132 GW, with data centers as the dominant driver. As previously explored by Relae, data center energy capacity in the US is projected to increase from 25 GW to 120 GW by 2030, characterizing it as the first wave of a longer demand surge that electrification of buildings and transport will reinforce.

Policymakers are responding in real time, and regulatory responses like Texas SB6 are already rewriting the rules for how data centers connect to the grid. The pace of change is fast, but the siting, infrastructure, and interconnection decisions made today will have lasting consequences: they will determine whether data centers are grid assets or grid liabilities, and the financial difference between the two is measured in billions.

New Solutions for Data Center Demand Response: Texas SB6 as a Test Case

The rapid nature of this new wave of load growth means traditional approaches to managing the grid may not be sufficient. In the past, lead times on new sources of electricity demand allowed utilities to procure supply-side resources in advance. The scale and immediacy of data center deployment is revealing limitations of this approach. States, utilities, independent system operators (ISOs), and regulators are actively pursuing novel approaches for grid management in response.

Texas has become a focal point of US data center expansion, with SB6 as a leading policy response. The bill, which took immediate effect upon Governor Greg Abbott's signature on June 20, 2025, is the most significant restructuring of large-load interconnection rules in ERCOT's history. It requires large loads over 75 MW to install remote curtailment equipment as a condition of interconnection, so grid operators can disconnect them during declared firm load shed events, and creates a separate voluntary demand response program, procured competitively with at least 24 hours’ notice, that those loads may elect to join. If data centers are adding substantial new load to the grid, this reasoning goes, they should also contribute to grid stability by demonstrating load flexibility—reducing power draw at times of peak demand.

While demand response programs currently exist, incentivizing voluntary curtailment from data centers is challenging. In an AI compute arms race, the value of uninterrupted compute time far exceeds any available curtailment payment, such as via PJM’s capacity market mechanisms. 

Approaches to bridge that gap are coming to fruition: EPRI’s DCFlex program is working with hyperscalers and utilities to develop the technical protocols, measurement standards, and contractual frameworks that would make large-load demand response a routine grid service. Innovators like Emerald AI have demonstrated a 25% power reduction across a 256-GPU cluster over three hours during an Arizona grid stress event, while preserving compute service quality, helping to bridge the valuation asymmetry between energy and compute. 

Hyperscalers are already putting a flexible load commercial strategy into action. In March 2026, Google announced  1 GW in demand response contracts with multiple US utilities, including Entergy Arkansas, Minnesota Power, and DTE Energy.

The conversation has shifted from whether data centers can be flexible to how that flexibility gets structured and deployed. 

Putting a Dollar Value on Data Center Load Flexibility

For developers, investors, and grid operators navigating data center growth, the challenge has been making high-stakes siting and interconnection decisions without a clear picture of what load flexibility is actually worth, what inflexibility costs the system, or how curtailment requirements will reshape the regulatory landscape. 

Prior research has established that flexible data center load can absorb substantial grid stress. For instance, research from Duke University’s Nicholas Institute found that 22 of the largest US balancing authority areas could absorb approximately 98 GW of new flexible load if 0.5% of that load’s annual energy is curtailed, or 76 GW at a stricter 0.25% curtailment level. 

Relae’s analysis goes further by quantifying the economic cost at each increment of flexible load growth, and the precise threshold at which that flexibility stops being optional. We zeroed in on ERCOT, a region with high data center load growth and immediate regulatory stakes, determining the value of implementing flexibility and, conversely, the system risk of failing to do so.

How We Modeled It 

Assessing the impacts of load growth and flexibility solutions requires a systems-level analysis, best achieved via power market modeling. At Relae, we deploy our in-house power system modeling framework to navigate this complexity. 

Our toolkit includes CD-PyPSA-USA, used for this analysis, which is built on the Python for Power System Analysis (PyPSA) platform. This grid model simulates how power networks operate and evolve over time by solving for the least-cost optimization of the entire power system. Critically, our model is customized to explicitly represent complex, real-world dynamics, including data center load flexibility, co-located generation, and various policy constraints.

For this analysis, we simulated ERCOT operations under a range of data center growth scenarios. We modeled loads from 5 GW up to 40 GW in 5 GW increments, pairing each with sufficient on-site gas generation to cover roughly 70% of data center energy needs–a conservative estimate on the approach developers are taking today. 

We ran each scenario under two conditions: no load flexibility (“flex00”, the baseline) and 25% emergency curtailment capability (“flex25”), consistent with solutions exhibited by Emerald AI. This approach allowed us to determine the system's response to step-changes in electricity demand.

What Our Analysis Reveals

Before doing the analysis, we expected to see load flexibility become more important as more data center load is added to the grid, preventing forced load shedding with high lost-load costs. But we didn’t know how large this effect would be, or what amount of new data center load would start to trigger it.

Our analysis found that without flexible load management, forced load shedding first appears at 30 GW, small in scale at first (4.4 GWh over 3 hours) but growing sharply as load increases. At 35 GW, we observe 50 GWh of shedding across 39 hours. At 40 GW, shedding reaches 158 GWh across 81 hours, equivalent to nearly three times ERCOT’s average hourly energy consumption, with an economic cost at VoLL of approximately $5.5 billion.¹

By enabling on-demand data center load flexibility, forced load shedding is eliminated in every scenario we tested. The same 40 GW case instead sees 165 GWh of controlled, short-duration curtailment spread across 86 hours, less than 1% of hours in a year. Further, the average demand response in these hours was less than 5% of the nameplate data center load, with the largest event reaching 14% of data center load. The grid stays balanced, consumers remain connected, and data centers deliver substantial value to the system via flexible loads.

Load Flexibility Is High Value and Presents an Opportunity for Storage

Our modeling puts a dollar figure on what flexible load is worth. At a VoLL of $35,000/MWh, each hour of demand response in the 40 GW scenario delivers approximately $64 million in avoided consumer welfare losses. Over a full year, that adds up to $5.5 billion, achieved through an average of just 5% demand response across the 86 hours of curtailment needed to eliminate all forced load shedding.

That value points directly to an opportunity for storage. In our model, demand response functions as a “ghost battery” at the data center’s grid node, absorbing grid stress exactly as a physical battery would, without any electrons needing to flow. That virtual battery can become a real one. On-site battery storage allows a data center to dispatch stored energy during grid stress events rather than curtailing workloads, maintaining compute continuity while relieving grid pressure.

The implications point in two directions. 

• For the hyperscaler or data center operator, physical storage converts a compliance obligation into an uptime guarantee: the curtailment event becomes a battery discharge, with negligible impact to the compute stack. 
• For the storage developer, co-location with large data center loads represents a high-value deployment opportunity with a clear commercial case. The avoided welfare costs per curtailment hour our model quantifies is the value a well-positioned battery asset, co-located at a data center node, can credibly claim to preserve.

The Data Centers of Tomorrow Need to be Grid Assets

The data center buildout underway is large enough to reshape grid reliability across entire regions, and the regulatory environment is beginning to reflect that scale. Texas SB6 is the most prescriptive example to date: it requires new large loads above 75 MW to install remote curtailment equipment operable during firm load shed events.

Our modeling quantifies what load flexibility is worth across this landscape: data centers with curtailment capability can provide significant value and avoid billions in consumer welfare losses annually. For developers and investors, designing that capability in from the start can convert a compliance requirement into a long-term grid asset.

The federal picture has moved in the same direction since this analysis was published. In May 2026, NERC issued a rare Level 3 Alert on computational loads; in June, FERC ordered six RTOs and ISOs to revise or justify their large-load interconnection rules, and in July, FERC directed NERC to develop mandatory computational-load reliability standards by the end of the year. We covered what that means for grid modeling and interconnection in Inside NERC’s Level 3 Alert on data center loads. Flexibility is no longer only a Texas statutory question; it is becoming part of the federal reliability framework.

Frequently Asked Questions

What is data center load flexibility, and how does it work?

Load flexibility is a data center’s ability to reduce the power it draws from the grid on short notice, during the small number of hours when the system is under stress. In practice, that means shifting or pausing deferrable compute, drawing on on-site batteries or generation, or pre-cooling the facility ahead of a peak. The point is not to consume less overall—it is to move a thin slice of demand out of the hours when the grid can least afford it.

What does Texas SB6 require of large data centers in ERCOT?

Texas Senate Bill 6, signed June 20, 2025, makes remote curtailment equipment a condition of interconnection for new loads of 75 MW or more in ERCOT, so utilities can disconnect them during declared firm load shed events. Separately, it creates a voluntary, competitively procured demand response program those same large loads can elect to join, with at least 24 hours’ notice. The mandatory piece is the disconnection capability; paid participation in demand response is a choice.

What does data center inflexibility cost? How much curtailment avoids it?

Relae’s ERCOT modeling found that without flexibility, forced load shedding first appears at 30 GW of data center load and reaches 158 GWh across 81 hours at 40 GW—roughly $5.5 billion a year in consumer welfare losses at ERCOT’s $35,000/MWh value of lost load. Enabling curtailment eliminated forced shedding in every scenario we tested, at an average of under 5% of data center demand across 86 hours, less than 1% of the year. Each hour of demand response in the 40 GW case is worth about $64 million in avoided losses.

Is data center load flexibility proven today, and how does it compare to on-site batteries?

Data center load flexibility is past proof of concept and into commercial deployment: Emerald AI cut power to a 256-GPU cluster by 25% for three hours during an Arizona grid stress event without degrading compute service quality; EPRI’s DCFlex initiative is building the protocols and contracts, and Google has signed 1 GW of data center demand response with US utilities. 

On-site batteries reach the same result from the other direction—instead of curtailing workloads, the facility discharges stored energy, which is why our modeling treats demand response as a “ghost battery” at the data center’s grid node. For operators who cannot pause compute, storage turns the same compliance obligation into an uptime guarantee.

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.