Power & Sites · United States

AEP Ohio's 180-day rule puts supplier-failure risk on data centers, and Duke's tariff case is next

Credited to Rackline · Thomas Brennan · dis.puc.state.oh.us · 18h ago

Photograph · Rackline · Thomas Brennan

In AEP Ohio's territory, a data center that buys power from a competitive supplier now carries the cost if that supplier fails. Ohio regulators confirmed that on Sept. 30, when they refused to revisit the rule. The next tests are a Duke Energy Ohio data center tariff case, with comments due next Thursday, Oct. 22, and pending cases at AES Ohio and FirstEnergy.

The Public Utilities Commission of Ohio denied rehearing requests from Google, the Retail Energy Supply Association and Interstate Gas Supply in Case No. 26-113-EL-ATA, leaving in place interim rules it approved on Aug. 5.

Under the tariff language quoted in the entry, a data center that switches or returns to AEP Ohio's standard service offer, "either voluntarily or involuntarily," "must provide the Company no less than 180 days' advance written notice prior to returning to the SSO." A new data center that wants to start on default service must give the same notice. AEP Ohio then runs a separate procurement for that load, and the customer pays its full cost.

A data center that comes back without notice gets no soft landing. AEP Ohio will "obtain collateral or security to ensure full payment from the Customer and procure the necessary energy through PJM spot market purchases" until the separate procurement can be held, and the data center alone pays the full cost, including AEP Ohio's administrative costs. "The Company shall not take on any obligation to procure capacity on behalf of the load returning to SSO service," the tariff says.

The word "involuntarily" is what matters for procurement. A supplier default is not the customer's choice, but under the rule it lands the customer on spot-priced power and a collateral call, with no utility capacity behind it. The commission did not dispute the concern. It acknowledged the risk of a supplier default raised by Google and still ruled that "any temporary stay associated with the Interim Relief is far outweighed by the prospect of ratepayers having to bear the costs for an energy demand that they did not create."

For developers and their lenders, the practical effect is that a supplier's credit becomes part of the data center's own risk. A campus that once could treat default service as a backstop now has to treat it as a 180-day procurement, which puts more weight on the supplier contract and on what happens if the supplier cannot perform.

Choosing default service has a cost as well. IGS and RESA argued the 180-day notice amounts to a minimum stay. The commission accepted the effect but found that "on balance, the alternative of unfair cost shifts to customers outweighed the industry-specific inconvenience of a de facto minimum stay for data center customers in AEP Ohio's service territory." Either way, a data center's choice between a supplier and default service is now harder to reverse in AEP Ohio's territory.

Google had argued that AEP Ohio's Aug. 28 compliance tariff left unresolved who procures capacity for returning load, which entity holds PJM load-serving entity status for it, and whether the collateral requirement was new. The commission said the changes "do not introduce any new concepts" and that AEP Ohio had carried collateral over from its existing data center tariff.

Two practical pieces are still to come. AEP Ohio told the commission it is working with an auction consultant to design the separate procurement, which it will file for review. That design will shape what a returning data center pays. And the long-term rules sit in a second phase of the case that the commission declined to schedule until a federal process on large-load capacity is complete, finding that "any premature determination of Phase II's parameters could contradict or interfere with the final federal framework." It encouraged Google to raise its concerns there, and said that until then "nothing precludes parties from discussing pathways forward, including formulating alternative proposals."

That leaves the interim rule in force for as long as the federal work takes. The commission described it as "a pragmatic approach to an ongoing, dynamic issue across the country," limited to AEP Ohio's data center tariff.

The commission also rejected Google's call for a statewide approach, writing that "to initiate a statewide solution now would be counterintuitive to the developing regulatory framework." It pointed instead to pending data center tariff cases for Duke Energy Ohio (Case No. 26-755-EL-ATA), AES Ohio (Case Nos. 25-958-EL-AIR et al.) and FirstEnergy (Case No. 26-697-EL-ATA).

For a developer comparing Ohio sites, that means the terms for falling back on default service may differ by utility territory, and they are being set case by case.

Duke's case comes first on the calendar. Duke filed its application on July 27. According to PUCO's docket, a Sept. 18 entry from the administrative law judge set comments for Oct. 22, with reply comments and motions to intervene due Nov. 12. The commission's AEP Ohio ruling gives parties in those cases a clear view of how it weighs supplier and customer arguments against what it called cost shifts to other ratepayers.

What to watch, in order: Duke comments next Thursday, Oct. 22, Duke replies and interventions on Nov. 12, AEP Ohio's filing of its procurement design, and the entry that will set the scope of Phase II.

Power & SitesUnited States

MISO asks FERC for a 120-day fast track for data centers that bring their own power

Credited to Rackline · Sarah Chen · elibrary.ferc.gov18h ago

Useful

The Midcontinent Independent System Operator has asked federal regulators to approve a fast-track review that would study a large new power customer and the generation built to serve it side by side, in about 120 days.

In an Oct. 2 filing with the Federal Energy Regulatory Commission, in Docket No. ER27-25, MISO proposed what it calls the Large Load Addition Resource Study, or LARS. It would be open to loads of at least 200 megawatts paired with new generation in the same local resource zone, the planning areas MISO uses to set capacity requirements. MISO asked FERC to approve the changes effective Dec. 2 so the first study period can open in the first quarter of 2027. Comments are due next Friday, Oct. 23, according to FERC's notice of the filing.

MISO called LARS "a narrowly scoped but powerful process," while conceding that "the benefits of LARS are not free."

MISO, which runs the grid across 15 states and the Canadian province of Manitoba, says it needs the tool because demand is climbing. Demand in its region grew about 0.5% a year from 2009 to 2024. MISO's updated Reliability Imperative report, released in March, projects growth of up to 2.1% a year through 2044, driven by data centers, manufacturing expansion and electrification.

The filing also answers FERC, which found on June 18, in Docket No. EL26-70, that MISO's tariff may be unjust and unreasonable because it lacks interconnection procedures for generators serving "electrically proximate" large loads. MISO says LARS fills that gap for pairs that do not qualify for its zero-injection agreement for co-located plants, which is pending at FERC separately.

The entry terms are strict, and MISO says that is deliberate. Only the utility or other load-serving entity that will buy transmission service for the customer can file an application, though independent power producers can take part as the generator. An application must carry a non-refundable $250,000 deposit and a milestone payment of $30,000 for every megawatt of interconnection service requested; that payment is fully refundable if the project withdraws before signing its interconnection agreement. The generator and the transmission owner must agree on a connection point before applying. The generator can request no more than 150% of the load's peak demand, and the application must show both the plant and the load in service within three years, subject to an existing grace period of up to three more. MISO said the requirements are meant to admit only projects that are "shovel ready" and commercially viable.

MISO will take 10 applications in each of three study periods a year, in the order they are filed, which caps the program at 30 pairs a year. The generation study is meant to produce an interconnection agreement in about 120 days. The parallel load study, built on MISO's existing expedited review, runs about 90 days. Entergy backed the concept in September comments posted on MISO's website but called the cap "arbitrary."

The pairing is meant to stick. To apply, the generator needs a binding contract of at least five years with the utility or the customer, but the plant and the load must stay paired for 15 years once the load starts taking service. If the load leaves or the contract ends inside that window, MISO will cancel the generator's interconnection agreement after a transition period of up to a year, longer if the plant is moving through MISO's regular queue. Parties that withdraw after the study still owe for assigned network upgrades. After 15 years, the generator keeps its interconnection rights even if the load goes away. MISO said 15 years is a standard term for large-load contracts and long enough to deter developers from invoking a large load "as a pretext to bypass" the regular queue.

The design moved quickly. MISO introduced the concept in July with a 250 MW floor, then cut it to 200 MW after what it called "significant stakeholder feedback that many emerging data center and large industrial projects in the 200-250 MW range can have significant system impacts."

Not everyone is persuaded. In September comments on an earlier draft, posted on MISO's website, the grid operator's environmental sector, a group whose members include the Natural Resources Defense Council, Sierra Club and the Union of Concerned Scientists, said LARS "is not responsive to FERC's Show Cause Order" because it treats generation anywhere in a resource zone as close enough. It said a fast track that prioritizes large-load customers over all other end users "introduces serious discrimination concerns," and urged MISO to drop LARS for an alternative design or let it expire after a year.

MISO argues that LARS projects are kept out of its regular queue studies and use none of the transmission capacity available in them, so any effect on later queue projects "would only be positive." It said it will file a cost-shifting agreement and transmission service terms for large loads by Nov. 16.

At least 20 parties had moved to intervene by Oct. 13, including Entergy, Vistra, Constellation and NextEra Energy Resources.

Source · elibrary.ferc.gov

Front Method

Power & Sites · United States

Can the AI labs pay for the boom?

Credited to Rackline · Elizabeth Harding · anthropic.com · 12h ago

Photograph · Rackline · Elizabeth Harding

The most revealing document about Anthropic this year was not one the company wrote. It sits on page 14 of a prospectus filed by SpaceX, which owns Elon Musk's xAI, and it says the maker of Claude "has agreed to pay us $1.25 billion per month through May 2029." Once a cut-price ramp-up in May and June 2026 gave way to the full rate in July, that came to $15 billion a year, paid to a company whose own chatbot, Grok, competes with Claude.

The money buys time on machines in the American South. SpaceX's prospectus describes Colossus as its flagship data center, on Paul R. Lowry Road in Memphis, Tennessee; Colossus II spreads across Memphis and Southaven, Mississippi, part of what the company calls a "coherent gigawatt-scale AI training cluster." Anthropic's share includes "approximately 325,000 NVIDIA GPUs."

Multiply that contract across the industry and you have the American economy's biggest construction project. Can the customers at the end of the chain, the frontier labs selling Claude, ChatGPT and their rivals, earn enough to pay for it? Our answer, for now, is no. The build-out is being carried by the cash flows of four of the world's most profitable companies and, increasingly, by debt. The labs' revenue is the justification for it rather than the means. That can work for a while. It cannot work for ever.

Start with the size of the bet. Alphabet's second-quarter call in July raised its 2026 capital-spending range to $195 billion-205 billion; Anat Ashkenazi, its finance chief, said she expects capex "to increase significantly in 2027." Meta, the same week, guided to $130 billion-145 billion, a figure that includes principal payments on finance leases. Amy Hood of Microsoft put calendar-2026 spending at about $175 billion after an accounting change that moves some data-center leases out of the capex line, and promised more in the fiscal year to June 2027. Amazon's written guidance, from its February results release, is still about $200 billion: Andy Jassy, its boss, said then that "we expect to invest about $200 billion in capital expenditures across Amazon in 2026." Its second-quarter release did not update that full-year figure. Add those four plans with care and they come to roughly $700 billion this year—an illustrative sum, not a single industry total. Meta's range includes finance-lease principal; Microsoft's $175 billion is after the lease reclassification; the figures are not like-for-like.

The justification, in writing, is customer demand already booked. In his 2025 letter to shareholders, Mr. Jassy said: "We're not investing approximately $200 billion in capex in 2026 on a hunch." Of the AWS portion, "we already have customer commitments for a substantial portion of it."

The suppliers' books show where the money lands. Nvidia's data-center revenue was $89 billion in the quarter to July 26th, up 117% on a year earlier, according to its results on August 26th, and it guided to total revenue of $108 billion, give or take 2%, for the quarter now ending. Oracle reported $664 billion of remaining performance obligations at the end of August. Google Cloud's backlog reached $514 billion in the second quarter, Sundar Pichai told analysts.

Now set the labs' revenue beside that. Anthropic said in May that its run rate, which extrapolates recent sales over a year, had crossed $47 billion. On October 9th Sarah Friar, OpenAI's finance chief, wrote on LinkedIn: "We ended September with an annualized revenue run rate of ~$50 billion, up 77% in just three months." SpaceX's AI segment, which includes Grok and X, recognized $2.56 billion of revenue in the second quarter. Annualize xAI's quarter and add the two run rates—mixing run rates with recognized revenue, and figures months apart—and the three together are selling something near $110 billion a year, against a single year's illustrative capex of roughly $700 billion by just four buyers.

The comparison is unfair in one way: data centers last for years and revenue recurs, so a year's revenue need not match a year's capex. Microsoft has just stretched the assumed life of its data-center buildings from 15 years to 25. But chips wear out much faster than buildings, and roughly two-thirds of Microsoft's capex last quarter went on short-lived assets, mainly CPUs and GPUs, Ms. Hood said. Those have to earn their keep within a few years. The revenue that must pay for them is, for now, mostly a run rate.

That is why Anthropic matters beyond its own valuation. Its $965 billion post-money price, set in its Series H on May 28, is the market's most explicit wager that lab revenue will grow into the infrastructure built for it. And Anthropic has gone quiet. Its run rate stood at $14 billion in February and above $47 billion by May; since the Series H release, almost five months ago, it has published no new figure. On June 1 it said it had confidentially submitted a draft S-1 to the Securities and Exchange Commission. No public filing has followed.

Nobody outside can say what Anthropic earns, what it keeps after paying for chips, or what it owes Microsoft and Google in any given year.

What is known is the bills. Beyond the Memphis contract, Anthropic has said it is committing more than $100 billion over ten years to Amazon Web Services; spread evenly, that is at least $10 billion a year. With SpaceX, it carries at least $25 billion a year of compute spending on disclosed terms, roughly half its May run rate. Its $30 billion Azure purchase, agreed in November 2025, came with no term, and its April deal with Google and Broadcom for multiple gigawatts of TPU capacity from 2027 came with no price, though Anthropic called it "our most significant compute commitment to date."

No lab publishes a gross margin, so assume, crudely, that compute eats half of revenue. Anthropic would then need about $50 billion a year to carry its two contracts with known terms, and the May run rate is just short of that. Stretch the Azure deal over five years, a term nobody has disclosed, and the figure rises to about $62 billion. On that arithmetic a company valued near $1 trillion is running at about compute break-even, provided the run rate turns into revenue.

OpenAI looks healthier on paper, and the calendar helps it. Against its $852 billion valuation of March 31st, Ms. Friar's September figure implies a multiple of about 17, below Anthropic's roughly 20 times on its May number. Its AWS and CoreWeave contracts with disclosed terms come to roughly $19 billion a year, which on the same half-of-revenue assumption needs about $38 billion of sales. But the $250 billion of Azure services it agreed to buy in October 2025 has no stated term; if it ran ten years, the break-even would rise to about $88 billion. The two firms do not calculate their run rates alike, and Anthropic's latest number is more than four months older than OpenAI's.

The money also travels in circles. Amazon put $8 billion into Anthropic convertible notes between 2023 and 2025 and bought $5 billion each of Series G and Series H preferred stock in the second quarter of 2026. Microsoft and Nvidia pledged up to $15 billion between them alongside Anthropic's Azure deal. Microsoft owned about 27% of OpenAI when its recapitalization closed in October 2025. Much of that equity returns to the investors as payment for chips and cloud capacity. The cloud firms are financing their own order books.

They are also booking gains on them. Amazon's quarterly report for the period ended June 30th carries its Anthropic preferred stock at about $92.5 billion and its convertible notes at about $97.9 billion, after $50.5 billion of upward adjustments in the second quarter alone; those are estimates derived from private rounds. Microsoft booked $6.5 billion of net gains on OpenAI in the fiscal year to June. Profits that rest on private marks of customers' shares are a thin cushion if the customers stumble.

Increasingly the cushion is debt. Meta, which does not sell cloud capacity, raised $24.9 billion of long-term debt in the second quarter, according to its results. In October 2025 it set up a joint venture with funds managed by Blue Owl Capital to build and own its Hyperion campus in Richland Parish, Louisiana: Blue Owl holds 80%, the partners committed to fund about $27 billion of development costs, and Meta will lease the buildings on four-year initial terms. Meta also gave the venture a residual-value guarantee for the first 16 years, a capped payment if it walks away. Structures like that move the borrowing to outside lenders, partly bonds sold to PIMCO and others, while leaving Meta on the hook if it decides it needs less space.

The build-out is now large enough to show up in the national accounts. In its third estimate of second-quarter GDP, published on September 30th, the Bureau of Economic Analysis said the biggest contributors to the upward revision in private fixed investment were nonresidential structures, "led by commercial and health care, mainly data centers." BEA does not publish an "AI" line in the accounts. The closest related category, private fixed investment in information-processing equipment and software, ran at an annual rate of $1.65 trillion in the second quarter; the contribution of information-processing equipment alone to GDP growth slowed to 0.19 percentage points from 0.79 in the first.

It is also showing up in prices. Lisa Cook, a Federal Reserve governor, said on September 28th that companies "have only spent a small fraction of the $2 trillion in announced plans," and that electricity and water costs "are each up around 5 percent over the past year, potentially attributable in part to AI." She voted with her colleagues to raise interest rates by a quarter-point in September, with inflation at an estimated 3.8% in the year to August. "In the short term," she said, "AI appears to be adding inflationary pressures to the economy."

The grid is being planned around the same expectation. PJM, which runs the grid across much of the mid-Atlantic and Midwest, forecast in January that its summer peak would reach 222,106 megawatts in 2036, an increase of 65,733 megawatts in ten years, with data-center load growth written into the outlook for 14 of its zones. ERCOT, the Texas grid operator, said in an April CEO update that transmission-service providers' preliminary RFI submissions for non-crypto data-center loads rise from 7,401 megawatts in 2026 to 228,420 megawatts by 2032—requested large loads, not firm interconnected demand. Power plants and transmission lines are built on 30-year assumptions. The revenue they are being built for is measured in months.

Further down the market, the price of intelligence is falling. Mistral, a French lab, raised €3 billion in September at a post-money valuation of more than €21 billion, selling open-weight models customers can run on their own hardware. On the companies' own price lists, read on October 15th, Anthropic's flagship, Claude Fable 5.1, costs $10 per million input tokens and $50 per million output tokens; DeepSeek-V4-Pro, whose weights are open, costs $1.32 and $3.96 at peak hours. If cheap models handle most corporate work, revenue per chip will fall just as the chips arrive.

What happens if the money stops? It need not stop to hurt. A listing that prices Anthropic below its Series H valuation would turn up promptly in Amazon's other income, because those marks would then follow a market price. A pause in lab spending would hit Nvidia's order book, then the utilities and lenders that financed the buildings, then the investment line that has been propping up GDP. Ms. Cook's worry runs the other way, too: a slowdown would ease the price pressure she described, but it would also expose how much of recent growth rested on one sector.

A sudden stop looks unlikely in the next year. The hyperscalers' guidance is set, their cloud backlogs are long, and Amazon says it already has customer commitments for a substantial share of this year's AWS spending. The risk is slower and duller: capex keeps rising on the strength of contracts and marks, while the revenue meant to justify it arrives as run rates, from companies that publish when they choose.

For the build-out to pay, several things must be true. Lab run rates have to become audited revenue at something like their stated pace. Compute must take a falling share of that revenue—cheaper chips, or prices that hold up against open-weight rivals. Enterprises, not just labs and hyperscalers trading capacity with one another, must become the paying customers. And the debt now financing the buildings must find tenants who stay.

Over the next 12 to 24 months, watch five things. First, the clouds' third-quarter results at the end of October, and their first formal capex guidance for 2027. Second, Anthropic's public S-1, which should show recognized revenue, losses, customer concentration and the missing Azure and Google terms, and Amazon's next filing, which will show whether it has marked its stake up again. Third, Nvidia's next quarter against its $108 billion guide. Fourth, the BEA's advance estimate of third-quarter GDP, and whether data centers keep carrying investment. Fifth, the Fed: if AI spending keeps pushing up electricity and construction costs, the build-out will be paying higher interest rates of its own making.

The labs have persuaded the world's largest companies to back them against revenue not yet recognized, and to take payment in their own future sales. The economy is leaning on that arrangement too. The first audited numbers will tell investors whether $965 billion bought a company or a promise.