The AI Bubble Debate Misses the Point: The Bottleneck Is Physical
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.
Power & Energy
Relae provides independent advisory for large corporate buyers, power providers, and infrastructure investors making high-stakes decisions about clean firm power, grid constraints, data center energy optimization, and long-term investment strategy. Our insights help you evaluate solutions that can be deployed reliably, responsibly, and affordably, so you can navigate an evolving energy landscape with confidence.
1. Numbers based on earnings releases from Alphabet, Amazon, Meta, and Microsoft as of Q1 2026 reporting; note that Alphabet and Amazon have since raised their full-year 2026 guidance further.
2. Model assumptions include: 167-megawatt illustrative inference data center located in Virginia; ~93,200 Nvidia B200 GPUs; ~US$6.7 billion invested capital; 1.25 power usage effectiveness; ~1.5 terawatt-hours annual electricity consumption; US$100/megawatt-hour power; 50% debt ratio; US$3.50/GPU-hour initial topline pricing with a 3% annual escalator.
What to Read Next
Carbon Capture for Natural Gas-Fired Power Generation: An Opportunity for Hyperscalers
Key Takeaways
- AI-driven data center demand is outpacing grid capacity, and hyperscalers are bringing more natural gas, which already supplies about 40% of US electricity, online to meet their needs.
- Pairing carbon capture and sequestration (CCS) with natural gas lets data centers source firm power today while cutting plant-level emissions up to 95%—without waiting on multi-year renewable interconnection queues.
- The Google-Broadwing deal demonstrates real progress and commitment toward natural gas with CCS as the first major commercial deployment of this exact pathway
Meeting Electricity Demand and GHG Emission Reduction Targets
Rapid growth in electricity demand across the US, driven by AI data center expansion and increased industrial electrification, is placing significant pressure on power grids. After decades of stable electricity load, demand has increased significantly since 2022 and is expected to rapidly grow for the foreseeable future. Natural gas currently fuels around 40% of US electricity generation. Its share is expected to grow in the coming years. However, unabated natural gas generation is not compatible with stakeholder targets to reduce greenhouse gas (GHG) emissions. Combining CCS with natural gas-fired generation is one pathway to meet growing electricity demand and achieve GHG emission reduction targets.
[cta]

The Role of Natural Gas in Electricity Supply
Natural Gas Generation Versus Renewable Generation Deployment
Electricity generators can provide multiple products to regional grids, generally providing two services: energy (power production) and reliability (consistent availability). Natural gas-fired plants can provide both, whereas renewable energy sources like wind and solar generate energy but offer less reliability.
As electricity demand rapidly grows, grids will need both energy and reliability to function effectively. However, the interconnection queue for renewable energy assets has a years-long backlog which is delaying their deployment. Grids will need additional reliability assets to support the large amounts of renewables (usually in the form of storage). Some jurisdictions are creating an alternate pathway for natural gas plants to bypass the lengthy interconnection queue which may allow for the rapid development of natural gas generators. Hyperscalers are also pursuing development of large behind-the-meter (BTM) generation of electricity from renewable and fossil sources, but these must also meet high standards for reliability.
The Case for Carbon Capture Deployment
Electric utilities and developers of data center infrastructure are planning to build substantial new natural gas generation assets in addition to maximal deployment of renewable electricity. CCS technology enables natural gas plants to deliver stable, continuous power while significantly reducing emissions by capturing up to 95% of emitted CO₂. Natural gas plants with CCS are viable options to deliver the lower-emission, reliable power needed to respond to rapidly emerging AI data center power demand growth. The 45Q tax credit, a key government incentive for CCS, was preserved and effectively strengthened under 2025's One Big Beautiful Bill Act. The Google-Broadwing deal, the first major commercial deployment of this exact pathway, was signed in October 2025 and serves as a useful proof point.
Benefits of Integrating CCS into Natural Gas Power Generation
Integrating CCS into natural gas-fired power plants provides several advantages for data center stakeholders:
- Reduced carbon emissions: Achieve emission intensities of approximately 80–120 kg of CO₂ equivalent per megawatt-hour (CO₂e/MWh), significantly below the current US grid average of approximately 340–420 kg CO2e/MWh.
- Reliable baseload power: Continuous, predictable electricity delivery.
- Compact infrastructure: Requires less land compared to renewable energy projects, simplifying data center siting near existing infrastructure.
- Cost: CCS integrated with new natural gas-fired generation can deliver low-cost decarbonization. Relae estimates $75-150/MWh, which is competitive in many markets with other firm baseload options such as new nuclear power or wind and solar with battery backup.
Seven Key Considerations for Implementing CCS
Stakeholders considering CCS technology must carefully evaluate seven critical factors:
1. Meeting Rapid Deployment Timelines
Traditional natural gas plants can be operational within roughly 18 months, provided they bypass interconnection queues for reliability purposes and have access to key equipment. Integrating CCS technology extends this by an additional 18–36 months. Designing plants to be "capture-ready" allows for quicker initial deployment and smoother CCS integration in the future. However, deploying a capture-ready plant without a commitment to build the carbon capture portion is inconsistent with serious climate action.
2. Sizing Plants Optimally
CCS is most economically and environmentally optimal at natural gas plants with capacities of 100 MW or greater. It offers significant opportunities for emissions reductions for the forecasted new data center load. CCS is not suitable for smaller or highly variable natural gas plants.
3. Selecting Effective Carbon Capture Technology
CCS technologies such as solvents, sorbents, membranes, and oxyfiring vary significantly in maturity, efficiency, and cost. Choosing the right approach requires thorough evaluations aligned with specific project requirements. These will vary by setting and configuration (e.g., turbine class, reciprocating engines, number of units, water availability, etc.).
4. Navigating CO₂ Transportation Logistics
The safe and efficient transport of captured CO₂ via pipelines, rail, or barges is critical. Aligning infrastructure planning with overall project timelines prevents delays.
5. Ensuring Safe and Effective Sequestration
If there is no CO₂ storage, there is no project. Permanent CO₂ storage in Class VI injection wells requires detailed geological studies and regulatory permitting. Early collaboration with experienced sequestration operators is essential to success.
6. Conducting a Comprehensive Life Cycle Analysis
Full life cycle emissions analyses, including upstream methane leakage, construction impacts, and CO₂ transportation, are critical for accurate environmental assessments and ensuring low-carbon electricity supply. Prioritizing low-leakage, third-party verified natural gas supply enhances positive climate impacts.
7. Performing Siting Feasibility Early
An early and quick feasibility assessment is critical to identifying promising opportunities and key barriers at candidate CCS sites. Important factors include available transmission capacity, the potential to expedite approval of interconnection for thermal resources, regulatory barriers, state and local incentives, the sufficiency of natural gas infrastructure, and water supply.
Frequently Asked Questions
How much longer does adding carbon capture take compared to building a natural gas plant alone? Traditional natural gas plants can be operational within roughly 18 months, provided they bypass interconnection queues for reliability purposes and have access to key equipment. Integrating CCS technology extends this by an additional 18–36 months.
Is a "capture-ready" natural gas plant a legitimate climate strategy if the capture portion isn't committed yet? Designing plants to be "capture-ready" allows for quicker initial deployment and smoother CCS integration in the future. However, deploying a capture-ready plant without a commitment to build the carbon capture portion is inconsistent with serious climate action. “Capture committed” is a better stance than “capture ready”.
How does the cost of natural gas-fired power with CCS compare to nuclear or renewables with battery storage? CCS integrated with new natural gas-fired generation can deliver low-cost decarbonization. Relae estimates $75-150/MWh, which is competitive in many markets with other firm baseload options such as new nuclear power or wind and solar with battery backup.
Has any hyperscaler actually deployed natural gas-fired power with CCS at scale yet? The Google-Broadwing deal, the first major commercial commitment of this exact pathway, was signed in October 2025 and serves as a useful proof point. Others are in development.
[cta]
How Relae Supports Data Center Decarbonization
Natural gas-fired generation combined with CCS is a proven solution for meeting the urgent electricity demands of data centers while significantly reducing emissions. Relae helps stakeholders navigate the complexities of CCS deployment through deep, science-backed expertise and strategic advisory services. Our experienced team provides comprehensive support throughout CCS project planning and execution, including technology selection, life cycle emissions analysis, infrastructure assessment, project viability, regulatory compliance, and risk management.
The $5.5 Billion-Dollar Case for Enabling Data Center Load Flexibility
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.
Voluntary Curtailment, Forced Outages, and the Value of Lost Load
This analysis makes an important distinction between three curtailment types: one voluntary and two involuntary electricity demand reductions.
- Demand response or load flexibility (voluntary reduction): This involves industrial consumers curtailing their power requirements in response to pricing or regulatory incentives. Participation is optional.
- Mandatory curtailment (targeted reduction): This is a required, controlled reduction of power draw by a specific consumer group (e.g., data centers under Texas SB6) when directed by grid operators during declared emergencies. This is a deliberate policy directive aimed at grid stability.
- Load shedding (forced outage): This is a non-targeted, involuntary outage event, such as a rolling blackout, where the system operator must cut power to prevent grid failure. These events affect all types of consumers, including residential and commercial electricity demand.
Grid operators in Texas use a value of lost load (VoLL) of $35,000 per megawatt-hour (MWh) to measure the welfare cost borne by businesses and households who lose power involuntarily. That figure, the standard benchmark applied by ERCOT in reliability and market design analysis, is what we use as the basis for valuing shedding events in our modeling. At $35,000 per MWh, even a small number of unplanned outage hours produces welfare losses in the billions.
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.
Reconciliation Bill Dramatically Shifts the Clean Energy Landscape
Key Takeaways
- Accelerated phase-out schedules for key clean energy and decarbonization tax credits will shorten the runway for project development, which could stall or cancel projects.
- Urgency is paramount, and qualified projects should expedite construction and operational timelines to secure eligibility for existing credits.
- A more complicated policy landscape requires concerted effort to navigate, including with the support of policy professionals.
Reconciliation Rolls Back Much of the IRA
On Friday, July 4, 2025, the President signed a sweeping reconciliation bill, H.R. 1, that will add at least $3.3 trillion to the national debt and marks a pivotal, contentious moment for US clean energy policy. The law was enacted through the complex legislative process known as budget reconciliation, requiring only a simple majority of votes in the House and Senate. The new law substantially modifies or terminates many of the Inflation Reduction Act of 2022 (IRA)'s clean energy incentives and has extensive implications for the economic viability of American energy and manufacturing projects.
In the Senate, three Republicans crossed party lines to vote against the bill, requiring Vice President JD Vance to break the tie. In the House, only two Republicans broke ranks to vote against final passage. While some of the more complex provisions of the bill, such as new foreign entity of concern (FEOC) restrictions, will require more time to fully assess, we've prepared a rapid run-down of key alterations to IRA incentives for carbon management, hydrogen, and clean fuel technologies.
What Is the 2025 Reconciliation Bill?
While the 2025 reconciliation bill is staggering in length, scope, and severity, containing provisions to cut Medicaid, reduce nutrition assistance, raise the debt limit, and cut taxes primarily for the wealthy, some of the most drastic sections of the bill modify tax incentives and other public funding for clean energy and emissions reductions.


Many of the incentives to deploy clean energy that were created or enhanced under the IRA will be phased out early or repealed altogether. Credits with accelerated phase-out schedules include the newly created 45Y clean electricity production tax credit, which will no longer support wind or solar projects after 2027, and the 45V credit for clean hydrogen production for which projects must now commence construction before Jan 1, 2028 (moved up from Jan 1, 2033).
Since the passage of the reconciliation package, there has been active litigation on several provisions, including an order from a federal district court to vacate IRS guidance that would have prohibited certain wind and solar projects from securing safe harbor. The table below provides a detailed breakdown of key changes to major tax credits between the original IRA, the draft that moved through Committees in the House, and the final text that was passed by the Senate and signed into law.
Major Tax Credit Changes in the Reconciliation Law
How FEOC Restrictions Threaten Clean Energy Supply Chains
Many clean energy tax credits include ambiguous language restricting projects connected to FEOC, complicating supply chains and creating new problems for developers of clean energy projects. The law also introduces a complex matrix of new definitions, such as "Prohibited Foreign Entities," which includes both "Specified Foreign Entities" and "Foreign-Influenced Entities."
The FEOC restrictions embedded in the reconciliation bill represent a seismic shift for clean energy developers. These new rules, designed to limit the influence of Covered Nations (China, Russia, North Korea, and Iran), will disqualify projects from receiving tax credits if they source components, minerals, or intellectual property from entities tied to these nations. In other instances, the partial ownership or investment of an entity with financial ties to a Prohibited Foreign Entity may also disqualify a project from qualifying for tax credits.
This FEOC language matters for developers and investors because of the resulting global supply chain disruptions, investment uncertainty, and compliance burdens. The clean energy sector is deeply reliant on global supply chains, especially for solar panels, batteries, and wind components, industries where China currently dominates. The IRA intended to counter this by moving the manufacturing and production of these supply chains to the US. Project developers must now thoroughly review their supply chains and capital providers, and may need to quickly pivot to compliant resources.
In February 2026, the IRS released interim guidance on the FEOC provisions to provide safe harbor guidance for clean energy manufacturing, investment, and production credits to help taxpayers gauge whether material assistance was provided by a prohibited foreign entity.
Other Major Rollbacks to the IRA
Beyond clean energy tax credits, the reconciliation package also repeals and rescinds many other IRA provisions. This includes a full rescission of all unobligated IRA appropriated balances at the Department of Energy's Loan Programs Office, and several other programs, including:
- The Tribal Energy Loan Guarantee Program
- Greenhouse Gas Reduction Fund
- Transmission Facility Financing
A complete list of rescissions of energy-related funding is outlined in Sections 60001-60024 and 50402 of the law. These rescissions represent tens of billions of dollars in lost climate investments made under the IRA, which would have provided funds to state, local, and Tribal governments, federal agencies, non-profits, and commercial project developers to reduce emissions and update critical infrastructure.
What Can Project Developers and Other Companies Do?
Developers will need to act quickly to meet updated commence construction and place into service requirements, though circumstances are technology specific (e.g., safe harbor updates to 48E and 45Y). Tax credits generally have advanced commence construction and operational deadlines, resulting in a strong first-movers advantage. Companies should also review their supply chains and revise equipment and material procurement sourcing plans as necessary to address restrictions presented in the reconciliation bill.
An executive order from President Donald Trump issued on July 7 will further complicate how companies proceed. In the EO, the President directs his administration to "strictly enforce the termination of […] 45Y and 48E […] for wind and solar facilities." The Administration will likely issue extremely strict interpretations of "commence construction" clauses and FEOC requirements in forthcoming tax credit guidance issued by the Treasury Department, though these moves are quite likely to face litigation.
The new restrictions being proposed by the Administration, including specific details on FEOC, qualified equipment, commence construction, and other reporting requirements, will require additional guidance from the IRS and provide an opportunity for engagement through public comment. It is important that impacted companies weigh in during these public comment periods, not only to help inform and influence the final rules issued by the Administration, but also to build an administrative record that could support litigation efforts to strike down the final rules.
Staying Ahead of Policy Changes
Given the rapidly shifting landscape of energy policy, it's paramount that companies stay abreast of the latest changes and dedicate resources to understanding how they may be affected. Policy professionals, including the experts at Relae (formerly Carbon Direct), can support organizations as they engage in the regulatory process, anticipate and prepare for new legislation, and navigate the requirements to access essential tax credits and incentives. Even under new constraints, expert guidance can help maximize impact and minimize disruption.
Frequently Asked Questions
How does the reconciliation bill change the timelines for major clean energy tax credits?
Most clean energy tax credits saw their windows shortened relative to the original IRA:
- The 45Y and 48E credits now terminate entirely for wind and solar facilities placed in service after December 31, 2027, with a separate phase-down (75% in 2034, 50% in 2035, 0% after) for other technologies.
- The 45V clean hydrogen credit's "commence construction" deadline moved from December 31, 2032 to December 31, 2027.
- The 45Z clean fuel credit now ends on December 31, 2029 (versus 2027 in the original IRA, but bonuses for SAF have been removed and new emissions-calculation methods favor corn ethanol).
- Notably, the 45Q carbon capture credit saw little change and retained transferability, with credit values for enhanced oil recovery and utilization raised to match secure geological storage.
What are the FEOC restrictions, and why do they matter so much for developers?
FEOC ("Foreign Entity of Concern") restrictions disqualify projects from tax credits if they source components, minerals, or intellectual property from entities tied to China, Russia, North Korea, or Iran. Even partial ownership or investment ties to a "Prohibited Foreign Entity" can trigger disqualification. The definitions are complex and still being clarified through IRS guidance, meaning developers need to review supply chains and capital providers carefully and may need to pivot to compliant sourcing.
What should project developers do now in response to these changes?
Developers should move quickly to meet the earlier "commence construction" and "placed in service" deadlines, since credits now benefit early actors. This includes reviewing and potentially restructuring supply chains and procurement plans to address FEOC restrictions, and closely monitoring forthcoming IRS/Treasury guidance.
How to Fix Load Forecasting for the AI Era
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]
AI Scale and Climate Commitments: A 2026 Outlook
The AI and Climate Execution Challenge
Data center energy capacity in the US is projected to increase from 25 GW to 120 GW by 2030—a fivefold increase. Hyperscalers are projected to invest $7 trillion globally in data center infrastructure through 2030, with approximately $2.8 trillion invested in the US.
While 2025 was defined by a 'scale at all costs' scramble for compute, in 2026, the new mandate is responsible scale: reconciling voracious power demands with aggressive net-zero commitments and rising energy costs.
Grid constraints determine the geography and velocity of growth, forcing companies into complex trade-offs between speed-to-market and “clean, firm” power, which can take years to develop. Evolving carbon accounting rules are shifting procurement strategies and infrastructure choices at this trillion-dollar scale, creating a “carbon debt”—embodied emissions that will stay on the books for decades. In 2026, the competitive advantage likely belongs to those who integrate power, hardware, and climate strategy from day one.
Powering AI: Grid Reliability, Constraints, and Interconnection
Grid infrastructure faces reliability challenges from aging systems and capacity constraints. Interconnection queues stretch three to five years for renewables, while large electrical load interconnection lacks consistent standards.
Federal Regulatory Response
The federal government is moving to standardize these processes, with a critical decision point in 2026. For companies planning data center deployments in 2026, understanding these regulatory shifts is likely essential to realistic timeline and site selection planning.
On October 30, 2025, the US Department of Energy (DOE) leveraged Section 403(a) of the DOE Organization Act to direct the Federal Energy Regulatory Commission (FERC) to issue a rulemaking to “ensure efficient, timely, and non-discriminatory load interconnections” for large (>20 MW) electrical loads.
By April 30, 2026, FERC is expected to issue a final rule on large electrical load interconnections for grid operators, providing federal regulations for approval pathways, timelines, and rates.
Public comments on DOE’s advanced notice of proposed rulemaking were due on December 5, 2025, and grid operators, utilities, NGOs, and customers submitted over 150 comments reflecting a wide range of perspectives.
While federal standardization should reduce procedural uncertainty, it doesn't create new grid capacity. Even with clearer approval pathways, the underlying supply-demand mismatch remains a primary gating factor for growth.
Bridging the Supply-Demand Gap
Data center energy demand is surging, but new clean electricity generation takes years to build. This mismatch between accelerating demand and slow-building supply is forcing the industry to pursue solutions on two timelines: near-term load flexibility strategies that unlock existing capacity, and long-term generation investments that build new power supply.
Load Flexibility: Near-Term Grid Access
Load flexibility is emerging as a possible path to faster grid connection. Oracle, NVIDIA, Emerald AI, and Salt River Project's joint research demonstrated 25% power reduction during peak hours through workload tiering. The demonstration shows that if data centers reduce consumption during peak times (roughly 1% of the year), it unlocks 126 GW of currently constrained capacity that could be available now.
Large power loads increasingly face incentives or mandates to demonstrate flexibility as part of interconnection agreements, making this an access requirement, not an optional efficiency measure. For example, Senate Bill 6 in Texas mandates that data centers and other large loads must reduce their consumption during certain grid peak times. Many other state legislatures are passing legislation that will impact data centers.
Storage has shifted from smoothing renewables to enabling multiple strategies: making intermittent renewables firmer, providing grid reliability services, and supporting 24/7 matching. Storage may emerge as a solution to allow data centers to reduce grid consumption during peak hours while maintaining operations.
Relae helps clients design load flexibility strategies under evolving regulatory frameworks: evaluating behind-the-meter generation options, sizing storage for peak reduction scenarios, and structuring interconnection configurations that preserve optionality across accounting methodologies.
Clean Firm Power: Long-Term Generation
Hyperscalers remain committed to clean, firm generation that’s reliable: power that's both low-carbon and dispatchable 24/7. Natural gas with carbon capture and storage (CCS) is emerging as a critical bridge technology. Google's 400 MW CCS power agreement with Broadwing, expected online in 2029, demonstrates commercial demand at scale.
In our analysis evaluating CCS pathways, commercial viability depends on rigorous assessment of permitting timelines, capital and operating costs, storage geology, vendor compatibility, and 45Q tax credit optimization. Execution has been most prevalent where technology intersects with regulatory approval and storage access.
Hyperscalers are also investing across geothermal, nuclear, including Small Modular Reactors (SMR), hydrogen, and fusion. Long-duration energy storage has also been an area of focus. Each has different risk profiles and opportunities across technical maturity, permitting, commercial viability, emission accounting methodology, dispatchability, and political support.
Evaluating these pathways requires multi-dimensional frameworks. Each technology faces distinct challenges: SMRs struggle with execution complexity, geothermal with extended development periods, and hydrogen with production-dependent carbon intensity. Tax credit eligibility (particularly 45Q for CCS and 45V for hydrogen) significantly impacts project economics.
Both power generation and data center Infrastructure site selection require integrating environmental and social vulnerability data to avoid community conflicts that delay or stop projects.
Power Accounting Rules Determine Clean Energy Procurement
The Greenhouse Gas (GHG) Protocol extended the public consultation period for proposed scope 2 guidance changes to January 31, 2026. The results will determine clean energy procurement strategies and the carbon value of load flexibility for the next decade.
The proposed shift in electricity emissions accounting could increase clean energy procurement costs for buyers. The accounting methodological debates matter for hyperscalers: 24×7 energy matching versus carbon matching. The issues of deliverability (being located in the same grid region) and additionality (being new, rather than repurposed, generation) are also hotly debated.
These different frameworks strongly influence whether natural gas with CCS, nuclear, geothermal, or battery-backed renewables are considered optimal for a site, and whether load flexibility has carbon value.
Companies need to model scenarios across advanced power emissions methodologies, evaluating portfolio costs and carbon performance before final standards are published in 2027. Companies are also signing forward renewable energy certificate contracts (RECs) and structuring power purchase agreements (PPAs) now to preserve optionality across scenarios.
AI Infrastructure Emissions at Scale
Data center construction creates substantial scope 3 emissions, and their relative importance depends on grid carbon intensity. For facilities powered by average-carbon grids, scope 2 operational emissions dominate. But for data centers powered by very low-carbon electricity (renewables or nuclear), scope 3 embodied emissions can represent 40% of total lifetime greenhouse gas emissions.
In AI data centers, IT equipment drives the majority of embodied emissions. Chips and memory account for 67%, followed by structural materials at 17%, with server power supplies, aluminum, and other components comprising the final 16%.
Direct procurement of low-carbon materials faces constraints: limited supply, geographic concentration, and contracting complexity. Environmental Attribute Credits (EACs) provide an interim pathway by decoupling environmental benefits from physical materials, but require rigorous quality standards and verification to ensure real emissions reductions.
Our high-quality EAC criteria, developed with Microsoft, establish standards that separate market-making from greenwashing. Levelized Cost of Carbon Abatement frameworks make materials decisions comparable to power decisions, treating infrastructure decarbonization as portfolio optimization, not separate workstreams.
Carbon Removal: Addressing Residual Emissions
Complete supply chain decarbonization by 2030 isn't feasible. Despite aggressive efforts to procure clean power and reduce construction emissions, residual emissions will remain significant. For hyperscalers with net-zero commitments, carbon dioxide removal (CDR) has shifted from an optional component to a structural necessity. Microsoft remains the world's largest CDR buyer, and Google increased purchases 14-fold from 2023 to 2024.
CDR credit quality varies widely. Companies must apply science-based principles to evaluate credits. Our Criteria for High-Quality CDR, developed in collaboration with Microsoft, establishes six science-based principles for evaluating credits—critical as emerging hyperscaler and other corporate demand high-integrity supply.
AI and Climate: Looking Ahead
The window for strategic maneuvering is narrow. The AI infrastructure buildout is happening now, and the decisions made in 2026 will impact a company’s cost structure and carbon profile for years.
Companies treating power, infrastructure, and decarbonization as separate workstreams will face compounding constraints. The winners of the AI era will be those who integrate power, infrastructure, and carbon strategy into a single, cohesive system.
The New Geothermal Energy: How EGS Unlocks Clean, Firm Power at Scale
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.



