Sizing battery capacity for self-consumption vs backup: they're different jobs

I got an email a few months back from a reader in Ballarat who'd just had a 13kWh battery installed and was furious it only ran her fridge, a couple of lights and the router for about five hours during a grid outage. She'd been told the battery would "cover blackouts." Technically true. Practically, she'd sized for the wrong job.

This is the single most common mismatch I see in battery installs, and it's rarely the installer's fault entirely — it's a framing problem. Home batteries get sold as one product but they're actually doing two quite different jobs, and the sizing math for each one barely overlaps. Get clear on which job matters most to you before you sign anything, because a battery optimised for self-consumption is a mediocre backup unit, and a battery optimised for backup is often oversized and underused for daily solar shifting.

Self-consumption is a daily arbitrage problem

If your main goal is soaking up midday solar so you're not exporting it for a few cents a kilowatt-hour and then buying it back at night for thirty-plus cents, you're solving an arbitrage problem. The battery's job is to capture the gap between what your panels produce and what your house uses during daylight, then discharge that stored energy across the evening peak.

The right size here is a function of your daily export surplus, not your total load. I usually tell people to pull twelve months of interval data from their inverter app or retailer portal and look at a typical sunny weekday — say, a 6.6kW system on a north-facing roof in a decent climate zone might export 8-12kWh on a clear autumn day after the house has taken what it needs. That's roughly your target usable capacity, not the full 40kWh some installers will try to spec "for future-proofing."

Oversizing for self-consumption alone is genuinely wasteful. A battery that's only ever cycled to 60% because your daily surplus doesn't fill it is dead capital sitting in your garage, degrading slowly on the shelf whether you use it or not. LFP cells handle partial cycling well, as I've covered in our LFP vs NMC chemistry piece, but there's still no financial upside to capacity you never touch. Get the charge and discharge windows dialled in properly — I wrote a separate piece on setting those windows correctly because a mis-set discharge window will quietly waste the storage you do have.

Backup is a duration-and-priority-circuit problem

Backup sizing asks a completely different question: not "how much solar surplus do I have" but "what absolutely needs to keep running, for how long, and does it matter if the sun isn't shining that day." A grid outage in July with three overcast days back to back is a different scenario to a five-hour summer outage with a full battery and six hours of strong sun the next morning.

This is where critical-circuit panels matter more than raw kWh. Most residential battery systems, including the Powerwall 3 and the BYD Battery-Box Premium HVS, can be wired either as whole-of-home backup or as a partial "essential loads" backup through a sub-panel or backup circuit. Whole-home backup sounds appealing until you realise your ducted heat pump, electric oven and pool pump will drain a 13kWh battery in under three hours on a hot day. Essential-loads backup — fridge, some lighting, router, a few power points, maybe a well pump if you're rural — stretches the same battery to a day or more.

My own set-up runs a partial backup circuit. Fridge, office, and the EV charger are all off the priority panel, but the ducted system isn't, because I'd rather ration comfort during an outage than lose the battery in two hours flat. That's a judgement call every household needs to make honestly rather than deferring entirely to whatever the installer defaults to.

Why one battery struggles to do both jobs well

A battery that's perfectly sized for daily self-consumption — say 10kWh usable, matched to your typical surplus — will get you through maybe half a day of essential-loads backup if the outage happens in the evening when the battery's already partly discharged from the day's cycling. That's the scenario that caught out the Ballarat reader. Her battery had done its job perfectly for eleven months of the year, then failed to meet an expectation nobody had actually sized for.

Conversely, a battery sized generously for multi-day backup resilience is going to sit mostly full for self-consumption purposes on a normal day, because you're deliberately reserving headroom for an outage that statistically isn't going to happen that week. That reserved capacity isn't earning you arbitrage savings; it's insurance. Which is fine, as long as you know that's what you're paying for.

The honest fix is to decide your priority first, then size around it, accepting the trade-off on the other axis. If self-consumption savings are the main driver, size tight to your daily surplus and treat any backup capability as a bonus, not a guarantee. If resilience is the main driver — bushfire-prone area, unreliable rural grid, someone in the house on medical equipment — size for duration on essential circuits and treat the self-consumption savings as secondary.

Reserve settings are where the two goals actually fight

Most hybrid inverter and battery systems, including the 8kW-class hybrid inverters now standard on new installs, let you set a minimum state-of-charge reserve — effectively a floor the battery won't discharge below during normal daily cycling, kept in case of an outage. This single setting is the actual lever between "maximise savings" and "maximise resilience," and it's one most homeowners never touch after install day.

Set the reserve at 0-10% and you're optimising almost entirely for self-consumption; you'll squeeze every dollar out of daily cycling but a night-time outage catches you with an empty tank. Set it at 30-40% and you're keeping meaningful backup headroom every single day, at the cost of that reserved energy not doing arbitrage work for you. There's no universally correct number here — it depends on how exposed your area is to outages and how much that resilience is worth to you personally. I'd actually push back a little on the "set and forget" advice some installers give: reserve settings are worth revisiting seasonally, tighter in stable weather months, looser heading into storm season.

What this means for battery capacity you actually buy

Rather than picking a battery size off a brochure, work backwards from both numbers and take the larger one, understanding what you're overpaying for on the smaller axis. If your self-consumption target is 10kWh usable and your essential-circuit backup target for a realistic outage duration is 14kWh, buy toward the 14kWh and accept you'll rarely draw it that deep on a normal day. If they're roughly the same, you're in luck and can size once.

Where the numbers diverge a lot — say 8kWh self-consumption target against a 25kWh multi-day resilience want — that's genuinely a two-battery-bank conversation, or a case for asking your installer directly whether a modular system (stacking additional battery units later) makes more financial sense than buying all the capacity upfront. Our decision framework piece goes into this trade-off in more depth if you're still at the "should I even bother" stage.

It's also worth checking what your state or territory actually incentivises before you finalise a size — the Cheaper Home Batteries Program and state-based schemes like the NSW VPP Incentive can shift the economics meaningfully depending on whether you're enrolling the battery in a virtual power plant, which usually assumes it's cycling daily rather than sitting in reserve. A battery locked at 40% reserve for backup peace of mind may not be much use to a VPP operator wanting your capacity available for grid events. Worth asking the installer directly how VPP enrolment and your reserve setting interact, because the two goals can genuinely conflict.

Getting a straight answer out of your installer

When you're getting quotes, ask the installer to show you two numbers separately: expected daily self-consumption capture in kWh based on your actual production and consumption data, and expected essential-circuit backup duration in hours at your reserve setting. If they can only give you one number, or if the quote just lists a kWh figure with no explanation of which job it's solving, that's a red flag worth raising — our piece on what a proper battery quote should itemise covers what a fully worked-through quote actually looks like line by line.

The Clean Energy Council's consumer guidance on battery storage is a decent independent reference point if you want a second opinion beyond the installer's sales pitch, and it's worth a read before you sign. The core idea, though, is simple enough to hold in your head at the kitchen table: solar self-consumption and backup power are different jobs with different maths, and the battery that's perfect for one is rarely perfect for the other. Decide which one you're actually buying before the capacity gets locked in.