On a mild, windy afternoon in late spring, South Australia’s grid can run for hours on nothing but wind and rooftop solar, with gas generators idling in reserve and the local coal fleet a memory going back to 2016. That’s not a projection. AEMO’s own market data has recorded periods where wind and solar have supplied effectively all of the state’s operational demand, and South Australia has held the informal title of the most solar-and-wind-saturated grid of its size in the world for several years running. The state government leans on that record constantly. It’s earned, but it’s also a narrower achievement than the headline suggests, and the distinction matters if you’re trying to work out what SA is actually proving for the rest of the National Electricity Market.
The record everyone quotes, and why it’s about power, not energy #
Let’s be careful with that number, because this is exactly the kind of thing that gets mangled in translation. When AEMO reports that renewables briefly supplied “100 per cent” of South Australia’s demand, that’s an instantaneous measure — a snapshot in megawatts, taken at a five-minute dispatch interval, published through the mechanism AEMO documents in how the NEM dispatches power every five minutes. It says nothing on its own about what happens across the other 8,700-odd hours of the year, and it says nothing about energy — the total megawatt-hours consumed over a day, a month, a year.
Those two figures move very differently in South Australia because the state’s wind and solar resource is genuinely excellent but not remotely constant. On the annual energy measure — the one that actually determines emissions and fuel cost — SA’s renewable share is still hugely impressive by any comparison, comfortably ahead of the rest of the NEM, but it sits well short of 100 per cent once you average across a full year including the still, overcast weeks in winter. AEMO’s Electricity Statement of Opportunities and its Quarterly Energy Dynamics reports both track this gap explicitly, and it’s worth reading the underlying tables rather than the press release line.
A worked example, because the arithmetic actually explains the state’s problem #
Take a round-number, illustrative day rather than a specific AEMO record (the real figures move around too much to quote precisely without checking the live dashboard):
- Underlying demand at midday: roughly 1,400 MW
- Rooftop solar output at midday: roughly 1,000 MW
- Grid-scale wind and solar: roughly 700 MW
- Operational demand seen by AEMO’s control room: roughly 400 MW — or negative, on a very sunny, low-demand Sunday
That last line is the state’s actual distinctive problem, and it’s a power problem, not an energy one. When rooftop solar alone can exceed total demand, AEMO has to find generators willing and able to turn down, and it has to keep enough spinning machinery — or its synthetic equivalent — on the system to hold voltage and frequency steady. Too much inverter-based generation and not enough synchronous plant is a system strength issue, covered in general terms in negative electricity prices at midday: what they mean for you, and South Australia has been living with a sharper version of that problem for a decade.
Hornsdale, and what a battery is actually being paid to do #
The Hornsdale Power Reserve, built by French developer Neoen near Jamestown in 2017, is still the reference point for battery storage in Australia, even though bigger projects have since overtaken it in capacity. It’s worth stating the units properly: Hornsdale was commissioned at 100 MW / 129 MWh, later expanded. The MW figure is what it can inject or absorb in an instant; the MWh figure is how long it can sustain that before it’s empty. A 100 MW battery with 129 MWh of storage can run flat out for a bit over an hour, not all day — which is precisely why big batteries in the NEM are built for fast frequency response and price arbitrage rather than bulk overnight supply. That distinction is the whole argument in pumped hydro vs big batteries: which firms the grid?, and South Australia is the state where you can see it playing out in real dispatch data rather than in a slide deck.
Since Hornsdale, the state has kept adding storage — Neoen and others have expanded projects around the mid-north and beyond — and batteries developed by firms like Akaysha Energy have become part of the standard NEM toolkit, discussed in Akaysha Energy: BlackRock’s big battery bet on the NEM. But no battery fleet built to date, in SA or anywhere else in the NEM, is sized to firm a genuinely low-wind week on its own. That’s still gas, or imports.
Interconnection is the quiet infrastructure doing the heavy lifting #
South Australia’s ability to run at extremes — near-zero net demand at midday, heavily gas-reliant on a calm winter evening — depends on the Heywood interconnector to Victoria and the older Murraylink line to New South Wales. ElectraNet and TransGrid have also been building Project EnergyConnect, a second SA–NSW link intended to widen that door further; construction has run behind the original schedule and cost estimates have grown since the project was first approved, which is a familiar story for NEM transmission builds generally. When SA’s wind drops out, that capacity has to come from somewhere, and interconnection is the somewhere. It’s the least glamorous part of the state’s renewables story and, in my view, the more instructive one — the wind and solar numbers get the headlines, but the transmission and system-strength engineering is the actual laboratory work.
ElectraNet has also installed synchronous condensers at Davenport and Robertstown specifically to hold system strength up as coal and gas plant have retired. It’s unglamorous kit — essentially spinning mass with no fuel attached — bought purely to keep the grid stable enough for inverter-based generation to keep operating safely, and it’s a direct legacy of the state-wide blackout of September 2016, when storm damage to transmission towers triggered a cascading failure that AEMO’s subsequent investigation attributed to a chain of protection settings and wind farm ride-through behaviour rather than simply “too much renewable energy,” as it was characterised at the time in some political commentary.
The hydrogen detour and the 2027 target #
South Australia’s stated ambition, under the current state government, is net 100 per cent renewable electricity by 2027 — an energy-basis target, not the instantaneous one, which makes it considerably harder to hit than the headlines about record days suggest. The state’s earlier hydrogen ambitions, including plans for a hydrogen-fired power station near Whyalla, have moved more slowly than first announced, and the timeline has clearly slipped from the original pitch. That’s not unusual for early-stage hydrogen projects anywhere in the world; electrolyser costs and offtake certainty have been the sticking point nationally, not just in SA, a pattern visible in the broader gas-and-hydrogen debate covered in Santos: Moomba’s carbon bet and the Narrabri gas standoff. Whether Whyalla’s hydrogen plans still make commercial sense at their original scale is a genuinely open question, and South Australia’s own energy department publishes progress updates worth checking rather than taking the original 2021 pitch at face value.
What the laboratory is actually testing #
I rode the last stretch of the Mawson Trail into Adelaide earlier this year, past the old stacks at Torrens Island, and it struck me how unremarkable the transition looks from a bike saddle now — wind turbines on the ridgelines north of the city, a battery compound behind a chain-link fence, nothing dramatic. That’s rather the point. South Australia isn’t proving that a state can run on wind and solar alone; nobody serious claims that, and the state’s own numbers, read on an energy basis rather than a power snapshot, confirm gas and interconnection are still doing real work. What SA is proving, slowly and less quotably, is what a grid needs to hold together once synchronous generation becomes the minority partner: enough storage to smooth the sharpest edges, enough transmission to borrow power when the wind drops, and enough purpose-built system-strength equipment to stop a storm turning into a blackout. Other states with weaker interconnection and thinner grids, this masthead has argued elsewhere in pieces like state renewable energy targets compared: the honest scorecard, will need a different mix of that same toolkit rather than a copy of South Australia’s. The record days make good headlines. The synchronous condensers at Robertstown are the actual experiment.
— Anjali Rao, Grid & Storage Correspondent
Photo by hector espinoza on Unsplash