Ratepayer Protection Act Passes House: Data Centers

Correction, 2026-10-12: Coverage depends on PURPA retail-rate jurisdiction and its retail-sales threshold, not utility ownership alone. Financial-assurance and cost-recovery provisions are proposed standards for consideration, not requirements every utility must adopt. Legislative status and unrelated reporting below remain as of September 28, 2026.

The House passed the Ratepayer Protection Act, H.R. 9340, on September 16 by a vote of 417 to 3, under a motion to suspend the rules and pass the bill as amended, according to the House Clerk’s record of roll call 312. The bill would create a federal standard on who pays when a data center needs new generation, transmission, or distribution upgrades to interconnect. The introduced version covered any large load at 100 MW or more. The version the House passed covers data centers only.

As of the September 28, 2026 legislative-status check, it was not law. The Senate had received the bill on September 17 and placed it on its legislative calendar on September 24, with no floor action recorded in that check.

That is the headline. The mechanism underneath it is what a procurement team needs to read correctly, because the bill does not do what the headline implies.

What the Ratepayer Protection Act Actually Requires

The bill amends Section 111(d) of the Public Utility Regulatory Policies Act of 1978. PURPA 111(d) uses a “consider and determine” process. If enacted, this bill would require covered state regulatory authorities and nonregulated electric utilities to consider the proposed standard and issue a determination; it would not require adoption. For this purpose, nonregulated status depends on whether a state authority has ratemaking authority over the utility’s retail sales, not on municipal or cooperative ownership. Virginia, for example, has electric cooperatives subject to SCC rate jurisdiction. PURPA title I also has a sales threshold: retail sales exceeded 500 million kWh in any calendar year after December 31, 1975 and before the immediately preceding calendar year; wholesale sales are excluded. Congress has used that same mechanism to push net metering, time-of-use rates, and integrated resource planning onto state agendas since the 1970s, and the adoption record on all three diverged sharply by state: some commissions adopted the standard, others opened a docket and declined, and a few let the question sit for years.

Read the bill correctly and the forecast is not “data centers will pay their own way nationwide.” It is closer to “covered regulators and nonregulated utilities would have to consider data-center cost allocation on the record, within a defined window, if enacted.” The passed text gives them one year after enactment to begin consideration or set a hearing date and two years to finish it. It also exempts a utility where, before enactment, the state already implemented a comparable standard, the regulator or utility already held a proceeding to consider one, or the legislature voted on it. That is a procedural mandate with carve-outs, not a rate mandate.

The standard itself is worth reading in the bill’s own words. The proposed standard would call for a rate charged to a covered large-load customer to recover “the full, incremental cost of any generation, transmission, or distribution upgrade necessary to serve the load of such large-load customer.” Three categories, named individually. Shorthand like “grid upgrades” throws away the only word that decides anything below the transmission system. A generation or transmission upgrade is somebody else’s capital plan. A distribution upgrade is the substation, the feeder rebuilds, and the direct-connect facilities, and whether those come out of a utility’s rate base or a developer’s capital budget is the difference between a municipal utility absorbing a 100 MW interconnection and a municipal utility invoicing it.

The threshold sits with the definition rather than the standard. In the House-passed text, a large-load customer is a non-residential consumer whose facilities require electric energy primarily to operate information technology infrastructure for data storage and computation, and carry, in the aggregate, a peak demand of 100 MW or more at a single site or campus. Both conditions have to be met.

A note on citation, because it decides what you can rely on. An earlier version of this article quoted the introduced text and noted that the committee substitute had not yet been published. The House-passed text is now public on GovInfo, and every quotation in this article has been checked against it. The three-category cost language and the financial assurance language carry through unchanged. That legislative-status statement reflects the September 28 check. The October 12 correction addresses coverage and conditional language; it is not a new status check of every claim in this article.

The bill is sponsored by Reps. Gabe Evans (R-CO) and Kathy Castor (D-FL), with committee action led by Chairmen Brett Guthrie and Bob Latta. It was introduced June 18, cleared full committee 52 to 0 on July 21, was reported to the House on September 10 (H. Rept. 119-814), and passed the House 417 to 3 on September 16.

Large-Load Financial Assurance Under the Proposed Standard

Cost assignment decides who pays eventually. It does not decide who carries the risk if the load never arrives. H.R. 9340 handles the second question in a separate provision.

Under the proposed standard, if adopted, before making any generation, transmission, or distribution upgrade necessary to serve a large-load customer, “an electric utility shall require the large-load customer provide to the electric utility financial assurances or contributions to cover the cost of such upgrade.”

Read that against the failure mode it is written for. A cooperative or municipal utility builds a substation against a signed 100 MW interconnection request. The project cancels. The steel is in the ground, and the cost lands in a rate base with a few tens of thousands of customers to spread it across. Cost assignment alone does not help there, because there is no longer a large-load customer to assign anything to. Assurance before construction does.

The cost-recovery language carries the same idea past energization. It applies, in the bill’s words, “including in the event of such large-load customer terminating a contract or other agreement with the electric utility pertaining to the sale of electric energy, or otherwise ceasing the purchase of electric energy from the electric utility.” Under an adopted version of that standard, a developer’s departure would not by itself remove the cost-recovery obligation.

For a buyer this is the provision that changes a specification decision rather than a rate filing. If your commission adopts a standard carrying an assurance requirement, the first question on a large-load interconnection stops being what the customer will pay and becomes what the customer has posted and when it releases. That question gets settled in a term sheet, before you commit to a transformer slot, not after.

Why Congress Narrowed the Federal Large-Load Standard to Data Centers

The lopsided votes, 52 to 0 in committee and 417 to 3 on the floor, line up with one fact: the bill codifies a standard that a voluntary coalition had already accepted. Seven companies signed the original Ratepayer Protection Pledge in March, agreeing, in the words of White House AI and crypto adviser David Sacks as quoted in the White House’s March 5 release, “to cover the costs of all new power generation required for their data centers”: Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI. Legislating an existing voluntary commitment is low-friction, which helps explain near-unanimity on a bill touching utility cost allocation, usually contested territory.

The narrowing from “any load at 100 MW or more” to “data centers only” is the detail worth tracking. It exempts new industrial and manufacturing loads at the same size threshold: steel mills, chemical plants, EV and battery factories. An aluminum smelter and an IT/data-center campus at identical load would fall on different sides of this proposed standard’s definition. Expect state-level fights over the definition of “data center” wherever a hyperscale facility sits next to, or inside, an industrial campus.

Ratepayer Protection Pledge Signatories: Who Signed, and Why the Mix Matters

Two days after the committee vote, on July 23, the White House announced that more than 200 additional utilities, data center developers, cooperatives, and states had joined the pledge, which it says now covers 80 percent of all power delivered to US homes and businesses. The White House release gives no category breakdown. POWER magazine’s tally puts the total at 187 organizations across three stakeholder groups plus 23 governors, and its cooperative and public power group is the largest bloc on its list: 105 entities in POWER’s article text and 108 entries in its published list, which also includes public power trade groups. Basin Electric, Dairyland Power, and Hoosier Energy are among the cooperatives POWER lists. Treat the bloc sizes as one outlet’s count and the ordering as the finding.

That ordering inverts how the pledge usually gets described. This is not a hyperscaler-and-IOU arrangement with public power watching from the outside. The segment that signed it most heavily is the one with the smallest rate base to absorb a mistake, and Basin Electric, Dairyland, and Hoosier are generation and transmission cooperatives, which is the level where wholesale large-load contracts actually get written. Our read on what a cooperative fleet does when data center load lands on it covers the operating end of the same question.

Two limits are worth holding onto. The pledge is voluntary and nonbinding throughout, so between it and a consider-and-determine mandate, the operative requirements for a specific interconnection depend on the applicable commission order or utility policy and agreement. And the 23 governors POWER lists are all Republicans, which makes the geographic durability of the commitment a live question rather than a settled one.

The Large-Load Cost-Allocation Hole the Act Doesn’t Close

The Ratepayer Protection Act operates at the retail, state-jurisdictional level. It says nothing about network-upgrade costs allocated regionally, at the wholesale level, through a Regional Transmission Expansion Plan. That gap is not theoretical. Days before this committee vote, consumer advocates from Delaware, Illinois, Maryland, and Ohio filed jointly at FERC in docket EL26-67, and Pennsylvania’s Office of Consumer Advocate filed separately in the same docket, arguing that the commission’s own large-load show-cause order for PJM leaves exactly this question open.

The scale is worth citing directly. Joseph Bowring, president of PJM’s Independent Market Monitor, told Utility Dive that data centers account for $6.3 billion, or 38 percent, of the $16.4 billion in charges from the 2028/2029 capacity auction, and $29.4 billion, or 46 percent, of the $63.6 billion across the last four base auctions. A state can adopt the full Ratepayer Protection Act standard for its retail rates and still see the network upgrades serving that same data center socialized across every ratepayer in the RTO footprint. Congress is answering the state question. Nobody with authority has answered the RTO question yet.

What This Means for Equipment Procurement

If a state opens a docket and adopts a version of the standard, the purchasing decision moves, not just the invoice. A utility recovering costs through its rate base buys to its own specifications, its own vendor list, and its own multi-year cycle. Developers paying directly buy to a construction schedule, negotiate their own supply agreements, and are often more willing to pre-buy long-lead transformers, switchgear, and HV breakers, or fund a manufacturer’s added capacity outright to hold a slot. That last move has become its own procurement instrument, and we covered how a production slot reservation now substitutes for the purchase order on exactly this class of buyer.

That shift is already underway in states with codified large-load tariffs, from Oregon’s Schedule 96 to Duke’s proposed rate class to Pennsylvania’s but-for framework. The Ratepayer Protection Act adds federal pressure behind a trend already in motion rather than starting a new one. For municipal and cooperative buyers competing for the same transformer and switchgear slots without a 100 MW customer of their own, the practical effect is more counterparties in the queue carrying fewer procurement constraints and more schedule urgency than a rate-regulated utility.

The September 28 status check identified two items to watch. First, the Senate. H.R. 9340 sat on the Senate legislative calendar (Calendar No. 684) after its September 24 placement, and a same-titled Senate bill from Sen. Jon Husted, S. 5028, had been in the Energy and Natural Resources Committee since July 16 with no further action in that check. A “consider” standard that never clears the Senate stays a House-only signal. Second, whether the EL26-67 rehearing requests move FERC to address wholesale network-upgrade allocation, because that is the half of this problem a state commission cannot touch no matter how the Ratepayer Protection Act resolves.

Our utility procurement intelligence guide tracks every large-load cost-allocation template, federal and state, shaping the 2026 buying window.

The Feeder is our free weekly read on the regulatory and procurement signals that move equipment demand. We break down what each Congressional, FERC, and state action means for buyers before the orders show up in your lead times. Subscribe to The Feeder and stay ahead of the buying cycle.

Related Reading

• FERC Puts 6 RTOs on the Clock: Large-Load Show-Cause Order
• Who Pays for Data Center Grid Upgrades? FERC’s June 18 Decision
• Pennsylvania Large Load Tariff: ‘But For’ Cost Allocation
• AI Ratepayer Protection Meets State Utility Law
• Cooperative Utility Data Center Load Meets a 936 MW Fleet
• Equipment Production Slot Reservation Replaces the Order