Pennsylvania moves to price grid access for data centers

The Pennsylvania Public Utility Commission has opened a proceeding on ratemaking, return on equity and curtailment, and the three subjects belong in one docket because together they answer a question the commission has not had to face at this scale: what grid access costs, and who loses it first when there is not enough to go around.

The commission's release describes the action as a review of reliability and cost allocation, with the number underneath it coming from PJM's July 2026 capacity auction, which the commission said identified a system-wide reliability shortfall of 6,831 megawatts for the 2028-29 delivery year — the market marking a hole in the supply stack two years before it opens. The commission's report traced that hole to supply: too little new generation is being built while baseload plants keep retiring, it said, leaving policymakers, regulators, utilities and generation owners to hold the bulk electric system together, and it called the impact of data centers and other large loads "significant," in Pennsylvania and across PJM's footprint alike.

Pennsylvania's existing regulations already give utilities a basic framework for managing load during pre-emergency and emergency conditions, but what the commission now wants to rewrite is the order, criteria and applicability of curtailment, the point where a reliability document turns into a capital-markets one. Curtailment order is a ranking of who keeps power when the system runs short, and for a developer underwriting a decade-long obligation the gap between first-to-shed and last-to-shed is not an operating detail but the difference between a financeable project and an unfinanceable one.

The commission's Law Bureau and Bureau of Technical Utility Services will present a tentative order outlining proposed changes to the state's load management framework at the commission's Oct. 1 meeting, and if the commission approves it, the order goes out for public comment with a target vote at its Jan. 28 meeting.

The ratemaking half of the docket may matter more. A working group will weigh whether return-on-equity questions are better handled through periodic, sector-specific cost-of-capital proceedings than litigated utility by utility in individual rate cases, along with ways to improve transparency around ROE in those cases, and it will also examine the return on equity used for Distribution System Improvement Charges.

The first question is the consequential one: litigating ROE inside each rate case ties the allowed return to one utility's petition and one record, whereas a periodic cost-of-capital proceeding sets a standing price for utility equity and lets every utility that comes forward asking ratepayers to fund load-driven capex negotiate against a number the commission already owns. That is a quiet transfer of leverage from the petitioner to the regulator, and it is the part of this docket other state commissions will likely read closely.

The commission tied the ratemaking review to a "rapidly changing landscape" for utility costs — data center development, inflation and infrastructure needs — that it said had increased pressure on costs and contributed to a significant number of proposed rate increases. Chairman Steve DeFrank, who submitted the motion, set the standard as making sure "every dollar recovered from customers is carefully scrutinized while ensuring utilities can make the prudent investments necessary to provide safe, reliable service."

Naming data center development among the drivers of proposed rate increases is a choice with consequences past the arithmetic: the next Pennsylvania rate case is where the cost of serving large loads gets argued, with interconnection agreements sitting in the record.

Read together, the two halves of the docket are the commission's response to a problem it cannot build its way out of on the timeline the load demands: cost allocation and curtailment are the instruments a regulator reaches for when the steel is not there. As this publication has argued, grid access is now the asset and generation and compute are derivatives of connection consent; Pennsylvania is where that thesis gets a rulebook, and rate design is the price of consent.

The docket lands in a PJM that has been clearing interconnection queues faster than it has been building, where the greenlights clear the queue, not the site, and the binding constraint has moved to construction schedules and permitting calendars — a gap the commission's report names directly in its warning about insufficient generation construction. It also lands in a supply mix that is meeting load growth with gas: EIA's forecast has PJM answering its load with gas while solar's growth goes to ERCOT and MISO. Curtailment rules are what a regulator writes when the generation it needs will not arrive on the schedule the load requires.

Reliability has been running the same way regionally. Peak outages fell sharply from Elliott and Uri, and the coordination work since has centered on transfers rather than new steel; a state curtailment framework is that logic one level down, extracting more from the system already standing and deciding in advance who gets less of it.

Public comment would open only after the Oct. 1 meeting, and the commission has set Jan. 28 for its vote. Between those two dates, the developers holding Pennsylvania load in their pipelines will learn where the commission intends to rank them — and that ranking, more than any single rate increase, is what this docket is actually producing.