sign up log in
Want to go ad-free? Find out how, here.

Report ordered by the Government finds that the power cuts on one of the coldest nights of the year 'need not have happened' as there was enough load to keep the lights on

Public Policy / news
Report ordered by the Government finds that the power cuts on one of the coldest nights of the year 'need not have happened' as there was enough load to keep the lights on

A report for the Government on the power cuts that left 34,000 households without electricity on one of the coldest nights of the year has found there was no need for any disconnections.

Energy and Resources Minister Megan Woods on Thursday released the report of the Investigation into electricity supply interruptions of 9 August 2021, headed by former Cabinet Minister Pete Hodgson and assisted by specialist technical advisor Erik Westergaard.

"The investigation sets out a number of key findings - crucially that turning off the power for more than 34,000 households for up to two hours, simply need not have happened, and there was enough discretionary load to keep the lights on," Woods said.

"The report also found that Transpower, as the electricity system operator, had inadequate visibility of the extent of that discretionary load and that the Electricity Authority must review and strengthen its oversight of the system operator."

The report states that "forced disconnection" of household electricity "was entirely avoidable".

"...Turning off any householder’s electricity, apart from their hot water cylinder, simply need not have happened," the report says. 

"...Importantly, we make the finding that no household need have suffered a power cut even if the system operator had not deployed the demand allocation notice. We find that there was no need to issue that notice, and that the system operator did so in order to further honour an equity rule embedded in the electricity code. We find that rule to be ill-conceived, and in need of prompt revision."

Specifically, rule 7 in Schedule 8.3, Technical Code B of the Code states: “To the extent practicable, the system operator must use reasonable endeavours when instructing the electrical disconnection of demand to ensure equity between connected asset owners.”

The report says that In practice this means that Transpower as system operator ought not ask a particular electricity distribution business (EDB) to do more than its fair share.

"However, during the grid emergency some EDBs had spare discretionary load available to offer (typically hot water cylinders that could be turned off using ripple control), and whereas others had none."

Woods said she was writing to the chairs of Transpower and the Electricity Authority requesting they consider the recommendations in the report that relate to them and provide her with quarterly updates on progress.

“I do not want households to be put in this situation again. Kiwis deserve more. By implementing the report’s 18 recommendations I believe we will be better placed in future.

“It is also clear that the market requires much greater demand side participation if greater electrification and decarbonisation is to happen. In future better use of hot water control, known as ripple control, and other technologies are likely to be key to this.

“The investigation’s other findings included praise for the way system operator staff acted during a challenging evening, and for the responsiveness of electricity distribution businesses to the event.

“I would like to thank the investigation team for its comprehensive report. It drew on and analysed data and information from Transpower, the Electricity Authority, and other key information."

As part of the investigation more than 25 interviews were conducted with a range of market participants and key stakeholders.

Here are the key findings in the report:

Our findings are summarised as: 1. We find that forced disconnection of household electricity was entirely avoidable. We find that the demand side had enough discretionary load to maintain the system, but that the system operator had inadequate visibility or up to date awareness of that resource. Turning off any householder’s electricity, apart from their hot water cylinder, simply need not have happened. Ensuring that the system operator has accurate real time awareness of the size of each electricity distribution businesses’ discretionary load is a central recommendation.

2. Importantly, we make the finding that no household need have suffered a power cut even if the system operator had not deployed the demand allocation notice. We find that there was no need to issue that notice, and that the system operator did so in order to further honour an equity rule embedded in the electricity code. We find that rule to be ill-conceived, and in need of prompt revision.

3. The demand allocation notice was remarkably faulty. It required electricity distribution businesses (EDBs) to limit their load illogically. Its issuance caused considerable confusion. Transpower apologised in the days that followed. Trust between EDBs and Transpower has been damaged and will need to be restored.

4. We nonetheless find that the system operator staff acted capably and professionally during a challenging evening. They got us through, notwithstanding inadequate information and a faulty allocation tool. Their skill and commitment avoided the next stage of system defence, known as Automatic Under Frequency Load Shedding (AUFLS), which would have seen 16 per cent of New Zealand’s electricity load shed, automatically.

5. EDBs were also for the most part very responsive and engaged throughout the evening. Generators maximised their output where practicable. There was a lot of cooperation and goodwill evident throughout the system, and throughout the event, though we identify a touch of complacency too.

6. Transpower was criticised for inaccurately under-forecasting what turned out to be a record load. We do not support that conclusion, noting that competing forecasters were even less accurate. Nor do we place any importance on 9 August being a new record level of demand; it was the second such record this winter. We do however propose a rule change regarding wind energy estimation.

7. We investigated whether Genesis’ third Rankine at Huntly or Contact’s Taranaki Combined Cycle plant should have run. Claims of an undesirable trading situation and Code breaches have been lodged, and will be investigated by the Electricity Authority (EA). That scrutiny should be untainted by any detailed findings from us.

8. Accordingly, we have constrained our observations; forecast prices seemed to provide insufficient incentive to restart either plant. Some statements from generators immediately afterward were unhelpful. In coming years slow-start thermal plants will exit the system altogether.

9. We have examined planned generation outages and find nothing exceptional.

10. Some generation plant – Tokaanu and Waipipi – sharply reduced output at crucial times during the evening. Though that was an unusual coincidence, and had a material effect, we do not find the level of failure exceptional.

11. We find that the electricity system’s arrangements for generation shortfalls that may last for part of a day are very much less mature than arrangements for instantaneous and short outages (spinning reserve), and that that immaturity was at play on 9 August. We call this issue ‘managing multi-hour shortfalls’. We think it will become an increasingly important issue to address.

12. We find that while the market has matured steadily over 25 years, it has not yet matured sufficiently or with sufficient alacrity, though we are pleased to acknowledge the change to real time pricing in the market a year from now. Of all the market making or deepening options, we think it is time to revisit a cap market to support risk management.

13. Our review of the demand side was fruitful. Our first finding is our most important. There is significant opportunity to harness the discretionary load under ripple control, which is a long-standing demand side innovation, now at risk of decay. There is also growing need to harness new demand side opportunities – electric vehicles (EV), smart appliances etc.

14. We sense the demand side will shortly develop as a more important, accessible, digitised player, along with battery technology, when it comes to the increasingly important task of shifting load from peak to off-peak. We have some suggestions, principally around one or more ‘multi-hour shortfall’ products. We note a very varied contribution to demand side activity from large directly connected users, and recommend how that too might be addressed.

15. There is some urgency in progressing these things. The likely changes to transmission pricing will reduce the value of ripple control to EDBs, and therefore reduce the case for continued investment. We envisage ripple control and replacement technologies being at the heart of a transition to a richer demand side participation in the market over the next decade. Improved efficiency and improved security are both possible if done innovatively.

16. A widely held view is that the nature and standard of communication in the system is neither modern nor responsive. In particular, Transpower has room for improvement. We acknowledge that improved communication was reportedly very evident in a grid emergency event the following week. Elsewhere in the system, the ownership of customer relationships is contested under conditions of stress. Other analyses of some of these issues have already been made public. We endorse findings to date, and offer some additional ones.

17. Medically dependent customers are a sensitive issue among industry participants, especially retailers. Because all households suffer unplanned electricity outages, the most important thing for a medically dependent customer is a back-up plan, developed with their clinician. The most useful piece of information for a medically dependent customer, and for everyone, is the likely length of the outage. That was mostly unavailable on the night of 9 August.

18. The EA must review and strengthen its oversight of the system operator, and by implication Transpower. We find that self-assessment, whilst informative and useful, is inadequate. A regulator and a statutory monopoly have an unusual relationship, which must be determinedly developed at more than one level. Thus, Transpower must both be challenged to be a fully compliant and responsive player, and also be supported to continue innovation and leadership in our globally unique system.

19. We asked whether our system’s security settings are appropriate or whether they should be strengthened. The context includes the greater electrification of the economy, increased reliance on a continuous electricity supply, more adverse effects from climate change, and more intermittency occasioned from a transition away from fossil fuelled thermal generation towards new renewables. This question requires more analysis than we can command. What we can do is reiterate a key finding of our review, namely, that the market requires much greater demand side participation. We believe that this will be essential if goals of greater electrification and decarbonisation are to be achieved.

20. There is room for improvement in standard setting for appliance monitoring and control. Standards should ensure there is the ability for meters or phones to communicate with appliances, or EVs. We consider that both the EA and the Ministry of Business, Innovation and Employment (MBIE) have a role in proactively identifying opportunities where standard setting would be in the public interest.

21. The future will be characterised by reducing marginal costs for new generation, especially solar, contrary to the inexorably increasing marginal costs of recent decades. Rising social and political demands to progress decarbonising the economy will dominate. Various digital technologies will enable innovation, though only if the regulatory environment is conducive and timely. Accordingly, the considerable policy work already underway within MBIE and the EA assumes significant importance.

22. We have made a number of recommendations to address these findings. These are set out in each section of this report and collated for ease of reference in Annex G.

We welcome your comments below. If you are not already registered, please register to comment.

Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current comment policy is here.


Will somebody actually be held accountable?


This is NZ - I think we all know the answer.


Megan Woods should be sacked the way she jumped to her incredibly amateur conclusion about what Genesis did.


So Transpower did a terrible job managing the network and even ripple control is decaying in effectiveness.  The other conclusions are to muddy the waters but do make it clear that Transpower also has no plans for the future of the network.

There's certainly a lot of similarity with Council managed pipe networks.


Time for the Tiwai Point smelter to close. The people need all the power they can get.


You don't live in Invercargill I take it.


...the market requires much greater demand side participation if greater electrification and decarbonisation is to happen.

The issue is that people are very used to hot water being available on demand when they shower or their EV charging up when they plug it in. It was probably lockdowns that saved us this year.


And it will be. Unless your hwc is too small for the house hold, an a grade cylinder will store the heat.

I would say it will be more sophisticated than the ripple control. The internet used to send price signals to the consumer, whose appliances and relay controls will decide how to react.


Interwebs 'control' of household appliances suffer from two intractable problems:

  1. Security against hacks.  The makers of appliances are notoriously naive about this possibility - everything from cameras to refrigerators has been hacked from afar.  And given that the appliances in question have burnt-in firmware, updating is to say the least, problematic
  2. The ability to 'control' demand in this way rests on a heroic assumption: that a sufficient number of controllable consumer demand units is in fact available to make a difference.  As with old but perfectly serviceable vehicles, there is tremendous incumbent inertia.

The short version: ain't gonna be a useful demand management option.


No need to replace an appliance necessarily. Something like a HWC, refrigerator or spa could be controlled by an external add-on device with an added thermometer if necessary. Many people do this already just with timers to make use of Hour of Power, night rates etc and with sufficient price incentive they pay for themselves very quickly. 


The "interweb" does not control the appliances. It sends the market info(basically the spot price) , and the consumer makes the choices , or has his control system set them automatically. MFD  sets out many ways how it is done. The consumer can elect to consume power as normal , but will pay a higher price. 

Many grid conect solar houses already use such devices to use as much power as they can while the sun is producing , the interweb or power company does not get any information other than how much power that house is using or producing in total .  


Interesting replies but both miss the second point I made.  Effective demand control needs large enough blocks of consumption available for such control, to make the demand reduction actually matter. 

That's exactly why large industrial users often contract specific loads to be switched off at short notice, with monetary concessions made for this service. Those loads are big enough to make a difference.

But try applying this to domestic users - whole suburbs, consumer classes or regions would need a critical mass of control-capable loads to be there, ready for the big red switch to be thrown.  And every such consumer would need full due diligence made, understood, and the potential load-shedding contracted for.  

That's why it hasn't been seriously attempted - easier to black out an entire sub-circuit when the wind stops, night falls, the batteries discharge, or a bunch of weeds clog the intakes....


It has been attempted (and still) is  in a way, with the example you gave yourself. Large users are charged by the kva of their peak demand , as well as a kwh they are rewarded for smoothing out their own peaks , and charged for breaking their agreed limit. no reason why ripple control cant be used as well . but i would say enough consumers would see the benefits to make price are viable tool. 

Overall the large size of the national grid smooths things out pretty well, with the know choke points to be improved.

Certainly better than wehen i was an aprrentice doing welding tests in Petone , if you were welding when 3pm stuck , it would be blown to hell as all the industries around shut down for smoko.


"We call this issue ‘managing multi-hour shortfalls’. We think it will become an increasingly important issue to address."

That's the key point. Ripple control is old-tech, and works. But we are going to attempt to electrify the 60% of our work that is done by fossil fuels. That would be an additional 150% generation - ain't gonna happen. The fudge is that electricity is 'more efficient', but if it is via hydrogen or some similar intermediate process - it isn't.

So we will be doing 100% less, give-or-take, overall. With orders of magnitude more demand on a much more loaded system. Methinks 'deserve' will vanish from the vernacular......


Unless there is some financial gain in it the asset owners of ripple control which I still believe are the lines companies, will let ripple control fall over. Some lines companies have sold the control of ripple control, not the assets, to retailers. I don't know how widespread this is. This is where the financial gain comes into place and under these circumstances the lines companies may upgrade there existing equipment  or even install ripple control where none has been in place before.

The control of ripple control is probably better served by Transpower, not withstanding Transpowers warts as an organisation. The EA could put their oar in as Transpowers minder.