Electricity customers across 13 eastern states and the District of Columbia face a tightening power supply this summer as PJM Interconnection, the nation’s largest grid operator, enters the peak season with reserve margins that barely clear minimum reliability thresholds. Federal regulators have opened formal proceedings targeting the speed at which data centers are plugging into the grid, a pace that outstrips the construction of new generating capacity. The collision between surging demand and slow infrastructure buildout has moved from industry conference panels to active federal dockets, with cost allocation and grid stability now at stake for roughly 65 million people served by PJM.
What is verified so far
The Federal Energy Regulatory Commission has taken two distinct, documented steps that frame the summer 2026 risk. First, FERC released its 2026 Summer Energy Market and Electric Reliability Assessment, a report that draws on NERC’s 2026 Summer Reliability Assessment to compare projected reserve margins against reference margin levels across U.S. grid regions. The assessment examines whether available generation can meet forecasted peak demand with an adequate safety buffer. For PJM, the comparison shows margins sitting close to the reference floor, meaning the region has little room for error if extreme heat drives air-conditioning loads higher than expected or if unplanned plant outages remove supply during critical hours.
Second, FERC ordered action on co-location issues tied to data centers running artificial intelligence workloads, instituting a Section 206 proceeding under dockets EL25-49-000 and ER24-2888-001 et al. That proceeding treats large-load and data-center interconnection, along with generator-load co-location inside PJM, as a formal reliability and cost allocation problem. In practical terms, the commission is asking whether the current tariff rules fairly distribute the expense of keeping the grid reliable when a single customer can add hundreds of megawatts of demand in a matter of months.
The co-location question is specific and technical but carries broad consequences. Some data-center operators have sought to pair their facilities directly with on-site or adjacent generators, effectively operating behind the meter. That arrangement can reduce the load visible to the grid operator while still relying on the transmission system as a backup. FERC’s proceeding signals that the commission views this practice as raising unresolved questions about who pays for the reliability upgrades the rest of the system needs when those backup calls come.
The Section 206 order also underscores that FERC is no longer treating data-center demand as a marginal or niche issue. By grouping large-load interconnections and generator-load co-location into a single, focused case, the commission is effectively telling PJM and its stakeholders that the existing toolkit for handling incremental industrial load may not be sufficient when that load arrives in gigawatt-scale clusters. The docket’s framing places data centers alongside traditional reliability concerns such as generator retirements and transmission congestion, elevating them from a planning footnote to a core stress factor on the grid.
What remains uncertain
Several pieces of the puzzle are not yet public in the cited federal materials. PJM-specific reserve margin tables for summer 2026, including forced-outage rate assumptions and the exact gap between projected and reference margins, have not been released in granular form through the FERC documents available. The summer assessment discusses reserve margin and reference margin comparisons at a regional level, but the underlying scenario modeling and stress-test parameters remain within NERC’s full report rather than the FERC presentation summary.
Equally absent from the primary dockets is direct testimony or data from transmission owners detailing the volume of queued data-center megawatts versus completed interconnection studies. Industry reporting has described a surge in PJM’s interconnection queue, with data-center applications accounting for a growing share, but the exact figures and timelines are not confirmed in the FERC filings referenced here. Without those numbers, the scale of the mismatch between connection requests and grid upgrades cannot be stated with precision.
Utility capital-spending timelines and permitting delay statistics for new generation inside PJM are also missing from the primary sources. While the general dynamic is clear, with new gas plants, battery storage, and renewable projects taking years to permit and build, the specific bottleneck data that would quantify how far behind supply additions are trailing demand growth has not appeared in the co-location docket or the summer assessment. Any projection about how quickly the gap could close depends on assumptions that remain outside the verified record.
The outcome of the Section 206 proceeding itself is uncertain. FERC has opened the case and identified the issues, but it has not yet imposed new rules or cost-allocation formulas. The proceeding could result in tariff changes that shift significant costs onto data-center operators, or it could produce narrower adjustments that leave the current framework largely intact. Stakeholder comments and technical conferences will shape the result, and no timeline for a final order has been confirmed in the available materials.
There is also uncertainty around behavioral responses. If FERC ultimately assigns a larger share of reliability costs to large loads, some data-center developers may alter siting decisions, seek more self-supply, or delay projects. Those reactions could, in turn, change PJM’s demand trajectory relative to what is implied in current planning documents. None of these potential shifts are modeled in the federal materials, leaving a wide band of possible futures for both peak demand and reserve margins.
How to read the evidence
The two FERC documents carry different evidentiary weight. The Section 206 order is a primary legal instrument. It establishes that the commission has found sufficient grounds to believe PJM’s existing tariff may be unjust or unreasonable as applied to large-load co-location. That finding is not a conclusion but a formal threshold determination that triggers a proceeding with enforceable deadlines and potential remedies. Readers should treat it as a confirmed federal action, not a policy suggestion or staff recommendation.
The summer reliability assessment is an analytical product. It synthesizes data from NERC and market participants to describe expected conditions, not to mandate outcomes. Its value lies in establishing the official federal view of where margins stand heading into peak season. When the report says projected reserves are only modestly above reference levels for a region, it is signaling that the system can meet typical peak loads but may be stressed under more extreme scenarios. That framing should be read as an early-warning indicator rather than a guarantee of shortages.
Taken together, the documents tell a consistent story even if they operate on different planes. The assessment highlights that PJM is running close to its reliability floor at the same time that demand from data centers and other large users is accelerating. The Section 206 order, meanwhile, shows that FERC is not comfortable leaving cost responsibility and interconnection rules unchanged in the face of that trend. The combination points toward a period of regulatory and planning adjustment, with reliability concerns providing much of the impetus.
Readers should also note what the materials do not claim. They do not assert that rolling blackouts are inevitable in PJM this summer, nor do they provide a project-by-project breakdown of which plants or lines will make the difference during heat waves. Instead, they frame the problem at a system level: reserve margins are thin, demand growth is concentrated and lumpy, and existing rules may not allocate the costs of reliability upgrades in a way that FERC considers sustainable.
For policymakers, the evidence supports a cautious but active stance. The legal proceeding offers a venue to recalibrate tariffs and interconnection rules so that rapid load additions do not erode reliability for existing customers without compensation. The analytical assessment, meanwhile, underscores the need for contingency planning, demand-response readiness, and clear communication with consumers if extreme weather pushes the grid toward its limits.
For customers in PJM’s footprint, the key takeaway is that federal regulators see the region as entering a more constrained era. The grid is still expected to function, but the margin for error is slimmer, and the rules governing who pays to widen that margin are in flux. How those questions are resolved over the next several years will influence not only electricity bills but also where and how fast the next wave of data centers, industrial facilities, and clean-energy projects can connect to the system.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
