The grid needs downtime to expand—but downtime is becoming scarce
PJM’s autumn reliability squeeze exposed a buildout problem as much as an operating one. The same shoulder-season outages that take generators and transmission equipment temporarily off line are also when crews perform maintenance and construction. Those outages are not incidental interruptions to the plan: they create access to existing infrastructure for repairs and upgrades, including transmission projects intended to move power into areas facing generator retirements and growing demand. But holding equipment out of service reduces what the system can use right now. PJM’s account of the October episode describes a system in which both the immediate reserve margin and the time available to improve the grid are under pressure. The central constraint is therefore not simply how much new capacity is announced; it is how to keep the system reliable while making the physical work that may improve its future options. PJM
PJM describes the demand challenge as increasingly continuous. It says much of the rapidly growing load is from data centers and other large customers whose consumption does not vary substantially with weather, while that demand is outpacing new transmission and electricity supply coming onto the system. That framing matters for construction schedules: a mild day no longer necessarily means a low-demand day on which equipment can be taken out of service with ample operating room. PJM also points to unusually hot or cold conditions arriving early and persisting later, further shrinking the historical shoulder-season reprieve. These are PJM’s explanations of the changing operating environment, not an independently supplied forecast of future load or proof that every proposed large-load project will materialize. PJM
October’s squeeze: reliability actions displace planned work
The October episode shows the trade-off in operational terms. In its October 13 update, PJM said unusually warm fall weather coincided with planned equipment outages needed for maintenance and construction. It extended a Maximum Generation Alert through October 14, retained a Capacity Advisory for October 13–14 and issued a Maintenance Outage Recall for October 14. These were measures for a particular period, not evidence that the same alert or recall remains in effect now. The update said the alerts notified system owners that conditions would be tight and enabled them to consider deferring or canceling maintenance and testing, keeping equipment available to serve demand. This response can protect near-term reliability, but the mechanics are inherently consequential for work schedules: postponing an outage can postpone the work that required it. PJM
The scale of unavailable generation illustrates why that operational choice was difficult. PJM reported approximately 61,000 MW of generation outages on Tuesday, alongside key transmission outages for maintenance and major construction. Its October 15 account says more than two-thirds of the generation outages were planned maintenance or construction, while about 15,000 MW were unplanned, according to preliminary data. PJM said that roughly 61,000 MW total was the level reached after it worked with member companies to return about 10,000 MW to service, and warned that canceling more outages risked leaving facilities unable to prepare properly for winter. The figures describe the conditions reported during that week, not a persistent present-day outage total. They also show why a simple instruction to recall all equipment cannot be a durable construction strategy: doing so may increase available supply now while compromising necessary preparation for the next seasonal peak. PJM PJM
DOE’s emergency order added short-lived flexibility to the reliability response. DOE said the order was effective October 13 through October 16 and directed PJM to dispatch specified units as needed; it also authorized backup generation as a last resort before or during an Energy Emergency Alert 3. PJM’s contemporaneous update described temporary relief from environmental permit restrictions and authority to direct data centers and other large-load customers to switch to backup generation if necessary. Those provisions were contingent tools for the covered emergency period, not a permanent substitute for new supply or transmission. PJM said it had not used large-load backup generation to date, and its October 15 account likewise said it had not exercised that authority. Authorization should not be confused with dispatch, and the evidence does not establish that a data center was actually directed to run on backup during this episode. DOE PJM
The pressure on outage windows is not limited to generation. PJM said the region’s planned autumn maintenance season typically runs from mid-September through early December, allowing transmission and generation facilities to be taken out of service so they can be ready for winter peaks. It also reported recalling or canceling transmission and generation outages to the extent possible and recalling all active generation maintenance outage tickets in its footprint effective noon October 14. That specific recall is a historical action with a stated effective time; it does not establish what outages are scheduled or active today. Its significance is structural: maintenance work, transmission construction and system adequacy draw on overlapping windows. A recall can put assets back into service, but it cannot make the deferred repair or upgrade unnecessary. PJM
Construction depends on an operating margin, not just a project list
PJM’s account links the outage squeeze directly to transmission construction. It says outages will continue to be necessary in shoulder seasons to build critical transmission projects, including upgrades or replacements of existing lines intended to bring energy to areas where generators are retiring and serve load growing across the footprint. The grid operator’s point is not that construction itself is optional; rather, the work requires equipment outages at the very times when system stress may make those outages harder to accommodate. The operational consequence is a scheduling bottleneck: if demand remains high or conditions deteriorate, operators may have reason to preserve equipment in service, while postponing the work leaves the physical buildout unfinished. The evidence does not identify specific project dates or quantify how much construction has been delayed, so it supports the existence of competing demands on the window, not a claim about a particular project’s schedule. PJM
That bottleneck also makes broad statements about new supply easy to overread. PJM lists process reforms for generation interconnection, a process to match new large-load customers with supply, a reliability backstop procurement process that received conditional FERC approval, and an accelerated process for large state-sponsored projects. Those are actions and proposals at different stages, not evidence that the resulting generation is already connected or operating. PJM itself says necessary generation and transmission are not expected to be placed in service in the immediate future. This qualification is central to schedule analysis: streamlining an engineering process or opening a procurement pathway may improve the prospect of future supply, but neither instantly creates replacement capacity nor removes the need to schedule construction outages on an operating grid. PJM
The immediate implication is a period in which reliability procedures and construction planning must coexist, rather than a clean handoff from emergency operations to completed infrastructure. PJM says demand-response customers were activated on nine different days in 2026, including October 13 and 14, and that the combined total across 2025 and 2026 was 15 days, compared with eight days across 2015–2024. Those counts point to more frequent use of a demand-side operating tool in the periods PJM describes; they do not establish a specific future activation rate or prove that demand response can replace network investment. In October, PJM also described recalling non-firm sales and considering deferral or cancellation of maintenance and testing. Such measures can help manage a constrained period, but their role is to navigate conditions—not to deliver the transmission upgrades and generation the grid operator says are still needed. PJM PJM
The buildout test is whether work can proceed without consuming the reliability margin
For project owners, utilities and large-load customers, the practical schedule risk lies between a planned outage and an outage that an operator can safely permit. PJM’s October reports document one case in which unusual fall temperatures, high planned and unplanned outages, and transmission work coincided; the emergency actions were bounded to stated dates, and no supplied follow-up establishes their status beyond those dates. The reports do not prove that all future shoulder seasons will be similarly stressed, nor do they give project-by-project construction calendars. They do establish why schedules that assume a reliably quiet autumn deserve scrutiny: the traditional maintenance window is no longer assured to be operationally quiet, in PJM’s assessment. Any schedule implication should therefore remain conditional on system conditions, outage coordination, and the readiness of replacement supply—not be stated as a fixed delay or inevitable reliability failure. DOE PJM PJM
The larger thesis is that construction time has become a scarce grid resource. Higher, persistent large-load demand and weather variability can reduce the room to take equipment out of service; recalling outages can recover near-term operating capability but risks pushing necessary maintenance and construction; and emergency permissions can offer temporary options without adding durable infrastructure. PJM’s proposed and conditional remedies may improve the pipeline, but the source evidence does not show that they have yet translated into energized projects. The buildout challenge is therefore not simply to announce more generation and transmission. It is to create a sequence in which equipment can be maintained, lines upgraded, and new supply connected without relying indefinitely on emergency procedures to bridge the gap. That is a schedule constraint with direct reliability stakes, not a claim that the October episode left the region in continuing emergency conditions. PJM