[HN Gopher] Miscalculation by Spanish power grid operator REE co...
       ___________________________________________________________________
        
       Miscalculation by Spanish power grid operator REE contributed to
       blackout
        
       Author : croes
       Score  : 97 points
       Date   : 2025-06-17 16:54 UTC (6 hours ago)
        
 (HTM) web link (www.reuters.com)
 (TXT) w3m dump (www.reuters.com)
        
       | tux3 wrote:
       | They've now presented the report, but the report itself hasn't
       | yet been made public.
       | 
       | The press release is here:
       | https://www.miteco.gob.es/es/prensa/ultimas-noticias/2025/ju...
        
         | kragen wrote:
         | That's fantastic information, thanks! When will the report be
         | made public?
        
           | diggan wrote:
           | Last I heard they were aiming for sometime between July and
           | October to have the full technical report ready and shared
           | with the parlament and the EU commission.
        
           | tux3 wrote:
           | My understanding is that it was supposed to be released later
           | today. It's 8PM in Spain right now.
        
         | riedel wrote:
         | Interesting read. I reminded me of some stories a German guy
         | who used to work for a Spanish utility told me (the guy lost
         | his job at one of the first closed German nuclear facilities
         | and now worked at the museum of a pumped storage power plant):
         | he told us that the Spanish utility he worked for waited far
         | too long to kick in gas power plants waiting for prices to rise
         | further making tons of money by risking stability. I also
         | learned from him that a lot German nuclear facilities are still
         | running there generators only as weight to stabilize the
         | network. (Actually he also told us that some utilities legally
         | abuse the pumped storage to trade energy between networks). So
         | it seems there are some strange incentives and necessities in
         | today's system , that might explain the failures to react in
         | time.
        
           | mschuster91 wrote:
           | > I also learned from him that a lot German nuclear
           | facilities are still running there generators only as weight
           | to stabilize the network.
           | 
           | We aren't. Pretty much the first things that get dismantled
           | is the turbines and machinery that has no radiation exposure
           | because you need space to maneuver, buffer, dismantle and
           | decontaminate everything else that's coming out of the
           | innards of the plant and actually is contaminated.
           | 
           | Maybe you can get some remaining use as reactive power
           | provider done during the ~1 year that it takes for the
           | bureaucracy to process the actual permit for dismantling [1]
           | - but even during that time, restarting it again is
           | practically impossible after the main coolant pipes have been
           | flushed [2] (that happened shortly after the plant had been
           | formally taken offline and all the uranium pellets were
           | removed from the core, although I can't find out how many
           | months that took - but it did happen in the same year as the
           | shutdown, so less than 8 months).
           | 
           | The only thing that's being kept connected to the grid after
           | that is the small portion that supplies the cooling for the
           | cooldown pools where the freshly removed uranium pellets
           | deplete enough material that they can be stored in CASTOR
           | containers [3]. In addition, the high-voltage switching
           | equipment is being kept alive in Landshut's case as the area
           | of the former NPP is planned to be the end point of the
           | SudOstLink national grid expansion link and the power is
           | supposed to be transferred from Ohu to the wider
           | regional/state grid using the existing tie-ins of the former
           | NPP [4].
           | 
           | If you want I can try to contact the utility handling the
           | teardown about specifics, I live a few kilometers from there
           | after all, but I doubt they're going to share much
           | information given the obvious threat of Russian espionage.
           | 
           | It's also a workplace safety issue. You absolutely do not
           | want electricity anywhere where you do not vitally need it,
           | particularly no high voltage.
           | 
           | EDIT: Thanks to @closewith, indeed Biblis A's genset was
           | converted for a "second life" as power regulator in 2011/2012
           | [5], but that's (at least from a cursory research) all I
           | could find.
           | 
           | [1] https://www.br.de/nachrichten/bayern/genehmigung-fuer-
           | rueckb...
           | 
           | [2] https://www.bayerische-
           | staatszeitung.de/staatszeitung/politi...
           | 
           | [3] https://de.wikipedia.org/wiki/Castor_(Kerntechnik)
           | 
           | [4] https://www.br.de/nachrichten/bayern/tennet-stellt-
           | suedostli...
           | 
           | [5] https://www.amprion.net/Presse/Presse-
           | Detailseite_2667.html
        
             | closewith wrote:
             | Biblis A is used for syn-con, which I presume is what the
             | person the GP spoke to was talking about.
        
               | mschuster91 wrote:
               | Ah, that explains things - I only looked at the history
               | of the last three in Emsland, Neckarwestheim and Ohu
               | because I assumed that everything else was far too much
               | dismantled to still be operational in any form or shape.
               | Thanks!
        
           | b3orn wrote:
           | > I also learned from him that a lot German nuclear
           | facilities are still running there generators only as weight
           | to stabilize the network
           | 
           | The term for this is synchronous condenser. According to the
           | system stability report 2023 [0] there were three of these in
           | Germany in 2023 (source is in german, the term you're looking
           | for is rotierender Phasenschieber). It is not clear if these
           | are converted generators, most likely they are not former
           | nuclear power plants, but likely located at the sites of
           | former coal power plants because of the existing grid nodes.
           | For the future this is a concept that is investigated to
           | provide reactive power, in that case from former coal power
           | plants.
           | 
           | [0]: https://www.bundesnetzagentur.de/DE/Fachthemen/Elektrizi
           | taet...
        
       | narag wrote:
       | Disclaimer: I can't really vouch for any of this, but let me tell
       | you what's being discussed in the press over here.
       | 
       | This is Government's version and it shows.
       | 
       | REE boss was put there by the ruling party. Still they try to
       | paint REE as a private operator now. It's not. Like many other
       | orgs, it's a mix of private and public capital, ultimately
       | controled by the government.
       | 
       | It's suspected that the blackout was caused because she overruled
       | technicians' opinion, looking for a solar % record.
       | 
       | Not "enough thermal power stations" carefully avoids mentioning
       | nuclear, that they want to erradicate.
       | 
       | Private companies are requesting the records of CECOEL
       | conversation to be published, some have leaked their logs. In one
       | of those, REE technicians readily admit that oscilations are
       | caused by not enough nuclear in the mix.
        
         | toomuchtodo wrote:
         | Doesn't have to be nuclear, can be anything that provides
         | reactive power and grid services, i.e. grid forming battery
         | storage. The problem is this service is needed on the Spanish
         | grid, but isn't paid for in the market. Pay for it and the grid
         | services will be provided. Resources below. Mind you, doesn't
         | have to be a Tesla battery, just has to be a battery with power
         | controls (grid forming) and some lithium cells for fast
         | response (vs sodium for longer duration discharge).
         | 
         |  _China launches world's first grid-forming sodium-ion battery
         | storage plant_ - https://www.ess-news.com/2025/06/03/china-
         | launches-worlds-fi... - June 3rd, 2025 ("With a total
         | investment of over CNY 460 million [$63.8 million] and
         | occupying 34k square metres, the Baochi plant is designed for
         | an installed capacity of 200MW/400MWh. Based on a dual daily
         | charge-discharge cycle, it can regulate up to 580 GWh annually
         | -- enough to power 270,000 households, with 98 per cent of its
         | energy sourced from renewables. The facility supports more than
         | 30 local wind and solar power stations, alleviating the impact
         | of intermittent supply and facilitating the integration of high
         | shares of renewables into the grid.")
         | 
         |  _Tesla Megapack in Texas Supports the Grid and Keeps the
         | Lights On [Gambit Energy Storage]_ -
         | https://www.tesla.com/videos/gambit-megapack - September 30,
         | 2021
         | 
         |  _Tesla's giant battery in Australia [Hornsdale Power Reserve]
         | reduced grid service cost by 90%_ -
         | https://news.ycombinator.com/item?id=17051066 - May 2018
         | 
         |  _Tesla Powerpacks Balancing the Grid in Terhills, Belgium
         | [video]_ - https://www.youtube.com/watch?v=EVHQFrGzThg - May
         | 14, 2018
        
           | spspeaker wrote:
           | Grid forming services are indeed paid for, according to this
           | very same article:
           | 
           | > Power plants "should have controlled voltage and, moreover,
           | many of them were economically remunerated to do so. They did
           | not absorb all the reactive power that was expected," Aagesen
           | said.
        
             | toomuchtodo wrote:
             | It is possible I misread or misunderstood something on the
             | topic with regards to compensation for grid services in the
             | article. With that said, before jumping to solutioning
             | (although I am a strong proponent of batteries based on
             | their performance, capabilities, and cost as of this
             | comment), I would like to see more information and data as
             | to which generators were signaled to change their output
             | when the grid started to fault and were unable or unwilling
             | to (failure to appropriately manage output and/or field
             | excitation to absorb reactive power, etc) and if
             | improvements to the electricity market are required.
             | Without root cause, we cannot effectively solve.
        
         | gred wrote:
         | > REE boss was put there by the ruling party.
         | 
         | Yeah, these public / private entities can get ridiculous. Her
         | background is legal and political: Secretary of Women and
         | Equality, Spokesperson for Housing, Minister of Housing, State
         | Secretary of Housing and Urban Construction, etc. But she's a
         | party loyalist and she took a few business classes, so it's all
         | good.
        
           | serial_dev wrote:
           | I mean it could be good enough, as long as she listens to the
           | experts and doesn't think that "these nerds are preventing me
           | from making history and breaking records" (assuming what
           | parent comment says is true).
        
             | gred wrote:
             | Sure, but even better would be if she had experience in the
             | industry to judge tactical advice she is given and set
             | strategic direction. Otherwise the best she can do is "not
             | get in the way".
        
         | 7952 wrote:
         | Well nuclear is a thermal power station.
        
         | mocmoc wrote:
         | 100% this. In the conversations , it's cristal clear
        
         | AshamedCaptain wrote:
         | > Private companies are requesting the records of CECOEL
         | conversation to be published, some have leaked their logs. In
         | one of those, REE technicians readily admit that oscilations
         | are caused by not enough nuclear in the mix.
         | 
         | And where are those leaks, if one may ask?
         | 
         | French-side the only generation station that actually was
         | ejected from the network that day was, precisely, a nuclear
         | power plant. [1] A non-specific "moar nuclear!" cry seems
         | therefore a bit hard to believe.
         | 
         | [1] https://www.edf.fr/la-centrale-nucleaire-de-golfech/les-
         | actu...
        
           | Etheryte wrote:
           | This doesn't make the point you think it makes. Nuclear power
           | plants are _supposed_ to go offline in unstable grid
           | conditions. This ensures they don 't overheat, equipment
           | isn't damaged, and that they can be taken offline in a safe,
           | rather than unsafe manner.
        
             | AshamedCaptain wrote:
             | Your point is not really a point either: _every_ power
             | plant can be ejected from the network in such cases, and
             | not just nuclear ones. The remark here is that it precisely
             | happened to a nuclear first.
        
               | cryptonector wrote:
               | Nuclear can't ramp up or down fast. As the grid loses
               | generation capacity the load on the base-load type
               | capacity becomes too variable and that capacity has to be
               | brought offline. It's that simple. Nuclear wasn't ejected
               | because there was too much of it but because there was
               | not enough capacity of all types, especially of the gas-
               | fired type that can most readily respond to solar and
               | wind drop-off.
        
           | cryptonector wrote:
           | This doesn't show what you think it does. It shows that there
           | was not enough gas-fired (and battery) capacity to offset
           | solar and wind drop-off. Nuclear was driven off because
           | nuclear is strictly base-load only and cannot ramp up fast
           | enough to cope with the variability of peak-load -- under
           | those circumstances nuclear _must_ disconnect from the grid
           | (or part of the grid must be isolated from the rest).
        
             | btilly wrote:
             | I would think that a trivial solution for that is to create
             | devices that can supply variable demand. (For example by
             | heating up water somewhere in the ocean.) If the power
             | created by nuclear exceeds base-line demand, create
             | artificial demand to level out the valleys. By
             | strategically wasting power, nuclear can supply a larger
             | share of the demand.
        
               | cryptonector wrote:
               | That's the point of batteries, yes. They don't have to be
               | chemical batteries. Sand (heated to charge, heat
               | extracted to discharge) works too.
        
         | cryptonector wrote:
         | Yeah, this is a miscalculation, a miscalculation of the -you
         | know- _political_ sort. It's not some junior employee's math
         | error. It was _policy_. It blew up in the policy maker's face.
        
         | robocat wrote:
         | > REE technicians readily admit that oscilations are caused by
         | not enough nuclear in the mix
         | 
         | Any argument using the words "nuclear" or "inertia" are
         | incorrect (I suspect often downstream of misdirection caused by
         | the incumbents).
         | 
         | A power network is supposed to be reliable, and it is up to the
         | regulator to create the internal incentives and restrictions
         | such that the network remains reliable as new generation is
         | added.
         | 
         | Other countries have successfully designed their electricity
         | market incentives so that solar provides synthetic inertia, and
         | manages reactive power, and so that voltage cutoffs (over or
         | under) don't destabilise the grid.
         | 
         | The regulator failed to design a stable network. If they were
         | told to run it in an unstable manner, it is up to them to cover
         | their arse.
        
           | cryptonector wrote:
           | > If they were told to run it in an unstable manner, it is up
           | to them to cover their arse.
           | 
           | With you up to this point. Screw that ass covering. Admit the
           | [policy] error and drop the idiotic policy and get back to
           | normal. No need to ass-cover.
        
         | cryptonector wrote:
         | Nuclear isn't the only issue. It's the lack of sufficient
         | online gas-fired and hydro and battery peak-load capacity.
         | There might not have been enough base-load capacity either, in
         | which case all the peak-load capacity being online would not be
         | enough -- in this case then yes, more nuclear (and coal) would
         | have helped.
        
       | ZeroGravitas wrote:
       | I think a gas plant lying about it's availability was eventually
       | discovered to be a root cause of the South Australia blackout a
       | few years back.
        
       | Rygian wrote:
       | For me the keywords are "capacidad de control de tension
       | insuficiente" (insufficient voltage control capacity) which was
       | already a known issue for years.
       | 
       | REE asked back in 2019/2020 to create a new real-time dispatch
       | mechanism for reactive power, because Spain still relies on fixed
       | dispatch based on emails, phone calls, even snail mail. There was
       | a pilot project in 2022 which was successful (dispatching in less
       | than 5 seconds) but proved to be very expensive, and therefore
       | incompatible with the European Union's directives on creating
       | market-ready regulation processes [1].
       | 
       | [1][ES] https://www.eldiario.es/economia/competencia-reconocio-
       | julio...
        
       | demarq wrote:
       | Wasn't it supposed to be Russia
        
         | b3orn wrote:
         | No, it was supposedly induced atmospheric vibration.
        
       | Dr_Knows wrote:
       | @narag excellent recap. "carefully avoids mentioning nuclear" --
       | the nuclear operator/s decided not to supply electricitiy on that
       | day.
        
       | cryptonector wrote:
       | > REE did not have enough thermal power stations switched on when
       | surge occurred
       | 
       | That's.. that's a miscalculation alright. But it's not likely
       | that it was just a small math error. It must have been policy to
       | live on the edge.
        
       ___________________________________________________________________
       (page generated 2025-06-17 23:01 UTC)