I’m planning to build a 48V off-grid solar system for my home. How do I correctly size my battery bank for 1–2 days of backup power?

Good sizing question — this catches a lot of people out.

Start with your daily consumption in Wh. Pull your electricity bill and divide annual kWh by 365, or better yet, stick a plug-in energy monitor on your consumer unit for a week. Most UK homes land somewhere between 8–15 kWh/day depending on what you're running.

For 1–2 days backup, multiply that figure by 2, then apply your depth of discharge limit. With lithium (LiFePO4), you can realistically use 80–90% usable capacity. With lead-acid, drop that to 50% or you'll murder the cells within a couple of years.

Quick example:

• Daily use: 10 kWh
• 2-day backup: 20 kWh
• LiFePO4 at 80% DoD → 20 ÷ 0.8 = 25 kWh usable battery capacity
• At 48V, that's roughly 520Ah at 48V

For a home system at that scale, I'd be looking at Fogstar Drift 48V server rack batteries or building a bank from Eve/CATL cells with a decent Daly or ANT BMS. Victron Cerbo GX ties it all together nicely for monitoring.

Don't forget to factor in inverter efficiency losses (typically 5–10%) and seasonal solar variation — a UK winter will absolutely punish an undersized bank.

What loads are you planning to keep running during outages? Heating, EV charging, and immersion heaters change the maths dramatically. Worth listing them out so we can give more specific advice.

Others here will have different takes on the Do

@CamperGeek ...or just accept that whatever number you calculate, your actual usage will be 40% higher once you've got reliable power and start boiling kettles at midnight like a feral gremlin 🫖

Don't forget depth of discharge matters enormously — your shiny new Fogstar Drift LiFePO4 cells can do 80–90% DoD all day, but if you're buying secondhand lead-acid off Facebook Marketplace like some of us definitely haven't, stick to 50% or you'll be buying replacements by Christmas.

Real number: usable Wh needed ÷ DoD% = actual bank size required.

@OffGridGeek is painfully right on that 40% figure — learned that the hard way when I wired up my shepherd's hut and suddenly had a coffee machine, a fan heater "just for five minutes," and an EV trickle charge all happening simultaneously.

One thing nobody's mentioned yet: depth of discharge matters enormously for sizing. With LiFePO4 (I run Fogstar Drift cells) you can comfortably use 80–90% usable capacity. With AGM, you're effectively halving your bank at 50% DoD.

So the formula becomes:

(Daily Wh × backup days) ÷ DoD% = minimum bank capacity

If you're on a boat or hut like me, also factor in Peukert losses — if you're drawing hard, you get less out than the nameplate suggests. Victron's battery sizing calculator handles this nicely if the maths feels daunting.

@ExFirefighter11 The 40% buffer is real, but there's another factor people consistently overlook: depth of discharge limits.

A 10kWh battery bank isn't 10kWh of usable capacity. With LiFePO4 you're typically looking at 80–90% usable, lead-acid it's more like 50%. So factor that into your sizing before you multiply by your backup days.

Quick formula I use:

(Daily Wh × backup days) ÷ DoD % = minimum nameplate capacity needed

For my static caravan setup I sized assuming 80% DoD on Fogstar cells, which meant buying roughly 25% more capacity than the raw Wh calculation suggested. Glad I did.

Also — 48V systems specifically, double-check your BMS can handle your expected charge/discharge currents. Victron kit plays nicely with most decent BMS units but it's worth confirming before you commit.

Good points all round. Can I ask a dumb question though — does the DoD limit change depending on battery chemistry?

Running a static caravan setup on LiFePO4 (Fogstar Drift cells) and I've seen figures anywhere from 80% to 100% usable capacity thrown around. Bit confusing when you're trying to nail down actual usable kWh for your bank size calculation.

Also — does temperature affect this significantly? My van isn't heated over winter so I'm wondering if I need to factor in cold weather degradation on top of the DoD limit, or if that's already baked into manufacturer specs somehow.

Feels like every variable you pin down reveals two more underneath it.