What is So Hard About Behind-The-Meter Power For Datacenters? Part 1

Live updates

1. 5d ago

   Behind-The-Meter Power Reshapes AI Data Center Infrastructure

   Data center developers are aggressively adopting grid-independent behind-the-meter power solutions to support artificial intelligence and hyperscale facilities, fundamentally altering energy economics and long-term cost structures. Research shows announced projects targeting behind-the-meter energy have surpassed 18 gigawatts, spanning multiple technology categories and regulatory hurdles. This shift moves operators away from traditional grid reliance toward integrated energy models. Meanwhile, analysts tracking global facility expansion note that American developers face growing local opposition, offering lessons from international markets like Ireland and New Zealand.

   Why it matters

   Traditional grid connections face median wait times ranging from 4.5 to 12 years, forcing developers to seek alternative power sources. On-site generation helps manage unprecedented power demand and rapid load fluctuations driven by artificial intelligence workloads. However, transitioning to standalone campus power introduces complex economic calculations regarding capital expenditures and long-term asset stranding risks.

   What is confirmed

   • Announced projects targeting grid-independent power for artificial intelligence and hyperscale facilities have reached over 18 gigawatts.
   • Behind-the-meter power strategies are altering energy economics, grid dependency, and long-term cost structures for artificial intelligence data centers.
   • Authors tracking global data center development point to international resistance patterns in places like Ireland and New Zealand to inform the domestic movement against data centers.

   What to watch next

   • Updates on the deployment status of the 18+ gigawatts of announced behind-the-meter power projects
   • Regulatory policy shifts regarding grid-independent power integration for hyperscale facilities
   Sources used for this update (5)

   1. www.chicagotribune.com — Chicago News - Chicago Tribune
   2. www.wired.com — What the US Movement Against Data Centers Can Learn From Everywhere Else
   3. www.jaredwatkins.com — Behind-the-Meter Power for Data Centers - The Infinite Unknown
   4. www.computeforecast.com — Behind-the-Meter Power Economics in AI Data Centers
   5. www.computeforecast.com — How Behind-the-Meter Power Is Reshaping AI Infrastructure
   confidence 100%
2. 8d ago

   Data Centers Push Behind-the-Meter to Escape Grid Bottlenecks

   Data center developers are increasingly turning to behind-the-meter microgrids scaling past 2 GW to bypass traditional grid hookups that face median wait times of 4.5 years, stretching to 8 to 12 years for large-scale operations. United States interconnection queues currently hold over 2,600 GW of projects. To manage unprecedented power demand and millisecond-level power fluctuations driven by artificial intelligence, operators are deploying on-site natural gas turbines, fuel cells, and multi-layered energy storage systems. However, this shift faces friction, including local community backlash over accompanying natural gas power plants and choices regarding capital costs and stranding risk.

   Why it matters

   Traditional grid connections create severe structural bottlenecks for massive artificial intelligence workloads, forcing developers to look for alternative power solutions. On-site generation helps deliver reliable and scalable power faster while addressing long interconnection delays. This approach shifts generation burdens onto private microgrids, altering energy management spending and drawing scrutiny from local residents.

   What is confirmed

   • US interconnection queues hold over 2,600 GW of projects with median wait times of 4.5 years, stretching to 8 to 12 years for large-scale data centers.
   • Data center behind-the-meter microgrids are scaling past 2 GW.
   • A proposed data center in Taylor has faced pushback because of an accompanying natural gas power plant.

   Still unconfirmed

   • Data center backlash is set to sway midterm elections after growing from local planning meetings into a political flashpoint.

   What to watch next

   • Evaluation outcomes from the Public Utility Commission of Texas regarding applications to net-meter large AI data centers behind existing plants.
   • Further decisions by developers on balancing capital costs for natural gas turbines and fuel cells versus waiting for traditional grid connections.
   Sources used for this update (9)

   1. www.forbes.com — The Rise Of AI-Ready Microgrids: What's Driving A New Wave Of Behind-The-Meter Innovation
   2. spectrumlocalnews.com — A look at data centers' behind-the-meter approach to get power
   3. jisenergy.com — Data Centers Pivot to Gigawatt-Scale BTM Microgrids
   4. www.strategicrevenueinsights.com — Behind-the-Meter Power: Build the Bridge, Keep the Grid
   5. datacenterpost.com — Bypassing the Grid: What Data Center Operators Can Learn From ...
   6. timestech.in — Data Center Power Management Market to Reach $107.06 Billion ...
   7. pv-magazine-usa.com — How energy storage can meet the power demands of AI data centers
   8. www.idahostatejournal.com — Data center backlash set to sway midterm elections
   9. finance.yahoo.com — Liberty Energy (LBRT) Q3 2026 Earnings Call Transcript
   confidence 100%
3. 11d ago

   Texas PUC Reviews Co-Location Deals as AI Power Demands Surge

   The Public Utility Commission of Texas is evaluating three applications to net-meter large AI data centers behind existing power plants, including nuclear, solar, and wind facilities. This shift toward behind-the-meter power allows developers to bypass grid hookups that can take 5 to 10 years. Current projects range from xAI's 1.2 GW installation to Venture Global's 11.5 GW capacity. Developers are increasingly deploying onsite gas turbines, fuel cells, and batteries to maintain timelines and shield ratepayers from costs associated with traditional grid connections.

   Why it matters

   Rapid AI growth has created massive electricity requirements that exceed current utility capacity. Onsite generation reduces reliance on aging infrastructure and avoids decade-long wait times for power access.

   What is confirmed

   • Grid hookups for data centers can take between 5 and 10 years.
   • AI data centers are utilizing onsite gas turbines, fuel cells, and batteries for power.

   Still unconfirmed

   • The Public Utility Commission of Texas considered three Senate Bill 6 applications on October 15, 2026, to net-meter AI data centers at a West Texas solar project, the Comanche Peak nuclear operator, and a Haskell County wind farm.
   • xAI has a 1.2 GW onsite power project.
   • Venture Global has an 11.5 GW onsite power project.

   What to watch next

   • The Texas PUC decision on the three Senate Bill 6 co-location applications.
   • Updates on the gas turbine backlog for AI data centers.
   • Confirmation of 2027 capacity acceleration projections.
   Sources used for this update (8)

   1. jen.jiji.com — Protests, Trump to AdnKronos: "In Italy and France it's the fault of immigration, I said so"
   2. www.thetechedvocate.org — Xbox Cloud Gaming Price Hike: The Unspoken Truth Behind Microsoft’s Latest Move
   3. pro.thestreet.com — The AI Boom Has an Electric Bill. Here’s Who Gets Paid.
   4. safeguard.sh — Gas turbines and on-site power for AI data centres explained
   5. www.texaspowercost.com — Texas PUC Weighs 3 Data-Center Co-Location Deals (Oct. 2026 ...
   6. newsletter.semianalysis.com — Datacenters 2027: The Year of Acceleration, +50GW in a Single ...
   7. www.strategicrevenueinsights.com — Data Center Onsite Power: Why AI Builds Its Own Plants
   8. www.texastechnologydigest.com — Texas 5,000 Acre AI Data Center Site Fast Tracks 500 MW ...
   confidence 90%
4. 14d ago

   Data Centers Turn to Behind-the-Meter Power Amid Grid Delays

   Data center developers are increasingly bypassing traditional utility connections to establish onsite power generation and microgrids, driven by the explosive rise of artificial intelligence and grid access timelines taking a decade or more. Seven states enacted legislation in 2025 enabling facilities to source their own power and shield ratepayers from grid costs. Meanwhile, Tesla secured a 260-unit, 1-gigawatt-hour Megapack order from Taiwan's Apollo Power for enterprise grid and AI data center projects, while developers tap high-yield markets to fund standalone power solutions.

   Why it matters

   Grid constraints and electricity affordability concerns are intensifying regulatory scrutiny, potentially slowing new grid-connected data center developments and extending implementation schedules. Because connecting to the public grid can take a decade or longer, onsite power is shifting from a temporary bridge to a permanent microgrid solution.

   What is confirmed

   • Seven states enacted behind-the-meter data centers legislation in 2025, allowing facilities to source their own power and avoid grid costs for ratepayers.
   • Tesla landed a 260 unit, 1GWh Megapack order from Taiwan's Apollo Power for AI data center and enterprise grid projects.

   Still unconfirmed

   • Grid constraints and electricity affordability concerns could slow the pace of new grid-connected data center development and extend implementation schedules, particularly for projects awaiting power and interconnection approvals.

   What to watch next

   • Whether additional states will enact legislation allowing data centers to source their own behind-the-meter power.
   • The pace at which data center developers transition from temporary onsite power bridges to permanent microgrids.
   Sources used for this update (10)

   1. www.teslarati.com — Taiwan power firm bets big on Tesla Megapacks as AI data center thirst grows
   2. www.multistate.us — State Lawmakers Look to "Behind-the-Meter" Policies for Data Centers
   3. www.datacenterfrontier.com — Behind-the-Meter Power Offers Control, Flexibility and Speed ...
   4. www.publicpower.org — Fitch Sees Data Center Developers Pursuing Behind-the-Meter ...
   5. octus.com — ‘Power Islands’ Tap High-Yield Market for AI | Octus
   6. elisawood.substack.com — Data Centers Are Rethinking Bridge to Grid. Here’s Why.
   7. jen.jiji.com — Germany, Saxony-Anhalt Parliament votes to stop sanctions on Moscow and aid to Kyiv
   8. jen.jiji.com — Totti and Vieri receive Euro-Mediterranean Excellence Award: "Roma-Real Madrid? Gasperini will play to win"
   9. haas.berkeley.edu — News Archives - Energy Institute - UC Berkeley | Haas School of Business
   10. temaetfs.com — The Data Center Boom and the Race for Power - Tema ETFs
   confidence 90%
5. 19d ago

   Data centers turn to on-site power to bypass grid bottlenecks

   Data center operators are deploying behind-the-meter power generation to ensure immediate resilience and supplement constrained grid connections. Current deployments include fuel cells and gas buildouts. Regulatory uncertainty regarding private power and grid access remains a challenge. Enverus Intelligence Research forecasts 25.5 GW of new U.S. industrial demand will be served behind the meter through 2030, with data centers accounting for 88%.

   Why it matters

   The rapid growth of AI workloads is driving data center operators to seek alternative power solutions. Grid capacity bottlenecks and the need for grid modernization are significant challenges. The trend towards on-site power generation is expected to continue, but regulatory clarity is needed.

   What is confirmed

   • Data centers account for 88% of 25.5 GW of new U.S. industrial demand served behind the meter through 2030.
   • The growth of AI workloads is driving data center operators to seek alternative power solutions.

   Still unconfirmed

   • AI companies are paying a premium to go off the grid

   What to watch next

   • Regulatory decisions on private power and grid access
   • Grid modernization efforts
   • Data center operators' on-site power generation deployments
   Sources used for this update (4)

   1. www.power-eng.com — Data centers are skipping the grid queue — and turning to on ...
   2. energyathaas.wordpress.com — Buying Time, Burning Gas – Energy Institute Blog
   3. heatmap.news — Dynamic Electricity Pricing Cuts Both Ways for EVs
   4. energysecurityfreedom.substack.com — Behind-the-Meter Generation for Data Centers May Not Be Such ...
   confidence 90%
6. 25d ago

   AI Growth Drives Data Centers Toward Private Behind-the-Meter Power

   Data center operators are deploying behind-the-meter power generation to bypass grid capacity bottlenecks caused by AI workloads. These on-site solutions provide bridge power and supplement constrained grid connections to ensure immediate resilience. Current deployments include fuel cells and gas buildouts, though operators face regulatory uncertainty regarding the boundary between private power and grid access. While some regions like Kazakhstan are accelerating digital infrastructure via projects like Data Center Valley, global developers struggle with LBNL queue delays and the need for grid modernization to deliver reliable capacity.

   Why it matters

   Public grids are struggling to keep pace with the rapid energy demands of artificial intelligence. This creates a bottleneck that forces developers to seek power independence to avoid long interconnection queues. The shift raises questions about climate accountability and how utilities will charge for continued grid access.

   What is confirmed

   • AI-driven growth is creating a power bottleneck that requires grid modernization for reliable capacity.
   • Data center developers are using on-site generation to supplement constrained grid connections.

   Still unconfirmed

   • LBNL queue delays are impacting behind-the-meter generation timelines for 2026 through 2028.

   What to watch next

   • Regulatory rulings on payment structures for data centers with private power and grid access
   • Progress updates on the Data Center Valley megaproject in Kazakhstan
   • Deployment timelines for the MV-to-800 VDC interface
   Sources used for this update (8)

   1. jen.jiji.com — Kazakhstan ready for large-scale AI rollout, PM Bektenov says
   2. www.datacenterfrontier.com — Data Center Private Power: Who Regulates Behind-the-Meter ...
   3. www.mintz.com — Grid Capacity Is Becoming the Data Center Bottleneck: Inside ...
   4. www.insidedeeptech.com — Behind-the-Meter Generation for AI Data Centers: A Full Guide ...
   5. www.uscardforum.com — BTM 与 AIDC 延期：SemiAnalysis 与 The Information 两篇文章解读
   6. www.emersonautomationexperts.com — Data Center Power Challenge: Managing Demand and Supply
   7. lowdown.today — behind-the-meter | Data Centres: Boom and Backlash
   8. datacenters.economictimes.indiatimes.com — Data centres eye on-site power for AI era resilience
   confidence 85%
7. 29d ago

   Data centers turn to behind-the-meter power amid grid strain

   Data centers are increasingly adopting behind-the-meter power solutions to meet their growing energy demands, driven by the rapid growth of AI and the need for reliable power. This shift is accelerated by the strain on public grids and regulatory considerations. Behind-the-meter solutions, including on-site power generation and energy storage, are becoming crucial for data center deployment.

   Why it matters

   The expansion of data centers is creating tensions within local communities due to concerns about the industrial footprint and strain on public grids. As a result, data center operators are seeking self-sufficient power solutions to ensure grid stability and meet their energy needs. This trend is driven by the increasing demand for AI and the limitations of the existing grid infrastructure.

   What is confirmed

   • AI data center electrical equipment crisis: new industry analysis finds AI training load patterns accelerate transformer aging two to three times faster than normal
   • Behind-the-meter solutions are accelerating data center deployment
   • Fuel cells are being examined as a behind-the-meter power solution for data centers, offering modular deployment and continuous on-site generation
   • Data center self-generation is transitioning from an operational preference to a regulatory mandate as grid operators protect utility ratepayers

   Still unconfirmed

   • Tesla Full Self-Driving release in Europe is set to be delayed by at least a few months

   What to watch next

   • Goldman Sachs' forecast for global behind-the-meter power generation capacity for data centers
   • Regulatory decisions on data center power generation and grid upgrades
   • Advancements in behind-the-meter power solutions and energy storage technologies
   Sources used for this update (14)

   1. www.teslarati.com — Tesla Full Self-Driving release in the EU gets delayed
   2. arstechnica.com — NASA has a Dragon dilemma, and there appear to be no good answers | Page 2
   3. www.fticonsulting.com — BTM Solutions are Accelerating Data Center Deployment | FTI
   4. www.techtimes.com — AI Data Centers Are Burning Worn-Out Grid Equipment and ...
   5. www.financialcontent.com — KBRA Releases Research – Fuel Cells as a Behind-the-Meter ...
   6. community.spiceworks.com — Snap! -- Warp Drives, Radioactive Crabs, Doomsday Delay - Spiceworks Originals - Spiceworks Community
   7. www.briefs.co — Accenture Pops After Big Q4 Bookings, AI Partnerships Spotlight Growth Push
   8. laist.com — Lifeline for local news
   9. computelaw.blog — On site and behind-the-meter power generation for AI data centers
   10. www.zerohedge.com — Bring Your Own Power Plant: Goldman Now Sees Behind-The-Meter ...
   11. marketchameleon.com — Vertiv (VRT) Targets the New AI Data Center Constraint ...
   12. www.andradesafe.com.br — Gas turbines for AI data centers: on-site power and BTM
   confidence 85%
8. 36d ago

   Community Opposition Impacts Data Center Deployment

   Data center expansion is creating political divisions within American communities, leading to new alliances that cross traditional party lines. This backlash coincides with ongoing efforts by operators to implement behind-the-meter power to avoid straining public grids. While operators prioritize grid stability, local populations remain concerned about the industrial footprint of these facilities. The tension highlights a growing conflict between the infrastructure needs of AI and the political will of the communities hosting the hardware.

   Why it matters

   Operators use behind-the-meter power to bypass public utility limitations. This strategy aims to reduce grid strain but often triggers local disputes over environmental and industrial impacts.

   Still unconfirmed

   • Debates over data center placement have become a political fault line in American communities

   What to watch next

   • Legislative changes to zoning laws for data centers
   • Public hearings on behind-the-meter power permits
   Sources used for this update (2)

   1. www.lpm.org — The data center backlash is reshaping American politics — one community at a time
   2. jen.jiji.com — Volleyball European Championship, today quarterfinals Italy-Finlandia: schedule and where to watch it live
   confidence 60%
9. 40d ago

   No New Data on Behind-the-Meter Power for Data Centers

   There are no new corroborated updates regarding the technical or environmental challenges of behind-the-meter power for data centers. Recent reports focus on unrelated sporting events and general consumer electricity billing. The core issues remain the risk of noise pollution, carbon-dioxide emissions, and the inaccuracy of site-specific Water Usage Effectiveness metrics in capturing the full environmental footprint of AI power demands. Operators continue to seek these power configurations to reduce strain on public grids, though public debate persists over the industrial impact of these build-outs.

   Why it matters

   AI data center expansion creates a conflict between operational efficiency and environmental sustainability. Behind-the-meter power allows operators to bypass some grid constraints but introduces localized pollution risks. The shift in water consumption upstream complicates how the industry measures its total ecological impact.

   What to watch next

   • Technical reports on carbon-dioxide emission levels from behind-the-meter sites
   • Updated Water Usage Effectiveness metrics for AI infrastructure
   • Policy changes regarding industrial noise pollution for data centers
   Sources used for this update (3)

   1. jen.jiji.com — Volleyball European Championship, today Round of 16 Italy-Denmark - Live
   2. jen.jiji.com — Juve-Atalanta, Nerazzurri protests for missed penalty on Rowe: what happened
   3. www.consumerreports.org — Why Is Your Electric Bill So High? How to Read the Fine Print.
   confidence 100%
10. 41d ago

    AI Data Center Expansion Faces Environmental and Public Scrutiny

    Data center build-outs for AI are accelerating, triggering public debate over their industrial impact. While operators pursue behind-the-meter power to reduce grid strain, the projects pose risks of noise pollution and excessive carbon-dioxide emissions depending on their cooling and power configurations. Previous research indicates that AI power demands shift water consumption upstream, meaning site-specific Water Usage Effectiveness metrics do not capture the full environmental footprint. Current discourse often focuses on these downsides, though some argue the prevailing public narrative is disconnected from the actual facts of the development.

    Why it matters

    Operators are targeting High Plains farmland for these expansions, which raises concerns regarding water rights and land use. The shift toward behind-the-meter power aims to bypass traditional utility constraints. This tension highlights a conflict between rapid AI infrastructure growth and local resource sustainability.

    Still unconfirmed

    • Data centers pose risks of noise pollution and excessive carbon-dioxide emissions depending on how they are powered and cooled.

    What to watch next

    • Publication of updated Water Usage Effectiveness metrics that include upstream consumption
    • Regulatory decisions on High Plains land use for data center zoning
    Sources used for this update (4)

    1. note.com — The hidden side of the 'increasingly severe' security environment. Its true content and the boundary of our daily lives
    2. jen.jiji.com — Sandro Mazzola, the last (unpublished) interview: "Inter is a mother to me"
    3. jen.jiji.com — Trump's axe on journalists: no White House access for CNN, Ms Now, and Politico. "Assault on rights, we'll continue our work"
    4. www.theatlantic.com — The Data-Center Debate Is Divorced From the Facts
    confidence 80%
11. 42d ago

    AI Data Center Expansion Targets High Plains Farmland and Upstream Water

    Data center operators are targeting High Plains farmland for expansion, raising concerns over land use, infrastructure costs, and water rights. While operators seek behind-the-meter power to bypass grid strain, new research from LBNL and Ceres indicates that AI power demands shift water consumption upstream. This suggests that site-specific Water Usage Effectiveness (WUE) metrics fail to capture the full environmental impact of these facilities, as the power systems supporting them withdraw significant water resources before electricity reaches the data center.

    Why it matters

    Surging AI power needs are pushing US electricity use toward record highs through 2027. To maintain resilience, operators are exploring independent power sources like natural gas and steam turbines. These shifts impact rural development and regional resource management.

    Still unconfirmed

    • Data centers are targeting High Plains farmland, which raises questions regarding water rights, infrastructure costs, and rural development.
    • AI-era data centers rely on power systems that withdraw and consume water upstream.
    • Site Water Usage Effectiveness (WUE) metrics do not tell the full story of AI water consumption according to LBNL and Ceres.

    What to watch next

    • Evidence of specific water table depletion in the High Plains region due to data center power systems.
    • Regulatory decisions on land-use zoning for data centers on agricultural land.
    • Updated WUE standards that incorporate upstream water withdrawal.
    Sources used for this update (6)

    1. jen.jiji.com — New malaria alert in Germany, one dead and one infected: both live near Frankfurt airport
    2. jen.jiji.com — The CSR and Social Innovation Exhibition returns to Milan from 6 to 8 October (2)
    3. jen.jiji.com — Ukraine, wine production in annexed Crimea drops by a quarter due to fuel crisis
    4. jen.jiji.com — Teacher stabbed in Rome, the 13-year-old's plan: at school with diving knives and phone on helmet to film the ambush
    5. hpj.com — Farming for data—land, energy, output
    6. www.techtarget.com — A shadow water problem: AI's power demand shifts consumption upstream
    confidence 80%
12. 44d ago

    PJM Proposes New Rules for Large Data Center Loads

    PJM Interconnection has proposed ride-through rules for large data center loads, aiming to address technical challenges as artificial intelligence facilities strain electricity grids. Federal projections indicate that surging power consumption by artificial intelligence data centers is pushing United States electricity use to record highs through 2026 and 2027. This rapid growth threatens global electricity systems and grid resilience. To secure reliable electricity independent of public transmission grids, operators are exploring behind-the-meter power solutions, including natural gas, clean energy supplies, industrial boilers, and steam turbines.

    Why it matters

    Data centers require massive, uninterrupted energy supplies that public transmission grids struggle to provide reliably. Behind-the-meter solutions help operators bypass grid congestion but introduce complex regulatory and technical hurdles for regional transmission organizations. Grid operators are responding with new operational requirements to maintain stability when massive single-site loads connect or disconnect.

    What is confirmed

    • PJM has proposed ride-through rules for large data center loads.

    What to watch next

    • Final approval or modification of PJM's proposed ride-through rules by regulators
    • Adoption of similar grid reliability requirements by other regional transmission organizations
    Sources used for this update (7)

    1. jen.jiji.com — UK, Burnham's father dies, premier cancels engagements for two days
    2. jen.jiji.com — NATO fighters shoot down drone in Lithuania. Trump: "Stop raids on refineries, Ukraine and Russia agree"
    3. heatmap.news — It’s Electric to Install 700 Curbside EV Chargers in New York City
    4. jen.jiji.com — "We will organize the deserters", Virginia Libero's viral message: the words of the young dem leader
    5. www.techtarget.com — PJM proposes ride-through rules for large data center loads
    6. jen.jiji.com — De Meo (Fi): 'Lack of true common defense policy. Extend mutual defense clause to hybrid attacks'
    7. jen.jiji.com — Veneto, Coldiretti-Philip Morris Italia: 55.8 million euros in value and over 9,100 jobs (2)
    confidence 100%
13. 48d ago

    AI Power Demands Surge Amid Grid Resilience Warnings

    Artificial intelligence data centers are driving up electricity use across the United States, pushing power consumption to record highs for 2026 and 2027 according to federal projections. The rapid growth of these facilities poses a rising threat to global electricity systems and grid resilience. To meet massive energy needs, data center operators are increasingly exploring behind-the-meter power solutions, including natural gas and clean energy supplies, as well as industrial boilers and steam turbines to secure reliable electricity independent of public transmission grids.

    Why it matters

    Data centers consume approximately 5 percent of electricity in the United States, significantly outpacing the global average. This surging demand has sparked international warnings from organizations like the United Nations Economic Commission for Europe regarding electrical system stability. Policymakers are responding by urging data centers to establish independent power supplies.

    What is confirmed

    • US power use is projected to beat record highs in 2026 and 2027 as artificial intelligence use surges, according to the EIA.
    • Around 5 percent of electricity in the United States is used for data centers, which is far more than the global average.
    • Artificial intelligence data centers pose a growing threat to electricity systems worldwide and threaten electricity system resilience.

    Still unconfirmed

    • Natural gas is winning the power race for artificial intelligence data centers.

    What to watch next

    • Updates on regional grid demand and supply growth through 2027 in ERCOT and PJM.
    • Further policy integration steps taken by the UNECE regarding electricity resilience.
    • Implementation details on European Union policies requiring data centers to build clean energy supplies.
    Sources used for this update (11)

    1. SemiAnalysis — What is So Hard About Behind-The-Meter Power For Datacenters? Part 1
    2. Reuters — US power use to beat record highs in 2026 and 2027 as AI use surges, EIA says
    3. politico.eu — Ribera says EU data centers should build their own clean energy supply
    4. WP Intelligence — Natural gas is winning AI’s power race
    5. UN News — AI datacentres pose growing threat to electricity systems worldwide
    6. The Motley Fool — Want to Make a Fortune From the AI Power Boom? Buy These 2 Industrial Stocks
    7. Unric — Datacentres threaten electricity system resilience, warns UNECE, calling for strategic policy integration
    8. Hart Energy — Can Natural Gas Feed the Artificial Intelligence Beast? —Misra
    9. RTO Insider — EIA: ERCOT Leads in Demand Growth, PJM in Supply Growth Through 2027
    10. POWER Magazine — AI Data Centers Turn to Industrial Boilers and Steam Turbines for Power
    11. Our World in Data — Around 5% of electricity in the United States is used for data centers — far more than the global average
    confidence 100%