rss
23

Storm of controversy erupts over AEMO blackout report

Print Friendly

Photo: EPA/Sven Koopmann/PHOTOKA

 

Another storm of controversy about the role of wind energy is certain to erupt after the latest report about the state-wide blackout in South Australia by the Australian Energy Market Operator.

In its second update, AEMO has pointed the finger at settings on certain wind farms and fossil fuel generators in the events immediately before and after the state-wide outage last month, but the handling of the report has also raised questions about the actions of the market operator itself – both before and after the event.

The report dismisses suggestions – mostly from the Coalition and mainstream media – that it was the intermittent nature of wind energy that was the cause of the blackout. But it also underlines the failure of the market operator to make any preparations for the storm that it could obviously see spreading across the state.

The AEMO makes clear that it was major voltage disturbances – six in 80 seconds – caused by the collapse of three major transmission lines that led to the blackout. “Five transmission line faults, resulting in six voltage disturbances on the network, led to the SA region black system,” it writes.

But – not for the first time – AEMO’s press release and executive summary differs in emphasis to the detailed report, and focuses on the role of the so-called “self protect mechanisms” in wind farms rather than major voltage collapse that followed the collapse of the transmission lines.

Even though these self protect mechanisms are just a matter of software and are easily fixed, AEMO’s emphasis has horrified many in the wind industry, who suggest that the market operator is deliberately allowing wind to be blamed even though its report highlights a collapse of voltage that could have been the main cause of the outage. They also point to basic errors in its report, and its failure to take any preventative action as the storms approached.

“What we see in this report is a concerted effort to focus solely on the wind farms as if the system is perfect and the market perfect,” said one wind farm operator. “I think we are now at war with the system operator.”

In the summary, AEMO points to the role of wind farm self protect mechanisms as the straw that broke the grid’s back, noting that many wind farms were designed to switch off after riding through as few as two voltage events. (This is disputed by some wind farms, who suggest their limit was actually three).

These so-called fault ride-through limits relate to factory settings – what South Australia energy minister Tom Koutsantonis describes as a “software issue” – and these have now been fixed at most of the wind farms. Those wind facilities without such limits rode through the events without any problems until the inter-connector closed.

AEMO was also careful to point out that the role of wind energy in the blackout had nothing to do with its “intermittency” or its ability to generate through high wind events, as many wind critics had suggested in the aftermath of the blackout. (See our separate story “Blackout report blows away big myths about role of wind energy”).

“The most well-known characteristic of wind power, variation of output with wind strength (often termed ‘intermittency’), was not a material factor in the events of 28 September 2016,” it notes in the report.

However, the point of contention lies around AEMO’s assertion that it was the role of the self protect mechanisms and the loss of 445MW of output that caused the Heywood inter-connector to overload and trip, separating South Australia from the rest of the grid and precipitating the system-wide blackout.

Energy experts point to the detail of the report and the collapse of voltage that occurred before two of the biggest wind farms switched off, suggesting that the system was unstable and would have tripped anyway.

As the AEMO report noted, voltage after the collapse of the last transmission line fell to 40 per cent of the line’s rating. Given that most equipment, as the AEMO report notes, is designed to self protect if voltage falls by more than 10 per cent, they wonder why the market operator would expect a generator should ride through voltage collapse of such magnitude.

They point to this graph below, showing the voltage change immediately after a failed attempt to re-connect the third major transmission line to have collapsed after the storms. The energy experts say that this shows a massive voltage collapse that was happening anyway.

“To suggest that loss of generation is the final straw is misleading,” said one. “When you see this graph it is clear that the whole system had become unstable, and nothing was going to ride through that voltage collapse.”

aemo voltage change

“No power grid in the world is designed to manage the rapid consecutive collapse of three major transmission lines like the SA system sustained on 28 September,” said Kane Thornton, the CEO of the Clean Energy Council.

The wind industry is angry and frustrated that while the exact cause is unlikely to be known until the final report is completed in six months time, wind energy has been left to take the rap in the interim – and for a technical issue that is easily fixed and may not even have been as significant a factor as AEMO is making out.

And while the report does make clear that the issues with the wind farms are technical, and easily fixed, there is no doubt that the political and ideological debate will be fierce, and won’t be confined to the facts.

But the report also points to two other major problems.

The first is that fossil fuels are not a panacea as some would wish. The diesel and gas generators paid handsomely to provide black start services to the state both failed, one within 15 seconds, causing the blackout to last much longer than it would have done otherwise. AEMO refuses to name the failed generators, citing “confidentiality” agreements.

The three diesel generators that should have provided power to Port Lincoln also failed – two tripped almost immediately and the third had to be shut down.

And, as the report also notes, this is not the first time that the Heywood interconnector has been separated from the state. Three of the previous disconnections occurred when the main coal generator, the now closed Northern brown coal power station, tripped and caused the system to collapse. The other occurred when bushfires in Victoria caused multiple transmission line failures.

PreviousBlackouts copy

The other issue is one of culture. As grid operators in china, the US and Europe have noted on many occasions, the big issues with the transition to renewables are not really technical, but cultural. This will be a question asked of the AEMO, and an independent assessment of its own actions is now called for.

Its report basically admits that the market operator had no Plan B for the storm. Some suggest it didn’t even have a Plan A.

It made no provision for extra back-up generation or for locally based ancillary services that could have stabilised the grid at its point of crisis and avoided the black-out. These ancillary services normally come from gas generators, but could in the future be provided by battery storage.

And it didn’t scale down the amount of power coming through the interconnector. It meant, that like a central bank with interest rates close to zero, it had no levers left to pull in case something went wrong. Its report suggests it was more worried about price impacts than system security.

And it seems that AEMO simply didn’t believe that the system was at risk. That approach seems extraordinary given its own admission that the winds of 120kmh were forecast, more than the rating of some wind farms in the state.

“Given the available evidence in advance of the storm hitting, one has to wonder why on Earth they (AEMO) didn’t re-classify the risks of the loss of one or more transmission lines or generators as a credible contingency,” says Andrew Stock, from the Climate Council and a former senior executive with coal and gas generator Origin Energy

“It would have been very unusual for senior management in a very high risk operation, in the face of evidence of a forthcoming storm, not to take precautionary actions to pre-emptively mitigate perceived risks.

“It’s a bit like driving a performance car flat out into a storm, which most people would think would have been foolhardy. On what basis did they form those judgments?

“In my experience it is very unusual for senior management in a very high risk operation, in the face of evidence of a forthcoming storm, not to take precautionary actions to pre-emptively mitigate perceived risks.”

The CEC’s Thornton said no evidence has been provided to show that the system would have remained up and running if wind farms in the state had not tripped off to protect themselves in an unsafe electrical environment.

“There is obviously a lot to learn from all facets of the event, from the operation of the interconnector to the robustness of the transmission network, the re-start procedures, the planning prior to the extreme weather and the operation of power generators,” he said.

“The wind industry is committed to working with AEMO to look at current standards and operating procedures and whether these need to be refined.”

  

  • john.boland

    this article has hit the nail on the head. It is AEMO’s actions – and inaction – that have to be questioned. The trouble is that how are we to get that interrogation with AEMO performing the investigation?

    • Nick

      Are you suggesting AEMO should rely more heavily on natural gas during inclement weather? What’s the cutoff to determine when to preemptively turn off the wind plants in favor of the gas ones?

      • neroden

        No, he’s suggesting that AEMO pay more attention to transmission lines falling over and plan for it.

  • Alan S

    The silver lining to this storm cloud is that the technically illiterate anti-renewables lobbyists may hesitate before jumping to their usual simplistic conclusion.

    • Jason

      I’m guessing you haven’t met many technically illiterate anti-renewables lobbyists.

      A heady mix of determination and Dunning-Kruger.

  • phred01

    Libs are now changing tact Unfortunately the way grid connect works is there must a supply to synchronize to before the system output any power …….. that is the way renewables were designed to operate; in other words as slaves. This is true solar as well wind for obvious safety reasons. So loose contact with the master, they shut down.

  • David Osmond

    Worth noting that the March 2005 trip of Northern Power station occurred after the coal power station was unable to ride through a 2nd voltage dip caused by short circuit at a substation. http://www.neca.com.au/Files/NECA_Report_FINAL_for_14_March_2005.pdf

    • Analitik

      Yes and they were heavily fined for not meeting performance specifications.

      The AEMO report also contends

      Investigations to date indicate that information on the control system involved and its settings was not included in the models of wind turbine operation provided to AEMO during NEM registration processes prior to connection of the wind farms.

      It will be interesting to see if the wind farms that went offline will be similarly judged and treated as Northern

      https://www.aemo.com.au/Media-Centre/-/media/9027D5FB69294D408E4089249F38A36D.ashx

      • AllanO

        Analatik I suggest that the information referred to here was “not included” because AEMO never thought to ask for it in the first place at the time of registration. If they had recognised that ability to tolerate multiple consecutive ride-throughs was dependent on windfarm control system settings then they surely would have requested these details.

        You are in effect contending that AEMO mandated some requirement for this setting but that every single windfarm operator in the NEM chose to ignore it.

        I think it’s more likely the case that this feature of windfarm control systems was not something that AEMO (and possibly not the operators) had ever realised might be an issue. A technical oversight in the shift to a different generation mix.

        Of course if you can point to a pre-existing performance specification dealing with this particular item then I’ll happily stand corrected.

  • Steve Fuller

    If the transition to more and more renewables is to proceed as planned by all political parties, then we need to use events to identify what the technical issues are and plan accordingly.
    These are engineering not political issues. We simply need to engineer a system that is robust, secure and with the flexibility to cope with an increasing percentage of electricity supplied by renewables.
    The renewables revolution is already well underway and we need to have a system engineered to cope – get on with it AEMO.

    • Analitik

      What do you suggest the AEMO do?

      They are getting the wind farm operators to increase the number of line events they can ride through before shutting off but if the improvement for some farms is marginal, then they can only increase the credible risk factor for storms which makes load shedding (selective blackouts) likely to occur earlier or else set pricing for gas turbines to kick in during storms which will drive up electricity costs.

      • Peter F

        They can also make sure that there is more battery or gas actually on line as spinning reserves or perhaps have a fast response demand management system set up

      • Steve Fuller

        You are considering a small issue within a complex evolving system. The whole system needs to be re-engineered to meet the demands of the 21st century and the distributed clean energy revolution.
        We are being held back by 20th century paradigms that entrench the role of dirty energy.
        We need to unleash our exciting innovative energy revolutionaries to develop a sensible plan to make the transition reality asap.

        • Shadeburst

          I may be a climate sceptic but I would love alternative energy systems to work. I don’t enjoy inhaling the particulate matter of FF combustion any more than you do. Describing wind as exciting, innovative and revolutionary doesn’t hit my rational buttons. All the aerospace engineers that NASA fired have been working on wind turbine efficiency with only tiny improvements. Long, long ago Holland used to rely on windmills to keep the country above water, with incomplete success. Now it’s electric pumps. Wind turbine power fits into the category of nice try, but no cigar. KiteGen may be a better solution, cheaper and more reliable over a wider range of wind speeds.

  • Chris Fraser

    The situation with trip ride-throughs intrigues non power engineers (like me) The mechanism to trip out appears either to protect assets downstream in the reticulation system, or the generator itself. In either case, it appears that a capacitor or battery between the generator and the grid would act like a shield for them both, and enable more conscientious decision making about what to do with generator output. Less of this automatic, cascading, emergency-induced panic.

  • I’m sorry Giles but I have to respect AEMO’s view ahead of your own. These guys are experts in this field. I’m not sure of your expertise in electricity networks, which requires a deep understanding. If you really have that expertise then I apologise.

    • RobS

      Giles is basing his opinion on the AEMO report, the body of it, the issue is the executive summary which the media and right wingers will use for their talking points is quite different in its emphasis than the body of the report itself. The executive summary places far more emphasis on the impact of wind farms on the blackout whilst the detail of the report makes clear that this was a transmission and back up failure. Multiple transmission lines failed nearly simultaneously, it would have made no difference whether wind, solar, coal or nuclear power generators were at the other end of those lines, once they are lying on the ground power transmission ceased. Other generators, including multiple diesel, natural gas and yes some wind generators were then taken off line by software based system control processes because the scale of the failure caused too much voltage variation for them to integrate into the grid.

      Wind farms did not fail because they were wind farms, a few failed alongside all of the backup fossil fuel generators as a result of an unprecedented failure of the state’s transmission lines.

      • Nick

        Your summary of the report is wrong. Section 3.4.3 clearly states all thermal generation remained online until SA separated.

        The report also does not state the wind plants were isolated from the grid. While some lines were on the ground, the subject wind plants remained connected through unfaulted lines.

        Your analysis is incorrect.

  • Malcolm M

    Was the Heywood inter-connector upgrade partially responsible for the system black? Before the upgrade, low frequency would have initiated the load shedding control system. After the upgrade, frequency was maintained by short term high flows on the inter-connector, until it tripped through overload. Is there now insufficient low frequency signal to initiate the load shedding control system? Or is the load-shed system simply too slow? Instead there needs to be an additional automated load shedding system controlled from Heywood should flows into SA exceed a threshold. Such a system would need to operate with a response time of under 1 second, so load shedding occurs before the inter-connector trips out. It is much quicker to recover from load shedding than a full system black, because most of the generators would remain on-line.

  • Nick

    Giles, you say “Energy experts point to the detail of the report and the collapse of voltage that occurred before two of the biggest wind farms switched off, suggesting that the system was unstable and would have tripped anyway.”

    This seems to be in error as the biggest wind farms switched off about 16:18:15 when the voltage was about 1.0 pu on the figure you included. The voltage collapse then followed those plant trips. Can you please address this?

  • Adam J

    AEMO’s updated preliminary report states:

    “AEMO then worked with each of the operators of these wind farms and determined that their ‘voltage ride-through’ settings were set to disconnect or reduce turbine output when between three and six ‘voltage ride-through’ events were detected within a given timeframe. Investigations to date indicate that information on the control system involved and its settings was not included in the models of wind turbine operation provided to AEMO during NEM registration processes prior to connection of the wind farms.”

    In this case, I would ask why didn’t AEMO ask for this information during the wind farm connection process? Given that these wind farms are here today means that AEMO approved them and managed their connection to the grid. To blame the system outage on the under voltage settings is dull and if AEMO did their job correctly, they would have reviewed the protection stud, controller model and stability study for each wind farm in response to the contingency scenarios put forward by AEMO. I doubt that the events of the September 28 were nominated by AEMO as a credible contingency for the wind farm manufacturers to complete their protection and stability studies on. The technical solution is easy because now AEMO have a real life credible contingency to base any future protection and/or stability study on. Generator protection relays are there to protect the generator, whether it be synchronous, wind turbine or otherwise. The job of managing settings, identifying credible contingencies to base these on is the job of AEMO. I am not convinced that directing the blame to the wind turbines protection settings is a good one. Unfortunately, this links to your comments about this being more about the culture of business in Australia. South Australia has been brave in its push to increase renewable energy generation. It will take braver souls to keep going with it and not fall victim to the witch hunt mentality.

  • Phil

    I’m surprised the big Issues are not being dealt with re this event. But then again after witnessing the ineptitude of this industry in the last 10 years or so nothing surprises me

    The big Issues are ……….

    1) Where is the SLA ( service level agreement ) for Uptime. This is standard practise for most if not all critical networks. Is it 99% or 99.9999% uptime. It’s never mentioned.

    Large Storm events are treated as rare events ( a force majeuer in business terms) and exempt it seems from any planning for uptime. Storm events are normal , if an entire state goes out for a significant period then that is clear evidence of a poor network design or poor maintenance , or both. Or perhaps the Grid providers have to come clean and tell the consumer – “look we MAY go out in a big storm for up to 12 hours , most of the poles and wires are above ground and we cant do anything about it”

    2) Many consumers and businesses are paying for extra “greenpower” . Where is this money going and if it (greenpower) is weakening the grid in ANY way , shape , or form , then surely some class action can be taken against the perpetrators who put the greenpower in.

    Personally i believe that this industry will use events such as this to further gold plate an archaic network design full of people infighting to get the best profits for themselves rather than holistically setting SLA’s for the ENTIRE network and meeting them

    I hope they do because the additonal costs forced onto consumers will solve the entire grid problem as they leave it in droves.

  • Les Johnston

    The AEMO report has to be read with consideration of what it omits to include not simply what it does include. The report author has to protect its reputation and claimed competence first and foremost. How often does a Government regulator admit to failing to do its job competently? Taking a more critical assessment of the AEMO report adds weight to the claims presented by the claims above. “software problems” seems a little too simplistic.