A client in Alphington rang me last month wanting a quote for a 10kWh battery to bolt onto a six-year-old 6.6kW solar system she’d never really upgraded. She’d been told by a sales rep the battery would “pay for itself” in six years. On the numbers I ran off her actual NMI export data, it was closer to eleven, and that’s before you factor in degradation or the chance she moves house before the warranty’s out. The solar panels on her roof, meanwhile, paid themselves off in year four and have spent every year since just making her money.
That gap, between what a battery promises and what a battery does, is the whole story of this comparison. It’s also why I keep getting asked the same question at kitchen tables from Reservoir to Frankston: if I’ve only got one cheque to write this year, solar or storage?
What actually earns its keep on the roof #
Rooftop solar’s payback case barely needs defending at this point, which is exactly why it gets taken for granted. A well-specified residential system (correctly oriented, not shaded by next door’s gum tree, sized to the household’s daytime draw) still routinely pays back in three to five years on a typical east-coast electricity tariff. The economics work because solar does one simple job: it replaces grid electricity you were going to buy anyway, at the retail rate, during the hours the sun is up. No inverter trickery required. You’re not managing a resource, you’re just intercepting sunlight before the meter does.
The catch, and it’s a catch most quotes gloss over, is that the sums above assume you’re home using that power, or that your retailer still pays something decent for what you export. Neither is guaranteed anymore. AEMO’s own work on minimum operational demand has flagged for a few years now that the grid in states like South Australia and Victoria hits periods of the day – generally the midday trough – when there’s more rooftop solar on the system than the grid can comfortably absorb. Feed-in tariffs have been sliding in response, and some networks now apply export limits or even brief curtailment windows. I wrote about this squeeze in more detail in our look at minimum demand and the midday problem, and it matters here because it changes what “payback” means for anyone installing solar today versus five years ago. A system sized the old way, to maximise exports, is increasingly a system leaving money on the table.
Why the battery case is slower and more fragile #
A home battery doesn’t generate anything. It’s a timing device. Think of it like a rainwater tank bolted onto a garden that already gets good rain. It doesn’t make more water fall, it just lets you use the water that already fell at a time you actually need it, instead of watching it run down the driveway. The value of a battery is entirely the difference between what you’d have earned exporting that stored energy versus what you’d have paid buying it back from the grid in the evening.
That spread has narrowed on both ends. Feed-in tariffs have fallen, which cuts what you’re giving up by storing rather than exporting. And retail time-of-use evening rates, while still higher than midday, haven’t blown out the way some battery marketing assumes. Run the real arithmetic on a 10-13kWh battery against a household’s actual evening consumption pattern, not the brochure household that conveniently uses exactly as much as the battery holds every single night, and most systems are cycling nowhere near 365 days a year at full depth. Some of the installs I’ve inspected around Melbourne’s north are getting solid use maybe 250 days a year once you account for overcast stretches, holidays, and the fact nobody’s fridge magically times itself to a battery’s discharge curve.
On the Default Market Offer-adjacent tariffs most households actually sit on, which I’ve covered in detail in our piece on the Default Market Offer and what it caps, a well-run 10kWh battery is saving a typical household somewhere in the order of $600 to $900 a year. Against an installed cost that, even with state battery rebates applied, is still commonly sitting in the $8,000 to $12,000 range depending on state and brand, that’s a seven-to-thirteen-year payback window before you’ve thought about inverter replacement or battery degradation. Compare that to solar’s three-to-five.
The rebate layer changes the comparison, but not the ranking #
The federal Cheaper Home Batteries programme, administered through the Clean Energy Regulator’s small-scale certificate scheme, has genuinely moved the needle since it landed. It knocks a meaningful chunk off upfront battery cost through STC-style incentives, and several states layer their own battery rebates or zero-interest loans on top. Run those together and you can pull a battery payback down into single digits in the better-subsidised states.
But subsidies don’t change the underlying physics of the trade, they just discount the price of buying into it. Solar still wins the race to break even because it’s solving a cheaper problem – converting free sunlight into electricity you’d otherwise have paid for, while the battery is solving a harder, more marginal problem of moving electricity a few hours sideways. You can subsidise a battery heavily and it’ll still generally lag a well-sized solar array on raw payback speed, because the solar array was never competing on thin margins to begin with.
Where the comparison gets genuinely interesting is for the household that’s already got solar on the roof and exhausted most of the easy daytime-shifting behaviours – running the dishwasher at noon, pre-cooling the house before the evening peak. For that household, a battery isn’t competing against solar, it’s competing against nothing, because the alternative is just paying the evening rate. That’s a different and generally better case for storage, and it’s the one I’d actually encourage people to run numbers on rather than the blanket “get a battery” pitch most installers lead with.
Export limits are quietly doing some of the battery’s job for free #
Here’s something installers see that doesn’t show up in payback spreadsheets: a lot of networks now cap how much a household can export back to the grid, often somewhere around 5kW for a standard single-phase connection, sometimes dynamically throttled lower during high-solar periods. If your system is generating more than that cap during the midday trough, that excess isn’t getting paid for anyway, capped or not. It’s being clipped at the inverter or curtailed by the network regardless of whether you own a battery.
In that situation, adding a battery recovers power you were already losing to curtailment, which is a genuinely stronger case than the standard “store and shift” pitch, because you’re not giving up an export payment, you’re giving up nothing. If you’re on a tight export limit and oversized solar, a battery’s effective payback improves because its alternative isn’t a feed-in tariff, it’s zero. Worth getting your installer to actually pull your export data before quoting a battery, rather than assuming every kilowatt-hour stored is a kilowatt-hour that would otherwise have been sold.
The sizing mistake that skews both numbers #
I’ll admit I spent a good chunk of this year’s veggie patch planning season instead on spreadsheet work for a run of western-suburbs clients, and the pattern that kept showing up was oversizing – on both solar and batteries. A 10kW solar array on a household that uses 15kWh a day is going to export heavily and get clipped by falling feed-in tariffs. Pair that with an undersized 5kWh battery and you’ve built a system that’s good at neither job. The households getting the best combined payback are the ones sizing solar to roughly 1.2 to 1.5 times daytime consumption, not to the roof’s full physical capacity, and sizing the battery to actual evening load rather than the biggest unit the installer had on the van that week.
This is where the Consumer Energy Resources roadmap work matters more than people realise, not as policy trivia but because it’s quietly setting the rules for how export limits, dynamic connection agreements and battery dispatch will interact over the next few years. I went through what that means for a rooftop system in our explainer on the Consumer Energy Resources roadmap, and the short version is that sizing decisions made today should assume tighter, more dynamic export management, not looser.
Where this actually lands #
If I’m genuinely asked which cheque to write first and there’s only one cheque, it’s solar, every time, for a household starting from nothing. The payback is faster, the technology risk is lower: panels degrade slowly and predictably over 25-year warranties, batteries degrade faster and the chemistry is still maturing, and the economics don’t rely on a rebate programme that could be adjusted by a future budget.
A battery earns its spot once you’ve already got solar, you’re genuinely running into export curtailment or a hard connection limit, and you can stomach a payback window closer to a decade even after rebates. Anyone telling you a battery pays for itself in under six years on a standard tariff without those conditions is quoting the brochure household, not yours. The spread between feed-in tariffs and evening rates is the only thing that makes a battery worth anything, and that spread has been getting squeezed from both directions for a few years now, not widening the way some sales pitches still assume.
My honestly mildly unpopular opinion with some installer colleagues: I think a chunk of the current battery rush is being driven more by rebate timing anxiety than by sound household economics, and that’s worth sitting with before signing anything.
– Priya Nair, Solar & Distributed Energy Correspondent
Photo by Watt A Lot on Unsplash