Rules, Not Renewables, Might Explain the Iberian Blackout

https://spectrum.ieee.org/spain-grid-failure

by rbanffy • 1 year ago

47 31 1 year ago

spectrum.ieee.org

diggan5• 1 year ago

the four official investigations into the cause haven’t yet released their conclusions, and people are still waiting to know the causes of the blackout.

Officials themselves have said it'll take months (https://elperiodicodelaenergia.com/el-gobierno-estima-que-su...) before the report is done, basically saying it won't be until sometime between July and October that we'll know what happened. I kind of feel like there isn't much point in speculation about the reasons until the people who sit on all the data can actually give some definitive conclusions.

A bit like guessing why Google/CloudFlare/Amazon went down this time, five minutes into the downtime, and arguing about why it is/isn't DNS/BGP.

cogman104• 1 year ago

The problem the OP is talking about is renewable islanding.

To be clear, this is a renewable problem, not a rules problem.

The problem, put simply, is that renewables don't want to be reactive power sources. It requires additional equipment to properly determine what the frequency of the grid should be (some rely on flywheels, for example) and getting it wrong can be disastrous for the grid and equipment. Effectively, if renewables get it wrong they become a load on the grid rather than a reactive power source. Compounding an active problem.

So, what renewables do instead is follow the voltage wave. If the frequency drops, they cut power to avoid causing damage to the grid and other power suppliers. The thing more traditional spinning disk power suppliers will do is simply power through the down voltage. The disks have momentum which means even if they wanted to, they could just stop producing power. This somewhat self corrects a large reactive power surge.

Grids do not like doing this. This is why they'll charge more to industrial power users that add highly inductive loads (think electric motors, like a saw). They'll even incentivize having these businesses install capacitor banks to avoid adding inductive loads.

The solution is rules, but rules renewable producers have to follow. That's mandatory islanding equipment installs rather than just shutting down power.

scblock2• 1 year ago

Renewables can absolutely provide and consume reactive power through the use of power electronics equipment incorporated into the inverters. All modern utility-scale wind and solar power plants use this type of equipment, and most interconnection requirements in the US require significant reactive power capability as part of the interconnection process. This may require additional reactive power capabilities such as capacitor banks, DVAR units, or reactors at the project substation.

This means if grid voltage drops they will provide VARs to increase voltage, and vice versa. They can similarly react to shifts in grid frequency. Only after they reach certain power factor, voltage, or frequency limits will they disconnect from the grid, similar to other generators.

If in spain renewable energy projects were somehow exempted from needing to provide a wide range of power factor capability then yes this would be a rules problem.

They charge more for inductive loads because of low power factors and significant VAR consumption yes, though you can also do power factor correction on induction motors. And though most wind turbines use induction generators, they are designed to be able to supply power at unity power factor and through a large power factor range.

Also, islanding means operating as a standalone grid, which does not happen when the generators shut down, and systems are designed to shut down after a certain number of cycles without grid. This does not sound like islanding at all.

cogman104• 1 year ago

Good comment and I do not disagree. I wasn't aware that the US had significant reactive power requirements (I'm guessing for utility scale solar).

And though most wind turbines use induction generators, they are designed to be able to supply power at unity power factor and through a large power factor range.

Correct me if I'm wrong, but I was under the impression that wind turbines did an AC->DC->AC dance to ultimately get the right power factors and frequencies.

jauntywundrkind• 1 year ago

The problem, put simply, is that renewables don't want to be reactive power sources. It requires additional equipment to properly determine what the frequency of the grid should be (some rely on flywheels, for example) and getting it wrong can be disastrous for the grid and equipment.

I sort of agree with you, but the motive I'm not sure I agree with. "renewables don't want to be reactive" doesn't feel right.

The challenge here is technical/control. There probably is no heuristic you can make at your home to know how to nudge the power phase/reactivity, or to supply less.

I think renewables probably could be enormously helpful and probably wouldn't mind, wouldn't be very hard for them to help shape the power!! They are vastly more flexible at inverting power than any other power source, are very readily programmable.

But most renewables don't have the whole grid view, would be guessing wildly based on local conditions that don't provide adequate information to show what's really happening to the power network.

I hope over time the idea of the Virtual Power Plant related infrastructure starts building good protocols, starts being able to ask for what it wants from who it wants, providing some of the coordination necessary to take advantage of the enormous flexibility that renewable power, with its digitally dictated power delivery, bring. I think renewables would be amazing and happy to help here, if there were better information flow.

cogman104• 1 year ago

But most renewables don't have the whole grid view, would be guessing wildly based on local conditions that don't provide adequate information to show what's really happening to the power network.

This is how other power generators are working. You don't really need a whole grid view either to make these decisions based on your local interconnect. You do, however, need either some pretty fast electronics and inverters or a reliable system that responds correctly to frequency demands. Making it smarter isn't entirely helpful.

kmitz• 1 year ago

I think you are mixing up different concepts.

Power impacts frequency, reactive power impacts voltage. It doesn't make sense to "cut power when frequency drops".

eldaisfish2• 1 year ago

I can confirm. The parent comment here mixes several concepts and is mostly incorrect.

eldaisfish2• 1 year ago

This is not an islanding problem.

The vast majority of wind turbines and every solar PV inverter can provide reactive power. They are mandated to provide this in many parts of the US and Australia.

Frequency and reactive power are not directly connected. Active power influences frequency, reactive power affects voltage.

Reactive power does not change whether a renewable generator is a load or a generator.

Low-voltage ride through, including reactive power requirements have been in place for over a decade now, at least in Europe.

The comment above demonstrates a shocking level of misunderstanding.

guerby• 1 year ago

"not a rules problem"

I beg to differ.

The only difference between your run of mill solar/battery inverter and a grid forming inverter is ... software. Source NREL: https://docs.nrel.gov/docs/fy24osti/90256.pdf

Grid connected inverter software has to follow mandated rules defined by the grid managing organisation, otherwise it's simply illegal to connect it to the grid.

So if there are issues related to the behaviour of the grid connected inverters, except for bugs, mismanagement or hardware failure, it's 100% the fault of the grid managing organisation.

Inverters can provide grid stability if and only if the grid managing organisations allows it by putting in place rules and market mechanisms for it.

SideburnsOfDoom3• 1 year ago

renewables don't want to be reactive power sources.

But batteries can be. No-one is talking anymore about renewables alone without batteries as well.

MaxHoppersGhost• 1 year ago

0% chance European investigators find renewables at fault.

cogman104• 1 year ago

I see no reason why they wouldn't. Nobody wants an unstable grid and if it's renewables that ultimately killed the grid they'll want to know exactly why and how to prevent it in the future.

There are known problems with renewables, but there's also solutions that can be retrofit onto renewables to avoid those problems. Renewable power suppliers won't like those costs because it will cost money.

ErneX5• 1 year ago

The Spanish government blamed nuclear since the day of the blackout. Red Eléctrica has a government majority stake and they appointed its director (someone with zero energy sector experience by the way) and seeing how they cover every corruption scandal lately I wouldn't be surprised they try to sweep this under the rug as well.

Hopefully the EU does a decent investigation though.

SideburnsOfDoom3• 1 year ago

IDk what you're saying:

1) ... because renewables are clearly not at fault and can be discounted immediately?

2) ... because they're ideologically protecting faulty renewables?

1 is a stretch (I don't think they likely are at fault - the disinformation on this is obvious, but investigators have to investigate everything), and 2 is bilge.

ErneX5• 1 year ago

They changed the mix since then to have more presence of combined cycle, that should be a hint.

andy12_• 1 year ago

It really is an obvious hint. Anyone can check that since the blackout there is never less than 4000 or 5000 MW of nuclear at all times (compared to 3000 MW of nuclear in the days before the blackout).

Also combined cycle. On the day of the blackout there was very little, about 1600 MW. Now there is usually never less than 3000 MW of combined cycle even at peak hours of solar generation.

Even in the report they released they talk about how just 10 minutes before the blackout they were scrambling to put online more conventional generators to stabilize the grid (they never had the change to go online before the blackout happened). [1]

"At 12:22 h when the system recovers from EVENT 2, it is observed how the system voltage starts to rise, but within the operating margins. In view of this situation, the REE Control Center begins to couple reactors in several substations to counteract this rise. [Likewise, [...] the decision is taken to couple more conventional generation that complies with P.O. 7.4 mainly in the south, asking for coupling times. A CCC group is chosen in Andalusia, which gives 1 h and 30 minutes of coupling time to the grid, technical minimum at 14:00h. This measure was never implemented due to the zero voltage. Coinciding with this decision, a control center reported that a nuclear group was oscillating a lot with the oscillations and that it could trip. As a precautionary measure, advance notice was requested to couple to another CCC group in the North zone. Then, another control center indicated that a CCC group in Andalusia needed between 2 and 2.5 hours to couple and asked to modify the offer to make it cheaper. After this, the first control center confirms that the northern CCC group would couple at 15:00 h. [...] As a result of all the above events, the system was at a very different operating point from the initial one, with low damping and less flexibility in voltage control. [...] The corrective measures taken could not be fully implemented, e.g. synchronization of the two groups."

[1] https://d1n1o4zeyfu21r.cloudfront.net/WEB_Incidente_SistemaE...

jakub_g• 1 year ago

Relevant parallel thread:

"Miscalculation by Spanish power grid operator REE contributed to massive blackout, report finds"

https://news.ycombinator.com/item?id=44301186

SideburnsOfDoom3• 1 year ago

Nobody serious thought that this was likely a "renewables" problem.

The obvious bandwagoning was obvious.

Rules, Not Renewables, Might Explain the Iberian Blackout

Related Stories