The week starts with a finance term: backstop.

OpenAI is reportedly discussing a 10-GW Ohio data center campus with Nvidia support behind up to $250 billion in lease and construction debt. That is larger than most utility capital plans, tied to one customer, one campus, and one compute race. The power sector is now reading AI infrastructure through credit risk, turbine supply, transmission cost allocation, water use, emergency grid orders, and old nuclear sites coming back into public view.

The Lede

DATA CENTERS
Nvidia and OpenAI Discuss $250B Backstop for Ohio Data Center

OpenAI is in talks with Nvidia over a backstop of up to $250 billion to help finance a proposed 10-GW data center campus in Pike County, Ohio. The support would help OpenAI raise debt for lease and construction costs using Nvidia’s credit strength, while chip purchases would be discussed separately. The full campus could cost more than $500 billion, according to CNBC’s source.

The site was once a uranium-enrichment plant. SoftBank and SB Energy are developing the campus with the U.S. Department of Energy, and SoftBank is already a major OpenAI investor. Nvidia previously said it would invest up to $100 billion in OpenAI as part of a strategic partnership, though that was later reduced to a $30 billion contribution to OpenAI’s March funding round.

GridBrief Take: Nvidia’s exposure also shows how tight the AI infrastructure loop has become: the chipmaker may help finance the customer, the customer leases the campus, the campus needs power, and the power system becomes the limiting input.

Other Things to Check Out

• RealClearEnergy: Hyperscalers Are the Grid’s School Choice Moment
  A case for Consumer Regulated Electricity as the first real competitive escape hatch from the century-old utility monopoly model.

• Yahoo Finance: Data Centers Were Actually Making Electricity Cheaper
  New causal evidence finds data centers modestly lowered average retail rates from 2015 to 2024, with the warning that future supply constraints could reverse the effect.

• New York Times: Nuclear Power’s Palisades Test in Michigan
  Palisades remains the restart case to watch as Holtec tries to bring a retired U.S. nuclear plant back into service.

• Politico: Leaked Document Reveals White House Nuclear Waste Plan
  The White House is reportedly exploring a new push on nuclear waste storage and disposal, the unresolved policy fight beneath every serious nuclear expansion plan.

• World Nuclear News: UK Start-Up Nuclear Turbines Unveils Novel Reactor Concept
  A BAE Systems spinout is pairing a novel reactor design with high-temperature turbine technology aimed at lower-cost modular power.

ENHANCED GRID
FERC Eyes Incentives for Grid-Enhancing Technologies

FERC Chairman Laura Swett told the Senate Energy and Natural Resources Committee that the agency has created a task force on grid-enhancing technologies, or GETs. These include dynamic line ratings, advanced power flow controllers, and high-performance conductors. Swett said FERC cannot require utilities to use specific technologies, but it can direct transmission owners to analyze them and can consider GETs in transmission incentive proceedings.

The hearing also covered PJM governance, transmission competition, and FERC’s June show-cause orders on data centers and other large loads. Swett called PJM “probably performing the worst” among U.S. electricity markets and said FERC wants faster action on transmission services, interconnection studies, and large-load rules. Sen. Angus King pressed the utility incentive problem directly: utilities earn returns on capital spending, which can make cheaper grid optimization less attractive than traditional buildout.

GridBrief Take: GETs are the fastest test of whether regulators are serious about affordability. If a utility asks customers to fund new transmission while skipping cheaper line ratings, reconductoring, or flow-control options, FERC should make that choice visible and expensive to defend.

FERC
Ratepayer Advocates Push FERC on PJM Data Center Costs

Ratepayer advocates from Delaware, Illinois, Maryland, Ohio, and Pennsylvania told FERC that its large-load show-cause order for PJM does not adequately address transmission network upgrade costs. They argue FERC recognized cost-shift risks, then left unresolved the actual allocation rules that can place data center-driven transmission costs on existing customers.

The advocates want FERC to clarify that cost-recovery agreements are just and reasonable only when large-load customers pay the full cost of network upgrades needed to serve them. Pennsylvania’s advocate warned that cost shifts can occur through both PJM Regional Transmission Expansion Plan projects and supplemental projects. The filings also note that states have limited ability to sub-allocate costs to specific customers, especially where PJM zones cross state borders.

GridBrief Take: This is exactly where CRE and bring-your-own-power options become crucial. If large customers can finance dedicated supply and infrastructure outside the shared rate base, fewer cost-allocation fights need to be settled by FERC after the fact and rent-seeking becomes harder and harder.

NEW STUDY
Kansas Health Institute: Data Center Power Use Could More Than Double by 2030

The Kansas Health Institute says U.S. data centers consumed an estimated 183 TWh of electricity in 2024, more than 4% of national electricity use, and could reach 426 TWh by 2030. KHI also highlights water use: U.S. data centers consumed an estimated 17 billion gallons directly in 2023, with LBNL scenarios showing that use could double or quadruple by 2028.

The report frames data center development as a state and local policy problem involving economic development, land use, water planning, electricity reporting, ratepayer protection, and environmental assessment. KHI says current data center electricity is supplied 56% by fossil fuels, 22% by renewables, and 21% by nuclear. Local pushback is growing, while states are using reporting, environmental reviews, rate protections, and moratoria to slow or condition new projects.

GridBrief Take: KHI is right to put water, power, and local planning on the table. The mistake would be turning those questions into a presumption against growth. Data centers are usually quiet, high-investment neighbors with serious balance sheets and predictable demand.

THE WEATHER
DOE Emergency Order Covers 17-State SPP Region

DOE issued an emergency order allowing the Southwest Power Pool to dispatch specified generation and backup resources during a heat-driven reliability event. The order covers a 17-state region from North Dakota to Louisiana and runs through August 3. SPP requested the order after high demand, outages, reduced wind output, and limited imports tightened system conditions.

The order allows SPP to use backup generation as a last resort before declaring an Energy Emergency Alert 3 or during an EEA 3. Reuters reported that the affected region serves roughly 20 million people. SPP had narrowly avoided outages days earlier after a drop in imports and lower renewable output forced conservation appeals.

GridBrief Take: Emergency orders are useful in a pinch, but they are not a planning model.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• EPA clarifies islanded data center power rules. Reuters reports EPA says power facilities serving only data centers and not the broader grid are outside the Clean Air Act’s Acid Rain Program. That gives behind-the-meter projects another regulatory opening.

• AI infrastructure debt is getting more expensive. WSJ reports AI-related bond issuance reached $270 billion by early July, with Meta backing a $12.55 billion bond sale for an El Paso data center at a 7.5% yield. Compute buildout is moving into the credit markets quickly.

• Negative power prices are now an investment signal. Reuters notes Germany had 573 hours of negative wholesale prices in 2025, with similar issues showing up in California and Texas. The market is paying for flexibility, storage, and better demand timing.

• Four U.S. microreactor developers hit criticality. Antares, Valar Atomics, Deployable Energy, and Aalo Atomics reached zero-power fueled criticality under the DOE Reactor Pilot Program. Early milestone, but useful evidence that the nuclear startup world is no longer only pitch decks.

• Nuclear Turbines emerges from stealth in the UK. The BAE Systems spinout raised £15 million for a reactor concept using high-temperature turbine technology instead of a traditional steam-turbine system. The company is targeting behind-the-meter, microgrid, data center, and industrial uses.

• NRC approves Framatome TRISO fuel manufacturing. Framatome received approval to manufacture TRISO fuel at its Richland, Washington facility. Fuel availability is becoming one of the practical bottlenecks for advanced nuclear deployment.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

Bank of America put a number on the near-term supply problem: more than 230 GW of new generating capacity needed over five years, against 93 GW of expected accredited utility additions.

That gap is already changing behavior. Data center developers are moving toward on-site gas engines, Google is anchoring a 2.5-GW solar and storage project in Arkansas, and PJM’s market monitor is arguing that large loads should procure capacity outside the base auction. The common thread is practical, not philosophical: buyers with real load are looking for ways around slow supply.

The Lede

SUPPLY
BofA Sees a 100-GW Supply Gap

Bank of America forecasts more than 230 GW of new U.S. generating capacity will be needed over the next five years, while regulated utilities are expected to add only about 93 GW of accredited supply. Data centers alone could add roughly 125 GW of electric load and push overall electricity demand growth to a 4.1% CAGR from 2026 through 2030.

The supply gap is already pushing developers toward behind-the-meter generation. More than 7.5 GW of data center projects with on-site generation are under construction, with another 60 GW-plus in pre-construction. Large gas turbines are largely sold out through 2030, increasing demand for faster-deploying reciprocating gas engines.

GridBrief Take: BofA’s forecast is the speed-to-power case in one chart. Serious customers will not wait forever on interconnection queues, turbine backlogs, and utility planning cycles.

Other Things to Check Out

• National Review: Red-State Freedom Is Better for the Environment Than European Green Socialism
  Drew Bond argues that Europe’s top-down green industrial policy has delivered high costs and weak competitiveness.

• Reason: Trump’s 18-Month Deadline to Cut Electricity Prices in Half Has Arrived. Rates Are Up 18%.
  Jeff Luse uses the rate data to make the basic point: campaign slogans do not build generation, transmission, transformers, or gas turbines.

• New York Times: Britain Moves to Cut Electricity Taxes
  Andy Burnham’s new government plans to remove VAT from domestic electricity bills, a direct response to cost-of-living pressure.

• Bulletin of the Atomic Scientists: Data Centers Powered by Next-Gen Nuclear?
  A skeptical read on tech-sector nuclear announcements and the gap between press releases, reactor timelines, and actual deployment.

• Axios Pro: Bluecore Energy Launches With $10M for Nuclear on Barges
  Bluecore is targeting ports and data centers with small modular reactors on barges, starting with a test reactor at the Port of Long Beach.

• Bloomberg: Spain’s Renewable Overbuild Hits Investor Returns
  Spain shows the financial problem created when renewable buildout outruns transmission, storage, and flexible demand.

UTILITY RATES
Electric Rates Outpace Inflation

Lawrence Berkeley National Laboratory says U.S. electric rates increased 2.6% from 2024 to 2025 after adjusting for inflation. Since 2019, nominal residential rates rose 33%, commercial rates rose 26%, and industrial rates rose 27%, though inflation-adjusted average retail prices remain below 2010 levels.

Utility rate hike requests reached $18 billion in 2025, the highest level in decades, with regulators approving 64% of requested revenue increases from 2021 to 2025. LBNL points to transmission and distribution spending, wildfire mitigation, storm repair, commodity prices, tax policy, shrinking retail sales, and state-specific programs as major drivers.

GridBrief Take: The rate pressure is real, but the drivers are specific. Commissions should separate reliability spending from policy costs and monopoly cost recovery before blaming new load for everything on the bill.

SOLAR AND STORAGE
Google Backs Massive Arkansas Solar and Storage Project

Google signed a deal with Cypress Creek Energy to invest in and receive power from the Steel River Energy Center in Mississippi County, Arkansas. The project is expected to deliver 2.5 GW of solar generation and 2.9 GWh of battery storage by 2029, with Google serving as anchor investor and offtaker.

The first two phases will bring 1.6 GW of solar and 1.9 GWh of storage online. Cypress Creek says the project will use U.S.-made panels and mostly local steel. The project is expected to create 700 construction jobs per phase and generate about $300 million in local tax revenue over its life.

GridBrief Take: Google is putting capital behind new supply, which is exactly what large-load critics keep demanding. Solar and storage help, but firm capacity still has to be built around the AI load curve.

DATA CENTERS
Data Centers Could Reach a Fifth of U.S. Power Use

Recent forecasts keep moving data center demand higher. LBNL estimates data centers could use 9.5% to 15.3% of U.S. electricity by 2030, up from about 4.7% in 2024. EPRI projects 10% to 20% by 2035, Rhodium’s high-growth case reaches 18%, and BloombergNEF estimates data center power demand could hit 106 GW by 2035.

That load will not be served by one resource. Hyperscalers are pursuing nuclear restarts, SMRs, gas co-location, behind-the-meter generation, PPAs, storage, demand response, and workload shifting. Grid access and turbine supply remain the near-term constraints.

GridBrief Take: Nuclear is part of the 2030s answer, especially restarts and uprates. The late-2020s answer is more practical: gas engines, storage, flexible load, faster wires, plant life extensions, and private power.

DATA CENTERS, AGAIN
PJM Market Monitor Puts Data Center Capacity Costs at $6.3B

Monitoring Analytics says data centers accounted for $6.3 billion, or 38%, of the $16.4 billion in charges from PJM’s latest capacity auction. Across PJM’s last four base capacity auctions, data center-related charges totaled $29.4 billion, or 46% of total capacity charges.

Market monitor Joseph Bowring says large loads should contract for their own generation, or use separate auctions to procure capacity under long-term contracts. PJM’s board is developing a backstop auction proposal for data centers that could be filed with FERC this month for a September auction.

GridBrief Take: A separate capacity path can work if it gives large loads a clean way to buy what they need. PJM should also leave room for CRE, bring-your-own-power, co-location, and self-supply structures that reduce the amount of capacity PJM has to procure.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• VPPs move toward the data center market. Reuters reports that Sunrun, Tesla, Renew, and other virtual power plant operators are exploring contracts with data center customers. VPPs can reduce peak demand and defer some grid investment, especially in regions where interconnection timelines are too slow.

• SPP issued a Level 3 energy alert. The Southwest Power Pool warned of potential rolling blackouts after using some or all operating reserves during peak conditions. SPP has been developing faster large-load service options, and this heat event shows why flexibility and new capacity need to arrive together.

• Uniper targets data centers in strategy shift. The German utility plans about €5 billion in investment by 2030 and is targeting data centers at power plant sites. Europe’s utilities are discovering that existing generation locations, grid connections, and industrial land have new value in a compute-constrained economy.

• FERC pushes reliability standards for large computational loads. FERC directed NERC to propose reliability standards for data centers by year-end. The key issue is how large computing facilities ride through voltage disturbances and avoid sudden load drops that can create grid stability problems.

• New York’s hydropower line outage shows delivery risk. The Champlain Hudson Power Express line has been offline during high summer demand, forcing New York to rely more heavily on oil-fired generators. Clean energy procurement does not help reliability if the delivery system fails during peak conditions.

• National Grid buys into U.S. data center power. National Grid is investing $1.75 billion in Joulent, a developer focused on supplying power to data centers. The investment supports a Chevron-linked West Texas project intended to provide 2.67 GW to a Microsoft data center by 2028.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The grid is running into the machinery of its own reforms. PJM needs new firm capacity. Developers need turbines. Turbine queues are backed up. Interconnection rules were tightened to stop speculative projects. Now one serious gas project says those same rules could delay more than a gigawatt of dispatchable capacity.

That is the recurring theme this week. The country needs more power and faster grid access, but every solution runs into a process built for slower load growth. Columbia’s energy policy team says low-cost tools like GETs and demand response can cut pressure before customers get hit. Texas wants a 765-kV backbone for Permian load growth, while landowners argue they were not properly notified. FERC is forcing RTOs to rewrite large-load rules. DOE is offering $17.5 billion to revive large reactor construction. Everyone agrees speed is needed. The fights are over who gets charged, who gets delayed, and who gets to say no.

The Lede

GAS POWER
PJM Opposes Waiver for $2B Ohio Gas Plant

PJM told FERC it opposes a waiver request from Advanced Power Services that would let the company keep its Chestnut Run gas project in PJM’s fast-track Reliability Resource Initiative review. The project, located in Carroll County, Ohio, is a roughly $2 billion gas-fired plant owned by ArcLight-backed Advanced Power. The RRI process was built for shovel-ready projects and initially drew 51 projects totaling about 11.8 GW.

Advanced Power entered the RRI process using GE Vernova’s HA.03 turbine model and said the plant could be online by May 2030. Turbine backlogs have pushed that model out by at least two years. The company now wants to use GE Vernova’s HA.02 model, which would reduce the plant’s maximum output by about 550 MW while keeping the project close to its original schedule. PJM says the requested changes violate RRI rules that barred changes to project size and Capacity Interconnection Rights. Advanced Power says denial could delay more than 1.2 GW of firm, dispatchable capacity by about two years.

GridBrief Take: PJM is defending the integrity of a fast-track process created to stop queue games, and that concern is legitimate. The problem is that turbine scarcity is now colliding with resource adequacy. A waiver should come with strict conditions, updated milestones, and no special treatment on cost allocation. PJM also needs a process that can distinguish a real supply-chain adjustment from a developer trying to rewrite its bid after seeing the market.

Other Things to Check Out

• AMAC: How America Should Respond to China’s Industrial Dominance Playbook
  Drew Bond on how China uses subsidies, scale, and supply-chain control to dominate strategic industries.

• Reason: America’s Nuclear Industry Doesn’t Need Cronyism to Thrive
  A pro-nuclear, anti-cronyism case for licensing reform and competition over federal favoritism.

• Reuters: Sodium Batteries Enter the AI Energy Race
  Sodium-ion storage gets a closer look as data centers push demand for cheaper, less lithium-dependent batteries.

• MPR: Minnesota Studies New Nuclear
  Minnesota lawmakers agreed to study new nuclear after decades under a construction moratorium.

• Bloomberg: Spain’s Renewables Boom Hits Investor Returns
  Spain’s renewables buildout shows the financial risk when generation grows faster than transmission, storage, and flexible demand.

REPORT
Columbia: Use GETs and Demand Response Before Building the Expensive Stuff

Columbia University’s Center on Global Energy Policy released a commentary on electricity affordability and load growth that puts data centers inside the larger rate problem. The researchers identify several cost drivers: inefficient infrastructure planning, utility incentives that reward capital spending, underused flexibility, slow permitting, storm costs, wildfire costs, and supply chain inflation. Load growth from data centers compounds those pressures.

The report highlights near-term tools that already exist. Dynamic line ratings, advanced conductors, power flow controls, and other grid-enhancing technologies can increase capacity on existing infrastructure. Demand response can reduce peak demand, especially if large loads shape usage during constrained hours. Columbia also points to structural reform: better cost allocation for data centers, faster interconnection, permitting that connects price signals to actual supply, and utility regulation that rewards system optimization rather than only capital deployment.

GridBrief Take: Before regulators approve another giant capital plan, they should force utilities to show what dynamic ratings, reconductoring, demand response, and flexible large-load service can do first.

TRANSMISSION
Texas Landowners Push Back on 765-kV Buildout

Several hundred Central Texas property owners asked the Public Utility Commission of Texas to pause review of the proposed Bell County East to Big Hill 765-kV transmission line. The line is part of Texas’ high-voltage buildout to support projected load growth in the Permian Basin and broader ERCOT needs. Landowners say utilities failed to properly notify affected property owners after route changes.

The filing says more than 1,300 directly affected property owners were denied a chance to attend public meetings before their land was included in the project footprint. Oncor says it began public meetings in summer 2025, modified routes based on feedback, and responded to thousands of public comments. PUCT staff recommended 765-kV lines based on a 2024 ERCOT study because they can move more power over longer distances with fewer losses. Hearings were held June 8-12. A final PUCT decision is expected in September.

REGULATION
FERC Opens the Large-Load Rule Fight

FERC issued show-cause orders to CAISO, ISO New England, MISO, NYISO, PJM, and SPP after finding their large-load interconnection rules may be inadequate. The orders direct grid operators to address five areas: efficient study processes, cost-shift prevention, transparency into upgrade costs, co-location and behind-the-meter generation, flexible load services, and study processes for generation serving nearby large loads.

The commission chose regional proceedings rather than one national rule. That gives each RTO room to deal with its own grid conditions, but it also creates six separate timelines and six separate answers for developers. FERC’s orders also ask grid operators to evaluate alternative transmission technologies such as dynamic line ratings, advanced conductors, power flow devices, and synchronous condensers before defaulting to traditional network upgrades. RTOs and ISOs have 60 days to respond, with a possible extension process.

NUCLEAR
DOE Offers $17.5B for Large Nuclear Components

The Trump administration announced up to $17.5 billion in conditional loans to help utilities and energy companies purchase long-lead components for up to 10 Westinghouse AP1000 reactors. The loans would support up to five projects, each with two 1.1-GW reactors. The goal is to begin construction by 2030 and reduce timelines by ordering parts such as reactor vessels and steam generators in bulk.

The AP1000 is the only large advanced reactor design operating commercially in the United States. Vogtle Units 3 and 4 came online in Georgia in 2023 and 2024, but the project took about 15 years and cost around $35 billion, roughly double initial estimates. The DOE plan would require Westinghouse and utility partners to commit $500 million each per project before accessing federal funds. Seven utilities have expressed interest, according to Energy Secretary Chris Wright, though DOE has not named them.

GridBrief Take: Large nuclear will not return in the United States without solving the first-project risk problem. Bulk ordering can help rebuild the supply chain, but taxpayers should not become a permanent overrun account for one favored vendor. The cleanest nuclear policy is faster licensing, standardized designs, private offtake, credible cost-sharing, and no assumption that Washington can outguess construction risk better than balance sheets can price it.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Reuters says data centers expose old power-policy failures. Douglas Arent argues that data centers are being blamed for a rate problem rooted in utility incentives, slow infrastructure, and poor cost allocation. His Columbia-linked argument tracks several recent GridBrief themes: demand growth can lower average costs when the system is built well, and it raises bills when planning and tariffs are bad.

• Sodium-ion batteries get a data center push. Reuters reports that automakers and storage suppliers are accelerating sodium-ion battery production as data centers and grids demand more storage. GM is working with Peak Energy on Michigan production by 2028. Sodium-ion is still early, but lower material cost, wider temperature tolerance, and less cooling could make it useful for stationary storage.

• ICF sees 445 GW of new capacity by 2030. A new ICF report says the U.S. could add 445 GW of capacity by 2030, mostly solar and storage. The report also says effective capacity is much lower than nameplate and that high-growth regions like Texas and PJM have little spare margin after 2027. The headline number is large. The reliability number is smaller.

• Data centers are ready to trade flexibility for speed. Utility Dive reports that large-load customers are increasingly willing to negotiate curtailment, modular design, backup power, and flexible service if it helps them connect faster. That is where the large-load market is heading: firm service for customers willing to wait, conditional service for customers willing to move.

• DOE coal orders now raise cost-allocation fights. E&E reports that a utility is pressing DOE over the cost of keeping a coal unit online under emergency authority. This was inevitable. Keeping plants open by federal order may help reliability, but someone still has to decide who pays for fuel, staffing, maintenance, and opportunity costs.

• Newcleo advances its Italian test facility. NucNet reports that Newcleo installed the main vessel for a non-nuclear precursor test facility in Italy. The project supports the company’s lead-cooled fast reactor program. It is still a test facility, but advanced nuclear progress increasingly comes through supply-chain and demonstration steps before any reactor order.

• European heat brings power-system stress back into view. Reuters’ heatwave tracker points to price spikes, power flows, cooling limits, grid alerts, and import dependence as the numbers to watch. Europe’s summer reliability problem is no longer just peak demand. It is also how weather affects generation, transmission flows, and thermal plant operation at the same time.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The grid story is moving from demand forecasts into household budgets. Utilities need capital for data centers, factories, electrification, reliability upgrades, wildfire hardening, gas volatility, and transmission expansion. Customers see the result as a higher bill.

Fitch just moved its North American utilities and power outlook to deteriorating. The warning is not about weak demand. It is about the politics of recovering investment costs when residential electricity prices are already rising and 36 states have gubernatorial elections this fall.

The Lede

POWER OUTLOOK
Fitch Turns Negative on Utilities

Fitch Ratings revised its mid-year 2026 outlook for North American utilities and power to deteriorating from neutral, citing rising affordability concerns and growing political and regulatory risk. The ratings agency still sees strong demand fundamentals, with electricity demand expected to grow 2.0% to 2.5% annually through 2030, driven by data centers, electrification, and industrial reshoring.

Customer bills are rising into that investment cycle. EIA reported average residential electricity prices at 18.8 cents per kWh in March, up 10.2% from a year earlier. Utilities are planning roughly $240 billion in capital spending in 2026, with Fitch expecting annual sector capex to rise by a low- to mid-teens percentage rate from 2026 to 2030. PJM remains the clearest pressure point: capacity prices rose from about $29/MW-day for 2024/2025 to more than $329/MW-day for 2026/2027, increasing total regional capacity costs from $2.2 billion to more than $16 billion.

Grid Take: Fitch is describing the first real credit-market warning from the ratepayer backlash. Utilities can still benefit from load growth, but only if regulators believe new demand is paying its own way. The next few years will reward utilities that can separate growth capital from legacy customer bills, use large-load tariffs cleanly, and show commissions that higher capex is tied to measurable reliability, capacity, and customer value.

Other Things to Check Out

• Latitude Media: Data center jobs aren’t at servers. They’re in energy.
  A useful corrective to the local jobs debate. Hyperscale data centers are capital-intensive and designed to run with few permanent workers. The larger employment story often sits upstream in generation, transmission, construction, equipment, and the energy infrastructure built to serve the load.

• C3 Journal: Closing the Local Project Loophole
  A strong case for competitive transmission. The piece argues that AI and data center growth will require faster power-line deployment, while current rules still allow many incumbent utilities to build local transmission projects without competition. If transmission spending is going to rise, competitive pressure should discipline cost and schedule wherever possible.

• Detroit News: Fix the system behind Michigan energy bills
  The link was not accessible, but the topic fits the moment. Michigan’s utility debate is increasingly about monopoly incentives, capital spending, reliability performance, and whether customers are getting value for repeated rate increases.

• Bloomberg: Portland General Electric to raise data center rates
  Bloomberg’s video was not accessible, but the underlying Oregon story is important. Portland General Electric announced a 29% rate increase for data centers while cutting residential rates by 1.3%, following Oregon’s push to protect smaller customers from large-load costs.

• Fast Company: Curio’s nuclear waste-to-fuel bet
  Curio’s NuCycle process aims to turn spent nuclear fuel into usable products, including uranium hexafluoride that can be reenriched, industrial gases, and platinum metals. The company says the remaining waste would be only 3% to 4% of the original mass and require sequestration for about 300 years rather than tens of thousands.

Major Stories

NUCLEAR
Sweden Picks Rolls-Royce SMR for New Nuclear

Sweden’s Vattenfall selected Rolls-Royce SMR to supply small modular reactors for a new nuclear project at Ringhals, choosing the British company over GE Vernova. The project would include three 470-MW reactors, producing about 12 TWh of electricity per year, roughly 6% of Sweden’s annual power consumption.

The first reactor could enter service in the mid-2030s, depending on permitting and regulatory review. Sweden has not built a new reactor in more than 40 years, but its government expects electricity demand to double by 2045 and sees nuclear as part of its energy security and net zero strategy. The project company, Videberg Kraft, is 80% owned by Vattenfall and 20% owned by major Swedish companies, with the government planning to take a 60% stake pending parliamentary approval.

Grid Take: Sweden is using nuclear policy to answer a specific industrial question: how to serve a much larger power system without depending entirely on weather, imports, or gas. The Rolls-Royce choice also gives SMRs a serious European test case.

REGULATION
San Francisco Adds an All-Electric Renovation Rule

Starting July 1, San Francisco will require most major home renovations that replace core mechanical systems to go all-electric. Covered projects must use electricity for space heating, cooling, water heating, cooking, and clothes drying. The rule applies when renovations replace major systems and meet thresholds for extensive wall, ceiling, structural, or expansion work.

The city estimates the rule will affect the equivalent of about 261 single-family homes and 293 multifamily units per year. City officials say all-electric renovation can reduce emissions and may lower costs by avoiding new gas piping. Contractors and UCLA researchers have pointed to older-building constraints, including electrical panels, wiring, service upgrades, contractor familiarity with load management, and costs for lower- and moderate-income households. Higher project cost alone is not grounds for an exemption.

Grid Take: San Francisco is forcing electrification at the exact point where construction is already expensive, slow, and regulation-heavy. The policy may touch a small share of buildings, but it still shows the weakness of mandate-driven electrification. A cleaner pro-electrification strategy would lean harder on price signals, panel-friendly equipment, streamlined permits, and voluntary conversion when systems naturally reach replacement.

IN THE COURTS
Judge Restores $82.1M in Clean Energy Grants

A federal judge vacated DOE’s cancellation of $82.1 million in clean energy grants awarded during the Biden administration. DOE is expected to reinstate funding for 11 projects. The plaintiffs, led by the American Institute of Chemical Engineers, argued the grants were part of an October 2025 set of terminations targeting projects in states that voted for Kamala Harris.

The awards included funding for critical minerals recovery from electrolyzers and fuel cells, along with work aimed at reducing hydrogen production costs. DOE’s broader October terminations covered more than $7.5 billion in clean energy awards in blue states. Energy Secretary Chris Wright denied that politics played a role in the review process, while plaintiffs pointed to public comments from OMB Director Russell Vought describing nearly $8 billion in “Green New Scam” funding being cancelled in named states.

SOLAR POWER
Solar Heads Into Summer With 20% More Capacity

EIA expects utility-scale solar generation to increase 19% this summer compared with last summer, reflecting a 20% increase in capacity. Wind generation is forecast to rise 10%, hydro 5%, and nuclear 1%. Coal generation is expected to decline 2%.

This summer is expected to be hotter than last year, with cooling degree days up 3% from June through September. EIA expects the increase in generation to come almost entirely from renewable sources. Coal consumption declined 11% in the first quarter compared with the same period last year, helped by a warm March and April and lower natural gas prices, though EIA says hotter weather and higher gas prices could raise coal burn over the summer.

Grid Take: Solar is now a major summer capacity story because it is getting built quickly and showing up when air-conditioning load is high. That does not solve evening ramps, winter peaks, transmission congestion, or multi-day reliability risk. The lesson is narrower and more useful: when permitting, economics, and supply chains line up, markets can add real capacity fast. The rest of the grid needs that same buildability.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Permitting reform polls well. New Blue Rose polling shared with Heatmap found 60% of voters support making it faster and easier to build energy infrastructure. Another 62% said presidents should not have unilateral authority to cancel already-approved projects. That lands directly in this week’s DOE grant fight: energy capital can live with strict rules, but it struggles with rules that change after the election.

• Steel River lands $3.5B for solar and storage. Cypress Creek secured $3.5 billion for the first two phases of the Steel River Energy Center in Arkansas, one of the largest solar-plus-storage projects in the U.S. The first two phases include 1.63 GW of solar and 1.9 GWh of batteries, with the full project expected to reach 2.45 GW of solar and 2.9 GWh of storage by 2029. The energy transition keeps arguing on cable news while the financing desks keep building.

• Long-duration storage gets a real utility test. Energy Dome and Salt River Project announced a 19-MW, 10-hour CO2 battery project in Arizona, expanding a collaboration that includes Google. Arizona is the right proving ground: hot peaks, fast load growth, data centers, and a grid that needs more than four-hour batteries to get through the evening.

• Dutch grid moratoriums show the cost of slow wires. The Netherlands has expanded local bans on new grid connections for generation and large loads, with Utrecht adding new residential connection limits in July. Dutch data center power demand could reach 15% of national electricity use by 2030, while grid projects can take up to 12 years. Build the grid late, ration connections later.

• PJM gets a fast-track generation lane. FERC approved PJM’s expedited interconnection track for up to 10 large, shovel-ready projects per year through 2027. Eligible projects must be at least 250 MW, have state support for faster siting, and be capable of coming online within three years. PJM needs capacity quickly enough that the queue is becoming triage.

• Amazon finally reports data center water use. Amazon disclosed that its data centers used 2.5 billion gallons of water in 2025, down 2% from 2024 despite business growth. The company says its data centers are seven times more water-efficient than the industry average. Water is now a siting issue, a politics issue, and a trust issue, so disclosure is becoming part of the cost of building.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The data center debate is growing up, which means it is becoming less fun and more useful, at least in some places. Now the fight is moving to the place where civilization does its least glamorous work: tariffs, collateral, interconnection rules, exit charges, and who pays when a 600-MW dream arrives at the substation with a shovel and a press release.

This issue is about the architecture of responsibility. FirstEnergy wants FERC to make data centers pay for the transmission upgrades they trigger. SPP now has approval for a faster, non-firm path to connect large loads that can be curtailed during grid stress. Microsoft is trying to turn its “we’ll pay our way” promise into a Nevada tariff. GM wants EVs and batteries to become part of the energy business. And the Trump administration is putting real money behind coal because the power crunch has a way of dragging every idea into the light.

The Lede

DATA CENTERS
FirstEnergy Wants Data Centers to Pay for the Wires

FirstEnergy asked FERC to require data centers to pay for transmission upgrades needed to bring them online, rather than spreading those costs across existing customers. The proposal borrows from natural gas pipeline policy, where new customers can pay incremental rates for new infrastructure instead of asking legacy customers to finance the expansion. Under FirstEnergy’s plan, data centers would sign 15-year contracts, provide collateral, pay the normal zonal transmission rate for the existing system, and then pay an additional expansion rate for new facilities built because of them.

The instinct is understandable. Existing ratepayers should not get pickpocketed because a hyperscaler arrives with a 600-MW appetite and a consultant named Braxton. But the rush to drag FERC into every large-load negotiation feels like the beginning of a hundred short-sighted market distortions. Utilities and large customers are perfectly capable of negotiating cost responsibility, collateral, timelines, exit charges, and dedicated infrastructure without turning every data center into a federal regulatory regime.

Grid Take:
FirstEnergy is asking the right question, but probably inviting the wrong referee. The better answer is not a national data-center tollbooth managed by FERC; it is freer contracting, cleaner cost causation, and more ways for large loads to build or buy their own power outside the legacy utility maze. Right now everyone thinks data centers are politically vulnerable enough to kick around and maybe shake down a little. Data centers should take the lesson: remove themselves from the worst parts of the market where possible, push hard for Consumer Regulated Electricity, behind-the-meter generation, and other arrangements that let private capital build private power without asking the entire rate base for permission.

Other Things to Check Out

• The Martha Zoller Show: Drew Bond on Energy Independence
  Drew Bond gets at the distinction that matters: energy dominance is not the same as energy sovereignty. Producing more oil matters, but it does not solve the mineral, battery, semiconductor, defense-tech, and supply-chain vulnerabilities that now define energy security. The Strait of Hormuz still matters because the energy system of the future is not just barrels. It is wires, minerals, chips, batteries, factories, and allies.

• Reason: The Best Way to Keep Data Centers From Driving Up Electricity Costs
  Ari Shtein makes the right basic case: the answer to data center backlash is not moratoria, but more energy and better rules for building it. The piece points to Consumer Regulated Energy as a path for letting data centers and other large industrial facilities buy power from third-party providers rather than dragging every new project through the centralized grid. That is exactly the frame this debate needs. If demand is rising, suppressing demand does not make the system healthier. It just makes growth go somewhere else.

• Business Insider: Making Fusion Cool
  Commonwealth Fusion Systems is trying to make fusion feel less like a distant lab promise and more like a live technological race. The company is building a public campaign around SPARC, its fusion machine expected to turn on next year, and the challenge is wonderfully strange: marketing an invisible, pre-commercial infrastructure technology to a generation trained to suspect hype and expect institutional disappointment.

• ZME Science: China’s Truck-Sized Nuclear Reactor
  This one should be treated carefully. The claim is that Chinese researchers are testing a prototype 10-MW vehicle-mounted nuclear reactor, essentially a “nuclear power bank” for remote sites, ships, emergency backup, space systems, and possibly AI data centers.

• C3: How Data Centers Can Help the Grid
  Patrick Sullivan’s piece is a useful counterweight to the lazy “data centers equal higher bills” story. The Mississippi examples show how large loads can spread fixed costs across more kilowatt-hours, make use of surplus capacity, help finance new generation, and create revenue for reliability and storm-hardening investments.

Major Stories

FERC
FERC Approves SPP’s CHILLS Plan

FERC approved the Southwest Power Pool’s new “conditional high impact large load service,” or CHILLS, which allows data centers and other large loads to receive non-firm transmission service for up to seven years while they work toward firm service. The bargain is simple: large loads can get online faster using available transmission capacity, but SPP can curtail them when the system is constrained or in emergency conditions.

That is useful because it treats flexibility as the price of speed. The old model assumes a large new customer should wait for generation, network upgrades, studies, queues, and the slow public choreography of utility planning. CHILLS lets SPP say yes sooner without pretending the system has firm capacity it does not have.

Grid Take: CHILLS is directionally right because it lets large loads trade certainty for speed. But the real prize is not more FERC-approved cleverness; it is a market where private parties can contract, build, curtail, self-supply, and move quickly without waiting for every practical arrangement to receive a federal blessing.

COAL
Trump Puts $850 Million Behind Coal

The Trump administration announced up to $850 million to support coal power, including money to build two new coal-fired plants in Anchorage, Alaska, and Mt. Storm, West Virginia, modernize more than a dozen existing facilities, support coal exports, and preserve or add thousands of megawatts of coal capacity. The two new plants would total 2.85 GW and would be the first new U.S. coal plants to come online since 2013.

Being energy agnostic matters right now. The grid needs firm power, dispatchability, fuel security, and fewer ideological purity tests from people whose entire reliability plan appears to be a weather app and a sternly worded grant program. But subsidizing coal as a way to compensate for decades of subsidies and regulations that worked against coal is a strange kind of energy reparations. It may feel symmetrical, but it is still government putting its thumb on the market, just from the other hand. The better move is simpler and harder: stop forcing retirements, stop blocking new firm power, stop subsidizing favored resources into artificial dominance, and let the system build what can actually compete.

Grid Take: Coal should be allowed to compete fairly… not punished and not bolstered by government. If the country wants abundant power, government should get out of the way of markets rather than trying to balance yesterday’s distortions with tomorrow’s subsidies.

MSFT DATA CENTERS
Microsoft Tries to Turn “Pay Our Way” Into a Tariff

Microsoft filed a proposed Ratepayer Protection Tariff in Nevada that would require large-load customers to pay for infrastructure built specifically to serve their projects while preserving ordinary utility charges for broader grid services. The proposal would split costs into a Customer Contributed Share, paid by the large-load customer, and a System Benefit Share, which could be reviewed for inclusion in the rate base if the investment benefits everyone. It also includes contract demand, load ramp schedules, asset tracking, exit charges, bring-your-own-power provisions, and an expedited review pathway for projects that are fully funded upfront.

This is a more serious version of the tech-sector promise to pay its own way. Microsoft is trying to build a practical framework where large customers can directly fund or finance generation and transmission, protect residential and small-business customers, and move faster when they are willing to put real money behind the project.

Grid Take: Microsoft’s tariff is useful because it turns “we’ll pay our way” into something regulators can actually enforce. But the gold standard is still more private contracting and more CRE-style pathways that let large loads finance their own power instead of asking the legacy grid to become an AI concierge.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• ERCOT finds the data center tripwire. Texas grid operators say some large data centers and crypto facilities failed voltage-disturbance tests, with several groups of large loads capable of dropping more than 5 GW apiece under certain fault conditions. That is the kind of number that turns “large load” from an economic development category into a grid-stability problem with a hard hat on. The next data center fight will not just be about how much power they use. It will be about whether they can stay connected when the system sneezes.

• Texas transmission gets another $7 billion shove. Sempra says ERCOT-backed projects now require more than $7 billion in investment, much of it through Oncor, to serve new demand across the Dallas-Fort Worth area and the I-35 corridor. The buildout is meant to support roughly 16 GW of new load. Texas remains the place where the energy future shows up wearing boots: messy, expensive, fast, and usually already under construction before the rest of the country has finished forming a task force.

• EIA says the record-load era is here. EIA now expects U.S. power use to hit new highs in both 2026 and 2027, with commercial electricity demand passing residential demand as AI, data centers, and broader electrification keep climbing. This is the quiet statistical regime change under every loud policy fight. The American grid was built around homes, factories, and old commercial load. Now the load growth is increasingly commercial, digital, and concentrated enough to make yesterday’s planning assumptions look like they were written on a napkin during a blackout.

• Big Tech keeps warming to geothermal. Reuters reports that Google, Meta, Microsoft, and other large buyers are helping push next-generation geothermal toward lower costs through long-term deals and project finance. That matters because geothermal is one of the few clean-power options that speaks the language data centers actually need: firm, round-the-clock power without pretending noon solar solves midnight compute. It is still expensive and early, but the buyer pull is real.

• Not-for-profit utilities are turning to storage. Co-ops and municipal utilities are increasingly looking at batteries for reliability, price hedging, and avoided infrastructure costs as data centers and volatile power prices reshape the planning map. That is storage growing up. It is not just renewable garnish anymore. It is becoming a balance-sheet tool for utilities that need options before the next giant load request lands on the desk.

• DOE wants a fusion road map, while cutting the check. The Energy Department released a road map for developing fusion electricity, even as the administration’s fiscal 2027 budget would trim fusion funding. That tension is almost too on the nose: America wants the moonshot, the ribbon cutting, the private capital, and the technological miracle, but would still like someone else to pay for the boring middle years. Fusion remains a real race, but races are generally easier to win when the runners are fed.

• Switzerland and Japan reopen the nuclear file. Switzerland extended debate over lifting its post-2017 ban on new nuclear plants, while Japan is weighing replacement of aging reactors through the 2040s and 2050s. The pattern is now hard to miss. Countries that spent years imagining they could retire firm power are rediscovering that electricity systems have memories. They remember winter, industry, fuel imports, and the difference between a policy preference and a megawatt.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

This issue is about the scramble for capacity. Google is paying other customers to move load around in PJM. R Street is making the case that competitive transmission builds faster and cheaper than the incumbent model. DTE is buying Michigan-made batteries for a grid suddenly full of large-load suitors. Constellation's Three Mile Island restart got a critical FERC waiver. And DigitalBridge is buying a power platform because the data center industry has learned what every industry eventually learns: whoever controls the energy controls the future.

The Lede

VIRTUAL POWER
Google Funds 100-MW Virtual Power Plant in PJM

Google will fund a three-year, 100-MW virtual power plant in PJM through a new deal with Voltus, which will aggregate flexible load from homes, businesses, electric vehicles, smart thermostats, batteries, and industrial customers. Google has worked on making its own data centers more flexible, but its energy chief told Utility Dive the harder truth: when billions of dollars of chips are sitting in a facility, the company would rather pay other people to shift load than idle its own machinery.

That is the interesting part. This is not philanthropy or green branding. It is a large energy user paying for flexibility where the grid actually needs it, instead of waiting for new generation or transmission to arrive on the usual utility timeline. PJM is already dealing with rising capacity costs, shrinking reserve margins, and load growth from data centers. Google's VPP is a small answer, but it is the right kind: use existing infrastructure better, and stop treating peak demand as an unavoidable weather event with invoices.

Grid Take: This is what a more mature data center strategy looks like. The grid does not only need more supply. It needs customers rich enough, sophisticated enough, and desperate enough to pay for smarter demand.

Other Things to Check Out

• C3: A Consumer-First Grid Arushi Sharma Frank's piece on Texas is a great read and works well alongside the Google VPP story. ERCOT is moving toward a framework where large loads can accept curtailment in exchange for faster energization. The mechanism is Provisional Controllable Load Resource status, where developers commit to operational flexibility and enter a new batch interconnection process with defined study timelines, firm-load floors, and binding commercial commitments.

• E3: Data Center Grid Utilization and Cost-of-Service E3 argues that data centers do not inherently raise costs for everyone else, since high-utilization loads can spread fixed grid costs over more kilowatt-hours, but outcomes depend on system conditions, tariff design, collateral, exit fees, and whether incremental revenues exceed incremental costs.

• The Economist: Europe's Problem of Too Much Power The Economist looks at the strange new European problem of having too much electricity in some markets at some times. This is what happens when intermittent supply grows faster than storage, transmission, and flexible demand. Scarcity is not the only way to fail at energy policy.

• Bloomberg: China Wastes More Clean Electricity Bloomberg reports that China's emissions rose as more clean electricity went to waste. Building renewables is not the same thing as integrating them. Without enough transmission, storage, and dispatch flexibility, clean power becomes a stranded asset with better public relations.

• National Center for Energy Analytics: America's Electricity Grids at a Crossroads This report puts the current load-growth fight in a broader historical frame. The U.S. grid was built over a century into an enormous machine for reliable power, but that machine has frayed over the past 15 years just as demand is returning through data centers, reshoring, and electrification. The question is no longer whether demand will grow. The question is whether the institutions around the grid still know how to build.

Major Stories

REPORT - TRANSMISSION
R Street Makes the Case for Competitive Transmission

R Street's new report from Kent Chandler and Olivia Manzagol looks at whether competitive transmission projects can beat the incumbent utility model on speed and cost. The authors reviewed results from FERC Order 1000's removal of federal rights of first refusal, which forced significant regional transmission projects into competitive solicitations rather than handing them automatically to the incumbent. The first finding is familiar but important: the data are bad. Incumbent transmission cost and timing data remain too inconsistent and opaque to support the kind of clean comparison policymakers should already have demanded.

The available evidence still points somewhere useful. Competitive greenfield transmission projects generally came online faster than comparable incumbent projects in CAISO, MISO, SPP, and ISO-NE. The report also finds competitive projects appear to produce savings around 30 percent compared with incumbent projects, while often bringing in-service commitments, financial penalties for lateness, cost caps, and return-on-equity caps. That is not an ideological argument for competition. It is a practical one: when the incumbent knows the job is theirs by birthright, the ratepayer tends to inherit the family curse.

Grid Take: The country needs transmission badly enough that it cannot afford to treat incumbent monopoly control as a tradition worth preserving. If competition can produce faster projects, lower costs, and enforceable commitments, blocking it is not prudence. It is protectionism wearing a hard hat.

BATTERY STORAGE
DTE Buys 6 GWh of Michigan-Made Batteries

DTE Energy will procure 1.5 GW, or 6 GWh, of battery storage systems from LG Energy Solution Vertech in a $1.6 billion deal, with batteries manufactured in Michigan and deployed over the next two years. The storage will be used across eight projects. DTE estimates the work will generate $2.3 billion in economic impact while supporting roughly 1,800 jobs at LG's Holland plant and more than 350 additional construction and operations jobs.

The local manufacturing angle matters, but the grid angle matters more. DTE says the batteries will reduce strain on the grid and decrease the need to cycle generation up and down as demand fluctuates. The company is also trying to serve a new Oracle data center and is awaiting approval for a 1-GW Google data center deal. This is not storage as boutique climate policy. This is storage as industrial plumbing for a Midwestern utility watching large-load demand walk up the driveway.

Grid Take: Batteries will not replace firm power. But storage is becoming one of the basic tools utilities need to make large load less jagged, less expensive, and less politically radioactive.

NUCLEAR
Three Mile Island Restart Gets a FERC Boost

Constellation's plan to restart the Crane nuclear unit, formerly Three Mile Island Unit 1, got a significant boost when FERC approved a waiver allowing the company to transfer 760 MW of Capacity Interconnection Rights from its Eddystone plant near Philadelphia to Crane. The transfer could let Crane deliver more of its power when it restarts, potentially before the end of 2027, rather than waiting for transmission upgrades that may not be complete until December 2030.

The practical meaning is simple: a nuclear plant with a 20-year deal to serve Microsoft data centers in PJM may now be able to get its power onto the grid sooner. The market monitor objected, but FERC said the waiver solves a concrete problem and makes more efficient use of rights Eddystone cannot currently use. This is how the nuclear comeback will actually happen, if it happens at all. Not in speeches, but in waivers, interconnection rights, and procedural fights where the future waits in the hallway.

Grid Take: Restarting nuclear is hard because the plant is only half the problem. The other half is the grid around it, and right now America has a growing number of power plants that are useful before the paperwork says they are convenient.

M & A
DigitalBridge Buys Into Power

DigitalBridge, a data center and digital infrastructure investor, plans to buy ArcLight Capital Partners in a deal valued at up to $1.05 billion. ArcLight owns one of the country's largest private power generation portfolios, including about 20.8 GW as of June 2025, roughly 7 GW in PJM, and a 15-GW project pipeline.

The move is the clearest signal yet that the data center industry is no longer content to be merely a customer of the power sector. It wants expertise, assets, development rights, and options. DigitalBridge is also being acquired by SoftBank, and the ArcLight deal would bring power investment directly into the platform. The old joke was that the cloud is just someone else's computer. The new reality is that the cloud increasingly wants to own the power plant, too.

Grid Take: Data center investors are moving upstream because the grid bottleneck is now a business bottleneck. The next phase of digital infrastructure will not be defined only by land, fiber, and chips. It will be defined by who can secure firm power without waiting years in a queue.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Offshore wind litigation escalates: Seven states sued the Trump administration over its buyout of TotalEnergies' offshore wind leases, arguing the administration illegally terminated contracts already in development. The legal fight is now a referendum on whether the federal government can unilaterally extinguish existing lease agreements — which matters well beyond offshore wind.

• EEI's three-word summary: Labor, AI, affordability. Utility executives at this year's Edison Electric Institute conference spent the week wrestling with a workforce pipeline that can't build fast enough, AI load that's arriving faster than interconnection queues can process, and a rate environment where customers are already pushing back. The squeeze is structural, not cyclical.

• Samsung signs Vietnam's first corporate PPA: Samsung's Vietnamese subsidiary inked a direct power purchase agreement for 70 GWh of solar generation — the first of its kind in the country. Another data point in the quiet repositioning of hyperscaler supply chains toward energy-constrained emerging markets where securing your own electrons is becoming a competitive moat.

• Foxconn and Intel partner on AI infrastructure: The two companies announced a collaboration on next-generation AI infrastructure buildout. Neither disclosed financial terms. What matters is the hardware-to-power supply chain tightening: every partnership like this is another block of committed load looking for guaranteed megawatts.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The data center debate has finally reached the toll booth. For the past year, the loudest version of the argument has been cultural and local: too much water, too much land, too much noise, too much “why exactly are we paving over my county so a chatbot can write bad wedding vows?”

That fight is still real, but the sharper one is now financial. Data centers want grid access at a scale most utilities were not built to deliver quickly. Ratepayers do not want to become involuntary venture capitalists for hyperscalers. Generators want enough certainty to build. Regulators want growth, reliability, and affordability, ideally without being blamed for the physics. This issue is about the hard part of the AI buildout: turning demand into power without turning the grid into a subsidy laundromat.

The Lede

HIGHLIGHT ESSAY
How Will Data Centers Pay for Power?

This is not “news” in the narrow sense, but it deserves top billing anyway. Travis Kavulla’s “How Will Data Centers Pay for Power?” is one of the best pieces yet on the actual electricity problem beneath the AI boom: data centers are not just another customer class politely asking for service. They are a parallel industrial load trying to use a scarce public platform at a scale the platform was not designed to absorb quickly.

The essay matters because it moves past the bumper-sticker version of “make Big Tech pay.” That sounds easy until you get inside the utility machinery, where joint costs, monopoly incentives, average rates, embedded-cost pricing, and speculative interconnection requests all start breeding in the walls. Kavulla’s point is that data centers should be able to buy speed and certainty, but only under rules that protect existing customers and force new load to pay for the grid it actually triggers.

• Price grid access honestly. Data centers should not pay ordinary average rates if they are causing extraordinary new costs for wires, transformers, substations, and generation. Otherwise, the result is a quiet cost shift from hyperscalers to legacy customers.

• Treat interconnection like scarce capacity. The current queue can reward whoever files first, whether the project is real or just a paper claim on future power. Kavulla’s better model is an “open season” for grid access, borrowing from gas pipelines, where providers define available capacity and data centers bid for firm withdrawal rights.

• Let serious projects pay for certainty. If a data center truly values power at a particular site, it should be willing to back that demand with real money. That would help separate actual projects from queue confetti and give utilities a clearer signal for what infrastructure is worth building.

• Return surplus value to ratepayers. If data centers bid for scarce grid capacity and those bids exceed the cost of needed upgrades, the upside should not simply disappear into the utility fog bank. Done right, surplus revenues could flow back to existing customers.

• Make “bring your own generation” mean something. BYOG cannot just be a press release about clean energy while the legacy grid quietly absorbs the reliability risk. A serious version would require data centers to bring new generation, free up unused capacity, or operate flexibly as a condition of service.

• Reward flexibility, not just size. Data centers that can curtail, ramp, shift computing, or operate with backup power during stress should be more valuable to the grid than loads that demand perfect service at all hours and then send everyone else the bill.

Other Things to Check Out

• E3: Understanding the Drivers of Rising Electricity Rates and the Role of Data Centers
  A useful corrective to the cartoon version of the rate debate. E3 finds that many forces are pushing bills up, including inflation, gas volatility, resilience spending, grid modernization, market design, retirements, and load growth. In PJM, E3 estimates load growth accounted for about half the capacity price increase, with the rest driven by market design changes, retirements, accreditation changes, and other supply-side factors.

• VICE: The AI Boom Is Taking Over the U.S. Power Grid, and This Map Shows Where It’s Happening
  A more consumer-facing look at the national data center map, with 4,000-plus facilities spread across the country and heavy concentrations in Virginia, California, and Texas. Not the wonkiest piece in the stack, but useful for seeing how this issue is starting to reach normal people, which is where politics usually begins sharpening its little teeth.

• Financial Times: Is Gas Making a Comeback on the U.S. Power Grid?
  The FT looks at the surge in gas capacity in interconnection queues, with gas back in the conversation because data centers and industrial load need firm power faster than the existing clean-energy buildout can always deliver it. The uncomfortable reality: the AI boom is reviving arguments that many policymakers thought they had already retired, laminated, and placed gently in a drawer.

• Scientific American: NASA’s Plan for a Nuclear Reactor on the Moon
  A fun one for the nuclear file. NASA is looking at lunar nuclear power as part of the infrastructure needed for longer-term moon operations, with China and Russia also eyeing lunar reactors. The lesson travels back to Earth nicely: wherever humans want durable activity beyond the easy reach of the grid, firm power becomes the main character.

Major Stories

AUCTIONS
PJM Moves Up Its Backstop Auction

PJM now plans to hold its backstop reliability auction in September instead of March, accelerating a one-time procurement meant to help meet rising data center demand. The auction was originally part of a two-part process that included bilateral contracting between large loads and suppliers, but PJM’s board now says the backstop and its “connect and manage” large-load rules are too intertwined to keep on separate tracks.

The key sentence is the one about cost allocation. PJM urged states to immediately create frameworks to keep residential and existing customers from absorbing data-center-driven auction costs, but analysts still say it is unclear how those costs can be assigned only to hyperscalers. Translation: everyone agrees Big Tech should pay for the rescue boat, but nobody has yet figured out how to keep the ticket from being stapled to everyone’s utility bill.

Grid Take: PJM is trying to buy time because the normal market machinery is not moving fast enough. That may be necessary, but every emergency workaround becomes dangerous if the bill gets socialized while the benefits get privately contracted.

FLOATING DATA CENTERS
Quick Primer: How Floating Data Centers Work

Peter Thiel led a $140 million investment in Panthalassa, a startup trying to build wave-powered floating AI data centers. The company’s idea is to put compute offshore, use wave motion to generate electricity, cool the equipment with seawater, and move at least some AI infrastructure away from land, local grids, and county zoning boards with microphones.

• The basic idea: Instead of plugging a data center into a stressed land grid, the platform generates power from ocean-wave motion and runs compute directly onboard.

• The shape: Panthalassa has described large autonomous offshore structures, roughly “lollipop”-like, that convert wave motion into electricity through internal systems.

• The cooling advantage: The ocean is a giant heat sink, which helps with one of the least romantic but most important parts of AI infrastructure: keeping machines from cooking themselves.

• The connectivity problem: Offshore compute still has to move data. Reports say Panthalassa’s approach relies on satellite connectivity, including Starlink, which makes the project part energy concept, part telecom concept, part naval weirdness.

• The real question: Floating data centers avoid some land, water, and grid fights, but they create new ones around reliability, maintenance, storms, permitting, marine impacts, latency, and whether anyone actually wants their AI infrastructure bobbing around like a sovereign Roomba.

Grid Take: Floating data centers sound ridiculous until you look at the current interconnection queue. When land-based power gets slow, expensive, and politically radioactive, capital starts looking for stranger geography.

PJM ISSUES
PJM Paid Nearly $1 Billion to Money-Losing Plants in Q1

Reuters reports that PJM paid nearly $1 billion in uplift credits to power plants in the first quarter, a record payout tied to winter cold, fuel price spikes, and plants that were dispatched even though market revenues did not cover their costs. The first-quarter total of $990 million exceeded the combined annual uplift total for all of 2025 and dwarfed the $270 million paid in all of 2024.

This is the unglamorous underside of reliability. During stress events, PJM needs old, expensive, sometimes barely economic plants to run because the alternative is worse. Maryland’s Chalk Point, with gas units more than four decades old, received $205 million in uplift payments, the most of any single plant, because in a constrained pocket around Baltimore and Washington, D.C., an old plant can still be the thing standing between a bad day and a voltage-collapse seminar.

Grid Take: Uplift payments are the grid’s pawn-shop receipt: proof that someone waited too long, then had to pay whatever the system demanded under stress. The answer is not to sneer at old plants for being old; it is to build enough new supply and transmission so the old plants stop holding the system hostage.

DATA CENTERS
Data Centers Could Eat a Third of Commercial Building Power

EIA now projects that data center servers could account for 22% to 33% of commercial building electricity use by 2050, depending on the scenario. In its high-demand case, server electricity consumption reaches 818 billion kWh in 2050, more than 16 times the 2020 level. Across all cases, servers already accounted for an estimated 7% of commercial sector electricity consumption in 2025.

That helps explain why the commercial sector keeps becoming the center of gravity for electricity demand. EIA expects commercial building energy intensity to surpass its 2003 historical high in the early 2030s, with data centers pushing the sector in the opposite direction of the old efficiency story. For years, the commercial building story was LEDs, better HVAC, and smarter controls. Now it is also AI racks with the appetite of a small industrial republic.

Grid Take: Efficiency will matter, but it will not erase the load. The new commercial sector is not just offices, malls, and warehouses; it is compute infrastructure, and the grid will have to plan for buildings that behave less like buildings and more like factories.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• PJM got permission to pull data centers off the grid before blackouts. DOE granted PJM emergency authority to curtail data centers and other large loads with backup generation during this week’s hot-weather stress, but only as a last resort before rolling blackouts. That is a giant little sentence: data centers are no longer just “new load” in PJM. They are now a controllable emergency resource, which is what happens when the grid runs out of polite fiction.

• PJM moved the backstop auction up because March is too late now. PJM now plans to hold its backstop reliability auction in September instead of March to address rising data-center demand. The unresolved problem is cost allocation: PJM wants states to protect existing customers from data-center-driven auction costs, but nobody has fully solved how to keep hyperscaler capacity needs from washing through ordinary bills.

• PJM’s winter bill came due at almost $1 billion. Reuters found PJM paid a record $990 million in uplift payments to money-losing plants in Q1, more than the $764 million paid in all of 2025 and far above the $270 million paid in 2024. This is the grid’s most expensive kind of confession: when the system gets tight, old plants with ugly economics suddenly become priceless.

• NERC says the summer looks manageable, but the shoulders are getting weird. NERC says the U.S. power grid should have enough resources for typical summer demand, helped by new additions, but risk is growing in shoulder seasons as demand patterns, maintenance schedules, and data center interconnection delays make forecasting harder. The old calendar assumed spring and fall were the quiet rooms of the grid. AI load is dragging a drum kit in there.

• South Carolina’s data center politics are turning Republican-on-Republican. Rep. Nancy Mace is pushing for a statewide data center moratorium, joining a growing bipartisan pause-the-buildout movement driven by utility-cost concerns. That matters because data center skepticism is no longer just progressive environmentalism or rural NIMBYism. It is becoming right-flank ratepayer politics with a hard hat and a microphone

• Texas may be planning water as if AI never showed up. E&E reports that Texas officials are scrambling to prepare for AI-serving projects, but the state’s draft water outlook skips the data center surge. That is the Texas growth model’s blind spot in miniature: the power story gets all the attention, but the water ledger is sitting quietly in the corner with a knife.

• Storage just had a monster quarter. U.S. developers installed 9.7 GWh of energy storage in Q1, a first-quarter record and up 32% year over year, with SEIA pointing to data centers, volatile power prices, and gas-market disruption as demand drivers. Storage is not firm power, but it is becoming the grid’s shock absorber, and the AI buildout is buying every cushion it can find.

• California may fumble one of its best flexibility tools. Newsom’s revised budget funds California’s Demand Side Grid Support program through this year but zeroes it out in 2027, moving participants toward a utility-run framework that clean-energy groups oppose. The program dispatched more than 539 MW during a test last July, which is exactly the sort of flexible capacity everyone says they want until the budget scissors come out.

Chart of the Day

Natural gas is still doing the quiet heavy lifting.
NERC’s 2026 Summer Reliability Assessment shows fewer regions at elevated risk than last summer, helped by new solar, storage, and gas-fired capacity. But the attached chart tells the more important story: 2026 summer demand is still rising, with net internal demand up roughly 11 GW and total internal demand up roughly 15 GW from 2025’s actual summer peak around 900 GW.

That means the reliability outlook has improved, but not because demand cooled off. It improved because the system added enough resources to stay ahead of the load curve, and natural gas remains the flexible, dispatchable piece that keeps the whole thing from turning into a weather-dependent trust exercise.

Grid Take:
The clean-energy buildout matters, but summer reliability still runs through gas. The grid can add solar and batteries all day, but when demand climbs and the margin gets thin, dispatchable fuel is still the adult in the room.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

Americans want the benefits of AI, but not the machinery. Gallup now finds 71% oppose an AI data center in their area, making local server farms more politically toxic than nuclear plants. That is a remarkable achievement for a building whose main offense is using electricity, water, and enough concrete to make a county commissioner feel briefly important.

The deeper story is not just NIMBYism. PJM is admitting the grid has moved from surplus to scarcity, FERC’s chairman is questioning whether the nation’s largest power market is governable, and Eversource is openly resisting data centers to protect customers from higher prices. The AI boom has finally reached the physical world, where the future must compete with substations, gas turbines, transmission studies, utility bills, and the ancient American right to hate whatever is being built nearby.

The Lede

M&A
NextEra and Dominion Build a Utility for the AI Age

NextEra Energy plans to acquire Dominion Energy in an all-stock deal that would create what the companies describe as the world’s largest regulated electric utility by market capitalization. The combined company would serve about 10 million utility customers across Florida, Virginia, North Carolina, and South Carolina, with a $138 billion rate base and more than 130 GW of large-load opportunities in its pipeline. Dominion shareholders would receive 0.8138 shares of NextEra for each Dominion share, and the companies expect the deal to close in 12 to 18 months if regulators approve it.

The deal is a data center story wearing a merger suit. Dominion sits in Virginia’s data center alley and recently said its contracted data center capacity pipeline stands around 51 GW. NextEra brings one of the country’s largest generation development platforms, including gas, renewables, nuclear, and storage. Analysts immediately framed the merger as a possible return toward the integrated utility model, especially in PJM, where resource adequacy is becoming less a market-design problem and more a civilization-maintenance hobby.

Grid Take: The market is starting to reward utilities that can offer large customers a complete answer: power, wires, financing, permitting, and speed. If the AI load boom is real, the winning utility model may look less like a passive wires company and more like an infrastructure machine with a balance sheet.

Things to Check Out

• New York Times: Congress Looks at Annual EV Fees
  House lawmakers are considering annual fees for electric vehicles and plug-in hybrids to help fund roads as gas-tax revenue weakens. The politics are obvious: EV drivers use the roads, too, but Congress is now trying to retrofit a gasoline-tax system for a post-gasoline fleet without admitting the old funding model is wheezing in the garage.

• Wall Street Journal: Rick Perry’s Nuclear Fight in Texas
  The nuclear-data center gold rush is now producing its own Texas-sized corporate drama. Rick Perry’s Fermi America pitch captures the moment perfectly: Wall Street wants nuclear-powered AI, but building reactors remains harder than selling the dream to investors in a bolo tie.

• ABC News: Cuba’s Grid Crisis Grows Dire
  Cuba’s energy crisis is worsening as fuel shortages and grid instability produce widespread blackouts. It is a grim reminder that electricity is not just an economic input or climate-policy category. When the grid fails, everything else starts failing with it.

• RealClearEnergy: Two Actions State Leaders Can Take to Avert the Coming Blackouts
  Jon Sanders argues state leaders should stop forcing reliable plants off the system and remove barriers to new firm generation. It is a blunt piece, but the basic warning is hard to dismiss: demand is rising, dispatchable supply is under pressure, and hoping the math gets friendlier is not a reliability strategy.

• BGR: China’s Offshore Flying Wind Turbine
  China is testing airborne offshore wind concepts that sound ridiculous right until one of them works. Not every moonshot matters, but the global energy race is increasingly a contest between countries willing to build strange machines and countries still convening listening sessions about whether a substation has vibes.

• Our World in Data: China Added a Germany-Sized Grid Last Year
  China added roughly a Germany-sized amount of electricity generation in a single year, with most of the new generation coming from wind and solar. That is the scale problem in one chart: while the U.S. debates whether infrastructure feels too large, China keeps adding whole industrial countries to its power system.

Major Stories

LARGE LOADS
Pennsylvania Writes the Large-Load Tariff Playbook

The Pennsylvania Public Utility Commission released a “first-of-its-kind” model tariff framework for large-load customers, including data centers. The order is nonbinding, but it recommends applying large-load provisions to customers exceeding 50 MW individually or 100 MW in aggregate, and says utilities should charge those customers for system upgrades that would not have been needed “but for” their interconnection, even if other customers eventually benefit from the infrastructure.

That is the sentence every ratepayer advocate has been waiting to hear. The framework also covers collateral, deposits, stranded-asset risk, load ramping, minimum contract terms, exit provisions, and public-facing information on interconnection requests. It even allows large-load customers to self-construct some upgrades, which is a useful concession to the obvious reality that hyperscalers can often move faster than the utility process, a machine built partly from paper, partly from caution, and partly from the fossilized remains of 1970s hearing rooms.

Grid Take: Pennsylvania is trying to separate “economic development” from “please make grandma finance my substation.” That is exactly where the data center fight has to go next: not yes or no, but yes under terms that protect existing customers and still let serious projects build.

POWER OUTLOOK
Commercial Power Use Is About to Pass Residential

EIA now expects commercial electricity consumption to surpass residential use for the first time on record in 2027. The commercial sector, which includes hyperscalers, bitcoin miners, and cloud computing, is expected to reach about 1,530 billion kWh in 2026, roughly equal to residential consumption, before growing another 5.3% in 2027. Total U.S. electricity consumption is expected to rise 1.3% in 2026 and 3.1% in 2027.

Residential demand, by contrast, is expected to remain mostly flat, growing about 0.5% in both 2026 and 2027. But residential customers will still feel the shift. EIA expects average residential prices to rise to 18.2 cents/kWh this year, nearly 5% above 2025, with another 2% increase expected next year. The East Coast is expected to see some of the largest increases, with utilities citing fuel costs, transmission investment, extreme-weather hardening, and rising demand.

Grid Take: This is the quiet statistical handoff that explains the whole decade. The American grid was built around households and legacy industry; it is now being reorganized around compute, manufacturing, electrification, and commercial load that arrives in chunks large enough to make planners blink twice.

EPA GUIDELINES
EPA Cites AI Demand in Coal Wastewater Rollback

EPA proposed revisions to wastewater rules for coal-fired power plants, including rescinding treatment requirements tied to unmanaged combustion residual leachate from coal ash disposal sites. The agency argues the rollback could reduce electricity-generation costs by up to $1.1 billion annually and says growing electricity demand from AI infrastructure and data centers “cannot be met under the overly restrictive policies of past administrations.”

Critics, including Earthjustice, say the change would eliminate safeguards on wastewater containing arsenic, mercury, and other dangerous contaminants. That makes this a revealing story less because of the specific wastewater rule and more because of the political logic now forming around AI load growth. Demand is becoming the universal solvent in energy policy, used to justify more generation, slower retirements, faster permitting, regulatory rollbacks, and probably, by next Tuesday, a county commissioner’s nephew getting a consulting contract.

Grid Take: The AI power crunch is going to reopen fights people thought were settled, including coal regulations, gas permitting, transmission siting, nuclear licensing, and water use. The danger is that “we need more power” becomes either a blank check for bad policy or a veto against building anything at all.

NEW REPORT
NTU Offers a Ratepayer Playbook for the AI Race

The National Taxpayers Union released Speed to Power: How Electricity Ratepayers Can Win the AI Race, a wide-ranging policy guide from Nick Loris with contributions from Travis Fisher, Josh Smith, Katherine Wright, and Pete Sepp. The report argues that large data center demand can strain the grid, but it can also reduce residential rates by spreading fixed utility costs across a broader customer base and underwriting upgrades that benefit the shared system, if policy gets the incentives right.

The strongest parts of the paper are its practical framework: permitting reform, fair cost allocation, better load forecasting, more competition, data center flexibility, water-rights reform, and a proposal for Consumer Regulated Electricity, which would allow privately financed, physically islanded power systems to serve voluntary customers like data centers or industrial facilities. It is a useful antidote to the two lazy positions now dominating the debate: pretend data centers are free money, or treat them as invading machinery from a bad science-fiction sequel.

Grid Take:
The NTU paper gets the core question right: data centers are not automatically a subsidy scam or a ratepayer miracle. They become one or the other depending on whether policymakers force accurate cost causation, protect existing customers, and let private capital build where the legacy grid cannot move fast enough.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• PJM prices are now the political problem. PJM power prices jumped almost 76% year over year in the first quarter, with the market monitor pointing to data-center demand as a major driver. That turns the AI load fight from an abstract planning debate into the thing customers actually understand: the bill showed up, and it did not arrive wearing a tasteful innovation lanyard.

• The xAI gas-turbine fight is getting federal attention. DOJ says it may intervene in the NAACP’s lawsuit over Elon Musk’s xAI and its alleged operation of gas turbines in Southaven, Mississippi, to power a nearby data center. This is the next frontier of the buildout: not just whether data centers can find power, but whether emergency-style, behind-the-fence generation becomes a legal and environmental knife fight.

• Tahoe is learning what a crowded power market feels like. A California utility serving the Lake Tahoe area has been cut off from its power supplier and now has to find electricity in a market where data centers are soaking up available supply. That is the uncomfortable new order: small communities, utilities, and large-load buyers all reaching for the same scarce electrons, with the data centers often showing up with bigger wallets and shorter timelines.

• Ratepayer anxiety is no longer background noise. Utilities requested $9.4 billion in rate increases in the first quarter of 2026, while investor-owned utilities plan roughly $1.4 trillion in capital spending by 2030 to handle rising demand, much of it driven by data centers. The public may not follow capacity markets or interconnection studies, but it understands the monthly bill, which is where all these elegant policy theories go to be mugged

• Cyber risk is moving below the obvious targets. Utility Dive notes that the grid was designed to handle the loss of a major facility, but is increasingly exposed to many smaller resources failing at once. That matters in a more distributed, digitized, AI-adjacent grid, where “low impact” assets can still create high-impact trouble if enough of them start blinking red together.

• EPA is making speed part of the permitting fight. E&E reports that EPA has proposed allowing data centers, power plants, and other industrial facilities to begin some construction before receiving required federal air permits. That is the energy-abundance debate in miniature: build faster, but decide how much regulatory certainty has to exist before the excavators start chewing up dirt.

Generation Dive

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

Americans want the benefits of AI, but not the machinery. Gallup now finds 71% oppose an AI data center in their area, making local server farms more politically toxic than nuclear plants. That is a remarkable achievement for a building whose main offense is using electricity, water, and enough concrete to make a county commissioner feel briefly important.

The deeper story is not just NIMBYism. PJM is admitting the grid has moved from surplus to scarcity, FERC’s chairman is questioning whether the nation’s largest power market is governable, and Eversource is openly resisting data centers to protect customers from higher prices. The AI boom has finally reached the physical world, where the future must compete with substations, gas turbines, transmission studies, utility bills, and the ancient American right to hate whatever is being built nearby.

The Lede

POLLING
Americans Sour on AI Data Centers

Gallup’s first major survey on local AI data center construction found that 71% of Americans oppose building one in their area, including 48% who strongly oppose it. That is worse than nuclear power. In the same survey, 53% opposed a local nuclear plant, meaning AI infrastructure now carries a heavier local political burden than the technology that spent decades as the national mascot for public fear.

The opposition is broad and intense. Majorities oppose local data centers across party, age, income, race, education, and geography, with Democrats the most strongly opposed and the Midwest and South showing the highest total opposition. The reasons are familiar from local fights: water, electricity, noise, traffic, land use, higher utility bills, and a general suspicion that the benefits go somewhere else while the infrastructure lands next door.

Grid Take: The AI industry has a real case to make, but it cannot make it with server-rack glamour shots and sleepy talking points about jobs. It has to explain why load growth can support new generation, why large customers can improve system economics, and why the alternative is not a pristine meadow, but a slower, weaker future built somewhere else.

Things to Check Out

• National Interest: A New Regulatory Paradigm for Advanced Nuclear Energy
  Nick Loris makes the right argument here: advanced nuclear will not scale on enthusiasm alone. The piece walks through how NRC reform could affect cost, speed, and competitiveness, and why regulatory modernization may be the difference between nuclear as a load-growth solution and nuclear as another nice idea trapped in process.

• National Review: The Free-Market Case for Data Centers
  A timely counterweight to the Gallup numbers. The case for data centers is not that every project is perfect, but that banning new demand is a strange way to build prosperity in a country that still claims to like growth.

• Knowledge Problem: An attempt to remedy the consequences…
  Lynne Kiesling on capacity mechanisms, demand inelasticity, and the institutional choices that created many of the market problems regulators now try to solve with more rules. Read this before pretending PJM’s capacity fight is just a procurement issue.

• Department of Energy: The History of Electricity at the White House
  A charming reminder that every new energy technology begins its public life surrounded by fear, confusion, and someone refusing to touch the switch. The data center politics have ancestors.

• Cato: How America Can Unleash the Next Energy Revolution
  Video from Cato’s event with Energy Secretary Chris Wright. The useful frame here is not “energy dominance” as slogan, but supply expansion as policy discipline: more energy types, fewer artificial bottlenecks, and less romance about managed scarcity.

• RealClearEnergy: The Electricity Myth: Data Centers Aren’t the Villain
  A straightforward entry in the data center debate, pushing back on the idea that large-load customers are the sole villain behind rising power prices. Worth reading alongside the Gallup poll because politics and system economics are now badly out of sync.

Major Stories

RTOS
FERC Chair Says PJM May Be “Too Big to Function”

FERC Chairman Laura Swett used PJM’s annual meeting to say what many around the market have been whispering for years: PJM may be too large, too politically divided, and too slow to govern itself through the next era of load growth. She called PJM’s governance structure “unacceptable” and warned that confidence in the grid operator’s decision-making has “completely eroded.”

The timing matters. Swett’s broadside came just days after PJM released a unusually frank market-design paper arguing that the region has moved from managing surplus to managing scarcity, with data centers able to arrive far faster than new generation can be built. PJM laid out three broad reform paths: stabilize the existing capacity market, ration reliability among customer classes or regions, or shift more weight away from the capacity market and toward energy and ancillary-service markets.

Grid Take- PJM is becoming the test case for whether American electricity institutions can survive abundance. Without a governance structure that can move at industrial speed, every fix for generation, interconnection, transmission, and capacity gets fed into the procedural woodchipper.

DATA CENTERS, AGAIN
America Already Has 50 GW of Data Centers Online

FERC’s 2025 State of the Markets report found that the U.S. had more than 50 GW of operating data center capacity at the end of 2025, representing 24% compound annual growth since 2020. MISO saw the fastest growth, with a 43% compound annual rate, followed by ERCOT, SPP, and the Southeast at roughly 28% to 30%.

The facilities are also getting much larger. FERC staff said average data center size rose from about 25 MW in 2020 to nearly 80 MW for projects coming online last year, which turns data centers from ordinary commercial load into system-planning events. At that size, a project may require new generation, transmission upgrades, substation work, gas coordination, and much better forecasting about which loads are real and which are just queue confetti.

GridTake - The 50 GW number moves data centers into grid reality. The country can argue about AI hype all it wants, but the installed base is already enormous, projects are getting larger, and the regions with the fastest growth are where the next fights over gas, transmission, storage, and capacity markets will land.

DATA CENTERS, YET AGAIN
Eversource Draws a Line Against Data Centers

Eversource CEO Joe Nolan said he is “not interested” in data center development in the company’s service territory, arguing that the projects are “of no value” to residential customers and will only drive up energy prices. His comments came during the company’s first-quarter earnings call, where he also said Eversource is “resisting data centers” while relying on offshore wind, hydro, and other clean energy injections to help manage costs in New England.

The posture is striking because New England is already an expensive and constrained power market. Eversource reported that average market prices for Connecticut Light and Power’s wholesale sales rose to $112.71 per MWh in Q1 2026, up from $99.02 per MWh in the same period last year, while the company is also dealing with earnings pressure from FERC’s transmission return-on-equity decision. In that context, Nolan’s comments are more than CEO color. They are a regional strategy: defend existing customers by keeping new large load away.

GridTake - The concern about customer costs is real, especially in New England, where energy policy often resembles scarcity with a better press kit. But resisting load is not a strategy for abundance; it is a strategy for managing decline more politely.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• NERC treats data centers as an immediate reliability risk. NERC issued a rare Level 3 alert, its highest level, after instances of data centers unexpectedly dropping load or rapidly oscillating demand. The grid watchdog is requiring certain planners, operators, and balancing authorities to take seven actions by Aug. 3, which means compute load has officially graduated from “forecasting headache” to reliability paperwork with teeth.

• Congress circles transmission, then remembers Congress. House Energy and Commerce lawmakers held a hearing on grid upgrades and transmission as AI data centers drive the need for more power. Republicans focused on permitting bottlenecks under NEPA, the Clean Water Act, and the Endangered Species Act, while resisting Democratic efforts to give the federal government more authority over transmission siting and planning. In other words: everyone agrees the grid needs to move faster, then immediately disagrees over who gets the steering wheel.

• Electricity demand forecasts keep getting steeper. The National Electrical Manufacturers Association now expects U.S. electricity consumption to grow more than 55% by 2050, with the steepest growth concentrated in this decade. NEMA also expects data center energy consumption to rise 300% over the next 10 years, which is the sort of forecast that makes “just improve efficiency” sound less like a plan and more like a scented candle.

• PPL’s Pennsylvania data center pipeline keeps swelling. PPL says its “advanced” data center pipeline in Pennsylvania has grown to 28.3 GW by 2034, up from 25.2 GW just three months earlier. The company is also working with Blackstone Infrastructure on generation for data centers and reserving gas turbines, which is exactly what serious large-load planning looks like when it leaves the slide deck and starts hunting for machines.

• Storage is becoming data center plumbing. Fluence signed master supply agreements with two major hyperscalers, while its data-center-related order pipeline rose 30% from the prior fiscal quarter. That does not make batteries a substitute for firm power, but it does show how hyperscalers are starting to buy the flexibility products the grid increasingly needs.

• Ford finds a second life for batteries in the grid. Ford launched a new subsidiary to manufacture and sell batteries for the grid and data centers, targeting 20 GWh of annual production with first deliveries planned for late 2027. The EV slowdown is starting to redirect industrial capacity toward the power system, which is either clever adaptation or the most expensive corporate pivot table in America.

Quote of the Day

"Every nation has their own electricity and water. Every nation should have their own AI infrastructure." — NVIDIA’s Liron Freind-Saadon

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The AI power story is no longer theoretical. PJM is fast-tracking 11.8 GW of new resources. Texas is staring at 410 GW of load applications, about five times its current peak demand. And the American Capacity Project is trying to put a human number on what happens when states and localities say no to the infrastructure needed to serve new load.

That is the real story underneath the week’s grid news. Data centers are not just testing interconnection queues. They are exposing a basic governing problem: America has become very good at announcing demand and very bad at permitting the capacity needed to meet it.

The Lede

INTERCONNECTION
PJM Selects 51 Projects for Fast-Track Interconnection

PJM approved 51 projects for fast-track interconnection, totaling 11.8 GW of new capacity. The mix tells you almost everything about the moment: 69% gas, 19% battery storage, and 12% nuclear upgrades. The projects were selected from a much larger pool of applications and screened for readiness, site control, and the ability to actually get built rather than simply decorate the queue.

The driver is not subtle. Data centers have become the largest source of new load growth in PJM territory, with the RTO expecting roughly 30 GW of incremental demand from the sector by 2030. The normal interconnection queue is too slow for that world. PJM’s backlog still contains more than 2,000 projects, many of which will not be studied, financed, permitted, or built on a timeline that bears any relationship to the load now showing up.

Fast-tracking 11.8 GW does not fix the queue. It does, however, mark a useful change in posture. PJM is no longer pretending the old process can absorb the AI load boom with a few stakeholder meetings and some polite revisions to the spreadsheet.

GridTake - The gas-heavy mix will irritate the usual people for the usual reasons. But PJM is not writing a climate plan here. It is making a reliability call under ugly constraints. If the demand is real and the grid is tight, the choice is not between gas and a frictionless clean-energy buildout. It is between dispatchable capacity, delayed campuses, or a system that keeps making promises it cannot electrify.

Things to See

• EVENT TOMORROW: Cato: How America Can Unleash the Next Energy Revolution
  Secretary of Energy Chris Wright joins Cato’s Travis Fisher for a fireside chat on energy abundance, deregulation, and the next wave of American energy production. The event is May 13 from 2–3 PM EDT at the Cato Institute in Washington, D.C., with a reception to follow.

• CSIS — China's Nuclear Energy Priorities Under Its 15th Five-Year Plan China is planning to add 10+ reactors per year through 2030. The U.S. permitted two in the last decade. Read this and then think about what "energy dominance" actually requires.

• Latitude Media — Can Quantum Computing Help Solve the Load Growth Problem? The pitch: quantum optimization could cut transmission congestion costs by 30-40%. The honest answer: not before 2030, and probably not at grid scale. Worth reading to calibrate the hype.

• Heartland: Wind and Solar Focus Threaten U.S. and South African Economies
  Ronald Stein uses South Africa’s new resource plan as the warning label for variable-heavy grids: wind and solar may be cheap at the project level, but the system cost shows up in backup capacity, transmission, storage, and industrial risk.

• Financial Times: Can Europe Seize the Opportunity in Nuclear Fusion?
  The FT looks at whether Europe can keep pace in fusion as private capital, grid interconnection requests, and commercialization timelines start moving faster in the U.S.

Major Stories

DATA CENTERS
American Capacity Project Puts a Human Cost on Capacity Delays

The American Capacity Project’s new white paper, Capacity Delayed, Cures Deferred, argues that data center moratoria and electric-capacity delays are not just local land-use fights. They are constraints on the compute infrastructure now powering AI-enabled drug discovery, a pipeline that has doubled every 1.6 years for nine straight years and now includes more than 200 AI-originated drug candidates in clinical development. ACP estimates that moratoria, permitting delays, tax reversals, and interconnection barriers blocked or delayed more than $156 billion in planned compute infrastructure in 2025 alone.

The paper’s headline estimate: those delays destroyed or deferred about 18,480 megawatt-years of AI-capable compute and could defer 1.70 million healthy life-years over the next decade if the AI drug pipeline keeps growing at its recent pace. The argument is not that every data center cures cancer. It is that policymakers weighing a “temporary pause” on data centers are also, whether they know it or not, slowing the infrastructure behind faster drug discovery, clinical modeling, and medical innovation.

GridTake - This is the right argument because it makes “pause” defend itself as a policy choice, not hide behind the soft lighting of process. The model will get challenged, and it should, but ACP is putting the missing cost in the room: the cure that arrives late because America treated the infrastructure behind discovery as optional clutter.

INTERCONNECTION
Texas Is Stress-Testing the Grid Queue

ERCOT is now sitting on roughly 410 GW of load interconnection applications. Texas currently peaks around 85 GW. About 90% of the application volume is tied to data centers, much of it filed since early 2023. That does not mean Texas is about to add five grids’ worth of demand. It means the queue has become a speculative instrument.

A meaningful share of those applications are not mature projects. Some have not secured land. Some have not solved water. Some may be placeholders filed by developers who want optionality before they have certainty. Texas has limited penalties for queue withdrawal, so applications pile up because there is little reason not to file.

The real number is almost certainly much smaller, but still enormous. If only 10% of the queue materializes, ERCOT is dealing with a load shock larger than many states’ entire power systems. The Texas Energy Fund is helping bring new dispatchable capacity into the pipeline, and ERCOT has added new reliability tools, but neither solves the basic problem: the market is producing demand signals faster than the system can sort real projects from vaporware.

GridTake - The 410 GW number is less a forecast than a confession. It tells us the queue is no longer a planning instrument. It is a waiting room where serious projects, speculative filings, and PowerPoint empires all sit in the same chairs. Texas does not need to study every request as though it is real. It needs a way to make developers prove seriousness before they consume grid-planning oxygen.

CAPACITY MARKET
PJM Puts Three Capacity Market Reform Options on the Table

PJM is also floating three broad approaches to capacity market reform, and none of them are painless. One option would patch the current Reliability Pricing Model with changes to demand curves and offer rules. Another would allow prices to better reflect scarcity, which is economist-speak for “stop pretending reliability is cheap when it is not.” A third would move away from the standalone capacity market and toward a model that relies more heavily on energy and ancillary services pricing.

This is not an academic debate. PJM’s capacity market clears billions of dollars across a 13-state footprint and helps determine which plants stay open, which projects get financed, and how much consumers pay for reliability. The market was designed for a more stable demand world. That world is gone. Coal retirements, electrification, industrial load growth, and data centers have turned capacity planning into a live-fire exercise.

GridTake - PJM will probably choose the least disruptive reform it can defend, then spend two years arguing over the details. That is how these things go. But the deeper problem remains: a capacity market that pays for theoretical reliability while the real grid gets tighter is not a market. It is a very expensive reassurance ritual.

DATA CENTERS
Stockland Files for 250 MW Data Center Campus in Melbourne

Australian developer Stockland has filed applications for a 250 MW data center campus in Melbourne’s outer suburban corridor, with operations targeted in phases beginning in 2027. The project fits a pattern now visible across the Anglophone grid map. Primary data center markets are saturated or politically constrained, so developers push into secondary cities where land may be easier but transmission is not.

Melbourne is not Northern Virginia. But the infrastructure math rhymes. Large loads arrive faster than grid upgrades. Developers can announce campuses in months. Transmission planning takes years. Permitting can take longer. The gap between those timelines is where most of the political pain now lives.

GridTake - Every region thinks it has more time than it does. Then the applications arrive, the load forecasts jump, and the grid operator starts using phrases like “reinforcement needs.” Melbourne is just another reminder that the data center buildout is not a U.S. story. It is a global capacity test.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• FERC's 2025 Summer Assessment projects above-normal risk in MISO South and portions of PJM East through August — a combination of higher-than-expected load growth and thermal retirements that outpaced replacement additions.

• PPL's data center pipeline in Pennsylvania has grown to 28.3 GW of active applications — roughly equal to the utility's entire current peak demand. PPL has begun requiring non-refundable deposits at application to reduce speculative filings.

• MISO is seeing the strongest data center growth of any RTO as of year-end 2025, per FERC's State of the Market report, with 50 GW of data centers online nationally and the bulk of new applications concentrated in MISO's Upper Midwest footprint.

• ReliabilityFirst's 10-year assessment flags resource adequacy risk across both MISO and PJM territories through 2034, driven by accelerating retirements and interconnection queue delays — a combination that creates a reliability gap even in optimistic build-out scenarios.

• Accelerate Infrastructure Opportunities closed a $630M fund targeting critical infrastructure expansion, with energy and data center infrastructure listed as primary deployment targets.

• Florida enacted legislation governing data center development — covering ratepayer cost allocation, water use permitting, and local zoning authority — the first state-level framework of its kind. Expect other high-growth states to follow.

Quote of the Day

"The queue itself is the problem. You can fast-track 51 projects and still have 2,000 more sitting there. At some point you have to redesign the process, not just run it faster." — PJM stakeholder comment, queue reform docket, April 2025

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

A pivot in New Jersey's nuclear strategy, potential withdrawal from major U.S. power grids, and newly issued alerts underscore the evolving complexities of grid reliability in the face of emerging technologies and energy demands. As executives and policymakers juggle these developments, understanding the implications is critical for navigating the future of energy.

The Lede

NUCLEAR
PSEG CEO: New Jersey Nuclear Opening Is Real, But Federal Money and Hyperscaler Deals Have to Come First

PSEG Chair, President, and CEO Ralph LaRossa said the nuclear outlook in New Jersey has materially improved after Governor Mikie Sherrill signed legislation in early April 2026 lifting the state's longstanding moratorium on new nuclear builds and establishing a nuclear task force to advance development. PSEG Nuclear already anchors the state's grid, owning 3,758 MW of capacity across New Jersey and Pennsylvania, producing roughly 40% of New Jersey's electricity at a 95.8% capacity factor, well above the national average.

The commercial logic behind new builds runs straight through hyperscalers. PSEG's pipeline of potential large load customers, overwhelmingly data centers, jumped 47% to 9.4 GW as of June 2025, but CFO Daniel Cregg acknowledged the limits of the pitch without better state-level incentives: "Absent the significant tax incentives in New Jersey, you have not seen the sizable interest in New Jersey." Long-duration offtake agreements from creditworthy buyers are what turn a nuclear business case from a sketch into a financing package, and without competitive tax treatment, New Jersey risks watching that load land in PJM territory where the subsidies are richer. The task force and the lifted moratorium clear a political hurdle; the capital stack, the federal production tax credit runway, and the hyperscaler term sheets are what actually determine whether a shovel goes in the ground.

GridTake
New Jersey just removed a ban it had no particular reason to keep, and the industry is celebrating a permitting process that does not yet exist for plants that have no financing. LaRossa is right that the moratorium lift matters, but the real news is that 9.4 GW of data center load is circling the state and may go elsewhere if Trenton cannot match what competing states are offering on tax incentives.

Things to Read

• Heatmap News — Exclusive: Local Opposition to Data Centers Explodes in 2026
  This report highlights how community pushback is reshaping the landscape for data center development.

• Canary Media — America’s nuclear dry spell is over
  Read about the significant revival of nuclear energy with new projects finally coming online.

• Heatmap News — Is This the End of the Utility As We Know It?
  This article delves into critical discussions on the evolving regulatory landscape for utilities.

• Austin American-Statesman — Bring Your Own Power: Texas Data Centers
  An exploration of the escalating energy demands from Texas data centers and their implications for local grids.

• Breakthrough Institute — Five 'Pro-Nuclear But' Myths
  Delve into debunking common misconceptions about nuclear energy that impede its acceptance in today’s market.

Major Stories

UTILITIES
AEP Weighs Exit From PJM and SPP as Data Center Queue Backlog Strains Its Patience

American Electric Power CEO Bill Fehrman put the grid establishment on notice Tuesday, saying the Columbus-based utility is actively weighing withdrawal from both PJM Interconnection and the Southwest Power Pool over what he called a failure to move fast enough on connecting new AI data centers. "We have to solve the speed to market issue," Fehrman said on AEP's earnings call, a line that doubles as a barely veiled ultimatum to the two largest RTOs in the country. AEP has signed agreements for 7 gigawatts of new data center projects and now holds contracts for 63 GW of new large load by 2030, up 13% from 56 GW just three months ago, against a five-year capital plan that just grew 8% to $78 billion.

The friction is straightforward: AEP is moving faster than its grid operators will let it. AEP Ohio alone holds binding data center contracts for 5,642 MW of load on top of 12,219 MW signed before its new Data Center Tariff took effect in February 2026, and the company needs interconnection timelines to match the pace of its customers. PJM, which coordinates wholesale electricity across parts of 13 states and the District of Columbia, and SPP together cover most of AEP's service territory, so any exit would be operationally complex and politically loud. Fehrman also told investors that "AEP is executing on our strategic plan at an exceptionally high level during a time of unprecedented opportunity," which reads less like reassurance and more like a company that knows it has leverage and intends to use it.

GridTake - A $72 billion utility threatening to walk away from the two biggest grid operators in America is not a negotiating footnote, it is a stress test of whether RTOs built for a slower era can process demand at AI speed. If PJM and SPP cannot fix their interconnection queues fast enough to satisfy one of their largest members, expect others to ask the same question out loud.

NERC
NERC Pulls the Alarm on Data Center Load Risk, Issues Rare Level 3 Alert

NERC issued a Level 3 reliability alert this week targeting what it calls "unprecedented challenges from a surge in large power consumers," mandating that transmission planners, planning coordinators, and transmission operators complete seven specific actions by August 3. The alert centers on a structural blind spot: of approximately 33,282 MW of currently operational data centers evaluated, 25,504 MW rely on inherent ZIP model representation in stability software, meaning grid operators have been flying largely without accurate behavioral data on how these loads actually perform during disturbances. NERC found that "entities generally did not have sufficient processes, procedures, or methods to address emerging computational loads," a candid admission that the interconnection process never caught up with the AI buildout. Mandated steps include developing detailed modeling data lists, establishing commissioning processes for computational loads, and installing dynamic fault recording devices to capture how data center facilities behave during system events.

The urgency is grounded in real incidents. A transmission fault in the Eastern Interconnection caused the simultaneous loss of approximately 1,500 MW of voltage-sensitive load, primarily data centers, and in 2024 and 2025, separate events each saw 1,000 MW or more drop off the bulk power system due to unexpected customer-initiated load reductions. Summer peak demand is projected to rise 24% across the bulk power system over the next decade, with data centers accounting for most of that growth.

GridTake - NERC is essentially admitting that tens of thousands of megawatts of new load were connected to the bulk power system without anyone fully understanding how they behave when the grid hiccups. The industry most aggressively demanding grid access is, by one expert's account, the least engaged with the regulators managing the consequences.

NUCLEAR
Nuclear Hits 41% of TVA's Power Mix as Output Surges Year Over Year

The Tennessee Valley Authority's nuclear fleet now supplies 41% of its power, up from 31% in the same period a year ago, with output climbing to 33,772 GWh in the first half of fiscal year 2026 from 25,901 GWh in H1 FY2025. Interim President and CEO Mike Skaggs, who is working with the federal government to sharpen TVA's nuclear strategy, put the position plainly: "TVA is a pro-nuclear organization. We operate one of the largest nuclear fleets in the nation, and nuclear is a good investment to meet our future energy needs."

TVA's ~8,000 MW nuclear fleet, spread across seven units at three sites, ranks third largest in the country inside a total generating portfolio of nearly 35,000 MW. The stronger nuclear performance helped push net income to $658 million in the first half of FY2026, up from $533 million in the same period of FY2025, even as power sales grew only 1% year over year. TVA also has 3,770 MW of new generation currently under construction. Coal is not entirely exiting the picture: CFO Tom Rice noted that "we are taking steps to continue operations of TVA's coal fleet, subject to all applicable permits and regulatory requirements," a signal that the portfolio transition remains a work in progress rather than a clean pivot.

GridTake - A federally owned utility running the third-largest nuclear fleet in the country posting a 30% jump in nuclear output and a 23% jump in net income in a single year is a data point that should embarrass the narrative that new nuclear is the only nuclear story worth telling. Meanwhile, the CFO quietly keeping the coal fleet alive while the CEO touts nuclear investment is the kind of institutional hedge that rarely makes the press release.

NATURAL GAS
NYISO Wants Brooklyn's Barge Peakers Running Through 2029

The New York Independent System Operator is recommending that two barge-mounted peaking plants in Brooklyn, Gowanus and Narrows, stay online through 2029, citing reliability concerns during summer heat events. Zach Smith, NYISO's Senior Vice President for System and Resource Planning, put it plainly: "Looking forward, we have limited dispatchable capability on the system, and that's going to lead to greater operational challenges that we're quite concerned about."

On paper, New York's supply picture looks comfortable: 40,937 MW of available resources against a forecasted summer 2025 peak of 31,471 MW, a 25.4% margin, with an Installed Capacity requirement of 39,147 MW backed by 37,682 MW of in-state generation and 1,769 MW in net external purchases. But headline megawatts obscure what the grid actually needs when temperatures spike and intermittent resources underperform, which is dispatchable capacity that can be called on command. Nearly 650 MW of New York City peaking capacity retired in 2020 under tighter air emissions rules, and the state has not replaced that flexibility with anything comparably controllable. Keeping fossil-fuel barges in service through 2029 is not a policy choice New York's climate commitments celebrate, but it is the choice NYISO's reliability math is forcing.

GridTake - Load growth is no longer a theoretical tailwind. It’s landing unevenly, and prices are adjusting in real time where the system is least prepared.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• NERC's issuance of a Level 3 alert reflects heightened reliability concerns in the grid, particularly due to the burgeoning load from data centers.

• The potential exit of AEP from major U.S. grids highlights ongoing integration challenges faced by utilities amid rising data center demands.

• TVA's decision to increase nuclear reliance signals a broader trend towards enhancing energy security with traditional generation amid new market pressures.

• The NYISO's recommendation to maintain peaker plant operations through 2029 accentuates the balance utilities must strike between reliability and transitioning to sustainable energy.

• Local opposition to data center projects is causing significant investment losses, indicating a growing public pushback against energy infrastructure that strains grid capacity.

• Policies around nuclear energy are shifting nationally, with states exploring greater financial support mechanisms to revitalize lagging projects.

• The regulatory framework around utilities is becoming increasingly scrutinized, as evidenced by discussions from prominent state leaders about utility cost challenges.

• Challenges in interconnection timelines are pushing utilities to consider unconventional solutions for meeting emerging load demands.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

Data center demand is moving from forecast to friction. It’s showing up in plant deals, turbine backlogs, utility spending, and customer bills. This issue: PJM’s first real capacity tug-of-war with hyperscalers, China’s overstated power advantage, a $1.4T utility buildout, and the supply chain quietly setting the pace of everything.

The Lede

CAPACITY MARKETS
PJM Monitor Warns Gas Plants Could Exit Market for Data Centers

PJM’s market monitor is pushing FERC to block a 1,267 MW gas plant deal, arguing the buyer could redirect those plants away from PJM’s capacity market to serve co-located data centers.

• Assets: 677 MW (IL) + 590 MW (OH)

• Buyer tied to a data center + energy campus redevelopment

• Concern: capacity pulled from PJM → tighter supply → higher prices

This comes as PJM tightens rapidly under data center pressure. Nationally, more than 780 GW of data center projects have been announced against a grid with ~760 GW of peak load. Most of that won’t get built, but even partial realization is enough to strain capacity markets.

ERCOT modeling gives a preview: adding 10 GW of large load without sufficient transmission can push prices up ~34% and drive congestion costs up nearly 5x.

GridTake
The old model assumed generation served the market. The new one is testing whether generation can be pulled out of it. If large loads can secure dedicated capacity, the shared system gets tighter and more expensive by default.

Things to Read

• New York Times — America’s Grid Is Not Ready for What’s Coming
  A wide-angle look at the grid finally making its way into mainstream coverage. Long, but useful for how the story is being framed publicly.

• The Economist — Global Energy Markets Are on the Verge of a Crisis
  A reminder that electricity markets don’t operate in isolation. Fuel constraints and geopolitics still sit underneath everything.

• Commonwealth Fusion Systems — First Fusion Company Files to Connect to PJM
  Filed to connect to PJM. Fusion isn’t commercial, but it’s entering the same queues as everyone else. That shift matters.

• Fortune — Americans Are Starting to Turn on Data Centers
  Public sentiment around data centers is evolving. Jobs still win, but rising electricity costs are starting to register.

• Washington Post — The AI Boom Is Reviving Nuclear Ambitions
  A nuclear plus AI narrative that captures both the ambition and the friction. The timelines remain the quiet issue.

• Times of India — New Fuel Cell Turns Coal Directly Into Electricity
  Experimental coal-to-electricity fuel cell. Interesting concept, early-stage reality.

• Yahoo Tech — AI Systems Are Advancing Faster Than Expected
  Another look at AI acceleration. Worth reading alongside the power demand conversation it implies.

• Herald Review — Want Lower Electricity Prices? Build More Power
  A simple thesis that keeps resurfacing. If supply doesn’t scale, prices will.

Major Stories

GLOBAL COMPETITION
Oxford: China’s Data Center Power Advantage Is Narrower Than It Looks

A new Oxford Institute report finds China’s edge in AI infrastructure is real, but mostly about execution speed, not superior energy economics.

• Data center demand: ~200 TWh (2023) → 300–600 TWh by 2030

• ~11 million racks deployed (up from 3M in 2019)

• Power prices: competitive with U.S., not dramatically cheaper

• Western relocation policies cut costs just 4–12%

The constraints are familiar: fragmented regional grids, limited long-distance transmission, and difficulty moving renewable power from interior regions to coastal demand centers.

GridTake - China builds faster. The U.S. allocates more efficiently. Both systems are hitting limits, just in different places. The advantage will go to whoever solves their bottleneck first.

TAX POLICY
Republicans Move to Extend Clean Energy Credits

A new House bill would extend key credits including 45Y (production) and 48E (investment), reversing accelerated phaseouts.

• $34.8B in clean energy investments canceled in 2025

• Developers citing policy uncertainty as a primary risk

• Low probability near-term, but politically alive

The argument is straightforward: shorten the policy runway and projects stall. The counterargument is less discussed: extending incentives continues to tilt investment toward intermittent generation.

GridTake - This keeps capital flowing where policy points it. The issue is that the system’s binding constraint right now is firm capacity, not incremental solar output.

UTILITY FINANCE
Utilities Plan $1.4 Trillion in Grid Investment

https://powerlines.org/wp-content/uploads/2026/04/0413_PowerLines-CapEx-Report-1.pdf

A PowerLines report finds U.S. utilities are planning $1.4 trillion in capex through 2030.

• Spending up 21% year-over-year

• $31B in rate increases requested in 2025 alone

• Electricity bills up ~40% since 2021

• Nearly half of spend going to transmission and distribution

Data centers are a major driver. 32 of 51 utilities explicitly cite large-load growth, with some forecasts doubling in just a few years. AEP, for example, now expects 56 GW of load by 2030, with nearly 90% tied to data centers.

GridTake - This is the largest coordinated grid buildout in decades. The question isn’t whether investment is needed. It’s whether the structure ensures it’s disciplined, or simply passed through.

POWER PRICING
EIA: Electricity Prices Up 9% Year Over Year

U.S. electricity prices rose to 14.36¢/kWh in February, up 9% from last year. Hot spots:

• Virginia: +26.3%

• Ohio: +21.9%

• Pennsylvania: +19.5%

All sectors saw increases, led by transportation (+23.6%). Demand rose across most states, while generation shifted modestly toward natural gas.

This is showing up most clearly in regions absorbing new load, particularly data centers and electrification growth.

GridTake - Load growth is no longer a theoretical tailwind. It’s landing unevenly, and prices are adjusting in real time where the system is least prepared.

SUPPLY CHAIN
Gas Turbine Backlog Hits 100 GW as Demand Surges

GE Vernova’s turbine backlog has reached 100 GW, reflecting a surge in demand for dispatchable generation.

• Up from 83 GW last quarter

• Prices rising 10–20%

• Lead times extending into multiple years

Gas remains the fastest scalable firm resource, but the supply chain is now a constraint in itself.

GridTake - The industry spent a decade underbuilding firm capacity. Now everyone wants it at once. The constraint has moved from policy to hardware.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Capacity is becoming bilateral — Large loads are starting to secure supply directly

• Announcements ≠ buildout — 780 GW of data centers will be filtered by power and interconnection

• China’s edge is speed — The U.S. advantage remains in markets, not execution

• Capex is accelerating fast — $1.4T doesn’t stay invisible on bills

• Price spikes are regional first — Load-heavy states are feeling it early

• Gas is constrained too — Turbines are now a gating factor

• Transmission still decides outcomes — Without it, price signals do the balancing

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

A lot of this week’s energy news boils down to a familiar problem: the headlines say transition, but the grid still says constraint. Ember says solar and other clean sources met all global demand growth last year. A federal court just reopened the lane for a big slug of U.S. wind and solar projects. New England ratepayers are still fighting over transmission returns from more than a decade ago. And Gallup’s newest polling suggests the public is growing less sentimental and more practical. That usually happens when energy policy stops being abstract and starts showing up on a monthly statement.

The Lede

NEW REPORT
Ember Says Clean Power Met All Demand Growth. That’s True, But Incomplete.

Ember’s latest Global Electricity Review is full of genuinely important data, but some of its biggest headlines need a little unpacking before people start declaring a new era. The topline findings:

• Global electricity demand rose 849 TWh in 2025, or 2.8%

• Low-carbon generation rose 887 TWh

• Solar alone added 636 TWh

• Wind added 205 TWh

• Fossil generation fell 38 TWh, or 0.2%

• Renewables reached 33.8% of global electricity, just above coal at 33.0%

Those are real milestones. Solar growth in particular is stunning. It is now the dominant source of new generation globally, and batteries are beginning to make daytime solar more useful in the evening hours. But there are a few caveats hiding underneath the triumphalism:

• A 0.2% fall in fossil generation is basically flat, not collapse

• Gas generation still rose by 36 TWh

• Nuclear rose only 35 TWh, but still remains one of the few major clean firm sources at scale

• Coal fell globally, but the U.S. actually saw coal generation rise 85 TWh in 2025 because gas got more expensive

• Asia is still the only region where coal remains above renewables, and it accounts for 82% of global coal generation

GridTake
Ember is right that solar is now the leading marginal source of new electricity. But that headline sits on a heavily shaped market. Wind and solar have enjoyed enormous policy preference, while firm generation has spent years fighting permitting drag, fuel politics, and in nuclear’s case, outright institutional hostility. Annual growth charts also flatter variable resources in ways real system operations do not. Even Ember’s own data shows how early the flexibility buildout still is: battery additions in 2025 were enough to shift only about 14% of the new solar generation added that year. The question is not whether wind and solar are scaling. It is whether a system built around favored intermittency and discouraged firmness can stay affordable and reliable once the hard hours arrive.

Things to Read

• Heatmap on the odd new courtship between data centers and wind farms. The piece gets at a real tension in the market: hyperscalers want speed, developers want offtake certainty, and wind still has trouble pretending it is dispatchable.

• Boston Globe on Massachusetts and the faint smell of a nuclear comeback. Less a real renaissance than an indication that in deep-blue states, the old anti-nuclear reflex is no longer as automatic as it used to be.

• Interesting Engineering on a world-first non-nuclear lead-cooled reactor. Still very much in the “watch this space” category, but it is another reminder that thermal innovation is not limited to fission.

• RealClearEnergy on how environmentalism’s “follow the science” posture has curdled into a politics of delay and anti-build reflex. A worthwhile provocation, especially for anyone who still thinks infrastructure fights are mostly about carbon arithmetic.

• Grist on the Ember and IEA reports, and on the broader idea that we’ve entered an “age of electricity.”

• Missouri Independent asks whether competition is the answer to Missouri’s rising utility costs. Answer: Yes.

Major Stories

POLLING
Gallup: Support for Solar and Wind Slips, Nuclear Gains Ground

Gallup’s latest poll shows Americans still prefer more emphasis on solar and wind, but the shine is wearing off a bit. Nuclear is the only major source clearly gaining support. The standout findings:

• 66% want more emphasis on solar

• 55% want more emphasis on wind

• 46% want more emphasis on nuclear, the highest Gallup has recorded

• 42% want more emphasis on natural gas

• 34% want more emphasis on oil

• 23% want more emphasis on coal

A few more useful cuts:

• Democratic support for more nuclear rose to 42%

• Republican support for more nuclear sits at 52%

• That makes nuclear the narrowest partisan gap of any major energy source in the survey

• Even so, 53% of Americans still oppose building a nuclear plant in their area, versus 45% who favor it

GridTake - This is what practical politics looks like. People love fashionable energy positions until the bill arrives. Then they start rediscovering the virtues of density, firmness, and actual output. Nuclear support is rising for the simplest reason in the world: voters are slowly remembering that electricity is not a moral accessory. It is infrastructure.

TRANSMISSION
New England Wants FERC to Move Faster on $1.5 Billion in Refunds

New England governors and utility regulators are urging FERC to reject a request by Eversource and Avangrid to delay about $1.5 billion in transmission refunds while litigation continues over allowed returns on equity. Key facts:

• The underlying complaint dates back to 2011

• FERC cut New England transmission ROE to 9.57% from 10.57%

• Eversource utilities owe about $880 million

• Avangrid’s New England utilities owe about $203 million

• FERC already extended the refund deadline to May 20, 2027

• Utilities want it pushed further to December 17, 2027

The utilities argue faster repayment would damage credit metrics, raise borrowing costs, and make future infrastructure investment harder. Regulators and governors respond that customers have already been overcharged for years and should not have to wait even longer for relief.

GridTake
Transmission has become the one part of the energy system almost nobody wants to criticize because the need is so obvious. That makes discipline even more important. If utilities can invoke urgency to defend premium returns forever, “build more transmission” eventually turns into “write a blank check.” Ratepayers are starting to notice the distinction.

RENEWABLES IN COURT
Court Blocks Trump Administration Moves That Put 57 GW at Risk

A federal judge in Massachusetts issued a preliminary injunction blocking five Trump administration actions that had sharply constrained wind and solar permitting on federal land and offshore. The core numbers:

• 57.2 GW of wind, solar, hybrid, and offshore wind capacity had been delayed, canceled, or put at serious risk

• Roughly $905 million had already been invested in affected projects

• The threatened value of federal tax credits tied to those projects ranged from $8.4 billion to $25.6 billion

The court found the plaintiffs were likely to succeed on claims that the administration’s actions were arbitrary and capricious. The blocked actions included:

• Interior’s expanded review memo

• Restrictions on pre-permitting consultation tools

• A land-order framework biased toward “high capacity density” projects

• An Army Corps memo with similar prioritization logic

• The Zerzan opinion that had effectively frozen new offshore wind

GridTake
This case is less about renewables than about process integrity. If the administration wants less wind and solar, it can try to change the subsidy stack, the tax code, or the broader market structure. What it cannot do, at least not cleanly, is invent a shadow permitting regime to slow one category of project while pretending it is neutral. That kind of administrative improvisation usually ends the same way: in court, and with more uncertainty for everyone.

POLITICS
Wright Walks Back Gas Price Forecast After Trump Rebuke

Energy Secretary Chris Wright spent Wednesday trying to clean up comments he made over the weekend suggesting gasoline might not fall below $3 a gallon until 2027. After Trump publicly called that timeline “totally wrong,” Wright told senators he had never said prices would stay above $3 until next year and had simply “left some uncertainty” in his answer.

The episode is more political than grid-related, but it’s still revealing. Energy prices have become sensitive enough that even a modestly cautious forecast from the administration’s own energy secretary now requires immediate cleanup. Wright also told lawmakers he does not know the future path of energy prices, while emphasizing that current pump prices remain below Biden-era levels.

GridTake
This is what happens when energy prices stop being background noise and become political weather. The administration wants to promise abundance, but real markets do not move on command, especially in the middle of geopolitical stress. That tension is going to keep surfacing.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Solar is no longer “emerging” — Ember’s data now makes it the central growth engine of the global power system, not a sidecar to it.

• Coal lost the symbolic lead — Renewables overtook coal globally in 2025, even if coal still matters enormously in Asia.

• Batteries are becoming price setters — not just grid accessories, but real market-shaping infrastructure in high-solar systems.

• The courts are policing fuel favoritism — agencies can lean, but they cannot simply rewrite permitting law on the fly to kneecap disfavored technologies.

• Transmission’s political halo is fading — the buildout case is strong, but regulators are starting to push back on the idea that any return can be justified in the name of urgency.

• Nuclear is climbing back into public respectability — just not yet into people’s zip codes.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

FERC is inching toward a federal framework for data center interconnection. MISO is projecting a 35% peak load jump by 2035, with data centers driving much of the increase. Maryland is trying to turn flexible load into real grid capacity. And a new paper from the American Capacity Project makes the deeper case that the country’s energy problems are no longer mainly technical. They are institutional, cultural, and self-reinforcing.

The Lede

REGULATION
FERC Moves Toward Federal Framework for Data Center Interconnection

FERC is preparing a June decision on how large loads, particularly data centers, connect to the grid, stepping into what has historically been a state-led domain.

At the center is DOE’s push for a standardized framework that could accelerate interconnection timelines and potentially reshape cost allocation. The proposal contemplates whether hyperscalers should bear 100% of upgrade costs tied to their projects, a sharp departure from the traditional “socialized grid” model.

The Commission is clearly aware of the stakes. Chair Laura Swett emphasized the need for a “legally durable” approach while also signaling willingness to push jurisdictional boundaries.

At the same time, states are pushing back. Regulators argue they are better positioned to manage interconnection, pointing to emerging tools like large-load tariffs, long-term contract requirements, and collateral structures that screen speculative demand.

The backdrop is increasingly political. Data centers are no longer just an economic development story. They are showing up in rate cases, local elections, and cost debates across PJM and the Southeast.

GridTake
This is not just a technical rulemaking. It is a jurisdictional test. If FERC leans too far into centralization, it risks slowing the very buildout it is trying to accelerate. States have been moving quickly with contract structures that force discipline on large loads. A federal overlay could standardize the process, but it could also flatten the experimentation that is currently sorting real demand from speculative demand. The more interesting path is not full preemption, but selective pressure. Set guardrails, let markets and states do the sorting, and step in only where the system breaks.

Things to Read

• Axios looks at emerging “blue energy” models pairing gas and SMRs in modular builds, a sign that developers are leaning toward hybrid systems rather than single-technology bets.

• NucNet highlights Fatih Birol’s warning that stable, continuous generation is becoming more valuable as demand growth accelerates, a subtle shift back toward firm capacity in global planning.

• C3 News makes the case that permitting reform needs to move beyond incremental fixes, focusing on statutory changes to timelines and judicial review that actually alter build speed.

• Washington Post captures growing local resistance to data centers in Virginia, where electricity prices and land use concerns are starting to reshape the political landscape.

• Pew Research notes that most new data centers are landing in rural areas, a reminder that the economic upside and grid strain are increasingly being exported away from traditional urban load centers.

Major Stories

LOAD GROWTH
MISO Sees 35% Peak Load Growth by 2035 — With a Large Asterisk

MISO now expects peak demand to rise from 121 GW to roughly 163 GW by 2035, a 35% increase driven largely by data centers.

The near-term signal is just as important. The operator expects 8–14 GW of data centers to come online in 2026–2027, which will serve as the first real test of how quickly projected demand materializes. Under its mid-case scenario:

• Data centers could consume ~20% of MISO load by 2030

• Rising to ~25% by 2040

• Regional growth rates cluster around 2.6%–2.7% annually

But the caveats are doing real work here. MISO explicitly flagged:

• No centralized data clearinghouse

• Opaque project pipelines

• Limited historical load data

• Uncertain commercial viability for AI

Utilities are already responding by tightening interconnection requirements. Larger deposits, longer commitments, and stricter contracts are being used to separate real projects from speculative queue entries.

GridTake
This is the first honest load forecast cycle in a long time. The numbers are big, but the uncertainty is bigger. If even half of the 8–14 GW expected in the next two years actually lands on schedule, planning assumptions across the Midwest will harden quickly. If it slips, the industry risks overbuilding toward a demand curve that never fully materializes.

RELIABILITY & POLICY
Wright Signals Openness to Permitting Deal as Coal, Funding Decisions Face Scrutiny

Energy Secretary Chris Wright faced a familiar set of tensions on Capitol Hill: reliability versus cost, speed versus process.

On coal, Wright defended DOE’s use of emergency authorities to keep plants online, arguing they are necessary to prevent reliability risks. Lawmakers pushed back, citing rising costs being passed to ratepayers. On permitting, there was a clearer opening. Wright signaled willingness to work with Congress on:

• Aligning pipeline and transmission permitting timelines

• Addressing methane regulation alongside infrastructure reform

• Moving toward a broader bipartisan framework

At the same time, DOE’s budget direction is shifting capital toward “firm” generation, including gas, nuclear, hydro, and geothermal, while cutting back on portions of prior loan programs.

GridTake
The politics are catching up to the physics. Reliability arguments are winning in the short term, but they are being financed in ways that are hard to defend over time. The more interesting signal is the permitting conversation. If transmission and pipelines start moving on comparable timelines, it changes the entire supply curve. Until then, emergency orders and stopgap measures will keep filling gaps that are fundamentally structural.

DEMAND-SIDE GRID
Maryland Pushes Flexible Load Strategy Toward 440 MW Target

Maryland utilities are moving forward with plans to aggregate up to 440 MW of flexible load through virtual power plant programs, EVs, batteries, and demand response. Breakdown of proposed capacity:

• BGE: ~188 MW

• Pepco: ~185 MW

• Delmarva: ~34 MW

• Potomac Edison: ~18–33 MW

The goal is to reduce peak demand and delay infrastructure expansion, particularly as transmission buildout lags. Regulators are cautiously supportive but flagged a key issue:
flexibility only matters if it is dispatchable, measurable, and locationally relevant.

That last piece is the hard part. Shifting load in the wrong place does nothing for constrained parts of the grid.

GridTake
Flexible load is moving from theory to procurement, but it is still being treated like capacity when it behaves more like probability. A megawatt of dispatchable generation and a megawatt of voluntary demand response are not equivalent, and the system knows it.

WHITE PAPER
The Scarcity Trap: America Didn’t Run Out of Power. It Stopped Building It.

A new white paper from the American Capacity Project argues that the current grid crunch is not just an infrastructure problem. It is a cultural and institutional one.

The paper contrasts two very different American energy systems. In the postwar era, rising electricity demand triggered a rapid supply response. Capacity was built ahead of need. Growth was treated as a directive. In the modern era, that reflex has inverted. Demand now triggers anxiety, delay, and procedural drag. Instead of building through pressure, the system manages around it.

The paper’s most useful framing is its side-by-side comparison of those eras. In the early postwar decades, firm capacity growth roughly tracked or exceeded demand growth. In the modern period, demand has resumed climbing, but firm dispatchable capacity has barely responded. The result is not just tighter markets. It is a different governing philosophy.

That philosophy produces what the paper calls the “Scarcity Trap”:

• demand rises

• supply response is delayed

• scarcity conditions emerge

• scarcity is then used to justify more caution, more restriction, and more delay

GridTake
This is what makes the paper valuable. Not the diagnosis that projects take too long, everyone in this space knows that, but the claim that delay has become self-justifying. Once a system gets used to scarcity, it starts treating growth itself as the problem. That is how a rich country with abundant fuel, capital, land, and engineering talent ends up talking like energy shortage is a permanent condition instead of a policy choice.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• FERC wants a larger role — The big June question is no longer whether Washington will weigh in on large-load interconnection, but how far it will push before states push back harder.

• MISO’s forecast is a filter test — The 35% load-growth number matters, but the real tell will be whether the expected 8–14 GW of data centers actually lands in 2026–2027.

• Speculative load is finally getting priced — Bigger deposits, longer commitments, and tighter contract structures are becoming the new admission fee.

• Flexible load is growing up — Maryland’s 440 MW proposals suggest demand-side resources are moving from pilot language toward operational planning, but only if they can be measured and dispatched like adults.

• Permitting is back in the room — Wright’s hearing made clear that reliability patchwork is not enough; sooner or later the conversation has to return to transmission and pipeline timelines.

• Scarcity is becoming a governing style — The most important idea in today’s issue is not that projects are delayed, but that delay is now being used to justify more delay.

• The vocabulary has changed — Growth used to trigger construction. Now it triggers talk of strain, burden, and customer protection.

• The queue is not the pipeline — A system can be full of proposed capacity and still be starved of actual delivery.

• Jurisdiction is now a market variable — Where authority sits between FERC, states, and utilities will shape not just timelines, but who pays.

• The real shortage is throughput — Not fuel, not capital, not ideas. Conversion.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

A few threads are starting to snap into focus. States are pushing utilities to get more out of existing wires before building new ones. Utilities themselves are gearing up for a massive capital cycle that will land, inevitably, on customer bills. Large power users are beginning to route around the system entirely. And in the UK, excess renewable supply is leading to a strange outcome: paying people to use electricity just to keep the grid balanced.

Different stories, same tension. The grid isn’t short on ambition. It’s short on alignment.

The Lede

TRANSMISSION
Colorado Tries to Squeeze More Out of the Grid

Colorado is moving a bill that would require utilities to evaluate and deploy advanced transmission technologies before defaulting to new buildouts. That includes dynamic line ratings, advanced conductors, and topology optimization, tools that can increase transfer capacity on existing lines by 10% to 40% depending on conditions.

The key change is procedural. Utilities won’t just be allowed to use these tools, they’ll have to show regulators they considered them in their planning and explain why they did or didn’t deploy them. The burden shifts from “we’ll try it” to “justify not using it.”

That matters because transmission timelines are now measured in decades. New high-voltage lines regularly take 8 to 12 years to permit and build. Grid-enhancing technologies can be deployed in months and often cost a fraction of a full rebuild.

But the ceiling is real. Even aggressive adoption only unlocks incremental capacity relative to what’s coming. The U.S. is facing load growth that requires thousands of miles of new transmission annually, not just better utilization of existing corridors.

GridTake
Colorado is trying to force better behavior through process, which is understandable, but rarely how infrastructure actually improves. The more interesting shift is happening elsewhere: large customers are starting to demand speed, flexibility, and cost discipline, and they’re willing to bring their own capital to get it. That kind of pressure does more to surface efficient solutions than any planning requirement. If advanced transmission tech really delivers, markets will pull it in quickly. If it doesn’t, no amount of regulatory nudging will make it stick.

Things to Read

• WSJ walks through the coming $1.4 trillion utility spending wave, driven by AI load, electrification, and the slow grind of grid expansion. The number is big, but the more interesting question is how much of it actually gets built on time.

• Cato makes the case that Texas is actually building its way through the power crunch, leaning on speed, market signals, and willingness to add dispatchable capacity while others debate frameworks.

• PV Magazine flags that European grid investment could nearly double by 2027, a familiar story of ambition running headfirst into permitting, coordination, and supply chain limits.

• Reuters + market chatter have nuclear equities running again, with developers eyeing IPOs and capital finally treating advanced reactors like something that might exist outside a slide deck.

• RealClearInvestigations takes a long look at a potential Alaska pipeline revival, a reminder that even in an AI-fueled energy cycle, the hardest problems in infrastructure haven’t changed much.

Major Stories

CAPITAL
Utilities Line Up a $1.4 Trillion Build Cycle

A new PowerLines review of 51 utility earnings calls shows at least $1.4 trillion in planned capital spending through 2030, up more than 20% from just a year ago. That number is doing a lot of work.

Electric bills are already up roughly 40% since 2021, utilities requested $31 billion in rate increases in 2025 alone, and regulators have been approving a growing share of those requests.

The spending is not evenly distributed. About half sits in transmission and distribution. The South alone accounts for more than $570 billion. And the drivers are remarkably consistent across companies:

• Data centers and load growth (32 of 51 utilities)

• Grid hardening and extreme weather (28 of 51)

• Aging infrastructure (16 of 51)

There’s also a structural wrinkle that matters more than any headline number. Utilities earn returns on capital expenditures, not operational efficiency. So the system naturally leans toward building things, even when cheaper solutions exist.

GridTake
The real story here is that the incentive structure hasn’t changed while the scale has exploded. In a slower-growth world, that bias was tolerable. In a high-load-growth, AI-driven environment, it becomes a pricing problem. The same regulatory model that once ensured reliability now risks overcapitalizing the system at the exact moment affordability becomes politically binding.

LARGE LOADS
Georgia Lets Big Customers Bring Their Own Power

Georgia regulators just approved a framework that lets large customers, think hyperscalers, industrials, bring their own clean generation to support their load.

Up to 3 GW of customer-identified resources can come online through 2035. These projects can even sit outside Georgia if they can deliver power into the system. The customer funds it, gets the environmental attributes, and receives credit for the energy.

It’s a subtle but important evolution. Instead of waiting for utilities to plan, permit, and rate-base new generation, large customers can move first and carry some of the execution risk themselves.

GridTake
This is what a system under time pressure looks like. When the queue is slow and the load is real, the market starts routing around the bottleneck. “Bring your own power” is a latency fix. The more this model spreads, the more the traditional utility planning cycle starts to look like a bottleneck rather than the backbone.

INTERNATIONAL POLICY
Free Electricity… Because There’s Too Much of It

In the UK, regulators are expanding a program that allows households to get free or discounted electricity during periods of excess supply, typically sunny, low-demand weekends.

This builds on existing “use less at peak” incentives by flipping the signal: now the grid will also pay people to use more when renewable generation is high and demand is soft. The system operator coordinates it, suppliers pass through incentives, and customers with smart meters can participate.

On one level, it’s clever. On another, it’s a quiet admission. When you have to pay people to consume electricity to avoid curtailing generation, you have a coordination problem.

GridTake
It works, but it’s clunky. A more flexible, market-driven structure would do this automatically. Instead, you get a kind of administrative choreography where the grid politely asks you to run your dishwasher at 2pm.

AI & HYPERSCALERS
The IEA Puts Boundaries Around the AI Hype

The International Energy Agency this week released a new report trying to quantify what AI-driven demand actually means for the power system over the next decade.

Their central estimate: global data center electricity consumption rises from roughly 485 TWh today to around 900–950 TWh by 2030. That’s a near doubling in under seven years, with AI workloads accounting for a disproportionate share of the increase.

But the report is just as much about constraints as growth. The pace of expansion is limited not only by power availability, but by bottlenecks in grid interconnection, transmission buildout, semiconductor supply, and physical infrastructure timelines. In other words, demand can scale faster than the system that feeds it.

On the supply side, the IEA does not land on a single dominant solution. Instead, it points to a hybrid buildout: renewables continuing to scale quickly, natural gas expanding to provide dispatchable capacity, and increasing use of on-site generation and storage to bypass grid delays.

GridTake
AI demand currently points to a hybrid system, not a pure one. The faster AI moves, the more pragmatic the energy mix becomes.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• Optimization gets political — States like Colorado are starting to force utilities to prove they’ve squeezed existing wires before building new ones. Efficiency is now a regulatory requirement.

• $1.4T is the new baseline — Utility capex plans aren’t creeping up, they’ve reset. Transmission, distribution, and load growth are all hitting at once.

• Load growth breaks the model — Traditional planning cycles weren’t built for hyperscale demand arriving in 3–5 year windows. Everything downstream is scrambling.

• BYO power goes mainstream — Georgia formalizing large-load self-supply is another step toward customers solving their own constraints.

• Rate design is the battlefield — The fight isn’t about whether data centers connect, it’s about who eats the risk if they don’t ramp as promised.

• Oversupply is real (sometimes) — The UK paying people to use power is what happens when generation shows up faster than coordination.

• Policy is simulating markets — Instead of fixing price signals, systems are layering incentives on top. Functional, but clunky.

• AI demand is elastic, infrastructure isn’t — Data centers can scale quickly. Transmission, generation, and interconnection cannot.

• Gas quietly backstops everything — Even in aggressive buildout scenarios, dispatchable capacity keeps filling the gaps.

• The grid is fragmenting — Some regions are short, others occasionally long. Transmission determines which reality you live in.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The power sector keeps running into the same problem from different angles: demand is rising faster than the system can comfortably absorb, and every proposed solution comes with its own bottleneck. In PJM, that means a one-time procurement to pull new capacity onto the system faster. In Washington, it means keeping one eye on long-shot technologies like fusion while trying to hold the current system together. In the background, cyber threats and equipment shortages are reminding everyone that reliability is no longer just about having enough megawatts on paper. It is about whether they can be built, protected, and delivered on time.

The Lede

RTO CAPACITY
PJM Tries to Pull 14.9 GW Forward

PJM is proposing a one-time procurement of 14.9 GW of new capacity to get ahead of projected shortfalls driven largely by data centers and broader load growth. The structure starts with bilateral deals between large loads and generators, with PJM acting as a matchmaker. If that falls short, PJM steps in with a centralized procurement to fill the gap.

The backdrop is stark. PJM is staring at a potential 50–60 GW capacity gap over the next decade. This effort is meant to accelerate projects that would otherwise get stuck in interconnection queues, permitting delays, or long construction cycles.

• Phase 1: bilateral contracting (more flexibility, tailored risk-sharing)

• Phase 2: centralized procurement (backstop if needed)

• Target in-service: June 2031

Expect pushback. States and utilities are already signaling concern about over-procurement and cost exposure to ratepayers.

Why it matters - This is PJM acknowledging that the normal market timeline is too slow for the load that’s coming.

Grid Take - This is a market trying to stay a market while acting like a planner. When supply can’t respond quickly enough to price signals, the system starts layering in coordination. The more often that happens, the harder it becomes to separate “market outcomes” from managed ones.

Things to Read

• C3 News – A sharp argument that nuclear’s real problem is economics, not reactor size. Worth it for anyone still assuming SMRs solve the equation on their own.

• Ember – The “twin fossil shock” reframes energy volatility beyond oil. Useful for understanding how gas and electricity are now central to macro risk.

• Vox – A clean breakdown of how EVs and batteries can act as grid assets. Not hype, but a helpful explanation of where flexibility could actually come from.

• Latitude Media – Gas is no longer a “bridge.” It’s being treated as core infrastructure by hyperscalers. This piece captures that shift well.

• Wood Mackenzie – A grounded look at whether fusion is nearing a real inflection point or just another cycle of optimism.

Major Stories

NUCLEAR
Fusion Gets Another Federal Push

The federal government is leaning further into fusion through a new partnership between Inertia Enterprises and Lawrence Livermore National Lab. The model is becoming consistent: public R&D meets venture-backed startups trying to commercialize fusion.

Inertia raised $450 million earlier this year, joining a growing pool of private capital chasing what is still a very uncertain technology. The 2022 Livermore breakthrough proved net energy gain in a lab setting, but sustained, commercial fusion remains a long way off. Still, momentum is building:

• Federal R&D infrastructure increasingly paired with private developers

• Billions in private capital flowing into U.S.-based fusion startups

• Policy interest rising alongside long-term demand projections

Why it matters - Even as the grid scrambles for near-term capacity, capital is still chasing technologies that could reshape the system decades out.

Grid Take - Fusion is not about solving 2030. It’s about not being stuck in 2040. In a system that is already short on firm capacity, even a small chance of a breakthrough starts to look worth funding.

GRID HACKERS
Cyber Threats Move Back to the Foreground

Grid operators are on heightened alert after warnings that Iran-linked hackers have targeted programmable logic controllers across U.S. infrastructure. These systems are deeply embedded in grid operations, controlling substations, generation, and distributed resources.

This isn’t theoretical. Federal agencies say disruptions have already occurred across multiple sectors. Utilities are treating it as familiar territory, but the scale and coordination are raising the temperature. Key reality check:

• PLCs are everywhere across the grid

• Many run on legacy systems not designed for modern threats

• Disruption at this level can impact both generation and distribution

Why it matters - Reliability is no longer just about having enough capacity. It’s about whether the system can be trusted to operate as intended.

Grid Take - The grid is being tightened by demand and tested by digital risk at the same time. As systems become more automated and interconnected, the cost of disruption rises. Reliability is now a hardware and software problem simultaneously.

FOSSIL FUELS
EPA Eases Coal Ash Rules

The EPA is proposing new “flexibilities” in coal ash regulations, allowing more site-specific approaches to cleanup and monitoring. The move would ease compliance requirements that have been in place, in various forms, since 2015.

There are roughly 775 coal ash sites across the country, many tied to legacy coal plants. The proposal would:

• Allow more discretion in cleanup standards and monitoring

• Ease certain closure and remediation requirements

• Reduce compliance burden for operators

Industry groups support the move as a way to keep costs down and maintain reliability. Environmental groups argue it risks delaying cleanup of hazardous waste linked to groundwater contamination.

Why it matters - This is part of a broader shift toward keeping existing generation online longer as new capacity struggles to come online.

Grid Take - When replacement capacity is slow and expensive, legacy assets gain value. But easing constraints doesn’t eliminate costs, it moves them. Often into longer-term risks that are harder to see and harder to price today.

SUPPLY CHAIN
Gas Turbine Market Hits a Hard Constraint

The biggest constraint on new gas generation may not be policy. It’s manufacturing. Gas turbine prices are projected to hit $600/kW by 2027, up 195% since 2019, as demand overwhelms supply. Global orders reached 110 GW last year, while manufacturing capacity sits closer to 60–70 GW.

Lead times are now stretching to six years, and order books are effectively sold out through 2027.

• Turbines = 20–30% of combined cycle project costs

• Procurement timing now dictates project viability

• Data center demand is a major driver of the surge

Manufacturers are ramping up, but bottlenecks remain in specialized labor and key components.

Why it matters - You can’t build what you can’t buy. Equipment availability is now a gating factor for capacity expansion.

Grid Take - The debate over gas used to be about policy and emissions. Now it’s about supply chains. The winners in this cycle won’t just be the ones with the best economics, but the ones who secured equipment early enough to matter.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Quick Signals

• PJM breaks the glass — A one-time 14.9 GW procurement signals urgency. When the grid operator starts matchmaking deals, timelines have officially become the constraint.

• Capacity gap gets real — PJM projecting 50–60 GW shortfall over the next decade. That’s no longer a planning exercise, that’s a build problem.

• Bilateral > centralized (for now) — PJM leaning on private contracting first, but keeping a centralized backstop ready. Market-first, planner-second.

• Fusion money keeps flowing — Another $450M into a young player. Still long-dated, but capital is clustering around the upside scenario.

• Public + private R&D converge — DOE increasingly pairing national labs with startups. Fusion moving from pure science toward early commercialization attempts.

• Cyber risk escalates quietly — Iran-linked actors targeting grid control systems. PLCs are the soft underbelly and they’re everywhere.

• Legacy systems = vulnerability — Much of the grid still runs on older control tech. Reliability now depends on systems that weren’t built for this threat environment.

• Coal gets regulatory breathing room — EPA proposing more flexibility on coal ash rules. Another signal that keeping existing capacity online matters right now.

• Gas turbine market breaks — Prices up ~195% since 2019, lead times at 6 years, and supply sold out through 2027. This is the bottleneck.

• Procurement beats economics — Gas projects now hinge on securing turbines, not fuel spreads. If you didn’t order early, you’re late.

• Data centers reshape everything — Driving nearly 100% growth in power demand this decade. Not just a load story, a supply chain story too.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The American energy debate is undergoing a quiet correction. Not a reversal, not a collapse, but a recalibration shaped less by ideology and more by cost, reliability, and lived experience.

This edition of GridBrief looks at that shift through five lenses: changing public sentiment, rising utility rates, the growing transparency of electricity costs, Texas’ bet on nuclear, and the continued stitching together of Western power markets.

The Lede

POLLING & ATTITUDES
Americans Pull Back from Renewables-First Framing

https://www.pewresearch.org/science/2026/04/03/americans-shifting-views-on-energy-issues/ps_2026-4-7-_energy_00-01/

New Pew polling shows support for prioritizing wind and solar is falling. Not disappearing, but clearly softening.

In 2020, 79% of Americans said renewables should be the priority. Today, that number is 57%. At the same time, support for prioritizing fossil fuels has risen from 20% to 42%.

The shift is being driven almost entirely by Republicans. In 2020, a majority of Republicans favored renewables. Today, 71% say fossil fuels should be prioritized, and just 28% say renewables. Democrats remain strongly supportive of wind and solar at 83%, though even that has edged down slightly.

Views on reliability and cost are moving alongside these preferences. Americans, particularly Republicans, are increasingly skeptical that wind and solar are reliable or inexpensive compared to other sources.

Why it matters - Energy politics is drifting away from abstract climate framing and toward tangible system performance. Reliability and cost are no longer secondary concerns. They are the center of gravity.

Things to Read

• The Economist on electricity prices and AI – A pushback on the now-common claim that data centers are the primary driver of rising electricity costs. The reality is more layered, and less convenient.

• ITIF on data center load growth – Four arguments for why AI demand may not overwhelm the grid. Worth reading as a counterweight to the more alarmist takes.

• BGR on photonic chips – A glimpse at a potential efficiency breakthrough. If compute shifts from electrons to light, the implications for energy demand are profound.

• New York Times on nuclear and geopolitics – Nuclear energy framed through global conflict and security concerns. Energy independence remains a strategic asset.

• The Dispatch (me) on NIMBYism and data centers – The national debate, grounded locally. What resistance looks like when it’s not theoretical.

Major Stories

UTILITIES & RATES
PECO Seeks $429M Rate Hike to Harden the Grid

PECO is asking Pennsylvania regulators for a $429 million rate increase, an 11% revenue bump, largely to fund reliability improvements and infrastructure upgrades.

If approved, a typical residential bill (700 kWh/month) would rise 12.5% to $180.45 by 2027, before offsets. With proposed cost-smoothing measures, that increase drops to about 11.1%.

The utility plans to spend $2.8 billion through 2027 on distribution upgrades, including vegetation management and equipment replacement, and nearly $9.8 billion through 2030 overall. That investment footprint is unusually large. PECO’s capital expenditure to net plant ratio sits at 88%, well above a peer median of 52.6%, signaling elevated financial and operational risk.

The request also includes a proposed 10.95% return on equity, aimed at attracting capital to fund that buildout.

Why it matters - This is the modern utility model in full view: spend heavily to harden and expand the grid, then recover those costs through rates. Reliability is not free. It is financed, slowly and persistently, through customer bills.

Grid Take - The debate often asks why prices are rising. This is why. The system is being rebuilt in real time, and customers are the capital stack.

RESEARCH & TOOLS
A New Tool Shows Why Your Electricity Bill Feels Like a Black Box

MIT researchers, working with Heatmap News, have launched the Electricity Price Hub, a tool that breaks down electricity prices and bills at a local level. The findings are less surprising than they are clarifying:

• Electricity rates are up roughly 33% over the past five years, adding about $35/month to the average household bill

• Monthly bill swings are wide, with a median difference of $92 between low and high months, and in some cases exceeding $200

• Prices and bills often diverge, depending on usage patterns and seasonal demand

More than half of Americans report that electricity bills put meaningful pressure on their finances.

Why it matters - Electricity pricing has always been complex. What’s new is visibility. As tools like this emerge, the gap between how the system works and how people experience it will shrink.

Grid Take - Once people can see the drivers, they stop accepting the story. Transparency won’t lower costs. But it will change the politics of who is blamed for them.

NUCLEAR
Texas Opens $350M Door to Advanced Nuclear

Texas is moving aggressively to expand firm generation. The state’s Advanced Nuclear Energy Office has opened applications for $350 million in funding to support nuclear development, including construction, licensing, and supply chain investments.

The program builds on earlier efforts to incentivize new gas generation and comes as ERCOT demand is projected to grow rapidly. The Energy Information Administration expects ~10% annual load growth between 2025 and 2027 in the region.

The funding can support everything from early site permitting to long-lead component manufacturing, with awards expected this summer.

Why it matters - Texas is not debating whether new capacity is needed. It is deciding what kind. Nuclear is being positioned as part of that answer, alongside gas.

Grid Take - This is industrial policy wearing a cowboy hat. Texas is treating electricity demand as inevitable and aligning capital accordingly.

MARKETS
Nevada Joins California’s Expanding Day-Ahead Market

Nevada regulators have approved NV Energy’s plan to join the Extended Day-Ahead Market (EDAM) run by the California ISO, with participation beginning in 2028.

The move connects Nevada to a broader regional market, allowing access to a more diverse resource mix and improving operational efficiency. The utility expects $93.1 million in annual savings, driven by lower production costs and improved market transactions.

Two competing market structures are now emerging in the West: CAISO’s EDAM and the Southwest Power Pool’s Markets+. Nevada’s decision strengthens EDAM’s position. The move also supports integration of distributed energy resources and aligns with federal efforts under FERC Order 2222.

Why it matters - Regionalization is one of the few levers that can improve reliability and reduce costs without building entirely new infrastructure. Sharing power across geographies smooths variability and reduces redundancy.

Grid Take - The West is slowly becoming a grid instead of a patchwork. Not through mandate, but through economics.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

The Conversation

Quick Signals

• Public opinion tilts back toward fuel security — Support for prioritizing renewables drops from 79% to 57% since 2020, while fossil preference doubles to 42%. The electorate is rediscovering tradeoffs.

• Republicans fully flip on energy mix — Just 28% now prioritize wind/solar, while 71% favor fossil fuels. That’s not drift. That’s a wholesale inversion.

• Democrats hold the line, barely soften — Still 83% pro-renewables, but edging down. Even the most stable bloc is feeling pressure at the margins.

• Rates up, reliability the justification — PECO seeks $429M (+11%) tied to outages, storms, and system upgrades. The new rate case template is resilience first, bill impact second.

• CapEx arms race intensifies — PECO’s planned spend equals 88% of its net plant, far above peers. Utilities are scaling investment faster than the system historically tolerates.

• Bills ≠ prices — MIT tool shows monthly swings of $92 median, over $200 in some cases. Volatility, not just increases, is what households feel.

• Electricity finally becomes visible — A “black box” system gets cracked open. Expect less patience for vague explanations of rising costs.

• Texas puts cash behind nuclear — $350M to jumpstart advanced reactors. Not rhetoric. Early-stage industrial policy.

• ERCOT demand growth stays blistering — ~10% annual growth (2025–2027). Few grids globally are facing this kind of load curve.

• Western grid quietly integrates — Nevada joins CAISO’s EDAM, chasing $93M/year in savings. Regional markets keep expanding while policy debates rage elsewhere.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The grid is doing something it rarely admits to doing: tightening. Not rhetorically, not in forecasts, but in the blunt arithmetic of supply and demand. Generation slips. Prices climb. Policymakers, staring at the same numbers, reach not for more supply, but for new ways to ration who pays.

Today’s GridBrief moves through that tension: a system producing less power at higher cost, a renewable buildout losing momentum at the margin, and a regulatory turn toward making large loads carry the burden of a system that hasn’t kept pace.

The Lede

MARKETS
Generation Falls as Prices Surge

January delivered a quiet but telling contradiction: less electricity, more expensive electricity.

U.S. generation declined 0.7% year over year, even as retail prices rose 9.5%. This wasn’t driven by extreme weather across the board. Outside pockets like the Northeast, much of the country saw average or warmer-than-average temperatures.

Natural gas prices tell part of the story. Henry Hub was up 86% YoY, pushing costs through wholesale markets and into retail rates. But even here, the response was uneven. Gas generation rose in some regions and fell in others. Coal continued its structural decline, down 13.2%, while nuclear ticked up modestly.

The result is something more structural than seasonal: a system where marginal supply is getting tighter just as inputs get more expensive.

Why it matters - The grid is no longer coasting on excess capacity. When generation falls during periods of rising demand and rising fuel costs, price volatility is no longer a risk, it’s the baseline.

Grid Take - For years, policy has treated generation as optional and infrastructure as inevitable. The market is now reversing that assumption. When supply doesn’t expand, prices do the work instead.

Things to Read

• Axios on nuclear and AI demand – A clean framing of what’s becoming unavoidable: data center growth is now a primary driver of new nuclear interest. Less climate narrative, more capacity math.

• BBC on grid security risks – A look at how fragile parts of Europe’s energy infrastructure remain. Not just geopolitics, but a reminder that reliability is physical, not theoretical.

• The Center Square on national energy tensions – A policy-ground view of how states and federal actors are diverging on grid expansion, cost recovery, and who ultimately pays.

• NucNet on France’s nuclear push – France is doubling down again. Not aspirational targets, but a clear signal that firm generation is still the backbone of serious energy systems.

• Axios Pittsburgh on rising electricity rates – A localized version of the national story: prices climbing, demand growing, and infrastructure lagging just enough to make everything more expensive.

Major Stories

RENEWABLES
Solar Installations Drop 22% as Policy Uncertainty Bites

After years of relentless growth, solar finally hit friction. Installations fell from 33.8 GW in 2024 to 26.5 GW in 2025, a 22% decline, according to FERC. The culprit wasn’t demand. It was policy and timing.

Developers pivoted toward “safe harbor” strategies amid shifting tax credit rules, delaying projects into future years. The fourth quarter alone saw installations drop nearly 40% YoY. Solar still led all sources in new capacity. But the signal is clear: when policy becomes unstable, even subsidized industries hesitate.

Meanwhile, the deeper issue remains unresolved. Solar adds capacity, but not necessarily capacity when it’s needed. As large loads arrive, the distinction between energy and firm power is becoming less academic and more expensive.

Why it matters - The grid doesn’t price capacity and energy the same way. A system dominated by intermittent additions still needs firm supply somewhere else. That cost is now showing up.

Grid Take = Solar didn’t fail. It paused. But the grid doesn’t pause. Demand is arriving on schedule, and increasingly it wants power that shows up on demand, not just on paper.

DATA CENTERS
States Move to Wall Off Data Centers with New Tariffs

Large load tariffs are spreading rapidly. 77 tariffs are now pending or active across 36 states, with 29 approved in 2025 alone. These tariffs aim to ensure data centers and other large users “pay their fair share” of grid expansion costs. In practice, they introduce a new layer of friction:

• Upfront infrastructure payments

• Minimum load guarantees

• Exit fees and long-term contracts

• Requirements to bring or flex capacity

Some early evidence suggests they reduce speculative demand. AEP Ohio, for example, cut its large load forecast by more than half after implementing a tariff.

But the deeper question remains unresolved: are these tariffs protecting the system, or quietly discouraging the very load growth that historically lowered costs?

Why it matters - Load growth has historically improved system economics by spreading fixed costs. If tariffs suppress real demand alongside speculative demand, they risk reversing that dynamic.

Grid Take - This is policy hedging against scarcity instead of solving it. Rather than building more capacity, states are deciding who gets access to what exists. That may stabilize rates in the short term, but it risks locking in higher costs over time.

REPORTS
The Berkeley Frame: Prices Are Rising… But the Story Is Messier

The new Lawrence Berkeley / Brattle update complicates the prevailing narrative in a useful way.

Yes, nominal residential prices are up 33% since 2019. But in real terms, prices are only ~3% higher than 2019 and still below 2010 levels.

Even more counterintuitive: electricity as a share of income remains near historic lows, around 1%, despite recent increases. At the same time, the report surfaces three tensions that matter more than the topline:

1. Cost drivers have shifted
   Distribution, transmission, and system hardening are now dominant drivers, not generation.

2. Load growth is accelerating
   Demand growth has jumped from 0.2% annually (2010–2019) to 1.1% (2019–2025), driven mostly by C&I.

3. Load growth can lower prices… until it can’t
   Historically, higher load spread fixed costs and reduced prices. But in constrained systems, it triggers new capital spending and higher rates.

Why it matters - Both sides of the data center debate are partially right. Load growth has lowered costs historically. But under current constraints, it can just as easily raise them.

Grid Take - The report quietly demolishes the lazy narrative. Prices aren’t simply rising because of demand. They’re rising because the system built to serve demand hasn’t kept up with it.

BUILDOUT
Texas Moves Forward with 765-kV Backbone Buildout

While some states debate tariffs, Texas is doing something more concrete: building. Oncor and LCRA have proposed 214–244 miles of 765-kV transmission, part of a broader effort to support Permian Basin load growth. The project carries a price tag of $1.6B–$1.9B, excluding substation costs.

The context matters. Oncor’s interconnection queue includes 255 GW from data centers alone, alongside additional industrial demand.

This is what scaling looks like: not debates over who pays, but steel in the ground sized for what’s coming.

Why it matters - Transmission is the only way to convert theoretical generation into usable power at scale. Without it, capacity exists only on paper.

Grid Take - Texas is treating load growth as inevitable and planning accordingly. Other regions are still deciding whether to accommodate it or contain it.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

The Conversation

Quick Signals

• Gas whiplash returns — Henry Hub up 86% YoY in January, reminding everyone that fuel volatility still runs through the entire stack.

• Demand strong, sales weak — Retail electricity sales down 1.7% even as parts of the grid flirt with peak demand. Efficiency and behind-the-meter are quietly reshaping load.

• Coal keeps slipping… mostly — Down 13.2% YoY, except in Florida, where it jumped nearly 19%. The “transition” remains deeply regional.

• Nuclear inches back — Generation up 2.1% YoY. Not flashy, just quietly doing the thing the grid increasingly needs: showing up.

• Oil still the winter fallback — Northeast again leaned on oil-fired generation during peak winter demand. The grid’s least favorite backup remains on speed dial.

• Solar backlog builds — Installations fell, but pipeline strengthens for 2026–2027. Delay now, surge later.

• Big load definition inflation — “Large load” now increasingly means 50+ MW, not 5–25 MW. The scale of new demand is rewriting the rulebook.

• Speculation getting filtered — AEP Ohio slashed projected large-load demand from 30 GW to 13 GW post-tariff. Not all demand was real.

• Texas thinking bigger — 765-kV lines aren’t incremental upgrades. They’re a signal that ERCOT expects sustained, industrial-scale load growth.

• Affordability paradox — Prices up, but electricity still ~1% of income nationally. The averages look calm; the distribution does not.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The grid is entering a more complicated phase.

For the past decade, rising demand has quietly helped stabilize costs by spreading fixed infrastructure across more load. Now demand is accelerating again, but the system it’s landing on looks very different. Queues are clogged, retirements are ahead of replacement, and reliability is being managed in real time.

Two new reports this week point in the same direction from opposite ends of the timeline. One looks backward and says load hasn’t been the problem. The other looks forward and says that only holds if we get the next phase right.

The Lede

MARKETS
IER Report: Load Growth Isn’t Driving Power Prices

A new analysis from the Institute for Energy Research challenges one of the most persistent narratives in energy policy: that data centers are pushing electricity prices higher.

Across all 50 states, there is effectively no relationship between data center concentration and electricity prices. States with the most data center activity have average prices that are nearly identical to the national average. Even more interesting, states with the fastest demand growth have seen lower price increases over time.

That pattern shows up clearly over the past decade. States with high electricity demand growth saw prices rise about 20%, while low-growth states saw increases closer to 40%. The underlying dynamic is simple. The grid is capital-intensive. When more energy flows across the same infrastructure, the cost per unit falls.

Why it matters - For the grid: Load growth has historically improved cost efficiency, not degraded it.
For policy: The case for restricting data center development on cost grounds looks weaker than advertised.

GridTake - IER is describing how the system has behaved when it still had room to absorb growth. The key question now is whether that condition still exists. Load is not inherently inflationary. It becomes inflationary when it forces poorly timed or poorly located investment.

Things to Read

• Politico on Southern Company’s CEO - A candid look at how one of the largest utilities in the country is thinking about data centers, nuclear, and grid expansion. Less spin than usual, more admission that the system is being stretched in ways utilities didn’t plan for.

• NucNet on NRC reform - The first real attempt in decades to rethink how nuclear plants get licensed. If this moves, it matters. Not because it guarantees new reactors, but because it starts to remove the procedural drag that has defined the industry.

• Washington Post on AI politics (Sanders vs. Claude) - A cultural and political lens on AI that’s worth reading just to see how quickly the conversation drifts away from physical constraints. The gap between rhetoric and electricity reality is getting wider.

• TechXplore on faster AI data links - A reminder that the bottleneck isn’t just generation. Faster interconnects mean more data movement, which means more energy density in fewer places. Efficiency gains upstream often translate into heavier loads downstream.

• NWI Times — my piece on the hometown fight - The national debate, but zoomed all the way in. Data centers, local resistance, and a region deciding whether it still wants to build things. This is what the abstract policy fight actually looks like on the ground.

Major Stories

SYSTEMS
Brattle: The Grid Is Underused, Not Undersized

The Brattle Group’s latest report flips the conversation from how much we need to build to how much we are failing to use.

On average, the U.S. power system operates at roughly half of its potential capacity across time. That gap creates an opportunity. If load growth is steered toward underutilized parts of the system and away from peak stress periods, the need for new infrastructure drops significantly.

Brattle models a scenario where demand rises 20–30% over the next five years. If that load is layered onto the system inefficiently, rates rise modestly. If it is integrated with a focus on utilization, rates fall. The difference is meaningful. Their modeling shows a swing of several percentage points on retail electricity prices, translating into over $100 billion in potential savings over a decade.

Why it matters - For the grid: There is substantial “hidden capacity” in when and where electricity is used.
For markets: Flexibility is becoming as important as generation.

GridTake - This is not a counterargument to building. It is a sequencing argument. Utilization can delay investment, reduce costs, and buy time. But it only works within physical limits. The risk is not that we optimize too much. It is that optimization becomes a substitute for expansion rather than a bridge to it.

NUCLEAR
Microsoft and Nvidia Try to Industrialize Nuclear

At CERAWeek, Microsoft and Nvidia outlined a joint initiative to use AI to accelerate nuclear plant development.

The effort focuses on streamlining permitting, standardizing plant design, and deploying digital tools like “digital twins” to simulate projects before construction begins. Early results from partner companies suggest that AI-assisted permitting could dramatically reduce timelines and costs by identifying inconsistencies, reusing prior regulatory work, and aligning applications with existing approvals.

The broader goal is to move nuclear away from bespoke, one-off engineering and toward repeatable, scalable deployment.

Why it matters - For the grid: Nuclear remains one of the few firm resources capable of meeting sustained load growth.
For the industry: The bottleneck is shifting from technology to process.

GridTake - This is an attempt to turn nuclear into something closer to manufacturing. It may speed up parts of the lifecycle, but it does not eliminate the harder constraints. Supply chains, skilled labor, and capital discipline still govern how fast reactors get built. AI can compress timelines, but it cannot yet pour concrete.

COAL & RELIABILITY
Indiana Coal Plants Stay Online Under Federal Orders

The Department of Energy has extended emergency orders requiring coal units in Indiana to continue operating past their planned retirement dates.

The plants were initially kept online during winter reliability concerns, and federal officials now argue the conditions that justified those orders still exist. During peak winter demand and periods of low renewable output, the units provided measurable support to the Midcontinent system.

Critics argue the extension is unnecessary and expensive. Estimates suggest the continued operation of these units is costing ratepayers hundreds of thousands of dollars per day. Federal regulators have now allowed those costs to be spread across a broader regional base.

Why it matters - For the grid: Retirements are colliding with real-world reliability constraints.
For policy: Emergency authority is becoming a recurring tool, not a rare exception.

GridTake - Indiana is not an anomaly. It is an early signal. The system is being forced to reconcile planning assumptions with operational reality. Instead of adjusting the model, policymakers are extending legacy assets to close the gap. That works in the short term. Over time, it raises a more uncomfortable question about how much capacity is actually being replaced versus assumed.

BUILDOUT
NextEra and Japan Bet Big on Gas and Nuclear

A major U.S.–Japan energy agreement is driving a new wave of large-scale generation projects, including nearly 10 GW of natural gas capacity in Texas and Pennsylvania and 3 GW of small modular nuclear reactors in the Southeast.

The projects are tied directly to expected demand from data centers and advanced manufacturing. They also rely on coordinated federal support, long-term offtake agreements, and access to infrastructure like transmission and fuel supply.

The scale is notable. Individual gas hubs are being designed to support multiple gigawatts of continuous load, effectively functioning as dedicated power systems for large industrial customers.

Why it matters - For the grid: Capacity is being built in anticipation of demand, not just in response to it.
For markets: Load risk is becoming a central factor in investment decisions.

GridTake - This is the other side of the utilization story. If demand materializes as expected, these projects look essential. If it does not, they become overbuilt assets in a system that is already struggling with cost allocation. The bet is not just on energy. It is on the accuracy of demand forecasts.

ADVERTISEMENT

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

The Conversation

Quick Signals

• Load growth is no longer a theoretical issue. It is showing up in tariffs, planning models, and capital allocation decisions at the same time.

• Gas continues to fill the near-term gap between demand growth and slower-moving alternatives.

• Emergency reliability measures are becoming more common, suggesting underlying capacity assumptions are being tested.

• Interconnection reform is emerging as one of the most important levers for unlocking new supply.

• The real divide in energy policy is shifting away from fuel types and toward a simpler question of whether the system can expand fast enough.

Chart of the Week

Utilization Changes the Cost Curve

Source: https://www.brattle.com/the-untapped-grid/

The Brattle analysis shows a clear split. Under a status quo approach, rising demand modestly increases electricity rates. Under a utilization-focused approach, the same demand reduces them.

The difference comes down to how the system is used. A grid that is built for peak demand but operated inefficiently carries excess cost. A grid that absorbs new load into underused capacity can defer investment and reduce pressure on rates.

The implication is straightforward but easy to miss. The cost of electricity over the next decade will be shaped as much by operational decisions as by physical buildout.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!

The grid story right now is not a lack of awareness. It is a lack of speed.

The United States knows it needs more transmission, more generation, and more rules for hyperscale load. It just is not building any of them fast enough. Over the last two days, that gap showed up everywhere: in the number of miles of wire we need versus what we actually built, in the 39 GW of large-load demand utilities are trying to connect, in new forecasts showing data centers could claim as much as 17% of U.S. electricity by 2030, and in an EIA scenario where gas fills the gap because everything else arrives too slowly. Even the environmental movement is adjusting, with NRDC backing a specific nuclear project for the first time in its history.

The Lede

TRANSMISSION
The U.S. Needs 5,000 Miles of Transmission a Year. It Built 888.

The transmission hole is no longer subtle. The U.S. needs roughly 5,000 miles of new high-capacity transmission every year from 2025 through 2035 to maintain reliability, but built just 888 miles in 2024. The comparison is highlighted in analysis from Americans for a Clean Energy Grid and Grid Strategies, based on DOE and FERC data. In response, DOE launched a $1.9 billion SPARK funding round focused on reconductoring and advanced transmission upgrades, betting that the fastest near-term capacity gains will come from upgrading existing corridors rather than waiting for entirely new lines.

That logic is not crazy. Grid-enhancing technologies could unlock more than 80 GW of incremental peak capacity, and utilities like TVA are already testing them across large portions of their systems. But even advocates emphasize that optimization must be paired with new construction, not used as a substitute for it.

Why it matters - For the grid: reconductoring and dynamic ratings are some of the few tools that can add capacity on a realistic timeline.
For your audience: the 5,000-versus-888 gap is the clearest expression of why “build more” keeps turning into “optimize harder.”

GridTake - The shift here is philosophical as much as technical. Federal policy is starting to treat optimization as a bridge for a system that cannot expand fast enough. That may buy time, but it also risks normalizing a grid that is perpetually stretched instead of structurally upgraded.

Things to Read

• New York Times on how data center developers are increasingly building their own generation. Interesting because it turns the “bring your own power” idea into physical reality.

• Wall Street Journal on battery manufacturers pivoting from EVs to grid and data center demand. A useful look at how AI is already reshaping industrial supply chains.

• Science News explores new findings on static electricity. A reminder that even foundational electrical phenomena are not fully understood.

• NucNet reports on Sweden’s Blykalla advancing plans for an SMR park. A signal that advanced nuclear development is still moving forward globally, even if slowly.

• Technology.org covers new advances in grid-scale storage chemistry. Worth watching as upstream innovation that could eventually reshape storage economics.

Major Stories

DATA CENTERS
Data Centers Could Consume Up to 17% of U.S. Electricity by 2030

EPRI estimates data centers could consume 9% to 17% of U.S. electricity by 2030, up from roughly 4% to 5% today. That estimate is about 60% higher than projections from just two years ago.

In Virginia, data centers already account for about 25% of electricity demand and could rise to as much as 57% by the end of the decade. Several other states could see data centers exceed 20% of total demand.

Other analyses show similar trends, with U.S. data center consumption potentially doubling or tripling in absolute terms by 2030.

Why it matters - For the grid: this is no longer an emerging load, it is becoming a defining feature of the system.
For your audience: the constant upward revisions matter more than the exact number.

GridTake - Data centers are crossing a threshold from large customers to system drivers. Once that happens, the grid doesn’t just serve them, it starts organizing around them.

GENERATION MIX
EIA Sees Gas Filling the Gap if Load Growth Outruns Buildout

The EIA says that if demand from data centers grows faster than expected and new capacity lags, natural gas generation could rise 7.3% between 2025 and 2027, compared to 1.7% in its base case.

Electricity demand is projected to grow 1.9% in 2026 and 2.5% in 2027, with the fastest increases in regions like ERCOT and PJM. In high-demand scenarios, ERCOT prices could rise nearly 79%.

EPRI modeling points to a similar outcome, with natural gas dominating near-term supply additions even as long-term scenarios include more renewables, storage, and nuclear.

Why it matters - For the grid: the system still defaults to the resource that can be built quickly and dispatched reliably.
For markets: load growth does not automatically translate into clean energy growth on short timelines.

GridTake - The gap between ambition and buildability is where gas lives. Until other resources can show up on the same timeline, the system will keep reaching for what it knows it can get.

NUCLEAR
NRDC Backs a Specific Nuclear Project for the First Time

NRDC has backed a specific nuclear project for the first time in its 56-year history, supporting the restart of a plant tied to a Google data center.

The group framed the decision in practical terms, noting the likely alternative would be increased reliance on coal and gas. In Iowa, coal generation rose 32% from 2024 to 2025, pushing its share above 25% despite strong wind penetration.

The support remains conditional, particularly around safety and regulatory oversight, but marks a clear shift in stance.

Why it matters - For the grid: demand is reshaping the politics of clean firm power.
For your audience: longstanding opposition lines are starting to blur under real-world constraints.

GridTake - This isn’t ideological conversion, it’s prioritization. When the choice is framed as nuclear versus immediate fossil expansion, the decision calculus changes quickly.

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

The Conversation

Quick Signals

• Transmission scarcity is now measurable in one brutal ratio: 5,000 miles needed, 888 built.

• Large-load policy is hardening fast, with 20 states already deploying tariffs and more on the way.

• Data center demand forecasts keep moving upward, not stabilizing.

• Near-term “AI power” still looks a lot like natural gas.

• Nuclear politics are shifting as reliability pressures override older alignments.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!