DOE Issues 7th Emergency Grid Order to PJM in 2026

The Department of Energy has handed PJM Interconnection the legal power to commandeer backup generators across the Mid-Atlantic, and it did so on 17 September 2026 with a single signature. Order 202-26-45, issued under Section 202(c) of the Federal Power Act, gives the grid operator authority to activate standby generation across its territory for a two-day window closing at the end of 18 September 2026, a measure designed to hold off uncontrolled outages during a late-summer reliability crunch.

This is not a one-off. It is the seventh Section 202(c) emergency order DOE has issued to PJM in the 2026 calendar year, each triggered by the same underlying strain: demand outrunning available supply as conventional plants retire, seasonal generation fades, and new large loads keep arriving on the system.

What follows here is the concrete detail behind the headline. You will see exactly what legal authority DOE just invoked, why more than 35 gigawatts of backup power sitting on U.S. soil normally cannot run without federal permission, and what Energy Secretary Chris Wright’s public framing of the crisis signals for the wider argument over baseload generation and who pays to fix the gap.

What DOE’s September emergency order actually authorises

At the moment Order 202-26-45 took effect, PJM gained three specific powers it does not hold under normal market rules. It can dispatch what the order calls “Specified Resources” as needed to keep the system stable. It can activate backup generation at large industrial and commercial loads as a measure of last resort. And it can direct its Transmission Owners and Electric Distribution Companies to help carry out those instructions.

PJM covers a large footprint: 13 Eastern states and Washington, D.C. The order applies across that entire region for the two-day window.

Here is the crux that makes the order’s logic click. The backup-generation authority can only be exercised immediately before or during an Energy Emergency Alert Level 3, PJM’s highest emergency classification, the stage that sits just ahead of uncontrolled outages. This is not a new everyday tool. It is a safety valve that can only be pulled when the grid is already at the edge.

The core provisions of Order 202-26-45:

• Authority to dispatch Specified Resources as reliability requires
• Authority to activate backup generation at large loads as a last resort, tied to the EEA-3 threshold
• Authority to engage Transmission Owners and Electric Distribution Companies to execute the directive
• Effective on issuance, expiring at the end of 18 September 2026

PJM submitted its formal application on 16 September 2026, and DOE responded the next day. For investors tracking grid reliability policy, the two-day window and the EEA-3 activation trigger are the details that matter, because they define which assets can be monetised under emergency conditions and which stay locked out. This order sits in a tight sequence with its immediate predecessors.

DOE’s Federal Power Act emergency authority under Section 202(c) allows the agency to order temporary grid interventions when it determines that an emergency exists affecting the reliability of electric service, with each order required to identify specific statutory findings before taking effect.

Why 35 gigawatts of backup power cannot be used without a federal order

Start with the puzzle. The United States has, on DOE’s own estimate, more than 35 gigawatts of untapped backup generation capacity accessible nationwide. Yet during an emergency, most of it sits idle. The reason is not that the hardware fails to work. The reason is legal architecture.

The structural conditions behind every 2026 emergency order trace back to the same underlying grid capacity limitations: retirements outrunning new dispatchable supply while demand from electrification and large industrial loads accelerates faster than capacity markets anticipated.

Consider first how PJM’s market is built. The standby and behind-the-meter generators at data centres, industrial plants, and large commercial sites are not enrolled in PJM’s central dispatch stack. Because they are not registered as standard generation or capacity resources, PJM has no routine authority to order them to run.

Then layer in environmental law. Many of these units operate under state and federal air permits that limit how often they can fire. That is precisely why Section 202(c) orders bundle two things together: authority to dispatch the units and a temporary waiver of the pollution limits that would otherwise keep them off. The 30 June 2026 orders made this explicit, authorising PJM to curtail data centres and waive plant emissions limits in the same package during a record heat wave.

The resources in question span a wide range. PJM’s 27 June 2026 application described exactly what it was seeking.

PJM sought authority to direct backup generation resources at large loads, including auxiliary, standby, directly connected, battery storage or other types, whether synchronised or not to the bulk power system, to operate as a last resort before declaring an Energy Emergency Alert Level 3.

Four distinct barriers keep this capacity out of reach under normal rules:

• Market enrollment: the units are not registered in PJM’s central dispatch fleet
• Environmental permitting: state and federal air permits cap routine operation
• Legal dispatch authority: PJM lacks power to compel operation without a 202(c) order
• Operational hierarchy: backup use is gated to the EEA-3 last-resort threshold

Here is what the 35 gigawatt figure actually reframes for you as an investor. The constraint on this grid is not a shortage of physical generation hardware. It is a regulatory and legal structure that keeps substantial capacity off the table until a federal emergency unlocks it. That distinction matters if you are evaluating backup generation assets, demand response programmes, or the large-load operators, including data centre owners, whose generators are now formally part of the emergency toolkit.

Secretary Wright’s diagnosis and the case against it

Energy Secretary Chris Wright offered a clear two-part explanation for the late-summer strain in his statement of 17 September 2026. He pointed to two named causes, and both are worth taking on their own terms.

Secretary Wright attributed heightened late-summer grid stress to reduced solar generation as daylight hours shorten into autumn, and to the premature retirement of conventional power sources driven by prior energy policies.

The solar mechanism is real and measurable. Solar output in PJM falls as day length shortens through late summer and early autumn, while cooling demand can stay elevated during heat waves. That creates a peak-hour mismatch that dispatchable thermal units, imports, or storage must fill. The North American Electric Reliability Corporation has previously warned that late-summer temperature surges threaten stability when dependable power sources are retired ahead of schedule.

A competing diagnosis reads the same events differently. The Center for Progressive Reform, in analysis published on 30 June 2026, argued that the repeated emergency orders reflect under-investment in clean firm resources, transmission, and demand response, not a flaw inherent to renewable energy. In the same post, the group warned that PJM’s emergency request “will increase air pollution,” because leaning on older fossil units and backup generators works against air-quality and climate goals.

There is a third framing worth noting. Reuters’ coverage of the 30 June 2026 emergency emphasised record heat and rapid data-centre load growth as the primary operational drivers, pointing to climate-driven demand and new load types rather than energy policy choices.

A federal court ruling issued the same week as Order 202-26-45 directly complicated the administration’s reliability posture, striking down DOE’s plant retirement authority and narrowing the legal toolkit available for managing conventional capacity exits from the PJM system.

For energy investors, this disagreement is not academic. Which diagnosis wins determines where a decade of grid capital flows: toward backup thermal capacity and plant-life extensions, toward storage and demand response, or toward transmission and clean firm capacity. Knowing the specific terms of the argument is what lets you read the next emergency order with full context rather than as an isolated headline.

A pattern of emergency orders, not a single event

The number that should anchor your read of 17 September is not the order itself. It is seven. DOE has issued seven Section 202(c) orders to a single grid operator in one calendar year, and each one tells a similar story about a system under sustained strain.

The 2026 sequence for PJM runs as follows:

1. Order 202-26-06, effective 26 January 2026: found a statutory emergency from sudden demand increases and generation shortages; authorised backup generation at large loads before EEA-3
2. Order 202-26-17, early 2026: determined an emergency existed and required specific generation and transmission arrangements
3. The heat-wave orders, effective 30 June to 3 July 2026: authorised PJM to curtail data centres and waive pollution limits during record heat
4. Order 202-26-35, effective 14 July to 21 July 2026: directed dispatch of Specified Resources and backup generation before or during EEA-3
5. Order 202-26-41, effective 1 September 2026: authorised dispatch of specified units and direction of backup generation at large loads
6. Order 202-26-45, effective 17 September to 18 September 2026: the current directive

Each order finds either a sudden demand increase, a shortage of electric energy, or a shortage of generation facilities, and each ties backup-generation authority to the EEA-3 threshold. The precedent runs back further still. In May 2026, a Section 202(c) order directed Constellation Energy to keep 760 MW of oil- and gas-fired peaking capacity at the Eddystone Generation Station in Pennsylvania running rather than deactivate it on schedule. Every order carries a sunset date tied to a specific weather or reliability event, which underlines that 202(c) is built for extraordinary intervention rather than systemic repair.

The Centralia situation illustrates how repeated emergency orders for single assets can accumulate without resolving the underlying reliability problem, a dynamic that mirrors the escalating frequency of PJM orders across the 2026 calendar year.

What the escalating frequency signals for grid reliability investment

Legal and trade analysis from the law firm DWGP has noted that the volume of emergency orders since May 2025 points to a clear conclusion: resource planning, capacity markets, and transmission investment have not kept pace with evolving demand and policy goals. That holds regardless of which side of the root-cause debate you favour.

The orders themselves do not close the gap they respond to. Each is time-limited and expires with the weather event that triggered it, leaving the underlying capacity shortfall in place.

For investors tracking energy infrastructure, sustained emergency-order frequency is itself a signal. It points to durable demand for backup generation assets, peaking capacity, and demand response programmes, and it shows which operators are absorbing the reliability risk through the back door of federal emergency law.

What the Mid-Atlantic directive changes, and what it leaves unresolved

For the next 48 hours, Order 202-26-45 does something concrete. It hands PJM a legal safety valve to activate backup generation and dispatch specified resources before blackouts begin, reflecting the administration’s stated commitment to using every available tool to keep the lights on.

The administration expressed commitment to utilising all available tools to prevent electricity outages, per Secretary Wright’s statement of 17 September 2026.

The order’s expiry at the end of 18 September 2026 is the detail that makes its limits concrete. What it resolves and what it defers fall into two clear categories:

• What this order resolves (short-term): a legal mechanism to keep PJM’s Mid-Atlantic system stable through the two-day reliability window
• What this order defers (structural): it does not permanently change market rules, does not close the underlying capacity gap, and does not settle whether plant retirements or renewable intermittency is the primary driver of grid stress

The forward variables worth watching are specific. Track the pace of conventional plant retirements in PJM, the trajectory of data-centre load growth, the outcome of ongoing capacity-market reforms, and whether Congress or FERC moves to formalise backup generation within the standard resource stack. PJM’s applications name data centres directly as the large-load category being mobilised, and the 30 June 2026 orders explicitly authorised curtailing them, which tells you how central hyperscale computing has become to the reliability equation.

Data centre load growth has moved from a long-run planning assumption to an immediate operational variable: PJM’s own emergency applications name hyperscale computing sites as the large-load category whose backup generators are now formally part of the emergency dispatch toolkit.

If you can hold the boundary between what an emergency order accomplishes and what it postpones, you are better placed to weigh grid reliability assets, capacity-market reform proposals, and the policy risk embedded in large-load infrastructure plays across PJM’s footprint.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. These statements are speculative and subject to change based on market developments and policy decisions.