How big a solar battery do you need?

"What size battery do I need?" is usually the first question people ask about off-grid solar, and the answer you most often find online is a rule of thumb. This guide shows you how to work it out properly, step by step, with a worked example you can copy. It is for anyone planning a battery for a cabin, shed, van, boat or backup, with no electrical background needed.

The short version

• Start with energy per day, in watt-hours (Wh): each appliance's watts times the hours it runs, added up.
• Battery size is energy too. Compare Wh with Wh. A battery's amp-hours (Ah) mean little until you multiply them by its voltage.
• You cannot use 100 % of a battery's label. Allow for the usable share (depth of discharge), small losses, ageing and cold. Together they can nearly double what you buy.
• The biggest question is how many days in a row the sun may not keep up. That depends on where you are and when you use the system.
• Rules of thumb for those days vary widely. One university course gives 3 days for one reliability level and 10 days for a higher one.
• Our quiz replaces the guess: it runs your load against many years of real hourly weather for your location and picks battery and panels that meet a reliability target.

Step 1: Work out your daily energy in watt-hours

Two words first. Watts (W) are how fast an appliance uses energy, like how fast water leaves a tap. Watt-hours (Wh) are how much energy it uses over time, like how much water ends up in the bucket. A kilowatt-hour (kWh) is 1,000 Wh, the unit on your electricity bill.

The US Department of Energy gives a simple formula: watts × hours used per day = watt-hours per day (the DOE then divides by 1,000 to get kWh). It also gives three tips worth knowing:

• The wattage on an appliance's label or nameplate is the maximum it draws. Many appliances use less most of the time, so the label figure is a cautious starting point.
• If only amps are listed, multiply amps by volts. The DOE notes that most US appliances use 120 volts; elsewhere, use the voltage printed on the label.
• A fridge is switched on all day but its motor cycles on and off. The DOE suggests dividing the time it is plugged in by three to estimate the hours at full wattage. Better still, use the yearly kWh on its energy label.

Write down every item you want to run, its watts and its hours per day. Include small things that run all the time, such as a Wi-Fi router.

Step 2: A worked example

Here is a small cabin load. The figures come from the sources named; the arithmetic is our calculation.

• Fridge. One ENERGY STAR certified top-freezer fridge of 17.6 cubic feet (about 500 litres) is rated at 360 kWh a year. 360,000 Wh ÷ 365 days = about 986 Wh a day. Use the figure on your own fridge's label (the yellow EnergyGuide label in the US, the energy label in the UK).
• Lights. ENERGY STAR says an 800-lumen bulb gives the light of an old 60-watt bulb. Two retail listings for 800-lumen LED bulbs show 8 W and 9 W. Four 9 W bulbs for 5 hours: 4 × 9 × 5 = 180 Wh.
• Laptop. Use the watts printed on your charger. In our example the charger says 65 W and the laptop is used 3 hours: 65 × 3 = 195 Wh. Because the label is a maximum, this is on the cautious side.

Total: 986 + 180 + 195 = 1,361 Wh a day, about 1.4 kWh (our calculation from the figures above).

Two notes. If your appliances run on household AC power from the battery, the inverter (the box that turns the battery's DC power into AC) uses some energy as well; our quiz adds it for you. And appliances that heat (kettles, heaters, hair dryers) or cool (air conditioners) use far more than anything in this list. The DOE's own kettle example is 1,500 W for one hour a day, 1.5 kWh, more than our whole cabin.

Step 3: Amp-hours versus watt-hours

Battery labels usually show amp-hours (Ah), which measure how much electric charge the battery holds. To get the energy in watt-hours, multiply by the battery's voltage:

Wh = Ah × volts

One maker's datasheet lists a 12.8 V battery of 100 Ah as 1,280 Wh, and a 25.6 V battery of 100 Ah as 2,560 Wh. Same amp-hours, twice the energy. So two batteries both labelled "100 Ah" can be very different sizes. Always compare in Wh. (A "12 V" lithium iron phosphate battery is 12.8 V nominal; "nominal" means its typical, named voltage.)

Step 4: Usable capacity and depth of discharge

Depth of discharge (DoD) is how much of the battery you empty before recharging. Taking 900 Wh out of a 1,000 Wh battery is 90 % DoD.

Deeper cycles wear a battery faster. The same datasheet rates its lithium iron phosphate (LiFePO4) batteries at:

• 2,500 cycles at 80 % DoD,
• 3,000 cycles at 70 % DoD,
• 5,000 cycles at 50 % DoD,

where the battery's life ends when its capacity falls to 80 % of new. A cycle is one discharge and recharge.

Our engine assumes you use up to 90 % of a LiFePO4 battery's rated capacity, and 50 % of a lead-acid battery's.

Step 5: Round-trip efficiency

Some energy is lost as heat on the way into a battery and on the way out. Round-trip efficiency is the share you get back. The datasheet above gives 92 % for its LiFePO4 batteries and about 80 % for an average lead-acid battery. In other words, for every 100 Wh the panels put into a LiFePO4 battery, you get about 92 Wh back.

Most of that loss is paid by the panels, which have to put in more than you take out. Our model counts a small loss each way: charging and discharging.

Step 6: Days without sun (autonomy)

On a sunny day your panels may run your appliances and refill the battery. On a grey day they may make only a fraction of that. The battery has to carry you through those stretches. The number of days a full battery can run your load on its own is called days of autonomy.

This is where rules of thumb get shaky. A Penn State university course on stand-alone solar gives two examples: about 3 days of autonomy for a system that meets the load 95 % of the time where the sun gives around 5 "peak sun hours" a day, and about 10 days for 99 % where it gives around 4. (A peak sun hour is one hour of full-strength sunshine; our sun hours guide explains it.) Going from 95 % to 99 % more than tripled the battery in that example. And both figures depend on the local climate.

Step 7: Ageing and cold

Two more things make a battery smaller in real life than on the label:

• Ageing. As we saw, a battery's rated life ends when it holds 80 % of its new capacity. If you want the system to still work in year eight, size for the old battery, not the new one. Our engine does this.
• Cold. The same datasheet shows a 100 Ah battery giving 80 Ah at 0 °C (32 °F) and 50 Ah at −20 °C (−4 °F). A battery in an unheated shed is a smaller battery in winter. LiFePO4 batteries also must not be charged below their minimum charge temperature (often about 0 °C / 32 °F). Our guide to LiFePO4 batteries in cold weather explains why.

Putting it together: the rule-of-thumb way

Back to our cabin, using 1,361 Wh a day and, as an example, 2 days of autonomy. Our calculation, step by step:

1. Energy for 2 days: 1,361 × 2 = 2,722 Wh out of the battery.
2. Loss on the way out: we take about half of the 8 % round-trip loss, so 2,722 ÷ 0.96 = 2,835 Wh. (Our quiz uses the same 96 % each way, 92 % for the round trip.)
3. Usable share: 2,835 ÷ 0.9 = 3,150 Wh rated capacity.
4. End of life at 80 %: 3,150 ÷ 0.8 = 3,938 Wh.
5. Cold battery at 0 °C (32 °F), 80 % capacity: 3,938 ÷ 0.8 = about 4,920 Wh.

That is about 385 Ah at 12.8 V, or about 192 Ah at 25.6 V. The simple answer "two days times 1,361 Wh" said 2,722 Wh. The careful one is about 1.8 times bigger, and it still rests on a guessed "2 days".

Why we size by reliability over many years instead

The weak point in every rule of thumb is that guessed number of days. In a sunny desert summer, one day may be plenty. In a Scottish winter, no affordable battery may be enough, because there may be weeks when the panels cannot keep up. More panels can also replace some battery: a bigger array refills the battery even on dull days.

So our quiz does not guess. It does this:

1. It takes many years of hour-by-hour weather for your location from PVGIS, the European Commission's solar database (for US places the record usually covers 2005 to 2015, for UK places 2005 to 2023; see the methodology).
2. It runs your load through every day of that record, for each combination of panel power and battery size, counting the panels' output, temperature, losses, the battery's usable share, efficiency, ageing and cold, including the days when a cold LiFePO4 battery cannot charge.
3. It counts how often the system would come up short. Reliability is the share of days in your season without an energy shortfall.
4. It offers three options: Minimum (90 % of days covered), Medium (97 %) and With margin (99.5 %). Battery sizes from half a day to seven days are tried; for each, the smallest panel array that reaches the target is found, and the battery is chosen where adding another half day would cut the panels needed by less than 5 %.
5. If even the largest battery and the panels that fit cannot reach the target, it says so honestly and suggests a backup such as a generator or grid charging, rather than an ever bigger battery.

The result also tells you how many days a year you might be short, in a typical and a worst year, so the reliability figure has a real meaning.

What to check before you buy

1. Make a load list in watt-hours per day, or use your electricity bill's kWh (the quiz accepts both).
2. Note the season you need power in. Winter-only and summer-only systems are sized very differently.
3. Compare batteries in Wh, not Ah, and check the voltage matches your system (12, 24 or 48 V).
4. Read the datasheet for the usable DoD the maker allows, cycle life and charge temperature range.
5. Decide where the battery will live: heated space or not.
6. Run the quiz and compare the three options.

What can go wrong

• Comparing Ah across voltages. A 100 Ah 24 V battery holds twice the energy of a 100 Ah 12 V one.
• Sizing for the average month. A system sized for the yearly average can fall short every winter.
• Forgetting the always-on loads. A small load that runs 24 hours a day can outweigh a bigger one used briefly.
• Counting the battery as new forever. It loses capacity with age.
• A cold battery that will not charge. Without heating, a LiFePO4 battery below its limit stores nothing, however sunny it is.
• A battery too big for the panels. A large battery that the panels can never refill gives less than it seems.

Questions people ask

Is bigger always better? No. Past a point, extra battery adds little reliability for its cost, because what runs it flat is a long dull spell that only more charging (panels, a generator, the grid) fixes. That is why we stop adding battery when the gain becomes small.

Can I size from my electricity bill? Yes, roughly. The quiz takes monthly or daily kWh and asks how much of it you want to cover, because bills often include heating, cooling or car charging that an off-grid system rarely runs.

How long will the battery last? Check the cycle life on the datasheet and the depth of discharge it assumes. Shallower cycles generally mean more of them.

Do I need an electrician? For a battery wired into a building's electrics, yes. For any system, if you are unsure about cable sizes and fuses, ask one. Batteries store a lot of energy, and a short circuit can start a fire.

Your next step

Put your appliances or your bill into the quiz. It sizes the battery and panels for your location from real weather, shows how many days a year you might be short, and lists every assumption so you can change it. The glossary explains every term used here, and the methodology page shows our full method and sources. If you are weighing an all-in-one unit against separate parts, read power station or DIY system.

Our results and guides are estimates and do not replace advice from a qualified electrician.