Stood in a garage in Buderim on the Sunshine Coast a few weeks back, watching a customer’s inverter throttle itself down to almost nothing on a cloudless afternoon. Full sun, clean panels, nothing wrong with the string. The inverter was doing exactly what its firmware told it to do under AS/NZS 4777.2, the grid connection standard every rooftop system has to comply with. Somewhere upstream, the local zone substation had more solar pushing into it than it knew what to do with, and the export limit on that connection was quietly clamping the household’s output back. The customer wanted to know why they’d spent six grand on panels that were being told to switch off in broad daylight. Fair question.
This is the midday problem, and it’s stopped being a curiosity for network planners and started being something installers explain at kitchen tables every week.
What minimum demand actually measures #
Minimum demand sounds like it should mean the quietest hour of the night, when everyone’s asleep and the fridge compressors are the only thing drawing current. For decades that’s exactly what it meant. Not anymore. On a mild, sunny weekend afternoon in spring, with air-conditioners off and rooftop solar flooding into the local network, the demand the grid actually has to serve from big power stations can fall lower than it does at 3am. AEMO calls this operational demand, and it’s been setting new record lows around the middle of the day in South Australia, Queensland and parts of Victoria in recent years, driven almost entirely by the roughly four million rooftop solar systems now bolted to Australian roofs.
Think of it like a stormwater drain built for a steady trickle that suddenly gets hit by every downpipe on the street at once, all draining at the same time on the one dry, sunny afternoon nobody expected. The pipe was sized for average flow, not for every house discharging simultaneously. The grid was built the same way, for demand that rose and fell gradually through the day. It wasn’t built for a wall of exported solar arriving between about eleven and two.
Why the middle of the day is now the hard part #
The old worry about the grid was always the evening peak, when the sun goes down and everyone gets home, switches on the airconditioner and puts the kettle on. That peak hasn’t gone away. But the harder engineering problem now sits at the other end of the day. When rooftop solar output is high and underlying demand is low, the big synchronous generators (coal units mostly, some gas) get pushed down toward their minimum stable output or taken offline altogether, because you can’t run a coal boiler at zero and there’s a floor below which they can’t safely operate. When there’s too much solar and not enough load, wholesale prices go negative, and generators start paying to keep running rather than shutting down. Our colleagues on this desk have covered why that price mechanic exists in more detail – see negative electricity prices at midday: what they mean for you, but the practical upshot for a rooftop owner is blunter: sometimes there is nowhere for your exported electrons to usefully go.
The other half of the problem is technical, not just economic. Coal and gas turbines physically spin in sync with the grid’s frequency and give the system inertia, a kind of mechanical stiffness that keeps things stable if something trips. Push those generators too low, or off entirely, and you lose that stiffness right when a fleet of rooftop inverters, which don’t behave the same way electrically – is doing most of the heavy lifting. We’ve written before about how this shows up as a genuine stability question, not just an accounting one, in rooftop solar grid stability: the honest read, and the fix on the generation side is the shift toward grid-forming inverters and batteries that can mimic some of that stiffness, covered in grid-forming inverters: why the grid suddenly needs them.
What this looks like from a switchboard #
For a household, the midday problem shows up as an export limit. Every solar connection agreement with a distribution network sets a cap on how much power a system can push back into the grid, measured in kilowatts, separate from the size of the panels themselves. In much of the country that’s still a flat number written into the connection agreement: 5kW export limit regardless of time of day, say, but the more crowded networks are moving to dynamic or flexible export limits, where the actual ceiling moves up and down depending on real-time capacity on that particular street’s transformer. SA Power Networks has run one of the more advanced flexible exports schemes in the country, and it’s worth reading in the context of the state’s broader renewables experience, which we’ve unpicked in South Australia’s renewable energy record: what it hides.
What that means on the ground: a system that could export freely at 7am might get clamped back hard by 12.30pm, then opened back up again by 3pm as the local street’s collective solar output tapers off. It’s not a fault. It’s the network trying to share a limited amount of headroom on that transformer among everyone connected to it, in real time, rather than handing out a fixed slice that sits unused for most of the day and runs out precisely when everyone needs it.
What a real household actually sees on the numbers #
Here’s where the glossy payback calculators tend to go quiet. A 6.6kW system oversized against a 5kW export limit isn’t wasting that extra 1.6kW. It’s there so the system still hits its rated output on overcast days and at the shoulders of the export window, when the sun angle is low. But if your retailer’s feed-in tariff has gone to near zero, or negative, for the exact hours your system produces most, the economics shift hard toward self-consumption rather than export. We’ve mapped how those tariffs actually move through the day in how rooftop solar feed-in tariffs actually work in 2026, and the short version is that flat feed-in tariffs are becoming rarer by the year. Most retailers now pay reasonably for morning and evening exports and next to nothing, or a charge, for the midday glut.
That’s the argument, honestly, for timing your big loads – the pool pump, the hot water system, the EV charger – to run when the sun is actually up rather than assuming any exported kilowatt-hour is worth the same as any other. It’s also the argument installers are making for batteries, though I’d push back a little on how that’s usually pitched. A battery doesn’t fix a network export limit; it just gives the household somewhere else to put the energy instead of sending it out. Whether that stacks up financially against simply accepting a lower export cap depends heavily on your own usage pattern, not a brochure figure, and we’ve run the actual comparison in rooftop solar versus a home battery: which pays back first.
The policy response, such as it is #
The Commonwealth’s Consumer Energy Resources roadmap, released through the Department of Climate Change, Energy, the Environment and Water, tries to get ahead of this by pushing standardised, interoperable rules for how household solar, batteries and EVs talk to the grid, rather than leaving each of the country’s dozen-plus distribution networks to invent its own export scheme. We covered what that actually changes on the ground in Australia’s Consumer Energy Resources roadmap: what it means on the ground. AEMO’s own operational forecasts, published through its regular market reports, now treat minimum demand as one of the primary planning risks for the grid, on par with summer peak: a fair reflection of how quickly the shape of the problem has flipped.
My veggie patch has the same rhythm, for what it’s worth #
Small aside, because it genuinely clarified the concept for me the first time a customer asked. My tomatoes get flooded with full sun between about eleven and two and do nothing useful with most of it. They can only photosynthesise so fast, the rest is just heat they have to shed. What they actually want is steadier light spread across more of the day. The grid’s in the same position with rooftop solar. It’s not that there’s too little sun, or too little capacity somewhere on the network. It’s that everyone’s harvest arrives at once, on the same few hours, and there’s a limit to how much any one pipe, wire or transformer can carry through at a single moment, no matter how much total capacity sits either side of it.
Where this is heading #
The honest trajectory is that export limits get smarter rather than looser. Flat caps are being replaced by dynamic ones that flex block by block through the day; feed-in tariffs are moving from flat to time-varying almost everywhere; and batteries, EV chargers and hot water systems are increasingly expected to shift load into the hours the grid actually wants it filled, rather than sitting passive. None of that makes rooftop solar a bad investment. It remains one of the better-returning things a homeowner can put money into, on the Clean Energy Regulator’s own installation figures. But the sales pitch that treats every exported kilowatt-hour as equally valuable, at any hour, on any street, stopped being true a while ago. Worth asking your installer straight: what’s my export limit, is it flat or dynamic, and what does my actual feed-in tariff look like at 1pm versus 7pm? If they can’t answer both halves of that question, get a second quote.
– Priya Nair, Solar & Distributed Energy Correspondent
Photo by American Public Power Association on Unsplash