Pennsylvania’s Power Grid Is Racing Against the Data Center Boom — And Right Now, It’s Losing
A new state-commissioned study warns that PJM, the grid operator serving 65 million Americans across 13 states, could face reliability shortfalls more than 130 times worse than its safety threshold by 2030 — driven almost entirely by explosive data center growth.
If data center demand keeps outpacing new power supply, PJM Interconnection — the largest grid operator in the United States — could see its loss-of-load expectation (LOLE) spike to 13.20 in 2030, against a reliability standard of just 0.1. That’s not a prediction of 13 straight days of blackouts. It’s a warning sign that the odds of localized power shortfalls somewhere across the PJM footprint are rising sharply, and fast.
The culprit, according to the Pennsylvania Public Utility Commission’s newly released analysis, is simple to state and hard to solve: data centers are being built and energized faster than power plants, transmission lines, and fuel infrastructure can be permitted, financed, and constructed.
What the Study Actually Found?
The report — produced by Synapse Energy Economics, Mondre Energy, and Aspen Technologies for the Pennsylvania PUC — modeled several scenarios for how PJM’s grid might perform through 2040. The most alarming one, called High Load and Low Supply, shows LOLE jumping from a manageable 0.28 in 2029 to a startling 13.20 just one year later, in 2030.
For context, PJM’s own resource adequacy standard sets the acceptable threshold at 0.1 LOLE — meaning the system should expect a loss-of-load event roughly once every ten years. A reading of 13.20 blows past that benchmark by more than 130 times.
Even in the study’s more conservative Reference case, LOLE still climbs — from 0.21 in 2029 to 0.59 in 2030 — a smaller jump, but one moving firmly in the wrong direction.
What’s driving it? More than 200 terawatt-hours of new data center load entering the PJM system between 2029 and 2030 alone. Total electricity demand across the grid rises by 229 TWh in that single year — and nearly all of it traces back to data centers. To put that in perspective: all of Pennsylvania used about 150 TWh of electricity in 2025. This one-year demand jump is larger than the entire state’s annual consumption.
Why 2030 Is the Tipping Point?
PJM’s own spokesperson, Jeff Shields, didn’t mince words when describing the dynamic at play.
“Electricity demand, driven primarily by data centers and other large loads, continues to grow at a faster rate than the ability to build new generation,” Shields said, adding that the mismatch could “potentially compromise reliability of the system over the next five years while also contributing to increased electricity costs.”
That’s the crux of the problem — and it’s not a supply-side failure in the traditional sense. The study’s generation assumptions don’t shift dramatically in 2030. What changes is the sheer scale and speed of new demand hitting the system, which is precisely why the timing matters so much.
Is 13.20 LOLE a Forecast — or a Warning?
Not everyone reads the number the same way.
Neil Osnato, founder of Persistence Analytics Group, cautions against treating the figure as a hard prediction of what 2030 will actually look like. Instead, he frames it as a stress test — a way to pressure-test the assumptions PJM and its member utilities are planning around.
“I would not dismiss the number as arbitrary,” Osnato said. “It comes from actual utility forecasting inputs.”
Importantly, the High Load and Low Supply scenario isn’t built on speculative hype. It uses utility-reported large-load forecasts, discounted to just 50% of their stated values, with overlapping projects filtered out and year-to-year spikes smoothed. Even after all that conservatism, the model still produces a 229 TWh jump in a single year.
Osnato’s bigger point: the real work now is verification. Utilities and data center developers need to determine how much of that projected load will actually show up, and when — by examining:
- Site control and permitting status
- Committed end-users and executed power purchase agreements
- Required transmission and distribution buildout
- Construction timelines and realistic load ramp-up schedules
“The 229 TWh jump is less a prediction than a very consequential assumption that deserves verification,” he said.
Notably, PJM has already begun discounting the amount of planned data center demand it factors into its own forecasts, precisely because so much of it remains unconfirmed.
The Core Problem: Speed Mismatch, Not Just a Power Shortage
This isn’t simply a story about not having enough power plants. It’s a story about timing.
Data center projects can move from paper forecasts to fully executable, power-hungry operations far faster than new generation, transmission lines, and fuel infrastructure can clear permitting, secure financing, and get built.
“A generation queue is therefore not the same thing as dependable operating capacity,” Osnato noted — a distinction that cuts to the heart of the issue. Plenty of new power projects exist on paper. Far fewer are actually online and dispatchable when the grid needs them.
As Osnato put it: “The key question is, how many dependable megawatts can actually reach commercial operation before the corresponding load becomes executable?”
Pennsylvania’s Shrinking Role as an Energy Exporter
The study also flags a structural shift specific to Pennsylvania. The state supplied about 23% of PJM’s total generating capacity in 2025 — a share the study projects will shrink to somewhere between 14% and 17% by 2040, depending on the scenario.
Under the High Load and Low Supply case, Pennsylvania’s electricity exports to neighboring states collapse — from 91 TWh in 2025 down to just 26 TWh by 2035. By 2040, the state is projected to flip from a net power exporter to a net importer.
Zoom out across the entire PJM footprint, and the numbers get more sobering still: the study estimates that 2% to 19% of PJM’s total electricity demand could go unmet between 2031 and 2040 under the stressed scenario — a range that illustrates just how quickly grid reliability can erode once large loads outrun dependable supply.
What’s Being Done About It?
To be clear, PJM and Pennsylvania aren’t standing still. Both are actively working both sides of the supply-demand equation, according to Shields. Current efforts include:
- Streamlining permitting for new generation projects
- Setting a price floor and ceiling in the capacity market to stabilize investment signals
- Establishing data center development rules tied to the White House Ratepayer Protection Pledge
- Aligning with a statement of principles issued by governors across the PJM region
“PJM believes these combined, focused efforts can preserve reliable electrical service in the near and long term without unfairly burdening regular ratepayers,” Shields said.
The study also credits several PJM initiatives already underway — the Reliability Resource Initiative, Reliability Backstop Procurement, and an Expedited Interconnection Track — as steps designed to accelerate dependable capacity onto the grid.
Whether those measures move fast enough to outpace data center growth is, in many ways, the defining energy question for the back half of this decade.
This study isn’t a doomsday forecast — it’s a flashing warning light. It tells PJM, Pennsylvania regulators, utilities, and data center developers exactly where the pressure points are before they become emergencies. The next few years will determine whether new generation, transmission, and permitting reforms can catch up to demand growth that shows no signs of slowing down.
The questions now aren’t abstract. They’re operational: How much of the projected 229 TWh surge is real? How fast can it be verified? And can dependable megawatts reach the grid before that demand becomes unavoidable?
For a region powering 65 million people, those answers can’t come soon enough.
