On a mild spring afternoon a few weeks back, South Australia’s instantaneous wind and solar output covered well over 100 per cent of the state’s own demand, with the surplus heading up the Heywood and Murraylink interconnectors into Victoria. That’s not a freak event anymore. AEMO’s own published data shows South Australia now sits above 70 per cent annual renewable share, a figure most of the rest of the National Electricity Market is still a decade away from touching. I’ve been following this grid since the system black inquiry in 2016, and the thing that strikes me now isn’t the renewable share itself. It’s how unremarkable it’s become to the people who run the place.
South Australia is the smallest mainland region in the NEM by demand, which is precisely why it got to be first. A grid drawing maybe 1,000 to 1,800 megawatts on a given day, with two interconnectors to lean on, is a far more forgiving place to try things than NSW or Queensland. Call it a laboratory if you like: AEMO’s engineers and the state’s own network operator, SA Power Networks, have used that word themselves in technical papers. The results, good and awkward, are now shaping rules being written for the whole NEM.
The mix today, and the number everyone misquotes
Let’s be careful with that number, because it gets mangled constantly in press coverage. South Australia is not “100 per cent renewable.” It is around 70-plus per cent renewable on an annual energy basis, per AEMO’s Quarterly Energy Dynamics reporting, with gas generation and interconnector imports making up the balance. Annual energy share and instantaneous power share are different things, and South Australia hits 100 per cent of instantaneous demand from wind and solar on a growing number of days each year: AEMO’s engineering teams track this as the “minimum operational demand” problem, and it happens often enough now that it’s a planning input, not an anomaly.
The gas fleet left standing (Torrens Island, Pelican Point, Quarantine, Osborne) increasingly behaves like a reserve fleet rather than a baseload one. It sits idle for days, then gets called hard during a wind lull or an interconnector outage. That’s a fundamentally different commercial proposition to running flat out, and it’s part of why gas economics in this state read so differently from Queensland’s. Readers who want the national version of that argument should have a look at our piece on whether the gas industry has been written off too early, because South Australia is the sharpest test case going.
Interconnection is the whole plot
South Australia’s renewable story only works because of two cables. Heywood links the state to Victoria and has been progressively upgraded over the past decade; Murraylink, the smaller DC link through the Riverland, does similar duty on a quieter day. Without them, a state this size with this much wind and solar would be curtailing constantly or running into frequency trouble on calm, sunny afternoons. With them, the state exports surplus and imports firming power when the wind drops.
That’s exactly why Project EnergyConnect matters so much here: the new interconnector to NSW, built by ElectraNet and TransGrid, effectively gives South Australia a third door. I’d argue it’s underappreciated outside the state: most coverage of interconnection focuses on Marinus Link and VNI West because they’re politically contentious, but EnergyConnect is already doing quiet, unglamorous work moving renewable energy both directions across the border. For the national picture on why these cables carry so much weight in the system design, see our explainer on who really pays for the transmission build-out.
Grid-forming batteries did the hard engineering here first
This is the part of the South Australian story that actually matters to engineers more than the renewable percentage headline. When you strip out synchronous generation, the big spinning turbines that physically hold grid frequency and voltage stable, you lose system strength. South Australia hit that wall early, because it was decarbonising faster than anywhere else in the NEM.
The fix has been a mix of synchronous condensers (essentially spinning flywheels with no fuel, installed at Davenport and Robertstown) and grid-forming inverter technology trialled at Hornsdale Power Reserve, the Neoen battery that made Tesla’s name in this market back in 2017. Hornsdale wasn’t just about arbitrage and frequency response in the early years. It became a genuine test bed for inverters that can form grid voltage rather than merely follow it. We’ve covered the mechanics of that shift in detail in our piece on why the grid suddenly needs grid-forming inverters, and South Australia is where most of the real-world data has come from. AEMO’s Integrated System Plan now assumes grid-forming capability will need to scale nationally; South Australia is roughly five years ahead on the learning curve.
A worked example worth sitting with
Take a calm, overcast winter evening – wind output drops to a few hundred megawatts, rooftop and utility solar contribute nothing after sunset, and state demand sits around 1,400 megawatts. The gas fleet picks up maybe 500 to 700 megawatts. The rest comes down Heywood from Victoria’s brown coal and gas fleet, or down the new EnergyConnect line from NSW. Flip the scenario to a breezy, sunny Sunday afternoon with demand down near 900 megawatts because half the state’s air conditioners are off: wind alone might deliver 1,200 megawatts, rooftop solar adds another few hundred, and the state is now exporting hundreds of megawatts it doesn’t need.
Same grid, same assets, two completely different jobs for the interconnectors within 24 hours. That swing is the entire reason battery storage in South Australia earns so much of its revenue from frequency and system services rather than straightforward energy arbitrage: the economics genuinely are different here to a battery sitting in, say, the NSW REZ pipeline.
The storage build keeps compounding
Hornsdale was 150 megawatts when it opened; the state’s battery fleet now runs well past a gigawatt of combined capacity once you count the newer projects layered on since, with more under construction. South Australia’s own government has leaned into this with its hydrogen and storage ambitions, some of which, like the long-mooted Whyalla hydrogen plant tied to the steelworks, have run into real commercial trouble as the steelworks itself went through administration. Our profile on Zen Energy and the retailer standing in Whyalla’s shadow covers that saga in more detail than I have room for here, but it’s a useful reminder that South Australia’s renewables leadership hasn’t made every associated bet a winner.
Worth noting too: the state’s rooftop solar penetration is among the highest in the world on a per-household basis, which creates its own minimum-demand headache on sunny shoulder-season days – a problem AEMO now manages partly through remote curtailment schemes that didn’t exist five years ago. That’s a genuinely new regulatory tool, and one other states are watching closely.
Where South Australia sits on the national scorecard
Compared against other states’ renewable targets, South Australia isn’t chasing a number anymore. It largely got there. The state government’s effective target was 100 per cent net renewable electricity by around 2030, and on current trajectories it’s going to hit that comfortably, possibly early, depending how you count imports and exports in the accounting. For a side-by-side against Queensland, Victoria and NSW targets, our state renewable targets scorecard lays out where each jurisdiction actually stands rather than what’s been promised.
My honestly mildly contrarian view: I think South Australia gets treated in national commentary as a success story that’s basically finished, and that’s wrong. The hard part – running a modern inverter-dominated grid reliably at the lowest possible system strength, without the comfort of nearby baseload – never actually finishes. It just gets harder as more synchronous plant retires elsewhere and other states start facing the same arithmetic South Australia solved for itself a decade ago. The lessons transfer, but the engineering problem doesn’t go away; it just moves state to state as coal exits accelerate.
What to watch next
Keep an eye on three things over the next couple of years: how EnergyConnect’s actual utilisation compares to its design case once it’s bedded in fully, whether the synchronous condenser fleet gets expanded further as more gas capacity retires, and what happens to gas reliability obligations in a state where the gas fleet runs at genuinely low capacity factors but still needs to be available within minutes. AEMO’s Energy Supply and Operational Reliability work on South Australia specifically is worth a read for anyone who wants the regulator’s own framing rather than a journalist’s gloss. It’s on AEMO’s website under the NEM reliability reporting suite.
I’ll admit there’s something almost chess-like about watching this particular board: South Australia made its opening moves a decade before anyone else, and the rest of the NEM is now replaying the same positions one state at a time, just with more pieces on the table and higher stakes. Whether they get there as cleanly is the open question.
– Anjali Rao, Grid & Storage Correspondent
Photo by Zoshua Colah on Unsplash