NERC’s latest reliability assessment is unreliable

From CFACT

By David Wojick

The North American Energy Reliability Corporation (NERC, rhymes with jerk) has just released its 2025 Summer Reliability Assessment (SRA). NERC is a quasi-federal agency under the Federal Energy Regulatory Commission (FERC, also rhymes with jerk).

NERC’s mission is to keep America’s grid reliable, which it has clearly failed to do. The nonsensical language in the SRA helps explain this failure. There is a deep fallacy that lets NERC systematically avoid saying how bad things really are.

This fallacy is the dangerously misleading use of the word “normal.” Here is a good example (out of many). Their basic finding is that “all areas are assessed as having adequate anticipated resources for normal summer peak load conditions.”

This sounds very reassuring, as does the whole report. The fallacy is that there is no such thing as “normal” summer conditions for a given day, week, month, or season. They really mean average conditions, and these are rare, not normal.

Here is an analogy to make the point. Suppose I work downtown and eat lunch at a lot of different places. Maybe twice a month I eat at Arby’s. It would be wildly false to say I normally eat at Arby’s since I eat there less than 10% of the time. “Normally” implies most of the time.

In the same way, average weather occurs less than 10% of the time, so it is wildly false to refer to it as normal. Moreover, the weather will often be worse than average as far as stressing the grid goes, sometimes far worse.

So NERC should truthfully say something like this: “All areas are projected to have adequate resources for normal summer conditions, but it is highly likely that conditions will be worse, including far worse.”

This is not reassuring at all, as it clearly calls for caution, which is how bad things really are.

The report assesses projected generating capacity versus demand by dividing the country into 23 different assessment areas. This sounds good, but in every case they use average weather, making the whole show completely unrealistic.

Here is a simple example of what they are ignoring that I happened to live through: Winter Storm Elliot in 2022. I am in the territory of PJM, the biggest US regional transmission operator.

NERC’s Long-Term Reliability Assessment for that year said PJM had a whopping 30% reserve margin based on average weather. Elliot’s bitter cold wiped that margin out, and PJM barely escaped blackout. Nor was Elliot’s cold extreme. Pittsburgh’s low was -5F, while 30 years ago it was -22F.

Summer is just as prone to far-from-average events, and reliability is about these events, not average events. We have lots of weather data and know how to do probabilistic analysis. NERC should be telling us the likelihood of serious grid-stressing events in each of the 23 assessment areas. These would not be happy numbers, and the message would be far from reassuring.

In weather forecasts, a slight chance of severe weather can be the most important forecast. NERC actually waves their hand at such by generically listing some of the things that can go wrong. For example, this: “…above-normal electricity demand, periods of low wind and solar output, and wide-area heat events that disrupt available transfers and generator availability could leave system operators short on supply in at-risk areas.”

Estimating probabilities for bad stuff like this in each assessment area could begin to give us a real reliability assessment. All we have now is page after page of what it will be like if everything is nice.

In addition, their inclusion of solar and wind to meet peak demand is ridiculous. Here is an upfront example under Key Findings: “Meanwhile, growth in solar photovoltaic (PV) and battery storage resources has accelerated with the addition of 30 GW of nameplate solar PV resources and 13 GW of new battery storage. The new solar and battery resource additions are expected to provide over 35 GW in summer on-peak capacity. New wind resources are expected to provide 5 GW on peak.”

Summer peaks are usually heat waves with peak need often between 4 and 6 pm when there is no solar output. Batteries might buy a few hours, but these heat waves are typically multi-day events making batteries worthless. They are also stagnant high-pressure, low-wind events with no wind output.

If NERC is assuming major solar and wind output to meet peak need, their findings are ridiculously unreliable. The combination of using wind and solar with merely average conditions makes this so-called reliability assessment highly misleading at best.

We need a realistic reliability assessment.