It is not hard to understand why the South Australian government does not want to take any chances over electricity supply. They feel badly let down – in the middle of a storm and in the middle of a heatwave – by the Australian Energy Market Operator, screwed by the profit motive of fossil fuel generators, and betrayed by the federal government.
And they feel exposed. By their admission, they are leading the world on the adoption of wind and solar, which is inspiring for most, except for when the lights go out. And it is highly political. What happens next summer could decide the fate of the next election, due to be held in March 2018.
So it’s no surprise that South Australia has taken action of its own: giving emergency powers to the energy minister, ordering a new back-up generator, announcing the country’s biggest battery storage tender, and proposing to introduce an energy security target, which sounds like a great idea to guarantee dispatchable generation, loosen the hold over the market of a few incumbents, and lower prices.
But has South Australia been sold a pup by the fossil fuel incumbents on the issue of inertia? As we reported on Thursday, South Australia’s draft rules for its energy security target appear to rule out battery storage as a source of dispatchable generation.
It favours what it calls “real inertia” provided by turbines, which, in South Australia right now, means only gas-fired generation. Synthetic inertia, often described as fast frequency response, and which would likely come from battery storage usually paired with wind and solar, does not qualify.
It’s a decision that will provoke howls of protest from the battery storage industry, and has implications for the owners and developers of existing and future wind and solar farms.
Some have already reacted, with Tristan Edis, from Green Energy markets, labelling it as a daft idea, poorly designed and entrenching a lack of competition – the very thing the state government should be trying to avoid. And there have been many similar responses in the comments section of our article on Thursday.
The issue of inertia, and what can provide it, and overall system security, looks like tracing the same evolution as the debate over renewable energy. Go back a decade, and many engineers were saying it was impossible to incorporate more than 10 per cent wind and solar into the grid without having the whole thing collapse around them.
Now, say the CSIRO and Energy Networks Australia, between 30 and 50 per cent wind and solar can be considered trivial, such are the technology developments and progress in grid management. The CSIRO and the networks, even the big gentailers like AGL, are freely talking about 100 per cent renewable energy within a few decades.
“And once we have the storage capacity to get us to 50 per cent (renewable energy), we will have the technology to go to 100 per cent (renewable energy),” AGL CFO Brett Redman said in a major presentation last week.
South Australia now finds itself at that pivot point. It’s already got 50 per cent wind and solar and has a whole bunch more coming on to the grid in the next 18 months. Having lost faith in the centralised authorities, it is now trying to manage its own needs. The fate of power system security, and the careers of politicians, depends on it.
In the case of inertia, it does seem to be a battle between “real inertia” – that provided by spinning turbines – and synthetic inertia, also known as fast frequency response, and generally provided by battery storage, often paired with wind and solar.
It is also a battle of two grid management philosophies: the old style, which is to respond to every problem or challenge by generating more of something and shoving it into the system, however slowly; and the new style, which is to think about management by being smarter, and more efficient. That’s where fast response and smart software come into play.
“Inertia is not the goal, frequency and voltage stability is,” writes engineer Peter Farley. “Arguing for more inertia is a bit like saying heavy cars are safer than small cars in some accidents so all cars should be as heavy as possible.”
Right now it is accepted that fast frequency response (FFR) and battery storage can play a significant role in the provision of inertia and the management of the grid. It fills in the gap between a major fault and the relatively slow response of the governors that control the gas turbines.
That means less “real inertia” is needed. This has been found in the ERCOT system in Texas, and by research from AEMO in Australia. But it is not thought that FFR can completely replace “real inertia”, or at least not yet.
This graph from an AEMO presentation (slide 19) last year neatly summarises the issues. The debate has moved from “synchronous only” to a mixture of synchronous and fast frequency response (synthetic or storage), and will eventually evolve even further.
This report was followed by another report released by AEMO in March, which followed a consulting report by GE Energy. AEMO’s conclusions were:
FFR and synchronous inertia are technically distinct services due to the timescales over which they act. This means a minimum quantity of synchronous inertia will continue to be required in the short to medium term. However, FFR can compensate for, and help to mitigate, the effects of reduced synchronous inertia on power system frequency control by providing a wider range of options for meeting the frequency operating standards (depending upon a co-optimised consideration of the availability and costs of both services). This suggests that enabling FFR services in the NEM may allow the frequency operating standards to be met with a lower level of synchronous inertia.
So, it seems, the key question is what is a “minimum quantity” of real inertia.
Right now, that is up for debate, just as the total requirement for inertia is also up for debate (see this submission from two weeks ago by AEMO). But the feeling is that the minimum quantity of “synchronous” or “real” inertia it is less than the 36 per cent that South Australia is demanding from July 1, rising to 50 per cent by 2025.
The fear for the industry is that South Australia is raising the drawbridge against synthetic inertia and battery storage too quickly, possibly because the gas generators have put the fear of god, and more blackouts, into them.
It’s understandable that the government wants to take no chances, but if it wants its legislation to be cost effective, and not create a new series of stranded gas generators, it needs to work out exactly what it wants to achieve.
The draft legislation, and the apparent decision to rule out synthetic inertia will have significant impacts on the owners and developers of wind and solar farms, which were looking to pair with battery storage to provide some of those market services, and to make their output “dispatchable.”
And Farley, like others, warns that just relying on gas turbines is not a panacea, simply because their response is often too slow – an issue raised, but not fully explored, in the September 28 “system black” in South Australia.
“There is a good argument to suggest that a system with a large proportion of gas turbines is possibly more vulnerable to short-term instability than a modern renewable based system with storage,” Farley writes.
“Probably the most expensive, least reliable way to supply grid stability is by adding gas turbines, if the primary purpose is to add inertia.”
The CSIRO and ENA report into a zero emissions grid looked at the issue of inertia and concluded that – while more analysis is needed – there are a range of technical solutions to achieve inertia and frequency management outcomes in a high renewables grid.
“While battery storage is forecast to provide the dominant new source of energy balancing, there are a diversity of potential solutions which could be employed as alternatives while still achieving zero net emissions depending on their changing economic potential.”
The report specifically looks at the scenario where South Australia is sourcing 80 per cent of its demand from local wind and solar.
“Far greater levels of monitoring and control will be required to allow active management of the distribution network to meet its increasingly complex operational needs. This will require consistent, open and flexible systems, with suitable communications and open standards to permit new market entrants to participate.”
Iain Macgill, from UNSW, says that real or synchronous inertia and synthetic inertia aren’t direct substitutes. “However, there is a relationship. Other things being equal, synthetic inertia and very fast frequency response equipment can generally reduce the need for inertia while not being a direct substitute.”
But he agrees with Farley, and others, that relying only on gas turbines is not fool-proof either.
“Some of the most sensitive equipment to high Rate of Change of Frequency (RoCoF) events in a power system (the thing that inertia helps reduce) may actually be particular types of thermal generating plant (hard to know given that we generally have so few high RoCoF events).
“So there might be particular circumstances where the synchronous generators are providing inertia, but also set the requirement for inertia, and are unable to deal with high RoCoF from insufficient system inertia should there be a disturbance. If they fall off-line they will further add to the frequency disturbance and we are on our way to collapse.
“Who knows – it might actually prove easier in the longer term to manage power systems with fully power electronics interfaces than a mix of old synchronous plant and new power electronics interfaced plant. Lots of interesting research questions needing to be addressed.”
To be sure, this legislation – as it is – could encourage renewable energy providers of “real inertia” such as solar thermal and pumped hydro into the system. But these won’t be built for several years.
Small grids can operate quite satisfactorily without turbines, but every scenario for a 100 per cent renewable grid envisages a certain amount of spinning turbines coming from hydro, pumped hydro, solar thermal or bio-gas. But they also envisage significant amounts of battery storage.