Skip to content

Aurora’s Blackout Analysis

August 14, 2019

By Paul Homewood


h/t Joe Public

The energy research outfit Aurora have published their initial analysis on last week’s blackouts.

There is nothing new, but it is a good summary:


This section is particularly relevant:




They make the point that two plant trips are extremely unlikely, unless connected. However, we need to remember that by 2021 Hornsea windfarm will have capacity of 2.6GW, even before Project Two is begun.

If that lot goes down, the Grid will be in serious trouble.

  1. Joe Public permalink
    August 14, 2019 11:46 am

    Checkout Emily Gosden’s article’s Times’ headline today.

    “National Grid needs more batteries to halt future blackouts, say experts

    National Grid will need to pay for more batteries and gas engines that can compensate for sudden drops in electricity supplies to prevent further blackouts, according to experts.”

    Then, check out what the experts said National Grid actually needed more of:

    Emily also needs reminding that it’s not National Grid who end up paying for those extra measures needed because of intermittents’ grid-destabilising effects, but electricity consumers.

    • Gerry, England permalink
      August 14, 2019 1:41 pm

      As a journalist she probably has no idea how National Grid is funded and so who ends up with the final bill. Goodbye more manufacturing.

    • It doesn't add up... permalink
      August 14, 2019 4:24 pm

      I’m sure the plans for 3GW/?1.5GWh of batteries that National Grid have for the future will be speeded up by these events. I predicted as much in my comments several days ago. More bills for consumers. Really, in line with the Helm Report which recommended that renewables should pay for the cost of their intermittency, the cost should come off the generous CFD and ROC payments to low inertia generators.

  2. August 14, 2019 11:59 am

    Much concern will be expressed, but not for the poor consumer, rather for the Green Gravy Train, which will inevitably continue to construct an ever more unstable system, requiring ever more spending on control mechanisms.

    Expect a deluge of greenwash about the reliability of the UK grid, not mentioning that this reliability was when wind penetration was low. We are now in uncharted territory, with 50+% wind power during storms.

    There needs to a limit set for wind power (around 50% of demand), both for system stability, and to ensure that baseload power stations continue to be economically viable.

    • A C Osborn permalink
      August 14, 2019 7:17 pm

      It is very difficult to continue to be economical when Wind & Solar have first call and subsidies and your plant has to run below optimum or idle.

  3. john cooknell permalink
    August 14, 2019 12:03 pm

    Oversensitive Protection Systems !!!!! Totally wrong.

    The protection systems are necessary to prevent you ending up with a load of melted copper that won’t be good for anything, ie. total disaster.

    The problem is the protection systems (Vector Shift) that have been allowed to be fitted to 60% of the windfarms and solar, do not work to the required connection specification, a triumph of marketing over common sense.

    To quote OFGEM “Because of these events, a series of studies of VS relay operation was undertaken, to identify their effectiveness.”

    “The studies show that VS relays are, in general, ineffective at detecting the existence of an island therefore they do not perform their function as expected.”

    Click to access dc0079_d.pdf

    • Gamecock permalink
      August 14, 2019 6:31 pm

      Agreed. ‘Oversensitive’ seems an exaggeration.

  4. swan101 permalink
    August 14, 2019 12:21 pm

    Reblogged this on ECO-ENERGY DATABASE.

  5. Dave Ward permalink
    August 14, 2019 12:30 pm

    I wonder what they mean by “Gas Engines”? It could refer to Gas Turbines, or reciprocating engines running on gas, rather than diesel or heavy fuel oil. In the case of turbines, even Open Cycle (the least efficient) are unlikely to be able to start from cold, and run up to full power in seconds without risk of thermal damage. Reciprocating engines can, so long as they have oil and coolant heaters with circulation pumps, all of which adds to the cost and uses power all the time (just like wind farms!). They’ll still need to synchronise with the grid before helping it recover, and I suspect that process will take much longer if the frequency is already fluctuating. Batteries & Inverters can respond in milliseconds, but anything of GW scale is going to cost an absolute fortune, and will only be able to help for minutes at best – just enough to get the “Real” generators running.

    Many of us have been saying for years that there is NO substitute for proper spinning reserve – not unless the public are willing to accept more frequent “Demand Management” power cuts, with all that entails in a modern computerised lifestyle. And how will firms like “Upside Energy” help National Grid balance the system when the Internet goes down the pan during a prolonged power failure?

    • Joe Public permalink
      August 14, 2019 1:04 pm

      South Australia, that state known worldwide for its state-wide blackout and its Big South Australian Battery AKA Hornsdale Power Reserve, will shortly be benefitting from 12x Wärtsilä 50DF dual*-fuel reciprocating engines capable of generating approximately 18MW of output each.

      *Natural gas is the primary choice; oil is as standby.

      • Dave Ward permalink
        August 14, 2019 1:23 pm

        “Fast-starting capability The new plant will be capable of reaching full output within minutes”

        The Australian installation is, no doubt, going to be fine to deal with expected (forecast) reductions of supply from AGL Energy’s “Large portfolio of renewable sources, including the nation’s largest wind and solar farms”. But it will be no use if the grid is rapidly losing frequency after the sudden loss of two large generators. In that scenario seconds count – in other words the sort of time it takes for governors to respond on existing, running generators, not ones which have to be started and synchronised….

    • It doesn't add up... permalink
      August 14, 2019 4:19 pm

      Upside Energy is really a research project funded by government on behalf of the greens in BEIS, OFGEM and National Grid who think that we will all leave our EVs connected up as emergency grid supplies and submit to power cuts by remote control in our Smart Meters. They are at the stage of proving that their technology “works” to control many devices simultaneously and promptly – which it does, as their tweet of their performance shows. But finding a handful of partners prepared to put 6MW on the line is not going to cut it when you are asking for several GW of “support”, especially if the duration stretches from under 5 minutes to hours or days during a lull in the wind when we come to depend on it.

  6. Gerry, England permalink
    August 14, 2019 1:44 pm

    So it will come down to who is telling the truth over Little Barford. Unless they fear for their share of the green gravytrain run by the government, I presume RWE will tell the truth about why the plant shut down. The focus then moves totally to wind generation and the problems its variation creates for grid security.

  7. JimW permalink
    August 14, 2019 2:12 pm

    Little Barford was only running at 50% capacity. Once Grid ramped up the wind, the CCGT supply requirement hit 23MW and per contract Barford switched off. Nothing unexpected or strange about this. Grid got it wrong, going for record wind on day Javid visited. Not enough spinning reserve to recover within seconds. Not rocket science!

    • It doesn't add up... permalink
      August 14, 2019 4:10 pm

      Forget the 23GW story. Grid demand was running at 30 GW, and didn’t drop below 23GW until after 11p.m. It’s a complete red herring, and the result of a journalist consulting PR in Germany about what was happening in the UK when they had no clue of how to get hold of anyone who knew the answer. In any case, when power stations shut down because their output is no longer needed, it is done via a gradual ramp down over several minutes – not throwing the main switch – which allows the grid to ensure that there is sufficient generation to meet demand. These power cuts where the result of instantaneous automatic switching of the entire output.

  8. August 14, 2019 2:31 pm

    Reblogged this on Climate-

  9. August 14, 2019 3:09 pm

    I am almost tempted to write to my MP and ask him to find out from DBEIS how the government intends to maintain grid stability and prevent the lights from going out when, with all the new wind farms under construction and in planning come into operation, the amount of asynchronous generation connected to the transmission gird and distribution networks approaches and then exceeds the demand.

    • Bertie permalink
      August 14, 2019 4:38 pm

      Why only “tempted”?

    • wheewiz permalink
      August 14, 2019 4:49 pm

      MP’s are the problem: they will never have a solution to the problems they created. Lock ’em up !

    • yonason permalink
      August 14, 2019 4:57 pm

      If he’s one of those who doesn’t understand the problem, he may not even understand your question.

  10. Rowland P permalink
    August 14, 2019 3:11 pm

    Just a minute! Isn’t there going to be a “smart grid” to solve all our electricity supply problems?

  11. It doesn't add up... permalink
    August 14, 2019 4:04 pm

    Well, it’s a good summary… apart from the bits they got wrong, like the fact that it was Hornsea that tripped first. I have emailed them with some details that I don’t think they had looked at, and it will be interesting to see what response I get.

  12. Ian Cook permalink
    August 14, 2019 4:04 pm

    Confused. Soon after the event, NG were quoted as saying the shutdown of the St Neots (as I think it was deemed at the time) gas plant was planned and automatic, based on supply and demand. So that wouldn’t be two going off line at the same time, it would be one unexpected. My suspicion is that the Hornsea shutdown was deliberate, probably due to the high levels of still fluctuating supply from wind farms. Deliberate, but panicked to protect the system from a more massive shutdown. And feeding misinformation through ‘other sources’ would be how you hide the above and maintain the myth that wind power is wonderful, just like all those fairies at the bottom of the garden.

    • It doesn't add up... permalink
      August 14, 2019 4:35 pm

      I’ve not seen any report that NG claimed that the Little Barford shutdown was planned and automatic. If you have, can you point to them? It was AFAIK a rumour started by Bloomberg who asked the man in the moon (well, RWE PR in Germany) what was happening in a power station somewhere on Pluto (well, “somewhere in England”), who gave a holding reply that was possibly misreported but certainly simply nonsense in the light of the facts.

      I think the Hornsea shutdown was quite accidental. Trying to think of possible causes, I have come up with the idea of a freak wave washing over part of the offshore platform where they have the substation transformer and connection to the cable to shore which managed to cause a short circuit that led to the whole thing being shut down automatically. That may be completely wrong, but I am inclined to rule out component failure, as a failed component has to be replaced, which is not something you do in a couple of hours, especially if it occurred offshore.

  13. August 14, 2019 4:20 pm

    Hornsea 3 is waiting for final approval and Hornsea 4 is being promoted as a new project.

  14. August 14, 2019 4:27 pm

    Generator outputs during the event are not published yet.

    • It doesn't add up... permalink
      August 14, 2019 4:55 pm

      We should have those on Friday, but only at half hour resolution. I am steadily collecting the daily reports for Hornsea and Little Barford so we can see what recent typical operations look like, For me, one of the questions that comes out of this is the lack of detailed information available to the public rather than industry participants. Compare with the kind of data that comes out of AEMO in Australia, which allows the output from every major generator and throughput on interstate connectors to be monitored at 5 minute resolution by the general public in real time à la Gridwatch at the aggregated level. Industry insiders there have 4 second data, which is pretty good given how big of an area AEMO covers (everywhere except Western Australia, remote parts of Northern Territories and small offshore islands that aren’t grid connected: it does include Tasmania and the Basslink for example).

      We do eventually get 1 second frequency data, which allows you to search for untoward events. The latest posted by National grid is for April – it’s an 84MB file with about 2.6 million lines of data (3600x24x30), and you need the right software to be able to even look at it. I did so, and found one event with a nadir at 49.611Hz, and another with a nadir of 49.700Hz. They haven’t published the later files, but we know there were at least three other events since then where the nadir was somewhere between 49.5-49.6Hz: 49.5Hz is the statutory minimum that triggers OFGEM investigations. So we are getting “near misses” with increasing frequency over the summer (when the risk is probably highest, due to high proportions of renewables in the mix caused by much lower underlying demand, but overnight outside summer with high wind can also prove risky).

    • It doesn't add up... permalink
      August 15, 2019 12:23 am

      I’ve taken a look at Little Barford over 1st-7th August. It spent a good part of that time at close to full load. You can pick various different modes of operation. There are days when it backs off slightly and increases again in a gentle curve that might help offset the daily solar peak. There are others when its output fluctuations are more jagged, as if balancing against more variable wind at the margin, or providing other ancillary services. Over several nights, the station is reduced to half output after 23:00 and resumes full output ahead of the morning demand peak, ramping up after 6 a.m.. On the 6th, the overnight reduction from the previous evening was a complete shutdown, with a cold start on the 7th. The fastest rate of change either up or down is 12-13MW per minute over the course of half an hour, but more commonly under 10MW/minute, with it taking over an hour to shut down or start up in a directed manner to or from full output. There are some other upward and downward spikes in output that might take considerable investigation of activity on the rest of the grid to explain, but they appear to be the result of grid instruction, as ramp rates remain contained.

      Hornsea wind farm is divided into three reporting blocks. Block 1 is very small, never reporting more that 27MW, and being shut down entirely on the 5th and 6th. Block 2 was also shut down on the 1st and 2nd, leaving Block 3 to provide a steady output of just under 300MW on the 1st. With all three operational, the highest output was just over 700MW. Block 3 was shut down in the early evening of the 7th. Installation and tie in work is obviously still proceeding. Makes you think that literally a monkey wrench could have dropped or been blown off by the wind from where it was lodged doing a tie in to create a short.

      We can await the detail on the critical periods with interest.

  15. Tom O permalink
    August 14, 2019 4:41 pm

    I thank the UK and South Australia for showing us the future. It is sad that we can’t learn from it, though. Now picture what will happen if the same thing happens in the UK, but more wide spread, when it is experiencing a prolonged cold period. A nice black startup lasting days while the temperature is hovering at -10c. Hmmm, not a pretty picture, is it?

  16. August 14, 2019 4:56 pm

    12pm Radio Humberside News
    “Jo Swinson is visiting the Hull Siemens windturbine factory”
    they play a clip of her saying
    “currently the UK has 30% RENEWABLE energy, we think that could easily be 80%”

    Is she mixing up just electricity with renewable with ALL energy
    and of renewables isn’t wind actually quite small on average ?

  17. It doesn't add up... permalink
    August 14, 2019 5:05 pm

    Some positive news. I have heard back from Aurora that they are looking at my submission to them, and will get back to me after they have analysed it.

  18. August 14, 2019 6:04 pm

    What kind of liars say there is a “ban” on onshore wind ?
    Emily cutNpastes the PR into the Times as usual

    • It doesn't add up... permalink
      August 14, 2019 7:22 pm

      That’s a request for more subsidies. One thing that truly shocked me today was finding out just how misguided the general public are on matters energy. From a YouGov survey:

      92% of “left wingers” and 78% of “right wingers” believe that we should pursue green energy. The Malthusians are also highly popular. Where is our modern day Jonathan Swift with a Modest Proposal, and talk of moonbeams being stored in cucumbers?

      • August 14, 2019 10:41 pm

        It all depends what questions are asked. If they asked whether electricity should cost a lot more and be less reliable, the results might be different.

      • August 14, 2019 11:53 pm

        does straight html work ?

      • August 14, 2019 11:55 pm

        bbcode ?

      • August 14, 2019 11:58 pm

        @IDAU Look we know surveys are BS
        And we know that just cos you present something in a complex sciency way, that doesn’t make it true either

    • Joe Public permalink
      August 16, 2019 7:22 pm

      SG: “What kind of liars say there is a “ban” on onshore wind?”

      Harrabin, for starters.

      He was even forced to correct his ‘correction’.

  19. Chris Morris permalink
    August 14, 2019 7:15 pm

    What is the actual situation? The annotated graph shows the losses of GT station loss as rating and the wind farm at rating. I thought little Barford was generating at less than half load, and the windfarm at about 1GW.
    The graph does not make sense as drawn.

    • August 14, 2019 9:47 pm

      As I gather, the two dropped about 5% of capacity,ie about 1.5GW. The graph is Hz

      • Chris Morris permalink
        August 15, 2019 7:02 am

        The size of that drop is not in dispute Paul. The top graph here says the main contributor was the Little Barford
        Several headposts ago, the information was it only had one unit on, so was generating at just 300MW. If that is correct, the annotations on the graph are wrong and the windfarm came out first.

      • It doesn't add up... permalink
        August 15, 2019 10:24 am

        I believe the annotations on the graph are wrong and I am awaiting both Aurora’s response to my analysis, and the data on actual generation such as it is. I’ve looked at how Little Barford operated over 1-7 August and described that above. It’s possible that it was operating at close to 700MW, with the apparent lower loss of CCGT generation at the aggregate level being the result of other CCGT stations raising output to compensate.

  20. Ian Wilson permalink
    August 14, 2019 7:24 pm

    In the 1960s stand-by gas turbine power stations were built with 8 Rolls-Royce Avons each. They were probably quite costly to run but could be brought into use very quickly I wonder why they were scrapped.
    As an aside, the 8 engines were grouped in two quartets, powering large turbo-gnerators which were ‘handed. I was told that for the first two sites the main contractor delivered both right-hand sets to one site and the two left-hand ones to the other.

  21. john cooknell permalink
    August 14, 2019 7:49 pm

    The Electric Supply to UK has had the benefit of unachieveable political strategy it was de-nationalised, de-regulated, de-skilled and then the main power stations were shut down.

    Then the Green political blob took over, costs are rising exponentially, reliability of supply is deteriorating, at great cost a useless “smart grid” was implemented, and its all a bit of a mess.

    Just shows how valuable and knowledgeable politicians are,and the Green blob media say it is a success. You could not make it up.

  22. August 14, 2019 8:07 pm

    The single line diagram of Hornsea is described here –

    The point of common-mode failure for the two 58 turbine strings isn’t obvious?

    • It doesn't add up... permalink
      August 14, 2019 8:53 pm

      I see two buses at 220V – one offshore via the interlink cables, and one onshore. Any fault/short on either bus would propagate across the whole system.

  23. Joe Smith permalink
    August 14, 2019 8:21 pm

    A proper electrical engineering explanation, missing from the Meejah guff we have to put up with nowadays
    We said renewables would cause grid problems, this is only the start
    Who will be held to account?

  24. August 14, 2019 10:27 pm

    It would have been run with one turbine string on each of the two 220 KV busbars. The single line diagram shows a circuit breaker for coupling the busbars. It doesn’t show the isolators used to select which busbar a string is connected to.

  25. GeoffB permalink
    August 14, 2019 11:00 pm

    St Greta will fix it. Don’t worry, she is the Messiah.

    • It doesn't add up... permalink
      August 15, 2019 12:40 am

      It seems quite windy tonight. I suspect her blue bucket may be put to good use, and probably has the real content of her message. At least it’s not a message in a (plastic) bottle, even if it is supposed to be an SOS to the world.

  26. August 15, 2019 1:42 am

    Should get an idea of generation outputs from HOWAO-3 and LBAR-1 at the time of the frequency excursion when this updates (data presently available up to 7/8/19) –

    The event was near the end of the 30 min period so should provide a reasonable estimate.

  27. August 15, 2019 2:32 am

    Regarding the 220 KV busbars and circuit breakers, the protection should ensure that a fault on one busbar does not affect the other. (and it’s connected generation).

    Little Barford, as far as I know, comprises two gas turbines and a common heat recovery steam turbine. It should be possible to lose one gas turbine without having to shut down the steam turbine, but a fault with the steam turbine might require both gas turbines to be taken off.

  28. August 15, 2019 2:48 am

    Paul, in a previous post, the construction of a huge battery storage facility was discussed –

    If it turns out that a similar battery will be required just to support Hornsea then the overall cost is going to be even higher.

    • It doesn't add up... permalink
      August 15, 2019 1:16 pm

      I see I made some prescient comments on that story.

  29. August 15, 2019 3:40 am

    “One of the most common causes of electricity network faults is lightning and we know that there was significant lightning activity in the East of England on Friday evening. One wonders if that might have had any influence on what happened on the electricity system.”

    I’d put my money on lightning and some combination of undervoltage and frequency…

    • It doesn't add up... permalink
      August 15, 2019 1:30 pm

      That’s actually a good write-up. One of my earliest thought was the possibility of a lightning strike , but I discounted it when I ran a check on weather stations in the Humber area which didn’t suggest any storms. By the time I thought I should have checked the Meteoradar presentation that shows rain and lightning in recent hours it was too late. Perhaps they have some archive data, even covering out in the North Sea as well as the onshore sectors at risk.

      I note that he quotes the output from Hornsea at 790MW and from Little Barford at 660MW. Perhaps he had access to data we don’t, though he doesn’t make any attempt to suggest the sequence of events.

  30. August 15, 2019 4:42 pm

    The latest

    “Batteries can produce synthetic inertia, but this still take c (sic) 0.5 seconds to respond so having synchronised units providing inertia instantaneously is still a must for operating the power system safely.”

Comments are closed.

%d bloggers like this: