The grid operator responsible for keeping the lights on across 13 states and Washington, D.C., is staring at a hole in its power supply that it cannot fill in time. PJM Interconnection, which coordinates electricity for more than 65 million people from New Jersey to Illinois, has flagged a capacity shortfall of roughly 6 gigawatts for the 2027 delivery year. To put that number in perspective: 6 gigawatts is enough electricity to serve the combined peak demand of about six mid-sized American cities, all at once, on the hottest afternoon of the year.
The gap means PJM’s annual capacity auction, the mechanism designed to lock in enough power plants and demand-reduction commitments years ahead of time, failed to secure the resources needed to cover projected peak load plus the safety margin regulators require. With the 2027 delivery window now less than two years away, the timeline for closing that deficit is brutally short.
Where the shortfall comes from
PJM’s capacity troubles did not appear overnight. Three forces have been converging across the grid operator’s territory for years, and they are now colliding.
First, coal and older natural gas plants are retiring faster than replacements can be built. Tightening environmental regulations, aging infrastructure, and unfavorable economics have pushed dozens of fossil-fuel generators toward shutdown across Pennsylvania, Ohio, West Virginia, and other PJM states. Each closure removes firm, dispatchable capacity from the supply stack.
Second, electricity demand is rising in ways PJM’s forecasters did not fully anticipate even a few years ago. The single largest driver is the explosion of data center construction, particularly in Northern Virginia’s “Data Center Alley,” which already hosts the densest cluster of data centers on the planet. The buildout of artificial intelligence infrastructure has accelerated power demand from these facilities, and PJM has publicly acknowledged that load growth from data centers is reshaping its planning outlook. Dominion Energy, the utility serving much of that corridor, has reported unprecedented interconnection requests from data center developers.
Third, new generation is stuck in a bottleneck. PJM’s interconnection queue, the line that proposed power plants and battery projects must clear before connecting to the grid, has been widely reported as backlogged with hundreds of gigawatts of pending projects, most of them solar and battery storage. The queue backlog means that even projects with financing and permits may not reach commercial operation before 2027. PJM has undertaken reforms to speed the process, but the reforms apply mainly to newer applications, not the legacy queue.
What the auction results revealed
PJM’s capacity market operates under rules approved by the Federal Energy Regulatory Commission (FERC). Generators, demand-response providers, and energy-efficiency programs bid into the auction, and PJM clears enough resources to meet its reliability requirement. When the auction clears at high prices, it signals scarcity and is supposed to attract new investment. When it clears below the cost of new construction, developers pull back.
PJM’s most recent Base Residual Auction, held in 2024 for the 2025/2026 delivery year, produced record-high clearing prices, nearly ten times the previous year’s level in some zones. Those prices reflected the market’s recognition that supply is tightening. But high prices in one auction do not automatically translate into new megawatts on the ground for the following delivery year. Construction timelines for combined-cycle gas plants typically run three to five years. Utility-scale battery storage can move faster, sometimes 12 to 18 months, but permitting, supply-chain delays, and interconnection hurdles still slow many projects.
For the 2027 delivery year, the auction did not attract enough committed resources to cover PJM’s forecast peak demand plus its target reserve margin, leaving the approximately 6-gigawatt gap. The precise breakdown of that shortfall, including how much reflects generators choosing not to offer capacity versus a genuine absence of physical resources, has not been published in enough detail for independent analysts to fully verify. PJM’s planning documents and FERC filings provide the framework, but the granular offer-stack data remains limited in public disclosures as of June 2026.
Which states face the most exposure
PJM’s footprint spans all or part of Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia, and the District of Columbia. Not every corner of that territory faces equal risk.
Northern Virginia and the broader mid-Atlantic region carry outsized exposure because of data center load growth and the retirement of older generation nearby. Parts of Ohio and western Pennsylvania face pressure from coal plant closures that have removed thousands of megawatts of capacity over the past decade. Southern New Jersey and the Delmarva Peninsula, served by smaller utilities with fewer local generation options, can be vulnerable when transmission constraints limit power imports from elsewhere in PJM.
State utility commissions in several PJM states have begun scrutinizing resource adequacy more aggressively. Pennsylvania’s Public Utility Commission and Virginia’s State Corporation Commission have both examined whether their jurisdictions have sufficient generation commitments, though neither has issued emergency orders related to the 2027 gap as of this writing.
What PJM can do before 2027
PJM is not without tools, but each comes with trade-offs.
Reliability-must-run (RMR) contracts allow PJM to pay generators that would otherwise retire to stay online through a critical period. RMR agreements keep megawatts on the system, but they can be expensive and may prop up aging, high-emission plants that states want to phase out.
Emergency demand-response programs compensate large commercial and industrial customers for cutting electricity use during peak hours. The U.S. Energy Information Administration publishes data on demand-response capacity showing that utility-level programs can collectively reduce peak load by meaningful amounts, though the scale varies widely by service territory. Expanding these programs could shave the top off demand during the most dangerous hours, but they depend on voluntary participation and cannot substitute for firm generation around the clock.
Expedited interconnection for projects already near the front of the queue could bring some new solar, wind, and battery capacity online before summer 2027. PJM’s queue reforms, adopted in recent years, aim to prioritize projects with site control, financing, and completed studies. Whether those reforms move fast enough to matter for the 2027 delivery year remains an open question.
Imports from neighboring grids offer a theoretical backstop. PJM is interconnected with the Midcontinent Independent System Operator (MISO), the New York Independent System Operator (NYISO), and other regional operators. During localized stress events, power can flow across these seams. But during a broad heat dome covering the eastern half of the country, neighboring regions face their own demand spikes and may have little surplus to share. No public planning document quantifies how much dependable import capacity PJM can count on during a simultaneous, multi-region heat emergency.
What a shortfall actually means for households
A capacity shortfall does not guarantee blackouts. It means the statistical cushion PJM maintains to absorb unexpected plant outages, fuel disruptions, and demand surges is thinner than its reliability standard requires. In a mild summer, the gap might never show up in real-time operations. In a severe one, it could force PJM to escalate through a series of emergency actions: issuing conservation appeals, activating demand-response contracts, calling on emergency generation, requesting imports, and, as a last resort, ordering utilities to implement controlled, rolling outages to prevent a cascading grid collapse.
Rolling outages are disruptive but deliberate. Utilities rotate power cuts across neighborhoods in short intervals, typically one to two hours, to spread the burden and protect critical infrastructure like hospitals and water treatment plants. They are a tool of last resort, and PJM has rarely needed them. But the 2027 shortfall raises the probability that such measures could be necessary during an extreme weather event.
The natural gas supply picture adds another variable. Gas-fired plants make up a large share of PJM’s generation fleet, and fuel availability during a prolonged heat wave or a winter cold snap can tighten quickly. If gas storage is low or pipeline constraints emerge at the same time electricity demand peaks, some gas plants may not be able to run at full output, effectively shrinking available capacity further.
What residents and businesses can do now
For people living in PJM’s 13-state territory, the most practical step is preparation, not panic. Utilities across the footprint offer demand-response programs that pay enrolled customers, typically businesses but increasingly residential participants, to reduce electricity use during grid emergencies. Checking with your utility about enrollment is worth doing before summer 2027 arrives.
Large commercial and industrial operations should review backup generation plans, confirm fuel supplies for on-site generators, and test load-curtailment procedures now rather than during a heat emergency. Communication protocols with local utilities and PJM’s emergency notification system should be current.
Residential customers can take smaller steps that add up across millions of households: improving insulation, shifting dishwasher and laundry cycles away from late-afternoon peak hours, and understanding how their utility communicates outage warnings. None of these actions alone closes a 6-gigawatt gap. But collective demand reduction during the handful of critical hours each year when the grid is most stressed can meaningfully lower the risk of forced outages.
PJM has not yet released a detailed mitigation plan for the 2027 shortfall, and FERC has not issued public orders specifically addressing it. Until those plans materialize, the available evidence points to a grid that is tighter than its operators want it to be, heading into a period when demand is growing faster than supply. The question is no longer whether the gap exists. It is whether the institutions responsible for grid reliability can narrow it before the weather tests their work.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
