The headlines are screaming again from mainstream media about the threats of blackouts, only this time it is not about having not enough coal, or too much wind, but because of the so-called threat of having too many rooftop solar panels on the grid.
It began last month with this headline in The West Australian, “Rooftop solar poses blackout threat to WA’s main power grid,”, and was followed on Thursday – following presentations and interviews at the Energy Networks Australia conference in Sydney – by similar headlines in The Australian and the Sydney Morning Herald.
Both stories relied on comments from the head of the ENA Andrew Dillon, and the head of energy software company Greensync, Phil Blythe. “Dumb solar threat to grid stability”, said The Australian. “Solar surplus set to overload the grid,” said the SMH, and its Fairfax stable mates.
And federal energy minister Josh Frydenberg joined in on Thursday morning while addressing the same conference, warning that the “exponential” growth of rooftop solar was “placing real challenges” on the grid, and citing the same stories, though he stopped short of mentioning blackouts when asked by Reneweconomy.
What’s going on?
Rooftop solar has been growing steadily, encouraged initially by generous feed in tariffs and later by ongoing but declining upfront rebates.
And now – aided and abetted by the plunging cost of solar technologies and the soaring cost of the grid – the uptake is accelerating, with a record 583MW added in the first five months of 2018, the country total standing at more than 7GW and predictions it could be close to 20GW by 2030.
But this has been forecast, and welcomed. Distributed energy, such as rooftop solar and the predicted uptake of battery storage and electric vehicles and demand management is forecast by the Australian Energy Market Operator and the CSIRO to account for nearly half of all supply within a few decades.
It is generally recognised, even by this government, that the energy system is rapidly evolving from a centralised, fossil fuel based grid, to a decentralised, renewable one – a grid that will be faster, cheaper, cleaner and more reliable than the dumb, dirty and ridiculously expensive grid we have now.
Our first reaction to the West Australian story was that it was a beat up, but it turns out that it’s pretty much exactly what AEMO had said in a submission to a WA parliamentary enquiry (pages 18 and 19) into the opportunities for micro-grid technology.
“A critical issue arising from uncoordinated DER growth (particularly rooftop solar PV systems), is that at some point, the total output from rooftop solar PV systems will be greater than the demand on the system (ie. on low demand sunny days),” AEMO wrote.“This excess generation can result in an inability to dispatch sufficient frequency control ancillary services to manage system frequency effectively. In a situation of high rooftop solar PV output, should invertors ‘trip’ en masse in an uncontrolled manner, with insufficient frequency control ancillary services online and available this can result in subsequent under frequency load shedding.“The worst case outcome of such a scenario is a total system blackout (our emphasis).“This point may not be far away, considering the projected increases in embedded generation. It is possible that alternative ancillary service arrangements may be required to deal with this problem. With more utility-scale synchronous generators expected to exit the market, this work needs to proceed with some urgency.”
This was the substance of the message being broadcast by ENA’s Dillon and Greensync’s Blythe at the network knees-up in Sydney on Wednesday.
But does it stand up to scrutiny?
Well, not really.
The question is centred on the quality and the ability of rooftop solar inverters to respond to frequency and voltage disturbances, and what happens when rooftop solar output exceeds grid demand – as is predicted in South Australia and Western Australia on occasions within the next 10 years.
These issues have been addressed by a new standard, AS4777, which was published in October 2015, and came into force in October, 2016, since which time nearly 2GW of rooftop solar has been added to the grid.
This standard requires these inverters to operate and not switch off when frequencies venture outside the nominated AEMO frequency bands – i.e. solar PV inverters will be helping keep frequency up and will never be the reason why under frequency is a cause of grid instability.
And in over frequency situations where PV generation might be exceeding grid demand, the solar PV inverters operate to what is called a droop curve. That is, they automatically reduce the power being injected into the grid as the frequency rises – this is solid state and happens within microseconds of the grid changing.
It’s much the same thing that has enabled the Hornsdale Power Reserve, the Tesla big battery, to react so quickly and accurately to frequency changes. Basically, digital electronics are better than steam engines when it comes to responding to change, as AEMO has acknowledged.
We asked ENA’s Dillon on Thursday about the new standards, and he conceded that the issue was actually with the previous crop of inverters, before the new standard. But even this claim is doubtful.
A previous AEMO study, “Response of existing PV inverters to frequency disturbances” looked at the quality of rooftop solar systems installed on the grid up to May, 2015, when 3.7 GW of rooftop solar had been connected to the grid.
It looked into exactly the scenario of solar PV exceeding grid demand and it found that “a large proportion” of inverters have frequency trip settings that are outside the frequency operating ranges for “system normal, credible contingency, and non-credible contingency events.”
As these graphs show, more than half the inverters were designed to deal with under-frequency events, and three-quarters for over-frequency events.
As the AEMO report noted:
“With this diversity in frequency trip settings, it appears unlikely that a mass disconnection of small-scale PV generation would occur during frequency disturbances.”
So, no “total blackouts”.
Inverter suppliers point out that most tier 1 inverters have had such capabilities since 2010 (read this technical explanation from one leading inverter manufacturer published in June, 2010), before the solar PV boom emerged. That appears confirmed by the AEMO survey.
As AEMO also noted:
The introduction of the new AS/NZS 4777.2–2015 will result in the standardisation of frequency responses for PV inverters (including battery storage) installed from 9 October 2016.
This will mean that new compliant inverters will not disconnect for frequencies in the range of 47 Hz – 52 Hz. For some inverters installed before 9 October 2016, the risk of disconnection during operational frequency bands will reduce as they reach the end of their life and are progressively replaced with inverters compliant with AS/NZS 4777.2–2015.
It should be noted that drifting below 49Hz is still a very serious contingency event, so the solar would only drop after a serious contingency event – it couldn’t be the cause of one.
In other words, the chance of a “total blackout” are so remote there is just as much chance of the country’s hospital system being suddenly overloaded because everyone in Australia was hit by a bus at the same time.
“It’s news to me,” was the response of the CEO of one major energy utility when approached by RenewEconomy about the blackout threat during the conference.
Another, Richard Gross, the head of Australia’s biggest grid operator, Ausgrid, is actually encouraging more solar into the system so it can spend less on network upgrades.
“The Ausgrid network has sufficient capacity for the expected uptake levels of rooftop solar on our footprint,” he said. “Our current solar penetration is modest compared to other distributors. The impacts are higher in more sparsely connected rural networks rather than in dense urban networks such as Ausgrid’s.”
So what’s this all about?
Well, get used to this new acronym: DSO.
It stands for Distributed System Operator, and its about the ability of AEMO, the networks, or some other party having visibility, and control, over the rooftop solar systems installed on the nation’s 2 million plus houses.
In network talk, its called orchestration. AEMO wants visibility because in a grid that gets nearly half of its supply from distributed energy – rooftop solar, battery storage, electric vehicles and demand management – it wants to be able to see what’s going on, and control it if need be.
Network operators and software providers like Greensync are hot on this idea because they can see potential revenues.
The networks, initially, for convincing the regulator to allow them to spend tens of millions of dollars on IT systems and research, and further out because they might see a role for themselves in aggregating this demand and playing in the wholesale or grid services market.
Greensync and co like it because they want to be the traders of this new commodity. As we reported on Wednesday, in our story devoid of blackout threats, we will get some idea of what this DSO and orchestration might look like, and who might control it, when AEMO and the ENA release a joint report next week.
And as one wise soul pointed out to RenewEconomy on the sidelines of the conference: “A lot of companies are hanging their hat on this. There’s a lot of money to be made for this, they all want boxes in houses, and they want it to be their box.”
And, this good person further noted, the best way to get things moving in Australia – and grab control of a citizen’s asset – is to spread alarmist rhetoric, confect crisis, and then look like you have a solution.
That’s depressing.
Climate and energy policy in Australia has been constantly derailed by baseless scare campaigns about new technology. Even Frydenberg had the gall to tell the conference , on the subject of the National Energy Guarantee, that policy had been “bogged down by the hyper partisan debate.”
This from the Coalition that has mastered the art of hyper partisanship and exaggeration. As Frydenberg (rightly) noted, the biggest losers have been consumers, and they will continue to lose if such scare campaigns continue.
We asked AEMO for some clarity on this and whether there had been any further studies, but hadn’t heard back by publication time.
Perhaps they were all too busy sticking sticky tape on the coal generators which have been tripping all over the shop in NSW these past few days and remain the biggest threat to power supplies.
(Indeed, if you got back to 2011 its was the Northern Territory’s Power and Water that requested that solar PV inverters have their frequency band points expanded – so that if large loads or gas generators tripped that PV can actually help stabilise the local grid.
Power and Water found that solar PV inverters had a wider tolerance to fluctuation in voltage and frequency than their gas and diesel generators. So when a load tripped or generator they needed the PV inverters to kick in or back off to help stabilise the grid)
Clearly though, we need to move beyond the period of hyperbole, hyper-partisanship and scare campaigns.
There is probably a really good case for more visibility on rooftop solar and battery storage, and even “orchestration” and control of those assets to provide flexibility, ensure excess power can be stored and demand is managed.Bbut the case needs to be made without such blatant exaggeration.
In any case, it’s what is going to happen anyway. This transition is unstoppable, and if policy makers and regulators are smart, and fair, they will encourage the uptake of battery storage and other smart software that will usher in a new digital and distributed age that will actually help, not hinder, grid stability.
Solar PV inverters have dynamic response capabilities that lower or raise the energy they generate to help maintain stable voltage. They do this in an instant. Batteries make this even easier.
As another observer noted:
“Change creates risk for the incumbent and uncertainty for those employed to support the past and the present.
“The interests of the incumbent tend to encourage the fears of the latter. Luckily for the planet, clean energy technology is proving to be both better and cheaper, which means like lava flowing down a hill – the pace may look glacial but the outcome is predictable, and eventually all will move on.”
We could just ask the incumbents and the regulators to not beat it up in the meantime.
Update: We received this from AEMO is response to our questions:
The impact of rooftop solar on the grid is currently small, however as the percentage of rooftop solar grows, the system’s ability to deal with this will be increasingly challenged, particularly if there is a contingency on the system (eg a generator or very large load tripping off suddenly).
The power system requires frequency control and other ancillary services that cannot be provided solely by rooftop solar, particularly under present arrangements. Once uncontrolled rooftop solar output approaches the system demand, there is no room left for the generators that provide these critical ancillary services. Moreover, once there is more generation than load, the frequency will increase until the rooftop invertor over frequency settings are reached. The power system must be kept in constant supply-demand balance, such that managing the power system becomes more challenging the greater the percentage of supply is uncontrolled.
While the Australian Standards have been updated so that rooftop solar changes its output more consistently to frequency changes and hence respond by changing output to contingency events, it will do so in a less refined manner than traditional generator governor and frequency control ancillary service response. The swings in frequency as a result of rooftop solar coming on and off during these contingency or excess generation events needs further analysis and will impact how we are managing on the power system particularly as rooftop solar penetration increases.
The “Response of existing PV inverters to frequency disturbances” report covers both the NEM and WA in terms of the invertor frequency trip settings however the consideration of system impacts was contained to the NEM – a much larger, more interconnected market that can manage frequency using a larger pool of reserves and strong interconnections. The WA micro-grid enquiry submission relates to the WA market – an islanded system where a significant disturbance impacting rooftop solar would have a greater impact on the system due to its high penetration level and lack of interconnection.
Storage?