Hacker News Reader

Comment by alexgmcm | Hacker News Reader

As someone who doesn't speak German and has only a cursory understanding of the electrical grid - how significant is this?

View on HN

The grid was overloaded significantly, but not critically. For everybody concerned with the grid, this is a major event because the grid is usually extremely stable. For everybody else, it's not significant. Grid operating as designed.

This graph shows the frequency drop we encountered: https://www.netzfrequenzmessung.de/bilder/2020_01_08_Abschal...

The red line at 49Hz is when things actually get switched off for the first time (12.5% of load).

One of the comments in the Twitter thread was "I've been measuring the grid for 10 years and have never seen anything like this".

In general, slowing frequency is an indicator of a heavily loaded grid. I don't have an appreciation how much the frequency can practically be off the nominal 50 Hz (in Europe) before it indicates grid failure.

does this happen if the heavy load is at region scale? because at small scale (within a village in my experience) the voltage would drop. I didn't measure the frequency though.

This is fascinating to me. What is the mechanism by which this happens? The added load causes an average slow down across the population of actual spinning generator bits?

Yes, the generators literally have to push harder and slow down.

That's exactly right. Extra electrical draw on the grid (or a loss of electrical production elsewhere) causes the turbines in power plants to physically spin slower due to increase load on each turbine.

An (imperfect) analogy would be if you and a friend were biking up a hill on a tandem bike. If a third person hopped on the handle bars (or your friend stopped pedaling) the bike's wheels would slow down, assuming you are putting a constant amount of power into the pedals.

Tangentially (or maybe tandemially :) this illustrates the problem of synchronizing an e.g. solar power unit delivering to the net. The person jumping on a multicycle in speed must synchronize their feet to the rotating pedals to actually add power.

Most power generators are rotating engines (there are exceptions like solar and and batteries that use power inverters).

Increasing load decreases frequency exactly the same way as anything rotating slows down when you increase load. Electric grid is like chain in a bike. Generator producing power leads a little. Loads follow little behind. They are not exactly matched. Like bike chain is tighter to one direction and looser to other under load.

Mass of rotating generators provide kinetic inertia into the grid so it reacts very little to small variations in the load. But if the load becomes too high and turbines can't speed up, frequency of the grid slows down.

In practice it depends a lot on how a bunch of specific power equipment is set up.

Remember that the A/C power grid needs to be synchronized across the entire continent, so each A/C generator that makes A/C power and runs at 50Hz needs to have the position of a physical object (the rotors) match the position in the current power cycle. If it's badly wrong, you have power-plant sized energy feeds fighting each other — imagine a train wreck, two massive oversized freight trains smashing right into each other. It's like that, but with electricity instead of a physical collision.

This can obviously cause major damage, so once you start deviating from the reference frequency by enough, it's safer to just cut off the interconnects and go offline.

Under 49Hz they start disconnecting consumers. If that does not stabilize the dropping, under 47.5 Hz they disconnect the power plants and the network fails. [1]

[1] https://www.netzfrequenzmessung.de/abschaltung.htm

Very. It means that the grid suddenly lost a significant amount of generating power.

Or added a ton of load faster than it could ramp up

The graphs show the frequency falling off a cliff, which is more consistent with a sudden loss.

It could be that, too. But I'm at a loss trying to think of how that much load could be added so suddenly.

I didn't feel like reading a low effort anecdote about that one time someone tried to run their house off of some rooftop solar and a golf cart battery and all his ceiling fans slowed down and everything died when the fridge kicked on so I figured I'd bring load based frequency reduction into the discussion on my own terms.

I can't think of any practical way to use that much power all at once either. On a small country size grid with a lot of renewables that can't ramp quickly I could see some foundries pulling it off.

Maybe somebody connected woodpecker [1] to the European grid? But I guess even that does not need enough power.

[1] https://en.wikipedia.org/wiki/Duga_radar

Wikipedia says that's 10MW

The drop was around ~4.5GW

I know your joking but it's been dead for more than 30 years :) Only the antenna remains (and the only one still standing was the receiver, not the transmitter).

Though people have hooked up amateur radio equipment to it AFAIK.

Reply on news.ycombinator.com