State regulators already wrote this bill. The Senate won't.

The House has passed the Ratepayer Protection Act, which would require state utility regulators and unregulated utilities to consider large-load standards recovering the full, incremental cost of any generation, transmission or distribution upgrade built to serve a data center or a comparable customer. The two conditions that matter more than the standard itself are financial assurances from the large load before those upgrades are made and guaranteed cost recovery if the customer exits its power supply contract early. Rep. Gabe Evans, a Colorado Republican, and Rep. Kathy Castor, a Florida Democrat, are the primary sponsors, and Evans describes the bill as one that “ensures large data centers pay for the infrastructure they require while giving states the flexibility to determine what works best for their communities.”

The Senate is unlikely to act before the midterms, and analysts cited by Utility Dive describe the measure as largely reinforcing a transition already underway as more states adopt large-load tariffs of their own. That view should shape how a developer reads the House vote: the bill arrives after most of the relevant rulemaking has already been done in the states, so the House has voted on a question the commissions answered first, and the answer is binding in a way the bill is not.

All but 13 states have a utility tariff on the books setting requirements for data centers and other large loads, and at least three of the 13 were weighing proposed requirements as of July, according to the Smart Electric Power Alliance's large-load tariff database. The state terms come in three parts: upfront payment for system impact studies, a ramp to full load within a set number of months, and an exit fee if the customer stops development or significantly reduces its service.

Those three parts are where the connection gets priced. Shifting the study cost, the ramp schedule and the exit exposure onto the developer changes the economics of the connection itself, so two parcels with identical transmission headroom can carry very different economics depending on what a commission attached to the service agreement. As this publication has argued, power rights are now a distinct asset class — the queue, the permit and the connection trade before the electron does — and the large-load tariff is the schedule on which that trio gets priced. Tariff risk belongs in the acquisition model next to basis and curtailment. A pro forma that treats the service agreement as a formality is underwriting a site the commission has not finished describing.

Congress has also written itself a two-year runway: regulators and unregulated utilities would have a year after passage to begin considering whether to adopt large-load tariffs or service requirements, and two years in total to reach a final decision. Because that schedule runs longer than the one states are already keeping, the bill's weight is as a floor and a signal more than a deadline. It also means the provision most likely to change a project timeline — the treatment of power supply costs — would be considered by commissions whose own dockets are further along. The bill's scope also matters: it reaches both state utility regulators and unregulated utilities, which is wider than a purely commission-facing mandate.

The provision with the longest tail is the treatment of generation, and ClearView notes that including power supply costs could be significant precisely because interconnection agreements typically cover only transmission and distribution infrastructure, leaving generation outside the paper that binds a large load to the grid. Requiring data centers to bring their own new capacity could extend a project's construction process by years, the firm argues, since building a power plant takes far longer than building a data hall.

If any version of that clause survives a future Congress, the effect is reclassification more than regulation. Generation built for a campus would move off the rate base and onto the developer's balance sheet, and what had been a contract transferring supply risk to the utility would become a construction project the tenant or its sponsor has to finance, permit, schedule and staff. For anyone holding land with a queue position, the uncomfortable implication is that schedule becomes the binding constraint before capital does, which is why the bill's assurance and exit-fee provisions read as credit requirements. A letter of credit is now part of what makes a site buildable; the sites without one are options rather than projects.

None of this settles whether the grid gets built, because the utility side of the ledger points the other way: utilities see significant growth in the infrastructure needed to serve loads that can use as much power as small cities, and some argue those customers bring prices down by spreading fixed costs across more ratepayers while spurring grid investment. A tariff that guarantees full cost recovery is a tariff that de-risks the utility's capex. A bill pushing states toward one is favorable to regulated utilities with large-load pipelines whether or not it becomes law. The complication is that the same clause cuts both ways: recovery is a comfort to the utility and a direct cost to the developer, and both sides are arguing about the same number across the states at once.

The durable result of the House vote is political: a bill the analysts expect to stall in the Senate still passed the chamber with a Republican and a Democrat as primary sponsors, which means the backlash against data center development and artificial intelligence now has a legislative vehicle, and the cost-recovery standard moving through it runs in the utilities' favor. The terms that will land in a project model next sit in the 13 states with no large-load tariff and in the three that were weighing proposals in July; the first commission to write an exit fee where none existed is likely to set the template its neighbors borrow.