The Unlucky Windstorm

Show Notes

Listen to our conversation with Jay Schaefer about the Friday, March 13th "Unlucky" Windstorm in the Great Lakes area, and all of the factors that culminated in this "sleeper" storm.

What is Lights Out?

Lights Out, brought to you by PowerOutage.com, takes you through the latest in power grid intelligence, storm recaps, outage analysis, and industry insights backed by the most comprehensive real-time outage tracking platform in the U.S and beyond. Whether you're a utility professional, emergency manager, restoration crew, or just someone who wants to know what's happening with the grid, this is your place to dive in.

speaker-0: You're listening to Lights Out, a poweroutage.com podcast. We cover all things power outage related in 15-ish minutes an episode. Published whenever we have something to talk about. Hey everyone, this is Matt Hope again with PowerOuttage.com. And once again, I have Jay Schaefer, our chief innovation officer, resident meteorologist, and ⁓ expert in all things weather and grid impact. Today we're gonna be talking about another one of the storms that we've seen so far here in 2026 that we here at poweroutage.com categorize as a major storm or a major event. And again, we categorize anything over 500,000. meters out and a meter represents like an individual parcel or so the number of actual people impacted is gonna be significantly more. So today we're gonna talk about an event that occurred on March 13th in 2026, a windstorm in the Great Lakes area. And this storm was not named by the National Weather Service or the Weather Channel. We'll start off with the outage impact of this storm. So Jay, how many people were impacted by power loss during this storm, what were the worst hit areas and how long were people without power?

speaker-1: Yeah. The March thirteenth windstorm, which occurred on Friday the thirteenth, so I like to call it the unlucky windstorm, produced about one point three million customers without power at its peak.

speaker-0: So we're gonna name storms now. If if they don't name ⁓ we'll name ⁓ I like it.

speaker-1: It's it's good to name them just to have a name to remember it by. Yeah.

speaker-0: March thirteenth doesn't exactly have a ring to it. Yeah. So you said one point three seven without

speaker-1: Power. Yeah, little over one point three million customers at peak without power. And that was concentrated predominantly over parts of the Midwest and Great Lakes. Ohio had the greatest impacts. Michigan and Pennsylvania were the three hardest hit states. But this storm had a history of producing outages actually the day before, if you track it across the country, all the way from Montana and Idaho as it came across the country. So it produced outages there, but the impacts were greatest as it wound up in Ohio and ⁓ Pennsylvania, Michigan.

speaker-0: What was unique about this storm?

speaker-1: This storm in a lot of ways looked like a sleeper in that there were high wind warnings and high wind watches out for it, but it tended to overproduce the wind intensity than was predicted a day or two ahead for a lot of forecast scenarios. So one of the factors that produced that was the really dry air behind a storm. And it turns out drier air, it's easier to get winds aloft from the surface to the ground when the air's drier. You could make momentum or or mix air. And get it more efficiently to the ground. So that was one of the mechanisms there. And there was also evaporative cooling. So when rain or snow evaporates from a cloud, that can produce ⁓ some effects in the atmosphere that also enhance mixing. And it was a pretty strong storm system and it hadn't moved across the country. So it was a little bit of a Goldilocks porridge of a lot of factors coming together to produce really high winds. And then the time of the day was another factor in there. So in the afternoon. The winds in the lower atmosphere tend to be strongest because that's when the atmosphere is most coupled with the winds above the surface and at the surface from daytime heating effects. So the winds were the highest later in the day. That was another factor that came together in the the wrong way.

speaker-0: So you mentioned the day before in Idaho and Montana area, the system was causing outages. Did those areas see lower wind speeds than what was seen once it reached the Great Lakes area?

speaker-1: They were a little bit lower, but in some places the winds were the same. In a lot of locations, the winds just cause more impacts, also in places that have a little bit less resilience or more tree cover as well. So that was a factor. We didn't really study that super closely, but North Dakota, the winds were high. Minnesota, the winds were higher. Those places tend to be more flat, wide open, and the infrastructure there tends to be a little more resilient to higher winds, the same wind speed. So there definitely were some high winds with this that were indicators that this storm could deliver some higher impacts. Okay.

speaker-0: So as far as the forecast for this storm in the white paper that you published, there were some differences in what was actually experienced versus what was forecasted.

speaker-1: Yeah, so I think this storm, unlike winter storm fern, that ice storm, we had pretty good detection of a high impact event, like medium confidence, maybe one or two days ahead. So you didn't have a longer lead time with this. Some of the higher end forecast scenarios definitely captured it, but it was definitely on the higher end of a forecast distribution. The other thing was this was a day I was trying to take off of work and Mother Nature did not cooperate with me. I personally was impacted by this storm. And have, you know, some stories to recount from chasing the wind basically just right down my road here as it came through during that evening. Yeah.

speaker-0: The internal play-by-play of this event ⁓ from you. It was definitely interesting to see and it was really cool to get your ⁓ real-time experience of the event, but obviously with your background too, it was quite informative. So what I'm hearing then is it was on the high end of the highest forecasts and not a lot of lead time for the ⁓ utilities and emergency responders to really prepare for the event.

speaker-1: That's right. There wasn't as much lead time. Now, this being in March with the longer days and we didn't have really cold air mass behind it and there was no snowfall with it, that made restoration a little bit faster. Also, the damage to the grid didn't require full system rebuilds, unlike an ice storm from like winter storm fern. So power restoration in the worst hit areas were generally on the order of four to five days and compared to like winter storm fern, which was like two to three weeks. So

speaker-0: Then were the high winds the primary hazard for the grid in this storm or were any other factors that played into the damage that did occur?

speaker-1: This was pretty much a pure wind event. The dry air and being the springtime actually was an important factor as the air mass came across the country, it dried out, and then plants quite haven't come out yet and everything hasn't greened up. When that happens, you tend to get more water vapor in the atmosphere and storms tend to get a little more humid. So that time of the year factor was another factor in this. And another secondary risk factor with this was because it was windy and dry, was actually wildfire potential. And this type of storm doesn't tend to occur with winds this strong and this dry in this area of the world too often. So it had been pretty wet previously to this storm. So we didn't see any real notable examples of that. But this was definitely the kind of risk storm where you've got to keep your eyes out for if there's gonna be a wildfire might spread rapidly from the winds.

speaker-0: To dive a little bit deeper than we normally might on these storms because there's a portion of the white paper that you published that I found really interesting, not having a meteorological background. It was still incredibly intriguing to me and easy to understand. And that was around the wind intensity. And essentially, like as the wind speed reached certain thresholds, there's like significantly different impacts on the grid. And I Love to just spend a little bit of time talking about what you saw from this storm specifically and talk to that piece of the white paper a little bit.

speaker-1: Yeah, sure. So the intensity or the peak value of the winds is really important for how much damage you have. And if you go back to your high school physics class, kinetic energy is one half mass times the velocity squared. So kinetic energy is proportional to the square of the wind speed. So as you go from 50 to 60 miles an hour, it's not a linear response in how much more energy there is in the wind. So small changes in the wind speed affects how much kinetic energy there is. 50 to 60 miles an hour might almost be a doubling of the amount of energy, even though it's only 10 miles an hour. Right? Now, when you integrate that energy over time, that's called energy flux. And that's more dependent on the cube of the wind speed. So extremely high sensitivity, right when you start to get damage from forty to fifty to sixty miles an hour, five miles an hour can mean the difference of a half million customers or no customers being without power, just like five miles an hour. Wow. And in this case, it probably was that sensitive. We we showed in the paper how how sensitive that is. So if you think about predictability, you gotta get that pretty dialed in too. So that was another point we made is like if you're five miles an hour off on the forecast. Well, your impacts are gonna be significantly different. Yeah.

speaker-0: And I'm actually I'm gonna read off some of those numbers because the exponential increase is so significant. And I this is the part that I found really interesting. So in that forty one to fifty mile an hour range, the median percentage of customers out is only one point one percent. Go up to the next ten miles per hour. So fifty one to sixty essentially over triples to three point eight. And then crossing that sixty-one mile per hour threshold, we get up to seventy miles per hour is thirteen point seven percent for the median percent of outages. So I mean the increases there are so extreme. I really loved this part of the year analysis and it was incredibly informative. So anything else that you think we should address on this specific event?

speaker-1: There's a lot you could learn I think from from this event in terms of the winds and and thinking about the risk of winds on the power grid for utility companies, for emergency management, and you know, forecasters in the National Weather Service or otherwise. So wind intensity is important, but the duration of the wind and the coverage of it over an area are are all important factors. And we don't really normalize and think about things that way. So one of the things we're doing in the the the research lab is coming up with some sort of index value to help utilities or ⁓ any location understand on some sort of zero to a hundred scale like how severe this wind event might be to help understand impacts a little better. So this one kind of caught us a little bit with our proverbial pants down, but a lot you can learn from it. We were able to pick up and get the power back on relatively quickly within less than a week in most areas.

speaker-0: So if you're interested in the white paper, you can see that at powerutage.us slash research. And I would say even if you don't have an interest in reading the entire paper, there's a really cool figure, figure one in the paper. And it allows you the ability to essentially slide back and forth and see a comparison of the peak percentage of customers out and the peak wind gusts over a county display for the states that were impacted. And it's really interesting to see the overlap between those two.

speaker-1: Yeah. And one other fact about this windstorm is if we looked at ten years of outage data for this event and compared it to other events, at least in Ohio, this storm event produced more outages than any other weather event going back to at least two thousand seventeen. And, you know, we don't name storms like this conventionally, but this is a good example of a type of weather event that might not be as well understood and wind causes more power outages, but there's a lot more we could do to really improve prediction for wind and impacts like this.

speaker-0: Awesome. Jay, thank you as always. I really appreciate you deep diving this with us and looking forward to the next one. Thank you for listening to another episode of Lights Out. This series is brought to you by powertage.com. Our producer is Ryan Loyacano. Our editor and music composer is Ian Richter. I'm your host, Matt Hope, and I'll see you next time.