In the end it took about 12 hours to restore most of the power in two countries. There are some knock-on effects from trains being in the wrong places, but a solid result overall.
Plant startup is only the first step.[1] It's load pickup.
Here is a PJM training module on load pickup during system restoration.[2] It gives a sense of how touchy the process is. Power network control has good control over generation and transmission, but limited control over load.
When load is turned on, there are transient loads with different time constants. There's a huge load for the first second as inductors, capacitors, and incandescent filaments start up. That tails off in under a second. There's a second load as motors wind up to speed. Ten seconds or so. Then there's "cold load", where everything in HVAC starts trying to get temperatures back to normal. Maybe half an hour.
There's no mention of computer control. Listening to this, you visualize people running around reading meters and throwing big switches. It's probably people looking at display boards and sending commands to remote big switches, but the concept is the same.
Botching this means voltage or frequency goes out of tolerance, protective devices shut things down, and the system operators have to start over.
More PJM training modules on related subjects.[3]
Unclear what caused this yet. Something caused enough system instability to trip protective devices, but there's no good info yet. Once everybody has a chance to compare all the logging data from different points, it will make more sense.
[1] https://pjm.adobeconnect.com/_a16103949/p622tuwooba/
There are costly means to compensate for the lack of spinning baseload but actually building these devices have been neglected, to no ones surprise.
On-topic: Wind power brings in about 20% of Spain's electricity, and that one fluctuates a lot indeed. I hope they put out a detailed post-mortem, I'd be an enjoyable read.
(also we're a bit early for a proper postmortem, but that's an inevitable reality of social media, got to get the talking points in before the facts)
I agree as well. And I think the gp (or the ggp?) post noticed that there is a lack of "spinning mass" to do frequency damping would have agreed as well. They could be merely suggesting to add just spinning mass synchronizers. These could be independent units, doing just that or maybe even convert some old power plants: "just" decouple the turbine from the generator.
Looking at power generation reports it's the plants relying on inertia (like nuclear) which were disconnected and haven't been reconnected since:
In a control system when you have an actuator that is much more responsive than the load, you can get into instable operating regimes where. On a small scale, when you are setting up servo motors, there is an inertia matching ratio for stability. In some of the systems I set up it was 10:1 load inertia:motor inertia. If you exceed that, you end up with elasticity between the motor and load that the motor can locally be above set point, and the control tells the motor to slow down, while the load has not yet reached set point. The motor is then too slow, and the control tells it to go faster, and you get the motor oscillating around the speed of the load.
On the grid you can have similar elasticity. An inverter many km away can get too far ahead in the AC cycle, then when that signal reaches the local inverter, the local inverter slows down, but the remote inverter has already slowed down, and by the time that slower wave gets to the local inverter, it is already too slow, and so the local inverter starts to speed up. If you get the frequency "right" you can end up with a positive feedback loop. With a high inertia rotating generator, the system is damped and slow enough to respond you don't change set point faster than the signal can propagate.
-react immediately
-use its full capacity all the time
Imagine you're trying to coordinate a choir of ten million singers that are scattered across a radius of a thousand kilometer, that all sing into their individual Ham radio. And you need it to come out in a perfect unison that sounds good for those who tune into the mixed broadcast.
"A problem of coordination" is perhaps correct, but it neglects the difficulties involved
Also we know the locations of the generators, so we can calculate the phase shift and control them accordingly. At thousand kilometres it's around 60 degrees anyway, so not catastrophically huge.
Then depending on how out-of-sync a given station is with the signal, they ramp production up or down accordingly.
Is the problem here that there is inadequate or slow feedback to the supply side regarding load and therefore you'd end up with under-voltage?
It's not about getting the generator to spin at 50hz, it's about getting it to spin at 50hz under dynamic load. Most generators are mostly fixed speed I think for the sake of efficiency, so the mechanism the grid has to regulate the frequency of power is largely by controlling the number of generators running at any given time, and how much of the grid is connected to them. Ramping up and down based on the current observed frequency of the grid is exactly how this all works. However it's more complicated because what if two or more power stations see a low frequency grid and decide to ramp up more generators? Well now the grid frequency is too high. If they them respond to that by ramping down, it'll go too low. Rinse and repeat and you get undesirable frequency oscillation, so there needs to be more communication across the grid than just responding to the locally observed frequency.
Connect a stopped, or sufficiently out of phase generator to a full scale power grid and the puny thousands of pounds of metal that make up a generator are going to get ripped apart by the sheer inertia of the power grid forcing it to match speed and phase in an instant. This is also a problem when attempting to connect two isolated grids, of they're too out of phase with each other, catastrophic physical damage will occur.
Any individual station that gets out of sync either works hard to catch up, or gets disconnected.
The problem, as I see it, is that the load signal is equal and equivalent to the sync signal and maybe they could be separated. (e.g. the FM modulated signal wouldn't have to be a sine wave, nor would it even have to be 50Hz)
> what if two or more power stations see a low frequency grid and decide to ramp up more generators?
Seems to me it would solve this problem - they both have a single invariant sync signal.
> Connect a stopped, or sufficiently out of phase generator to a full scale power grid and the puny thousands of pounds of metal that make up a generator are going to get ripped apart
Seems to me this problem is solvable with electronics, but I don't know enough to say how much out-of-phase is problematic. What if, as you say, some heavy load is switched on, the local nuclear turbines slow by a fraction for a few 10s of seconds, and so a home PV array starts leading by 0.01Hz (e.g. home PV at 50.00Hz, local power station at 49.99Hz, so after 10s the PV is fully 36º or 0.6rad ahead). Does this result in the inverter (or the small local windfarm) exploding? Is that a realistic scenario?
> there needs to be more communication across the grid than just responding to the locally observed frequency
Precisely - a grid-wide sync signal, modulated at high frequency.
More importantly, according to the REE, PV played a mayor role in the current and successful black start.
If the power is out, do you imagine that all the load that is out there disappears too?
If I'm in a blackout and put my multimeter into the grid socket, will the measurement be something that my local inverter can drive an AC load through with no problems?
And if that's not possible, how do we get an AC wave out to my 5000 local solar panel neighbors? And if we don't get any carrier wave to synch it to, then how do all these thousands of individual inverters decide on the unison synchronization necessary to all start jumping together?
And if they try and fail, will all the connected load accept a graceful out of synch mixed power noise at 0-220v until things properly latch together?
The problem isn't like playing a sine tune of acoustic sound that's merely audible to your local neighbor, who can dial in his own tune, and then daisy chaining this until it matches on a grid sized level.
The one thing that surprised me is how quickly rumors started about power being out in Portugal (true), France (true to an extent), Belgium (false), and the UK (false). Walking back home from work you heard the names of Trump being mentioned as well.
It's a bit scary to personally experience how quickly people can start panicking. On the other hand the bars that had some power had a great afternoon with tons of beer being served before it got too warm :-)
Refrigeration is sort of a scourge to those who enjoy food. We are so used to simply freezing our esophagus and not tasting anything! Food is kept cold because it lasts longer, not because it tastes better.
Sure, refrigeration makes restaurants and grocery stores possible. But it also presents challenges to cooking and it's very resource-intensive -- think about it -- basically everyone everywhere is always running their refrigeration -- that's a huge load on any system.
Unless you drink Schultenbrau, which tastes worse cold than at room temperature (/s, kind of).
The same goes for wine, by the way: cellar temperature is usually optimal.
Do they? But why? Why would they? Are they simply accustomed to tasting it that way? I should say "not tasting it" because chilling a beverage basically kills its flavours, and they never make it to your tongue correctly.
I would say that many people would be surprised and amazed if they tasted such things as Pepsi without ice, and without coming right out of the fridge.
Unfortunately, any lack of refrigeration really messes up supply chains and will eat into anyone's bottom line, so it's not a realistic goal.
Beer also does not require refrigeration throughout the entire supply chain, they happily sit in a warm warehouse waiting to be picked up by an unrefrigerated truck, only to be cooled at the outlet or consumer.
On the upside, more people at all levels are going to be more prepared if it happens again.
In short: Heat Inertia of large molten bodies is massive with good insulation. If the time is too long tho, only dynamite will dismantle a solid chunk of material again.
https://www.blindex.com.br/-/media/blindex/site-content/xx-n...
The Texas winter storm resulted in a lot of scrap at semiconductor lines due to power loss. There are industries that are completely dependent on grid power. You can't generate enough on site to back up 100% of your operation. Think about how much power one EUV source consumes.
The only reason Samsung is building additional factories in Texas is because the local utilities are effectively treating them as a critical load. The new plants are right by the ERCOT operation center and likely have access to the same cranking paths that the grid uses for black starts.
https://www.asml.com/en/news/stories/2024/a-sustainability-m...
There are thousands of tools. Material handling robots, compressors, HVAC, pumps, an entire data center within the factory, etc.
First hit off google.
Assuming they run them 24/7 that amounts to a power usage a bit over 1 megawatt, but they could be pulsed for all I know and use quite a bit more
They are also frequently situated close to reliable power sources such as nuclear or hydro, usually fed by more than one generating station.
A friend of mine lived close to a brick plant his power NEVER went out.
As a result, the government invested heavily in installing gas turbines all over the country.
When we had another blackout event in 2018(?), it only took an hour or so to get things back again.
The downside is that the choice was made to focus on smaller 10-20 MW projects, with the occasional 100+ MW installation.
Hence, despite new projects entering service every couple months solar only accounts for less than 3% of electricity production nationwide.
https://www.youtube.com/watch?v=uOSnQM1Zu4w
from Practical Engineering
Sounds simple, and I appreciate from experience a huge amount of prep and validation needs to be done in the background.
Because if not sequenced and interlocked, islanding behavior can result in back-powering parts of the grid that are supposed to be disconnected and down for maintenance, or creation of desynchronized islands that then cause more damage when reconnected.
This is absolutely a problem that operators need to solve. Radio synchronization is possible, neighborhood islanding could be beneficial, etc. There are entire village-wide micro-grids in remote places where everyone's inverters do perform these functions, it just hasn't been embraced in monopoly environments yet.
For instance, relays that close isolated parts back onto the main network would need to be phase-aware, to only reconnect if it's safe to do so. You could mandate that all island-mode inverters center themselves around 50.1 or 60.1Hz in the absence of outside influence so there's certain to be phasing opportunities. Distribution networks would need more switches to provide positive isolation during maintenance. And all of these things would need huge amounts of interoperability testing. It's possible, but it has a cost.
I’m guessing this could be more of a thing in the future to make the grid more robust.
Didn’t Australia have a similar issue a few years back, which Tesla built a large grid battery to resolve, not just for storage but for stability.
Dyson sphere program is the by far most forgiving in that regard. You can overdraw to something like 200% before it becomes really problematic, and even that quickly resolves itself the moment you get more power online
it's that in Factorio there are no problems with phase synchronization, and the devices, consumers, producers, so the machines don't get damaged from a lot of abrupt starting/stopping.
still, it's strange that operators and grid controllers have to do the restart manually.
I operate a microgrid facility in Hispaniola and have wonderfully cooperative users backed by a separately powered communications and control system. Even for us, a facility serving a small neighborhood and farm, a black start must be performed as a careful choreography of systems and loads…and we can just pick up a radio and tell people to turn the main breaker of their house on or off, and to leave their AC units off until we finish bringing everything up. In 12 years we’ve only had to do it twice, but even for us it’s a tedious process.
matkoniecz•8h ago
It dropped from about 32GW to about 8GW
source: https://transparency.entsoe.eu/generation/r2/actualGeneratio...
though admittedly https://en.wikipedia.org/wiki/Black_start mentions that partial shutdown may be also requiring a black start?
scrlk•8h ago
joha4270•6h ago
It looks to me like the author of this article has fundamentally misunderstood what a black start is. The fact that some power generation remained both on the peninsula, but also on imported power being available means that to my understanding this cannot be classified as a "black start". A lot of overtime, a lot of careful coordination yes, a black start, no
franga2000•5h ago