Not sure what's more concerning, the investigation by Red Electrica [1] or Ars' misinformation-grade report about it. Mods, it'd be better to update the thread URL to the real report. The investigation confirms initial suspicions that the blackout was caused by relying too heavily on renewables. Ars claims the report doesn't mention renewables at all and that accusations renewables caused it were "completely without merit", which is a lie.
Summarizing the investigation, the blackout started with
(a) a malfunction at a solar plant in Badajoz that caused it to desynchronize from the grid whilst remaining connected, causing what the report calls "forced oscillations" in frequency, followed by attempts by RE to stabilize the frequency oscillations. Those led to ...
(b) a sudden loss of 845 MW of renewables that they struggled to understand due to the large number of small <1MW sources that feed into the grid but don't report telemetry, but which seems to be related to poorly configured feed-in equipment. Then finally ...
(c) an inability to bring gas plants online to stabilize the grid, because the plants they needed were offline and required 1-2 hours to start up. The rest of the report is concerned with the exact sequence of trips that followed.
The malfunction wasn't a one-off mechanical failure, it's a structural problem with the way solar/wind are connected to the grid. As observed at the time by outsiders and confirmed by the investigation, similar oscillations had been occurring for days beforehand as well as in the morning. Solar/wind doesn't contribute to system inertia, they follow the lead of the turbine-based plants (gas, coal, nuclear). If you don't have enough of that they can start following each other instead, causing feedback loops and instability. Some of them have inverters with algorithms that try to simulate synthetic inertia, but:
> New technologies based on power electronic inverters are capable of adjusting their output within a matter of seconds. While this capability is highly beneficial for the economic optimisation of individual generating plants, it is not necessarily ideal from a power system stability perspective in general.
The investigation itself is poor. If I were PM of Spain I wouldn't accept this, I'd send it back for another try. It fails to find the root cause of the problem at Badajoz (only recommending that someone else investigates), spends far too long on the exact sequence of events after the blackout already began (not that important), and it claims on p16 that the problem wasn't caused by lack of system inertia despite the problem originating in attempts to fight desynchronizations from solar plants that don't correctly follow the grid due to lack of inherent inertia.
Most of their recommendations are also just ways to increase inertia e.g. by importing it from nuke-heavy France, or blame shifting and obfuscation like:
• Provide to the System Operator sufficient observability of self-consumption.
• Review and update of technical requirements for self-consumption installations.
And how are they going to do that? They're talking about rooftop solar installations, who is going to visit and audit them all and with what funding? How long would it take? What's their plan to stabilize the grid until the problems with all the solar plants are fixed?
I already have a blackout kit in the basement, after posting this comment I'm going to go order windup radios and other things it currently lacks because this investigation isn't good enough. For decades power industry greybeards have warned that too many renewables will destabilize the grid. That day has now arrived. Judging by this report and Ars's deceptive coverage of it, the response will be ever greater levels of reality denial. Spain's power supply will remain on a knife-edge for the forseeable future because all the engineers who pushed back on political fantasies of the green parties have been purged, with those who remain being forced into publishing anti-intellectual contortions like this investigation.
That the cell phone networks went down is a criminal negligence and should be prosecuted.
foxyv•7mo ago
When we see power outages in Texas the cell networks will stay up for about 5 minutes until everyone switches from WiFi to 5G for their internet and crashes the network.
mike_hearn•7mo ago
Summarizing the investigation, the blackout started with
(a) a malfunction at a solar plant in Badajoz that caused it to desynchronize from the grid whilst remaining connected, causing what the report calls "forced oscillations" in frequency, followed by attempts by RE to stabilize the frequency oscillations. Those led to ...
(b) a sudden loss of 845 MW of renewables that they struggled to understand due to the large number of small <1MW sources that feed into the grid but don't report telemetry, but which seems to be related to poorly configured feed-in equipment. Then finally ...
(c) an inability to bring gas plants online to stabilize the grid, because the plants they needed were offline and required 1-2 hours to start up. The rest of the report is concerned with the exact sequence of trips that followed.
The malfunction wasn't a one-off mechanical failure, it's a structural problem with the way solar/wind are connected to the grid. As observed at the time by outsiders and confirmed by the investigation, similar oscillations had been occurring for days beforehand as well as in the morning. Solar/wind doesn't contribute to system inertia, they follow the lead of the turbine-based plants (gas, coal, nuclear). If you don't have enough of that they can start following each other instead, causing feedback loops and instability. Some of them have inverters with algorithms that try to simulate synthetic inertia, but:
> New technologies based on power electronic inverters are capable of adjusting their output within a matter of seconds. While this capability is highly beneficial for the economic optimisation of individual generating plants, it is not necessarily ideal from a power system stability perspective in general.
The investigation itself is poor. If I were PM of Spain I wouldn't accept this, I'd send it back for another try. It fails to find the root cause of the problem at Badajoz (only recommending that someone else investigates), spends far too long on the exact sequence of events after the blackout already began (not that important), and it claims on p16 that the problem wasn't caused by lack of system inertia despite the problem originating in attempts to fight desynchronizations from solar plants that don't correctly follow the grid due to lack of inherent inertia.
Most of their recommendations are also just ways to increase inertia e.g. by importing it from nuke-heavy France, or blame shifting and obfuscation like:
• Provide to the System Operator sufficient observability of self-consumption.
• Review and update of technical requirements for self-consumption installations.
And how are they going to do that? They're talking about rooftop solar installations, who is going to visit and audit them all and with what funding? How long would it take? What's their plan to stabilize the grid until the problems with all the solar plants are fixed?
I already have a blackout kit in the basement, after posting this comment I'm going to go order windup radios and other things it currently lacks because this investigation isn't good enough. For decades power industry greybeards have warned that too many renewables will destabilize the grid. That day has now arrived. Judging by this report and Ars's deceptive coverage of it, the response will be ever greater levels of reality denial. Spain's power supply will remain on a knife-edge for the forseeable future because all the engineers who pushed back on political fantasies of the green parties have been purged, with those who remain being forced into publishing anti-intellectual contortions like this investigation.
[1] https://d1n1o4zeyfu21r.cloudfront.net/WEB_Incident_%2028A_Sp...