FERC approves PJM’s registry for its biggest electricity users

The US Federal Energy Regulatory Commission approved regional grid operator PJM’s large-load definition and registry on 9 October, effective 12 October, while rejecting its linked curtailment and bring-your-own-capacity package. PJM must submit cleaned tariff language within 30 days of the order.[1]

Behind PJM’s August filing was a concentrated growth forecast: of 32 gigawatts of additional electricity demand it expected between 2024 and 2030, 30 gigawatts came from data centres. These were forecast increases, rather than a count of campuses already connected. PJM wanted new large customers to bring matching new supply or face an earlier power reduction during a shortage.[2]

The accepted registry has a simpler job: make large demand visible. Its threshold is a 50 MW peak at one electrical site, combining affiliated facilities within a one-mile radius. Existing and future loads qualify, across industries.[1]

A campus has more than one power date

Records will cover phased in-service dates, peak demand, ramp schedules and backup-generator fuel and permitting limits. Individual details retain confidentiality safeguards; PJM will publish aggregate information.[1]

Imagine two proposed 300 MW campuses. One expects to draw its full capacity in 2027. The other plans 100 MW in 2027, another 100 MW in 2029 and the final 100 MW in 2031. Both can be advertised as 300 MW projects. In this hypothetical example, however, the first requires three times as much power as the second at the initial milestone. Treating both as fully occupied from day one would produce a different forecast from following their staged schedules.

The difference is consequential even if both developers ultimately deliver everything they promised. Timing changes when additional demand reaches the system. A planner needs a sequence of expected loads, with the uncertainty attached to each stage, to decide what must be ready first. The final campus size supplies only the last point in that sequence.

The separate dispute over who loses power first

PJM’s proposed Interim Resource Adequacy Service, or IRAS, went further. Its August announcement described directing utilities to reduce demand from new large customers whose needs were not covered by new supply, before reducing demand from customers already paid to participate in load management. PJM presented the approach as a way to protect existing consumers and said state authorities would retain responsibility for retail arrangements.[2]

FERC found parts exceeded its jurisdiction and PJM had inadequately justified treating new large loads differently. It accepted the registry separately, rejecting the rest, including bring-your-own-capacity provisions.[1]

Commissioner David Rosner’s concurrence explains why the market design matters. He objected to mandatory cuts before voluntary demand response had been exhausted. He also warned about excluding the affected loads from the capacity market’s demand curve: the market could appear adequately supplied while the region was forcing substantial demand off the system.[3]

A capacity market pays for resources to be available when needed. Rosner’s concern was that removing demand from the calculation could weaken the signal to build the supply needed to serve it. He supported giving ordinary customers priority over non-critical data centres in an emergency, but urged states to establish the order of curtailment and PJM to use voluntary reductions first.[3]

PJM spokesman Jeffrey Shields told Utility Dive on 12 October that the operator was reviewing the decision, including the responsibilities assigned to states, while seeking to protect existing customers from unfair costs or reduced reliability. That response leaves a revised approach to the larger power-sharing problem ahead.[4]

The registry decision authorises information gathering. Any replacement curtailment proposal must address the objections that defeated this one.[1]

Report an error or suggest a correction