Short Wave - These Drones Could Help Keep Your Lights On

Episode Date: November 12, 2024

One in four U.S. households experiences a power outage each year. Scientists at Oak Ridge National Laboratory are working on technology they hope will help fix electric grids: drones. They're betting that 2-ft. large drones connected to "smart" electric grids are a cost-effective step to a more electrified future.Listen to every episode of Short Wave sponsor-free and support our work at NPR by signing up for Short Wave+ at plus.npr.org/shortwave. Have an idea for a future episode? We'd love to know — email us at [e-mail]!See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

You're listening to Shortwave from NPR. Hey, short wavers. So a lot of you might know this, but if you're going to tackle climate change, we have to stop relying on fossil fuels. And if we're going to stop relying on fossil fuels, we're going to have to start using a lot more renewable energy. And in order to switch over to renewable energy, we're going to need to electrify everything. Which is great. Electrification can create a lot of efficiency and a lot of jobs. But as anyone who has experienced a blackout can tell you, the electrical grid is not perfect.

So a key part of a successful electrification will be solving the problem of power outages. Because get this, one in every four households experiences a power outage annually. That's why today I'm talking to Peter Fior. He's an electrical engineer at Oak Ridge National Laboratory. He lives in Fort White, Florida, and he's experienced his share of power outages, like Hurricane Helene back in September. Helene was a massive storm. And she came ashore about 40 miles away from where we live.

So at our place, our winds topped out at about 90 miles an hour, and we had trees that fell and all sorts of stuff like that. But why stop there? Because two weeks later, here comes Milton. Here it was piles and piles and piles of furniture and mattresses and blown out buildings and everything like that. This past hurricane season, Peter and his wife, lost power for days.

We're without power for quite a while. The internet crashed, so, oh, my goodness, communications was tough. And here's the thing. Every time a power outage happens, whether it's because of a fallen tree or a small animal or a severe weather event, somebody has to go and investigate. And a classic way that has been the norm, you roll a truck and you send some workers out there to see what the heck is going on. As told by being corroborated by others in the utility side, to roll a truck just to go see, was it a squirrel?

Was it a snake? Did somebody hit a pole? Was it a giant tree? $1,500. Minimum cost. That's just to investigate, not even to fix it. It's not just expensive. It can be dangerous. Line workers may have to survey miles and miles of line in search of the electrical problem,

or bushwhack their way through woods or swamp land to access it. When there's extreme weather or severe flooding, the risk of electrocution is extremely high. And that led us to this idea of, let's automate some of the inspections to help the line workers. Instead of having to send everything, let's send a drone and just have to do levels of inspection and report back to the system. And Peter says these drones, he and his team are testing out, they're mostly just standard drones. You know, the wireless charging quadcopter kind. And then if the system is smart, it may be able to. to resurrect the grid in a controlled manner.

On its own. You bet. So today on the show, the drones are coming. We take a peek at the drones. Oak Ridge National Lab scientists are betting on for the future of the electrical grid. I'm Regina Barber and you're listening to Shortwave, the science podcast from NPR. Okay, so Peter, what we're talking about is like drones that could help fix the electrical grid. And that's just like one part of what's called a smart electrical grid.

Can you say what you mean by that? like what tech is used to make a grid smart? So, okay, what's smart? What's new? Well, it's a combination of software and some instrumentation that's in place. I'm going to euphemistically refer to that as all the software that goes into running utility. I'll call it Skata. It could be outage management.

There are other things. I'm going to lump them together to Skata. What does Skata stand for? Sorry. Supervisory control and data acquisition. Got it. Yeah, it's got a couple pieces to it. So the data acquisition says, hey, what are you?

am I getting the data from sensors? So part of the smart grid is put sensors out in the field to see what's going on. And supervisory control is basically issuing commands across the grid, either to the substations or certain devices that are on poles to say open now, close there. They're automated reclosers and things called intel eructors, and that leads to the smart grid. And the main thing is, hey, I can put sensors out on the lines and have a better idea of what's happening. So the SCADA system is monitoring the system, the grid itself.

Okay, so this SCADA system is basically just like capable of checking in on itself, like determining what's going wrong or when something's going wrong, like if there's an outage or a power surge or anything like that. Yeah. Versus the not smart grid doesn't have not so smart grid. Yeah, the rudimentary grid, the remedial grid. I like the remedial grid. Okay, so okay, so you have like this smart grid. First, it's like, it's like, like, I think something's wrong. Let's go back to the first decision. How does it decide when and where to send a drone?

Again, in a regular, it's sort of mimicking what happens in a regular utility. So we have drones sitting on station at various locations. And the system says, hey, I got something going on over here. Maybe it's crackling, like a cracked insulator. And you have all the GPS locations of where the actual line is. So you can load that into the drone's location, this flight controller, and fly the drone there and do some inspections. Right. The drone uses GPS locations to go straight to the problem. What's the process of doing an assessment?

I think it's pretty cool, but come on. I'm invested in this thing. So what we have designed is the following, a two-stage process, like we're going into an ER. So as you go in, it's like, what do you see? Maybe a triage individual to help you right there. then if necessary, then you go see a specialist. We've done the same thing with the drones. Okay, so you have these triage drones, and then you have these specialist drones, and you send out the triage drone first.

Just to go figure out what's going on. We call that a scout. Got it. And the scout has a regular camera, an infrared camera, an ultraviolet camera, got an acoustic sensors because you can hear crackling and stuff like that. Oh, cool. And with that being the case, it flies to the location. and it looks at what's going on here.

Well, well, let's paint the picture a little bit more, though. You said that these drones are situated in, like, various places. So, like, where are they coming from? Substations. Have them live at a substation. Why? Because it's a jurisdiction of the actual utility. So typically they have a barbed wire fence and there's some security there.

But more importantly, the drones can get recharged. It lands onto a charging platform. And then it sits there. until it's told, hey, you know, why don't you go over here, go check stuff out? And you mean those substations that have like, you know, those platforms where like the, they're in between Paralyins, right? Yeah, the big electrical transformers are there. Okay.

A substation is going to take voltage at one level and then knock it down to voltage at another level for distribution into a community. So, again, let's walk through the scenario. The SCADA system back at headquarters for the control system, for the grid itself, says, hey, hey, hey, I got something going on over here, this particular section of our delivery service area. Send a drone. What are the GPS coordinates? Load them up. If the drone has enough charge in it, power, if the ambient conditions are okay, fly it over there. It checks out what's necessary using that base, that minimum configuration of sensors, reports back in and says,

you know what we really need? We need to dispatch a specialist, just like at the ER. I need to call in a specialist. And is it making that decision through like algorithms, like who or like what is analyzing the information coming in? Well, it's machine control. But with a human override. So, you know, a human can do that if they want.

But it's straight up mathematics, which is just glorious because you can analyze the signals that are coming in, the measurements that are coming in to determine should we send scout within a higher resolution infrared or ultraviolet camera or something else with these radio frequency detectors, et cetera. And then the specialist goes out and joins the party and then does measurements. And all of that gets bundled up. A decision is made through the beauty of math and some algorithms. And the machine out in the field can say, you know what?

It's all okay. You don't have to do a darn thing. Or it can recommend, yeah, you better send a truck on this one and go fix it. And then that's the purview of the utility then to actually do that. So, I mean, technically, they are, like, mathematically going through these if-then systems, and then they come up with a diagnosis, right? And you're saying that this is, like, all automated, right? These drones could potentially, like, save time and money, plus, like, prevent dangerous work, right?

Because I can have the drone go do an inspection, and I don't have to have the person climb the pole, because that can be kind of hazardous. Or alternatively, it's like, you know, Like, wait a second, if I'm out in those long lines out in Nebraska or Minnesota, northern the Minnesota, northernmost where I'm from, it's like I may have to have a truck, dispatch a truck with a one or two people in it and drive for 80 miles to go check out what's happening at this place where perhaps I could have used a drone. Yeah, perhaps it was a squirrel, like you said.

Yeah. And all we're talking about right now are just aerial devices, drones, airborne. airborne drones, but it could be in the water. There are certain cases where these drones with inspection can go underwater and look at the front facade of dams to see if there's cracking underwater of the dams themselves. Oh, wow. There's all sorts of capabilities that are brought where you don't have to put a human in there. This is not job replacement. This is line worker augmentation.

It really is. I mean, we're talking about the benefits of these drones of this smart grid system. But like, what are the limitations? Like, are there times and locations where relying on it, like, too much? Could be detrimental. From my perspective, with the grid. So what's smart about it again? It's got sensors and some cognition of what's happening in this deal.

And maybe it can actually respond. So notice what I've been talking about is sensing and measurements. And then the communication fabric to get the information to an org, either to an application or to a person so that they can make the decision about do we dispatch a truck? That's really about it. The major downfall with these drones, it's weight and power. So you've got to watch out there because if it gets too heavy, you're going to suck down the battery. But if it gets too light, the winds will blow it around.

So there's a limitation in ambient environmental conditions in which the drones can fly safely. You don't want to gust a wind to take the drone and push that into a transformer as you're measuring stuff on the transformer. Like you're the cause of the power outage. Yeah. Yeah. No, thanks. Right now, you said this is an experiment, right? You're working on this.

It sounds like utility companies are interested. When do you think or when do you anticipate this going to like wider, spreading wider across the U.S.? Again, national lab. So one of our core tenants is we do not compete with the private sector. Don't do that. But we can help design stuff and we can do a few demos and build a few products, not quite products, and get it to a specific level and then toss it over the fence and let the private sector pick that up. And our goal is to do that within one year.

So a year from, yeah, we're going to have, we are slated to have demonstrations at various locations across the U.S. this summer, spring, summer, into the fall. Wow. We have a working instantiation of all this stuff. And the goal then is to throw it to the private sector at the end of this fiscal year, federal fiscal year, and let it go. Wow. Yeah, it's aggressive, but the needs are there. And, hey, let's see if we can do it. The needs are there.

Well, Peter, good luck with all the demos, because that sounds amazing. Good luck with your one-year goal. That sounds intense. Thank you so much for talking to us. Oh, you're so kind. It's my pleasure. This episode was produced by Hannah Chin. It was edited by a showrunner Rebecca Ramirez.

The audio engineer was James Keely. Beth Donovan is our senior director and Colin Campbell is our senior vice president of podcasting strategy. I'm Regina Barber. Thank you for listening to Shortwave from NPR.