What Caused the Recent Power Outage in Spain, Portugal, and Southern France?

[Guineafowl]

Can anyone offer some insight into the recent grid blackout in Spain, Portugal and southern France?

Possible causes/factors reported include a cyber attack, freak winds wrapping HV cables together, humidity and high temperatures causing physical oscillations and failures of cables, or a lack of physical rotating mass.

I realise no-one knows for sure yet, but some comments on those possibilities would be interesting.

 

[phinds]

I realise no-one knows for sure yet, but some comments on those possibilities would be interesting.

Sure, let's all just spent time speculating on what might have happened. I mean, who needs to wait for the facts?

 

[Guineafowl]

I was hoping the discussion would evolve as the investigation progressed. There are members on here who know a lot about power grids, and they must be thinking about the event. I’d like to hear their thoughts.

Would you say your post obeys the rule against snide remarks, or aligns with the PF values of civility, patience, diplomacy, etc.? I wouldn’t, so please turn down the sarcasm.

 

[nsaspook]

I was hoping the discussion would evolve as the investigation progressed.

Would you say your post obeys the rule against snide remarks, or aligns with the PF values of civility, patience, diplomacy, etc.? I wouldn’t, so please turn down the sarcasm.

Yes. He was very respectful and correct IMO about not speculating without at least some verified engineering facts.

I'd like to know what really happened, and the root causes that triggered this mess.

 

[Guineafowl]

Yes, I appreciate it’s early days, but as I say, the facts will dribble in as we go. I’m pretty sure the information on here will be better than what I’m reading in the papers.

Edit: for example, my usual paper has a graph of Spain’s energy demand, showing it floating around the 30 MW mark. It’s clearly a few powers of ten out.

 

[phinds]

Would you say your post obeys the rule against snide remarks, or aligns with the PF values of civility, patience, diplomacy, etc.?

Yes, this is a science forum. As member @PAllen has said

Note, we are not called 'idle speculation forums without understanding or knowledge'.

I prefer to wait for the facts.

 

[phinds]

I'd like to know what really happened, and the root causes that triggered this mess.

Me too. It was QUITE a mess for a whole lot of people.

I hope it doesn't devolve into a finger pointing exercise but gets to the bottom of why it occurred. The only thing I have read so far is that a cyber attack has been ruled out, but there was no discussion (at least I didn't see any) of HOW it had been ruled out.

 

[Guineafowl]

I wouldn’t call the speculation of someone like Anorlunda idle, nor lacking in understanding or knowledge, if he chooses to contribute. If you prefer to wait for the facts, that’s fine by me, but please allow the thread to continue, as I’m interested.

 

[Gheed]

It is impossible to prove or disprove many theories popping up. There are no videos of a transformer failing or high-voltage line melting to go on. A hack in the system, how would you even know? If you can rule out one theory, it does not exactly narrow it down either. But if the forum rules do not allow it you can still have a discussion about power grid reliability and risks in general.

Maybe some things can be somewhat 'grounded' like what exactly are 'high temperatures' causing breakdowns or 'how much rotating mass is too little' or 'what forms of physical oscillation happen in HV grids'. Assuming it is not all just quack.

Reminds me of this from Feynman about 'Why' questions (and magnets)

 

[berkeman]

I wouldn’t call the speculation of someone like Anorlunda idle, nor lacking in understanding or knowledge, if he chooses to contribute.

As you can see from his Profile, @anorlunda has not been active at PF for the last couple of years. Likely sailing around the world again...

 

[Guineafowl]

As you can see from his Profile, @anorlunda has not been active at PF for the last couple of years. Likely sailing around the world again...

Ah, I hadn’t looked. I bet there are some other power engineers on here, but his name came to mind as the one who wrote the Insights article on what happens in the grid when you flip the light switch.

 

[DaveC426913]

I confess, I'm a bit surprised that they - apparently - don't even have a clue yet. At least, not one they're willing to share.

I'd think we'd at least have some facts like where it originated and what components stopped working.

 

[tech99]

It is interesting to think about factors affecting the reliability of modern power networks. One of the factors seems to be the amount of mechanical inertia in the system, because the old rotating alternators have been partly replaced with inverters running off solar and wind power. The frequency of a network needs to change slightly to match the load, and rotating machinery helps to iron out the response to fast glitches and overloads. Rather than scrap the old power plants, maybe keep the machinery spinning.

 

[Guineafowl]

It is interesting to think about factors affecting the reliability of modern power networks. One of the factors seems to be the amount of mechanical inertia in the system, because the old rotating alternators have been partly replaced with inverters running off solar and wind power. The frequency of a network needs to change slightly to match the load, and rotating machinery helps to iron out the response to fast glitches and overloads. Rather than scrap the old power plants, maybe keep the machinery spinning.

You mean run them ‘in reverse’, ie motoring? That sounds simple and effective.

It’s emerging that a power source of some sort failed in the south-west, then another, then the Spain-France interconnect. In general, a drop in frequency leads to generators shutting down, which leads to more frequency drop, which is a positive feedback loop.

The idea of load shedding makes sense, but why do generators shut down in these cases? Does supplying a lower frequency lead to more current draw, and some sort of runaway?

The papers talk of frequency drops ‘damaging electrical equipment’, even in fractional amounts. But most sensitive electronics are SMPS-powered and so shouldn’t notice, ditto brushed motors. Induction motors would run fractionally slower, but my milling machine motor runs quite happily, in short bursts, from 5-100 Hz. In a nutshell, why does such a tiny drop in frequency matter?

 

[Rive]

The papers talk of frequency drops ‘damaging electrical equipment’, even in fractional amounts.

At grid level small (very small) frequency (phase) changes what determines the power flow (intensity, direction).

 

[Guineafowl]

At grid level small (very small) frequency (phase) changes what determines the power flow (intensity, direction).

In a similar way, as I understand it, my solar inverter pushes power into the grid by slightly leading the frequency. But I still wonder what equipment would be damaged by running at 49 Hz instead of 50 - is it an overcurrent/overheating issue? Or is it not so much about the equipment, but the low-frequency region pulling power from other areas and destabilising the network?

 

[Rive]

Or is it not so much about the equipment, but the low-frequency region pulling power from other areas and destabilising the network?

Yes. I think in this context 'equipment' is grid-operator stuff, the power flow is in MW range, with frequency variations usually measured in mHz...

 

[Guineafowl]

Yes. I think in this context 'equipment' is grid-operator stuff, the power flow is in MW range, with frequency variations usually measured in mHz...

I can see how a big rotating mass could stabilise the grid if we’re talking mHz and mS.

What would happen to a) an alternator and b) an inverter if it continued online at, say, 49.5 Hz?

 

[Rive]

Are you asking about the difference between the spinning mass vs. electronics type generation on the grid, or about end-user equipment?

 

[Guineafowl]

Are you asking about the difference between the spinning mass vs. electronics type generation on the grid, or about end-user equipment?

The former - what would happen to a power plant alternator, and a solar or wind inverter, if they failed to go offline in the face of such a frequency drop?

 

[Rive]

When the good old spinning mass meets a grid frequency drop then it'll increase the output for a limited time till the inertia can support it, or the protection allows.

Wind is tricky: big turbines/generators are not a homogeneous bunch. In general, it's usually said that they do not provide much power increase in case of grid frequency drop.

Inverters (solar) - in general they have no reserves of their own, so no means to provide increased power. They just pump out the same power till frequency is within allowable range, and then shut down.

Though there is a catch for the latter: some new (old ones can't!) solar inverters can hold back if the line voltage is high (and that coincides with higher frequency periods), so some increase in power is possible when frequency drops (voltage goes down from excess to normal level). But it's not centrally regulated, thus kind of unreliable/tricky/limited.

 

[Guineafowl]

When the good old spinning mass meets a grid frequency drop then it'll increase the output for a limited time till the inertia can support it, or the protection allows.

Wind is tricky: big turbines/generators are not a homogeneous bunch. In general, it's usually said that they do not provide much power increase in case of grid frequency drop.

Inverters (solar) - in general they have no reserves of their own, so no means to provide increased power. They just pump out the same power till frequency is within allowable range, and then shut down.

Though there is a catch for the latter: some new (old ones can't!) solar inverters can hold back if the line voltage is high (and that coincides with higher frequency periods), so some increase in power is possible when frequency drops (voltage goes down from excess to normal level). But it's not centrally regulated, thus kind of unreliable/tricky/limited.

Although such huge blackouts are rare, would you say, with the advent of more renewables, that we might move from such huge grids to a more cellular, compartmentalised arrangement? The latter would seem more resilient (by analogy with firebreaks/bulkheads) and simpler to control. Or is it the opposite, that a huge grid has more ‘inertia’ and therefore stability, and this trumps the extra complexity?

I’m still not quite clear on the function of underfrequency protection for the spinning alternators or inverters. It seems counterintuitive to drop out power sources in the face of a power drop, so there must be a good reason for it. Is it as simple as lower frequency leading to higher current draw, and overheating?

I used to tend to a 30 kVA 3ph genset that would run a big house during power cuts. Now, if a big load came on, it would ‘dog’ the diesel engine down, frequency would drop, but as long as the overcurrent protection didn’t trip, it would simply ride out the surge and carry on. Why do grid power sources not do this, or is my example too stripped-down to be relevant?

 

[Rive]

would you say, with the advent of more renewables, that we might move from such huge grids to a more cellular, compartmentalised arrangement?

For the integration of intermittent energy it was a key point that with more sources connected the fluctuations are supposed to 'average out' => widespread, connected grids.

On the other hand, with the integration of intermittent sources the prices are going up, making small, local grids/solutions viable.

Honestly, I have no idea how will this turn out.

I’m still not quite clear on the function of underfrequency protection

I'm also not an expert at the required level to answer that in proper detail. But: usually it's all about the carefully controlled network of connected devices and grids. As long as there is adequate control, the whole thing is operated within limits, thus connecting (synchronising) the pieces is possible.
Once it's out of control (and out of frequency range is a good indication) then it's just out of control and keeping connected becomes a serious risk.

With one grid (part) dropping at 49Hz and another still keeping 50Hz you'll get some really serious issues at the connecting transformer very soon, right?

I used to tend to a 30 kVA 3ph genset

Hook up three of that in parallel and see what happens when one of them suddenly goes out of sync...

 

[Guineafowl]

Hook up three of that in parallel and see what happens when one of them suddenly goes out of sync...

Argh! Yes, I see what you mean; throw in an interconnect with another house as well, and some solar inverters…

Thanks for the replies. Although I’ll never be a power engineer, I’m clearer on just how difficult it all is.

 

[tech99]

The alternators in a network must all be at the same frequency, not close but the same. The changes that occur during load variations in normal operation are slow changes in the position of one alternator rotor relative to the others in the network. As the network demands more energy, we have to supply more steam to get into exact step again. If we cannot do this quickly enough, an alternator will lose sync and be destroyed.

 

[DaveC426913]

 

[nsaspook]

PSRC - IEEE PES
https://www.pes-psrc.org
PDF
IEEE TUTORIAL ON THE PROTECTION OF SYNCHRONOUS GENERATORS

Read this.

 

[Tom.G]

I think a technical report gets more into the subject.
Try this: 119.pdf "Practices for Generator Synchronizing Systems"

Many more on the general subject here: https://www.pes-psrc.org/reports

Cheers,
Tom

 

[Guineafowl]

^^ thanks both, I’ll have a read through that.

 

[Dullard]

It's more useful (for me, anyway) to think in terms of phase errors. Being at the same frequency is necessary, but not sufficient.