Do you need a solar battery yet?

Every few days another battery quote turns up in my feed. Someone posts the price, a dozen people tell them it’s a bargain or a rip-off, and the thread runs for fifty comments without anyone asking the question that decides it: how much of your own solar are you using right now?

I didn’t know my own number until I spent a year measuring. It came out at 47 per cent. It mattered more than anything on any quote.

The short answer

A home battery costs roughly $9,000 to $11,000 installed in Australia right now, with the federal discount already taken off.

Are solar batteries worth it at that price? For most homes, not yet — or more precisely, not first. There’s something cheaper to do before it, you almost certainly already own it, and doing it first doesn’t stop you buying a battery later.

The Australian Government’s own solar guide says you save the most by using the electricity your panels make, rather than exporting it. It also says plainly that a battery may not pay for itself within its lifetime.

Two numbers decide it for your house. I didn’t know either of mine: how much of my solar I already used instead of exporting, and how much power the house drew after dark. If a house goes quiet at night, there’s very little for a battery to catch.

The hot water tank is a battery

More than half of Australian homes heat water in an electric storage tank. Researchers at UNSW describe those tanks as thermal batteries, and note they often hold more energy than a home battery does.

A hot water tank really does store energy. It just stores it as heat, so the only thing you can spend it on is hot water. But that’s 15 to 30 per cent of a household’s power, second only to heating and cooling the house.

For decades those tanks ran overnight, on what your bill probably calls a controlled load. That made sense when night power was cheapest. It isn’t any more.

One thing to check first. A tank on a controlled load sits on a separate circuit from your solar, and the network decides when it runs, so it can’t use your roof whatever time it heats. Getting it onto your normal household circuit is step one, and it’s an electrician rather than a phone call. About a quarter of Australian homes have already done it.

Heating your water at midday instead of midnight is the closest thing to a free battery most households will ever get.

This has been tested on 18,000 homes

A South Australian trial moved water heating from night to day across 18,000 households for a year, funded by the Australian Renewable Energy Agency and run with AGL and UNSW. Close to half of all that water heating shifted into daylight. Only 0.3 per cent of households asked to opt out.

Eighteen thousand homes, a year of changed timing, and 54 people wanted it stopped. You don’t feel it in the shower.

A second study, of 9,300 New South Wales homes, gives the number the rest of this argument needs: an average household puts about seven units a day through its hot water system.

What a battery really costs per unit

A kilowatt-hour, or kWh, is one unit on your power bill. A 10 kWh battery holds about ten units.

The sticker price isn’t the number that decides it. What matters is what each stored unit works out to across the battery’s whole life — the total cost, divided by every unit it will ever deliver. Researchers at Imperial College London set out that calculation for exactly this case: a household managing its bill and making the most of its own solar.

So I read the Australian warranties for three of the most commonly installed batteries here, rather than take a comparison site’s word for it. Only one of the three publishes the figure the sum needs — the units it is guaranteed to deliver over its life. The other two promise a capacity percentage at year ten and nothing about energy at all. You cannot compare two quotes on price alone.

For that one maker’s 13 kWh model the figure is about 38,500 units. Divide $13,100 by that and each unit costs 34 cents.

But delivering 38,500 units inside ten years means pushing more than ten units through the battery every single day for a decade. Surplus every day, demand every evening, no cloudy fortnights, no winter. At a realistic eight units a day the same battery delivers about 29,000 units, and each one costs 45 cents.

Now the same sum on hot water

Seven units a day is about 25,500 over ten years — near enough the same energy the battery moves. You won’t catch all of it on solar; call it half, which is roughly what the South Australian trial managed.

A timer to do that costs about $250. Spread across the units it shifts, that’s 2 cents each.

Two cents, against forty-five.

A hot water system isn’t free. But you’re going to own one either way, so the only cost that belongs in this comparison is what you’d spend to change when it runs. A battery has no equivalent. Every dollar of it is a choice you didn’t have to make.

Two costs of doing it my way. Tanks lose up to a third of the heat you put in them. And a heating window starting at noon carries up to a 15 per cent chance of leaving you short on a winter day, so start it mid-morning.

What it looks like on my own bill

I’m in Canberra on a time-of-use plan. At the evening peak I pay 48 cents a unit; when I export I’m paid 4 cents. So a unit I store and use at dinnertime saves me about 44 cents — against 45 cents to put it there.

On realistic usage a battery roughly breaks even for me, in my best three hours, on one of the better tariffs in the country. Every unit that comes out later is worth less, and the cost per unit doesn’t change by the hour.

An ordinary electric tank — the old kind, with a heating element in it — shifted the same way is worth about $365 a year on that tariff. Except I don’t have one of those. I have a heat pump, which does the same job on about a third of the electricity, so there’s far less demand there to soak up my surplus. Mine is worth closer to $100 a year. Still a tenth of what the battery would charge me for the same favour.

A 48-cent peak is high and a 4-cent export rate is low, and that gap is what makes storage even close for me. On a flat tariff with a 10-cent feed-in it isn’t close at all. This is one Canberra home. The method travels better than the answer does: your evening rate, minus your export rate, against what a unit really costs you.

What hot water can’t do

Hot water is only ever part of the bill. Shift half of it and you’ve moved about 7 per cent of your consumption onto your own solar. A battery at eight units a day touches about 17 per cent, and unlike the tank it can absorb anything: the oven, the aircon, the lot.

So a battery reaches roughly twice as much of your bill as your hot water does, and costs about twenty times as much per unit to do it.

And you’ll still be exporting. Last year my roof sent nearly five thousand units to the grid. Move the hot water and that drops to around three and a half thousand. Better, but the surplus doesn’t go away. The sun is up when the house is empty, and there are only so many jobs you can move into the middle of the day.

That leftover is the real case for a battery. Not the discount deadline. Once you’ve moved what can be moved, whatever is still flowing out to the street at lunchtime is what a battery would catch. Which is why my answer is not first rather than never.

The strongest argument against me

The federal discount is shrinking to a published timetable. The Cheaper Home Batteries Program takes roughly 30 per cent off an installed battery, funded through certificates that step down every six months until 2030. They dropped in May 2026 and drop again in January 2027. The schedule is on the department’s page. If you’ve already sorted your daytime loads and you use a lot of power in the evening, waiting has a cost.

Grid prices moved the other way. They fell in most states from July 2026, which makes every battery take longer to pay for itself — though not in Canberra, where our own regulator put standing prices up 2.73 per cent.

Neither changes the order. The tank is cheaper whichever way those two move, and doing it first costs you nothing if you buy a battery later anyway.

For a typical home being quoted ten to fifteen thousand right now: sort your hot water first, watch what it does to your bill for one quarter, then decide about a battery with a real number in your hand instead of an estimate. You’ll either find you’ve solved most of it for a few hundred dollars, or you’ll buy the battery anyway — better informed, and no worse off.

Three things worth checking

Your phone will show you all of them.

How much are you exporting? Open your retailer’s or inverter’s app and look at a sunny weekday. If more than half your solar is heading out to the grid, you have surplus worth catching.

What does the house use after dark? Same app, evening hours. A home using 2 units between sunset and bed is a very different case to one using 10.

Is your hot water on a controlled load? Look for a second usage line on your bill at a cheaper rate. If it’s there, that’s the first job.

And if you are looking at a battery, ask for the warranty document, not the brochure. Ask whether it guarantees a number of units delivered or only a capacity percentage at year ten, and whether it still holds if you join a virtual power plant. Those answers are worth more than the discount.

To see the shape of your own solar day, our solar savings calculator will draw it for your postcode, with no sign-up and nothing stored. It runs entirely in your browser. The government also points people to SunSPOT, which is free and independent.

I build software that automates this sort of load-shifting, so I benefit when you decide better timing beats more hardware. The sources are linked and the arithmetic is yours to check.

Prices, rebates and tariffs in this article were checked in August 2026. The federal certificate value steps down again in January 2027.