24 gigawatts. That’s roughly where Australia’s rooftop solar fleet sits now, spread across more than four million homes and small businesses, and it is bigger than any single generation category in the National Electricity Market. Bigger than coal. Bigger than gas. Bigger than wind. On a mild, sunny Sunday with low demand, rooftop systems can be supplying more than two-thirds of the entire grid’s needs in some states, and AEMO has been warning for years that minimum demand troughs are getting lower and arriving earlier each spring. So when people ask whether rooftop solar is quietly breaking the grid, I understand the instinct behind the question. I just think it’s aimed at the wrong target.
The grid isn’t being broken by solar panels. It’s being exposed by rules written for a system that no longer exists.
The minimum demand problem is real
Start with what’s genuinely true, because there’s plenty of it. AEMO’s own analysis of the Integrated System Plan has flagged minimum operational demand as one of the hardest engineering problems in the entire transition, harder in some ways than peak demand ever was. We’ve covered the mechanics of this before, in our look at how rooftop solar and home batteries compare on payback, but the system-wide version is the one that keeps grid engineers up at night.
Here’s the mechanism. Rooftop solar isn’t dispatched by AEMO. It just happens, whenever the sun is out, regardless of whether the grid wants it or not. That pushes operational demand, the amount the big generators and the market actually have to serve, down toward zero in the middle of the day. When minimum demand falls far enough, there isn’t enough load left for synchronous generators to stay online and provide the grid strength, the frequency control, the inertia that keeps the system stable. Fewer spinning turbines means less of that free physical stability the grid used to get for nothing. AEMO has had to intervene directly at times, curtailing rooftop solar remotely in South Australia and issuing minimum system strength directions, because the alternative was risking a genuine stability event.
That’s not a hypothetical. That’s happened. And it will keep happening on the sunniest, calmest days until the market redesigns around it.
The duck curve everyone already knows about
None of this is news to anyone who follows the NEM closely. The duck curve, the midday solar trough followed by the evening ramp as the sun drops and everyone gets home and turns the kettle on, has been the signature chart of Australian energy policy for the best part of a decade. We’ve written about the mechanics in our explainer on minimum demand and the midday problem, and the shape of the curve itself isn’t controversial. What’s shifted is the steepness. The ramp AEMO now has to manage in South Australia and Victoria on some afternoons is sharper than almost anything a coal fleet was ever designed to follow, and coal plants are mechanically slow. They don’t like being asked to drop output at 10am and then claw it back at 5pm, every single day, for years.
That’s part of why gas peakers and batteries have become the preferred answer for the evening ramp rather than coal trying to flex. We’ve gone through that trade-off in our comparison of gas peakers against big batteries for the firming job, and the short version is that batteries are winning more of that specific slice of work every year, because they can move from zero to full output in seconds and coal simply cannot.
Who actually pays for the fix
Follow the money and the picture gets more interesting. Rooftop solar owners paid for their own panels, largely out of pocket, with some help from state and federal rebate schemes administered through the Clean Energy Regulator’s Small-scale Renewable Energy Scheme. They’re not asking the network for anything extra to generate that power. But the costs of managing the knock-on effects, the grid-forming inverters, the synthetic inertia, the system strength services, the curtailment schemes, those land on the shared network and ultimately on every customer’s bill, including the ones without solar. That’s the bit of this that doesn’t get said out loud often enough. A household with panels on the roof is, in effect, partly subsidised by the household down the street without them, once you account for the stability services the system now needs to keep everyone’s power running reliably.
I don’t think that makes rooftop solar a mistake. It’s been one of the cheapest, fastest decarbonisation tools this country has ever deployed, and households made a rational economic choice given the incentives on offer. But pretending the costs of managing it are zero, or that they fall evenly, is just not an honest way to talk about the numbers.
Grid-forming inverters are the actual fix, not a slogan
The engineering answer to the stability problem already exists, and it isn’t exotic. Grid-forming inverters, which can establish voltage and frequency references the way a spinning generator does rather than just following a signal from elsewhere, are the technology AEMO and the AEMC have both pointed to as the structural solution. We’ve explained the mechanics in our piece on why the grid suddenly needs grid-forming inverters, and the honest read is that this is less a research problem now and more a procurement and standards problem. Projects like the Waratah Super Battery are already doing some of this work at grid scale in New South Wales, acting as the shock absorber the system needs when large generators trip.
The gap is that almost none of Australia’s four million-plus rooftop systems are grid-forming, and most were installed under standards that predate the whole conversation. Retrofitting isn’t realistic at that scale. So the fix has to come from new large-scale assets, from tighter inverter standards on new installs going forward, and from smarter orchestration of what’s already on the roof, which is where the Consumer Energy Resources roadmap comes in. We’ve set out what that roadmap actually means for households in our explainer on the CER roadmap, and it’s a more useful document than most of the commentary around it suggests.
The curtailment conversation nobody likes
Here’s my mildly unpopular view on this, and I’ll own it. The industry spent a decade telling households that rooftop solar was a straightforward win, cheap power, lower bills, good for the planet, install and forget. That was true as far as it went. But it meant almost nobody was prepared for the idea that, one sunny Tuesday, their inverter might get switched off remotely by their network because the grid had too much solar on it already. South Australia has already lived through versions of this. It is not a fringe scenario anymore, it is a standing operational tool, and networks in other states are building the same capability into new connection agreements.
Consumers understandably find this frustrating. They paid for the panels; they expect to use them. But a grid that can’t absorb more solar at certain times of day isn’t a grid failing because of solar, it’s a grid telling the market, loudly, that storage and flexible demand are now worth more than another kilowatt of panel. That’s precisely why home battery economics have shifted so much in the last two years, a trend we track closely in our ongoing look at whether home battery payback stacks up yet. The market is quietly repricing rooftop solar’s marginal value downward in the middle of the day and upward in the evening, which is exactly the signal a well-functioning system should be sending.
So is it actually breaking the grid?
No, not in the sense of an imminent collapse. But it has broken a set of assumptions the NEM’s rules were built on, around who generates, who’s dispatched, and who pays for stability. The grid I watched operate a decade ago assumed big spinning machines provided inertia essentially for free, as a byproduct of generating electricity. Rooftop solar has quietly made that assumption false in large parts of the country for large parts of the day, and the market bodies are now racing to write new rules, new standards and new pricing signals to catch up. AEMO’s own Integrated System Plan modelling assumes rooftop solar roughly doubles again from here, which means this isn’t a problem the sector gets to solve once and move on from. It’s structural, ongoing, and it will shape every connection agreement, every tariff reform and every battery subsidy debate for the next decade.
Whether that counts as “breaking” the grid is partly a question of semantics. I’d call it something closer to a system that’s been pushed hard enough, fast enough, that the old operating manual no longer applies, and the new one is still being drafted in real time, clause by clause, mostly in AEMC rule change processes that get far less attention than they deserve. A bit like a bowler who’s taken the new ball and found the pitch doing more than anyone expected. You adjust the field, not the bowler.
– Marcus Wren, Editor
Photo by Anja van de Gronde on Unsplash