Hacker News Reader

Comment by pjc50 | Hacker News Reader

Report on a smaller event in the UK: https://www.ofgem.gov.uk/publications/investigation-9-august... (see linked pdf)

That started from a combination of a lightning strike and generator trip, but turned into a local cascade failure as lots of distributed generation noticed that the frequency was under 49Hz and disconnected itself. I suspect the Spanish situation will be similar - inability to properly contain a frequency excursion, resulting in widespread generator trips.

(I suspect this is going to restart a whole bunch of acrimony about existing pain points like grid maintenance, renewables, domestic solar, and so on, probably with the usual suspects popping up to blame renewables)

View on HN

> probably with the usual suspects popping up to blame renewables

Renewables were a factor in the blackout here in Brazil a couple of years ago: the models used by the system operator did not correspond to reality, many solar and wind power plants disconnected on grid disturbances quicker than specified. That mismatch led the system operator to allow a grid configuration where a single fault could lead to a cascade (more power was allowed through a power line than could be redistributed safely if that power line shut off for any reason), and that single fault happened when a protection mechanism misbehaved and disconnected that power line. The main fix was to model these solar and wind power plants more conservatively (pending a more detailed review of their real-life behavior and the corresponding update of the models), which allowed them to correctly limit the power going through these power lines.

If you want an excruciating level of detail, the final 614-page report is at https://www.ons.org.br/AcervoDigitalDocumentosEPublicacoes/R... (in Portuguese; the main page for that incident is at https://www.ons.org.br/Paginas/Noticias/Ocorr%c3%aancia-no-S...).

The problem is that some media have a tendency to present "operators did not properly model renewables" as "renewables to blame for blackout". The cause is poor modeling, but renewables get the blame as it reinforces a pre-existing "renewables bad" mentality.

This is why solar requires redundant data links to the operator in CA now. Mine has it. During instability they can command my system to accept larger voltage and frequency excursions than would normally be allowed to prevent a "pile-on" where an excursion causes solar to make the problem worse.

Is 50Hz really that important these days – i.e. more important than maintaining power? (Honest naive question.)

Yes. https://storpower.co.uk/the-day-europes-power-grid-came-clos...

Frequency is health of the grid. If it's an entire hz off of where it should be something is massively wrong. Most generating equipment will trip on a deviation that bad.

Frequency is effectively more important than voltage. Or rather it collapses first. Frequency reduction for a rotary generator means that more energy is being taken out than the rotary shaft energy is being put in, so it is effectively an early warning that voltage is about to drop.

Frequency coordination is absolutely critical, via phase coordination. A large generator must not get significantly out of phase. So frequency going out of spec triggers the generator to "trip" (disconnect).

As I understand it, it's not possible for a generator to get out of phase with the grid. Sudden changes in the grid phase instantly and catastrophically break generators, power lines, transformers, and so on, but that can only occur when you connect two grids together without synchronizing them. Which implies there are two separate grids to connect.

I don't know what specific threat is addressed by tripping generators offline when the frequency deviates by 1 Hz. Are they so mechanically fragile that is already damaging to them, or is it a precautionary measure because that kind of instability is likely to precede sudden frequency or phase jumps that are damaging?

> Are they so mechanically fragile that is already damaging to them, or is it a precautionary measure because that kind of instability is likely to precede sudden frequency or phase jumps that are damaging?

Both actually. A frequency mismatch between what the grid has and what the turbine is supplying causes significant thermal losses, so you got to trip the generator anyway rather sooner than later, but a significant frequency deviation is always a warning sign that something is Massively Broken and requires immediate attention to find the cause - too low a frequency means you need to shed load immediately, too high means you need to shed generator capacity immediately.

Yeah, I think it really helps to think about the grid as basically one giant spinning machine. Everything is connected with wires instead of driveshafts, obviously, but a low frequency means that all of those machines that are burning fuel to try to spin at the right speed can’t generate enough torque to spin the rest of the machine at the right speed. And on the other side, spinning too fast means that all of the machines are burning too much fuel and need to slow the heck down right now (which can be hard to do with, for example, nuclear).

That's the point though? The generator can't be lower or higher frequency than the grid. It can be applying its maximum power into a grid that is still short of power, which is probably a better situation than if it was not doing that. Tripping a generator that is supplying power into a grid with too much power (frequency too high) at least makes logical sense. Shutting down generators on a grid with too little power (too low frequency) makes no sense, unless the generator itself risks damage. Loads should be tripped offline when there's too little power, not generators.

One of the parent posts correctly identified there are thermal issues with operating at significantly off nominal frequencies for extended time periods. System operators will specify frequency bands and minimum time a generator has to stay online for each band. For example we have to stay online indefinitely for frequencies 58.5 to 61 Hz, and can trip after 60 seconds for 57.5 to 58.5 Hz, and can trip instantly below 57.5 Hz.

Some plants the system operator has the ability to send a signal to instantly open the generator breakers to shed generation if necessary.

>> It can be applying its maximum power into a grid that is still short of power, which is probably a better situation than if it was not doing that.

> One of the parent posts correctly identified there are thermal issues with operating at significantly off nominal frequencies for extended time periods

Yeah, that's very much the problem. Power sources that burn fuel to generate electricity generally don't like being run at 100% throttle for long periods of time. In a low frequency situation you've got potentially multiple countries of generators all running flat out trying to get the frequency back to where it's supposed to be. If they reduce throttle without having something else come online to take the place of the energy they were producing, the line frequency drops further.

Nuclear has its own issue with running full throttle: once a plant has been running at 100%, it takes time for it to throttle back down and might not be possible to immediately throttle back up. There's decay heat from the fission products and there's short-ish lived (up to 40-50 hours though) neutron poisons.

By thermal issues I meant electrical heating of windings in rotor,stator,transformer due to impedance changes with frequency.

I have only worked on one steam turbine and it had no issue running with the throttle valve wide open all day long.

Every system has a maximum continuous output capacity and I would expect the controls to limit the output so as not to damage the equipment.

I’m curious what sources you are aware of that can’t run at a maximum continuous output all day

There is no such thing as a frequency mismatch between the grid and the generator. There are frequency deviations from 60 Hz but turbine and grid deviate together

Absolutely. The frequency needs to be maintained between 49.8 and 50.2 Hz, or the situation starts spiraling out of control freakishly fast.

What I wanna know is if I could/should have a plug-in box that detects the frequency is drifting and the grid is unstable —-> high risk of imminent failure.

Might even give you clues about something big tripping offline.

Probably lots of false alarms, but if it an outage is particularly bad for you, good to know as soon as the system operators do.

Obviously over the internet could work too, but who wouldn’t love their own box?

no worries, the energy utility does that for you. they care, because if they don't, their hardware bursts into flames.

> What I wanna know is if I could have a plug-in box that detects the frequency is drifting

Yes, it's not that hard. There's smart meters and plugs that have frequency measurement built in.

You can even do it with an audio cable: https://halcy.de/blog/2025/02/09/measuring-power-network-fre...

Old school electric clocks used to (still?) keep time with the grid frequency. Guaranteed to always have N cycles per day. So if things are gradually failing, you might be able to see your clock keeping worse time.

That's not going to work, because the clock will mainly be showing phase drift.

Having the grid operate at 49.99Hz instead of a perfect 50.00Hz for a day will make your clock lose 17 seconds, but it's completely harmless. That's normal regulation, not a gradual failure. The grid chooses to compensate for that by running at 50.01Hz for a day, but that's solely for the benefit of those people with old-school clocks - the grid itself couldn't care less.

A failure means the frequency drops from 50.23Hz to 48.03Hz, probably within a single second. You'd notice as your clock stops ticking due to the resulting power outage.

> Probably lots of false alarms, but if it an outage is particularly bad for you, good to know as soon as the system operators do.

When the phase gets pulled down hard like what nearly happened in Texas and what probably happened here, it'll go from looking like the background noise of phase changes to catastrophic in just a few seconds. It isn't like you'll get warning an hour ahead of time. You'll probably notice your computer monitor going dark before your grafana graph refreshes.

If you have a switching power supply brick for your tiny USB device, not really.

If you have a large spinning inertial mass like a factory motor or a power generation turbine, it's extremely important. Imagine a manual car transmission, but there's no slip-clutch, you need to perfectly align engine with the wheels rotating at 300mph, and the inertial mass you're up against if it's not perfectly synchronized is a freight train.

That's why generators trip offline in a blackout cascade if the frequency deviates out of spec. The alternative is your turbine turns into a pile of very expensive shiny scrap metal.

if a piano is out of tune, you'll hear it, perhaps even flinch.

if a high power transformer goes out of tune... it melts or blows up (or both); it'll try to shut itself down before that. getting it back on becomes a problem if other transformers do the same thing, which is apparently what happened in the whole country.

Yes. The speed of AC electrical machinery is determined by the frequency and number of poles. This means generators connected to each other in parallel MUST rotate in synchronism. If the frequency changes the generator is now producing power out of phase which can cause a fault (a short circuit) and possibly damage the machinery.

Its like a three legged race. You and your partner have to run in synchronism. If either of you slows down or speeds up, the other can trip and fall over taking both of you out.

Yes, maintaining frequency is the most important thing, provided you want your rotating power generation equipment not to explode.

In reality, power generation equipment will disconnect itself if the frequency is too low/high to avoid catastrophic failure.

Reply on news.ycombinator.com