What causes power outages

January 13, 2024 1:32 PM

My power is out, so I've got time. What are common causes of power outages (besides wires coming down due to winds, trees, ice which I assume is probably the leading cause)?

I'm curious to hear about literal and specific things that commonly break I.e. 'a transformer might fail due to a load spike', not general causes i.e. 'too much demand', 'cold weather'. Just idle curiosity at what might break and what it takes to fix it.

posted by so fucking future to Technology (26 answers total)

The last time I had an outage in my neighborhood, a truck had run into a transformer. It made a very loud BANG!
posted by SPrintF at 1:50 PM on January 13

Transformers can also blow on their own. I have no idea why they do, but the one that fed my neighborhood when I was a kid would blow every few years.
posted by jonathanhughes at 1:58 PM on January 13

The things I can think of that have knocked out my power over the years, in what I think is fairly accurate order of frequency:
- Vehicles running into equipment
- Lines down due to wind, snow, ice, trees, etc
- Random transformer blowing
- Landslide
- Wildfire
- Animals interfering with equipment
- People deliberately damaging equipment

And here's a one-off, odd one for you, just for amusement: A small(er) town I lived in had to shut off the power because the river got too high and was threatening to flood a certain building. They've since re-routed stuff. That same weather event, which was flooding during melt-off after very heavy snow and rain, also closed three highways with landslides, some of which took weeks to clear - we had no grocery restocking or mail for a week, it required driving something like a ten-hour route from the major metro area instead of one, once they got the easy-to-clear side open. I was in the middle of a job interview when they shut it off!
posted by stormyteal at 2:00 PM on January 13

Yes, random transformers blowing up is a thing. They are terrifying, and spectacular.

But all those things stormyteal mentioned do happen. Lines down in any place that gets lots of snow, wind, and/or freezing rain.
posted by Windopaene at 2:03 PM on January 13

Sometimes the system just kind of shits itself - this was fun.
posted by LionIndex at 2:09 PM on January 13

Squirrel!
posted by scolbath at 2:22 PM on January 13 [7 favorites]

Yup, US infrastructure is kinda scary.

Overall, American infrastructure earned a C- in 2021, the bottom end of average.... Up from a D, but that ain't saying much.

Stockholders wanting to cut repair and maintenance personnel is another one. No cites here except from my newly retired Idaho Power neighbor, who cusses that when one region goes down--from weather or fire, they need to call in personnel from other regions, and if it's more than statewide, it can get ugly, because there just aren't enough people to get things up in a timely fashion. Plus, higher drain on outlying systems can cause a cascading effect. It will be interesting to see what effect electric power demand by growing EV use will have on spending to improve infrastructure and outages.
posted by BlueHorse at 2:31 PM on January 13

Best answer: The Cyber Resilient Energy Delivery Consortium (CREDC) is an industry PR operation from a local uni but layout the basic energy system failure points in nice interactive widgets: Wind and storage and a mini grid. I think you are looking for something more specific than these models, but it is helpful to understand that there is a significant level of redundancy built in and that most failures do not directly result in your service getting interrupted.

In terms of knocking out your point of service - what part fails is primarily a factor of the particular regional grid. Here in Chicago-land the primary failure is the local distribution power lines that deliver the power from the step down substations to the point of service. These lines are very old - some are original, as in the original service line when electricity was introduced. They are mostly above ground so between age and exposure these lines are the first thing to fail. Should we bury the new ones? YES. Are we going to? Eventually.

Second is the other old components of our local service system - these are the switching and transformer components that are not quite as old as the lines but are next to fail. Many transformers have the key component sitting in a big tub of liquid, so that can leak out and then it will quickly over heat and fail.

Of course all of the devices in the system can age and then fail, the issue with your local service is that these are spread out all over, up on poles in peoples back yards. The substation equipment is much easier to service and monitor. But it's 2024, so even much of the of these critical remote components can be monitored/tracked by the power company. So the real issue is that it just costs money to maintain, and the general thinking is that maybe nothing bad will happen. Which is why public utilities should just be public. Or you end up with catastrophe - like how the deadly heat wave in 1995 prompted ComEd (fined a whole 2.5 million for helping kill over 700) to make the regional grid more resilient to overdraw failure. So at least that is less likely.

In areas that have witnessed explosive population/industrial growth, especially coupled with low regulation the whole local power system is much more likely to fail under heavy load. These failures typically happen when a general overdraw triggers cascading failures through the local power grid. Newer/smarter substation will go offline to save its expensive hardware from total failure and the shunted load increases the strain on other already strained substations. This domino power failure has repeatedly happened in Texas for example and this is exceptionally expensive to remedy.

In difficult, remote and NIMBY regions like California the point of failure is the lack of local power supply to provide redundancy. The overland transmission lines are simply an essential and fragile component. I grew up in the mountains and any number of accidents would happen to the very long power lines that connected us to the grid. There was just 5 lines and any less than 4 mean total local failure.

Places like Florida will actually have great regional and local power infrastructure, but that is primarily because the rest of the country has paid to have it mostly replaced after all the major disasters. Seriously, a total $139 billion was spent just between 2017 and 2019 on disaster recovery and a significant portion of that went towards power systems. In those places power service won't get interrupted by physical limitation or a lack of basic forward thinking but by actual accidents - be it human or weather.
posted by zenon at 2:31 PM on January 13 [5 favorites]

jonathanhughes: I have no idea why they do

Overload, degradation of the isolation of the coil windings, degradation of the cooling oil, mechanical failure of the enclosure.
posted by Stoneshop at 2:46 PM on January 13 [2 favorites]

Squirrel!

Bird shit but indeed, a lot of them are caused by squirrels.
posted by Stoneshop at 2:54 PM on January 13 [2 favorites]

Best answer: The one that kept the grid operators I used to work with awake at night was frequency drop. The grid is a remarkably fragile object with three circuits running at as close to 60 Hz as possible, but each circuit running at a slightly different phase to even out power delivery. (Yes, I've just made all the EEs choke on their dinner, but three phase power is difficult to conceptualize.)

If a local load (could be something as large as a town, could be a large motor in a factory, could be a slightly grunty wind turbine (of a type no-one installs any more) coming in line. This can cause the frequency in one or more of the phases to dip very slightly. When frequency drops, current goes up considerably: and very quickly, too. If the grid protection isn't quite working as it should, this can briefly induce circuit-melting currents. If you're really unlucky, this can take out some very large switchgear. This kit is typically on a several year lead time. If the operator doesn't have a spare, that whole circuit is out until it can be replaced. This very, very seldom happens, thankfully.

Outage analysis reports usually have timelines measured in fractions of a second from "nothing's wrong" to "everything's off".
posted by scruss at 3:18 PM on January 13 [4 favorites]

After the wildfires, which were started by lightning striking power lines, the state of California was told to come up with a plan to deal with this. They responded by shutting off the power whenever there was a thunderstorm.
posted by Melismata at 3:26 PM on January 13 [2 favorites]

Practical Engineering has a nice collection of videos on the U.S. power grid and what can go wrong with it.
posted by Tell Me No Lies at 4:46 PM on January 13 [2 favorites]

Mylar balloons
posted by mbrubeck at 6:29 PM on January 13 [1 favorite]

Enough Teslas move into a neighborhood whose distribution transformer(s) are already operating near capacity, and fuggedaboutit. One reason why utilities are working to increase outreach to potential EV buyers, while others are working to implement systems that will use meter-sourced data to identify when an EV load has been introduced, and correlate such datasets with location info to identify which transformers are at risk.
posted by jerome powell buys his sweatbands in bulk only at 7:38 PM on January 13

Outage caused by squirrel crawling into the transformer on the utility pole. It got zapped and tossed several yards. The transformer was brand new, had just been replaced, in part because there was a squirrel drey in the previous transformer, causing it to fail, and thus causing a local power outage.
posted by SaltySalticid at 7:43 PM on January 13

Wingnut with a AR-15 is not a common cause of outages, but it does happen.
posted by zamboni at 8:09 PM on January 13 [2 favorites]

A big earthquake can knock out a whole grid for hours to days.
posted by potrzebie at 8:40 PM on January 13

We had a couple of dudes here in WA who AR-15ed a substation, so they could knock out power to a couple of places they were looking to rob...

Didn't go well for them. But, pretty easy pickins if you are going for that stochastic terrorism.
posted by Windopaene at 10:08 PM on January 13

Yes, I've just made all the EEs choke on their dinner

I can confirm. Sorry, but the first two thirds of what you wrote is a whole lotta nonsense. But fun to read! Starting with "If you're really unlucky" you're good.

AR-15

Gun guys use high-voltage transmission lines as target practice. Power companies use drones (formerly helicopters!) to inspect along the wire and routinely find bullets embedded in them. (Actually they may very well be using time domain technology to find that damage now, instead of needing a drone to fly the length.)

Seconding Grady's videos, as noted by TMTL above. Watch the whole series! He's really great, even though he's just a civil engineer :) (I kid, I kid)
posted by intermod at 10:09 PM on January 13 [2 favorites]

*cries in South African* The government not building enough power stations despite everyone warning them for decades that they needed to build more power stations, then scrambling to build power stations in a hurry and doing it badly, with breathtaking levels of corruption further complicating all stages of this process.

But if we exclude the complex clusterfuck which has led to permanent planned rolling blackouts here, my last unplanned neighbourhood power outage was caused by a local organization trimming some trees on their property and dropping branches onto the power lines. One of the previous ones was caused by the same organization not trimming the branches on their property, which led to them falling down by themselves during a storm (I live in a notoriously windy city), and there have definitely also been some transformer explosions.
posted by confluency at 1:38 AM on January 14

I learned about Texas's "energy island" in 99% Invisible's Grid Locked about the deadly February 2021 winter storm and power outages. Lots of fascinating/ terrifying details about their unique independent grid, scarcity driven systems, frozen infrastructure, and trying to prevent grid failure from frying the entire system.
posted by carrioncomfort at 2:45 AM on January 14 [1 favorite]

→ Sorry, but the first two thirds of what you wrote is a whole lotta nonsense

Oddly, the bits about large motors and wind turbines come from lived experience operating DX-connected generation plant in Southern Ontario. The large motor was in a local industrial facility, and it took the local utility a while, and some $$$ from us to hire a consultant, to find it and install the proper suppression. It kept the local circuit recloser busy (and our generation frequently offline) until they did. The wind turbines were old-style single-speed fixed-pitch types that no longer meet grid code for new installations. The thing that melted was our own disconnect, which we discovered (during commissioning) had a SF₆ leak, so an arc damaged the contacts.

The area had been host to one of Canada's largest coal power stations which had recently shut down. While this was great news for local farmers (no more melted cucumber plants if it rained at the wrong time) the local distribution certainly felt its loss.
posted by scruss at 5:36 AM on January 14

The longest power outage I've experienced at my current address in Chicago was because someone was driving a cherry picker truck with the bucket/arm not fully down. Our power lines run above ground along the alleys, and when they drove past alleys where the lines crossed the street to get to the next block, they yanked them down. Took down at least 3 poles, smashing into garages and one pole landed in someone's above ground pool. (It was a hot summer day and they were very lucky no one was in it.) We were without power for 10-12 hours. Some folks even longer.
posted by misskaz at 8:26 AM on January 14 [2 favorites]

A few things that haven't been mentioned:

Power transmission lines heat up under heavy load, and when they get hot they sag. If there's a heat wave the load and the atmospheric heat add up, and may cause the lines to sag so much they hit something like a tree and short out. (And perhaps cause a wildfire that itself causes other problems, as now seems to happen in California every year).

One of the components of the Texas outage in 2021 (mentioned above) was a failure to weatherproof the power generation equipment. A lot of it runs on natural gas, and the cold snap was severe enough that the equipment that supplies the fuel for the generators froze up. No gas, no power. The Texas outage was a cascade failure, though, meaning that when some of the gas-powered supply went offline, the remaining system had problems coping, and that caused other equipment to fail or go offline.

When I lived in Oklahoma, almost all the outages were pretty prosaic, though. Our winter outages were primarily (perhaps all) thanks to ice on the lines from freezing rain; summer outages were a mix of transformer failures (due either to internal failures or critters), or transmission line damage from wind or falling trees. In ten years in our house in DC I think there's been one random transformer failure in the neighborhood (which sounded like a bomb, heard for blocks around), some tree-into-line problems in windy storms, and, most recently, a lightning strike that directly hit some utility equipment about ⅔ mi (1 km) away from our house.
posted by fedward at 12:52 PM on January 14 [2 favorites]

Almost all of our power outages are cars running off the road and hitting the LPT boxes. About 6 per year!
posted by congen at 1:22 PM on January 14