By Jason Charles Guck, Business Development, Delta Edge CI
October is supposed to be the easy month. Cooling season is over, heating season has not started, and the grid uses the gap to take plants offline for maintenance it cannot do in July. That is the whole design of an autumn shoulder season.
On October 12, the Secretary of Energy signed an emergency order for the PJM region. It took effect on October 13 and runs through October 16.
Not in a heat wave. In the middle of the quietest stretch on the calendar.
In Bring Your Own Power I wrote that states had stopped courting large loads and started asking them to pay their own way. This week supplies the other half of that argument, and it comes from the federal side rather than the states. The question is no longer only who pays for new capacity. It is whether the capacity that already exists can be counted on in a month nobody worried about.
What the order actually says
Order No. 202-26-50 was issued under section 202(c) of the Federal Power Act, the provision that lets the Secretary direct generation during an emergency. PJM submitted its application on October 11 and the order followed the next day.
The determination is written plainly. The Secretary finds that a statutory emergency exists in the PJM region due to a sudden increase in the demand for electric energy, or a shortage of electric energy or of facilities for the generation or transmission of electric energy, and other causes.
Two things are authorised. PJM may dispatch specified units and order their operation as needed to maintain reliability. PJM may also direct backup generation resources to run as a last resort before declaring an Energy Emergency Alert 3, or during one. Transmission owners and electric distribution companies can be called on to carry the order out.
The trigger figure is in PJM’s own application. PJM forecast demand at or around 111,500 MW or higher on October 13. Against that, PJM’s filing describes roughly 61,591 MW of generation sitting out, the large majority of it planned autumn maintenance rather than anything breaking.
Read those two numbers together, because the gap between them is the entire story. This was not a supply failure. It was a scheduling collision between ordinary maintenance and a demand level that autumn used to be allowed to ignore.
Why 111,500 MW in October is the number that matters
A grid operator plans the maintenance calendar around an assumption, and the assumption is that shoulder-season demand leaves enough headroom to take large amounts of generation offline at once. That assumption has held for decades. It is why the autumn outage season exists at all.
What has changed is the floor. PJM’s territory covers all or parts of Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia and the District of Columbia. Across that footprint, load that used to rise and fall with the weather now carries a large component that does not. Data centres run at the same rate in October as they do in July. So does cold storage, so does continuous manufacturing, and so does every building running a load profile that no longer has a genuine overnight valley.
When the floor rises and the maintenance calendar does not move, the shoulder season stops being slack. It becomes another period that has to be managed.
The administration moved on this quickly and correctly. Keeping the lights on across fourteen jurisdictions while a maintenance calendar and a demand forecast run into each other is exactly what section 202(c) is for, and using it before a problem becomes an outage is the right instinct. The harder question is what it means that the tool is now being reached for in the quiet month.
What this means if you buy electricity
An emergency order aimed at generators does not land on your bill this week. It changes what you should expect from the next several bills, and from next year’s budget.
Your capacity exposure is no longer a summer-only question
Capacity charges are set by measurement taken during peak conditions, and in PJM that measurement has always pointed at summer afternoons. The mechanics are unchanged for this year. What is changing is the assumption underneath them. If shoulder-season conditions can produce an emergency order, the argument for treating capacity as a seasonal problem gets weaker every year. I covered how the measurement itself works in the piece on capacity tags, and the practical advice there does not change. The window for acting on it is getting wider.
Demand response just got more valuable, and more real
An order that authorises backup generation as a last resort before an Energy Emergency Alert 3 is an order that would prefer not to get there. Load reduction is the step that keeps a system away from that point, and it is the step customers control.
For operators already enrolled in a capacity or reliability programme, this is the year to confirm you can actually deliver what you committed to, because being called is no longer a summer-only prospect. For operators who have looked at demand response and passed, the economics are worth a second read. I set out how the programmes differ, and when enrolling is the wrong answer, in Demand Response 101.
Behind-the-meter generation changes category
This is the direct line back to Bring Your Own Power. A standby generator bought purely as outage insurance has been a cost line with no return. An order that contemplates directing backup resources to run is a signal that the regulatory treatment of that equipment is moving.
That does not make every generator an asset. Permits, fuel, run hours, emissions limits and interconnection terms all decide whether yours can participate in anything, and most of those answers are site-specific. It does mean the question is worth asking now rather than after the next order.
Budget the delivery side, not the commodity
Nothing in this week changes the price of natural gas. It points at the cost of keeping enough firm capacity available, and that cost arrives through capacity, delivery and reliability charges rather than through the energy line. An escalator anchored to a commodity view will keep pointing the wrong way, which is the case I made at more length in the 2027 budgeting piece.
What to do this week
- Pull your interval data for October 13 to 16 and look at what your own load did during the order window. If your facility ran flat through it, you have a demand response conversation worth having.
- Find your current capacity obligation and the notification window attached to it. Confirm somebody on site knows what a call requires and can execute it.
- List your behind-the-meter generation with its permit conditions and allowable run hours. That list is the precondition for every option that follows.
- Check whether your 2027 budget treats capacity and delivery as separate lines from commodity. If it does not, it will absorb this as a surprise.
The bottom line
A four-day emergency order in October is not a crisis. The system worked, the tool existed, and it was used early rather than late.
What it is, is information. The shoulder season has stopped being the part of the year when nothing happens, and the reason is a demand floor that no longer falls when the weather turns mild. Every operator whose planning still assumes a quiet October is planning against a grid that no longer exists.
The states asked large users to bring their own power. The federal order this week asked existing generation to show up out of season. Both point in the same direction, and it is the direction where the customers who control their own load do better than the customers who do not.
This is a fast-moving regulatory picture and the order described here covers October 13 to 16, 2026. Confirm anything that affects a budget against the current docket before building a number on it. If you want a clear-eyed review of your capacity exposure, demand response options and behind-the-meter position, that is work we do at Delta Edge CI.