Small cabin — 3 kWh/day
Lights, fridge, laptop, water pump. A realistic weekend-cabin load.
1 kW array (3 × 400 W panels), 7.5 kWh bank
Array watts and battery bank capacity from your daily load — with system losses and depth of discharge included, which is where most sizing goes wrong.
In watt-hours per day. A 100 W fridge running 8 h is 800 Wh.
Use your WORST month, not the annual average. UK winter ≈ 1, Sahara summer ≈ 7.
Controller, wiring, heat, soiling. 25–30% is realistic for a good install.
Watts per panel. 400–550 W is typical for modern residential panels.
How long the bank must carry the load with no sun at all.
Usable depth of discharge for LiFePO4 is taken as 80%, so the bank is sized at 7.50 kWh nameplate to deliver 6.00 kWh usable.
Charge controller must handle about 25.0 A on the battery side. Specify at least 25% above that — panels can briefly exceed their rating in cold, bright conditions.
Lights, fridge, laptop, water pump. A realistic weekend-cabin load.
1 kW array (3 × 400 W panels), 7.5 kWh bank
Northern latitude in December. See why sizing on the annual average leaves you dark.
2.67 kW array — nearly three times the summer figure
Same load, cheaper batteries. Compare the bank size you have to buy.
12 kWh nameplate needed to deliver the same 6 kWh usable
A modest mobile setup. Watch the note about bank voltage.
286 W array, 208 Ah at 12 V
When to use this: planning a cabin, van, boat or remote installation that has no grid connection, or checking whether an existing system is undersized for the load you keep adding to it.
There are two independent calculations, and both start from your daily energy use:
Array watts = daily Wh / (peak sun hours × efficiency)Bank Wh = daily Wh × autonomy days / depth of discharge
The array replaces what you use each day. The battery carries you through the night and through the days when the array cannot keep up. Sizing one without the other produces a system that either wastes generation or runs flat.
Panels are rated at 25 °C under 1,000 W/m² in a laboratory. Real installations never see that. The losses compound:
| Loss | Typical | Why |
|---|---|---|
| Panel temperature | 8–15% | Output falls ~0.4% per °C above 25 °C; panels reach 50–65 °C in sun |
| Charge controller | 3–10% | MPPT is ~95% efficient; PWM much worse |
| Wiring | 2–5% | Voltage drop, particularly on low-voltage DC runs |
| Soiling and shading | 2–5% | Dust, pollen, bird droppings, partial shade |
| Mismatch and tolerance | 2–3% | Panels in a string perform to the weakest |
A well-built system loses 25–30% overall. Sizing an array on the nameplate figure alone undersizes it by roughly a third.
A 100 Ah lead-acid battery does not give you 100 Ah. Discharging lead-acid below about 50% state of charge shortens its life sharply — cycle life roughly triples going from 80% depth of discharge to 50%.
| Chemistry | Usable DoD | Typical cycles | Notes |
|---|---|---|---|
| Flooded lead-acid | 50% | 500–1,200 | Cheapest, needs watering and ventilation |
| AGM | 50% | 500–900 | Sealed, no maintenance, tolerates cold poorly |
| Gel | 50% | 700–1,200 | Sensitive to charge voltage |
| LiFePO4 | 80% | 3,000–6,000 | Needs a BMS; will not charge below freezing |
To deliver 6 kWh usable, you need a 12 kWh lead-acid bank or a 7.5 kWh lithium one. Lithium costs more per nameplate kWh and usually less per delivered kWh over the bank's life.
Peak sun hours are not daylight hours. They are the equivalent hours at full 1,000 W/m² irradiance that deliver the same daily energy. And they swing enormously by season:
| Location | Summer | Winter | Ratio |
|---|---|---|---|
| Northern UK / Scandinavia | 4.5 | 0.5 | 9× |
| Southern UK / Germany | 5.0 | 1.0 | 5× |
| Mediterranean | 6.5 | 2.5 | 2.6× |
| South-west USA | 7.5 | 4.0 | 1.9× |
| Equatorial | 5.5 | 5.0 | 1.1× |
Size on your worst month, not the annual average. A system sized on the average will work beautifully from April to September and leave you in the dark every December. In high-latitude installations the honest answer is often a generator for the winter rather than an array sized for it — the December array would be idle for eight months of the year.
Daily load:
| Appliance | Power | Hours/day | Wh/day |
|---|---|---|---|
| 12 V fridge | 45 W | 10 (compressor duty) | 450 |
| LED lighting | 40 W | 5 | 200 |
| Laptop + router | 60 W | 8 | 480 |
| Water pump | 120 W | 0.5 | 60 |
| Phone charging, misc | — | — | 110 |
| Total | 1,300 Wh |
At 4 peak sun hours with 25% losses:
Array = 1300 / (4 × 0.75) = 433 W → two 250 W panelsBank (LiFePO4, 2 days) = 1300 × 2 / 0.8 = 3,250 Wh → 68 Ah at 48 V
Note how the fridge dominates. It is the only load running around the clock, and in most off-grid systems refrigeration is the single largest consumer. Replacing an inefficient fridge is usually cheaper than adding the panels to feed it.
Higher voltage means proportionally lower current for the same power. Current sets conductor size, and resistive losses scale with its square:
| System | Current for 2 kW | Suits |
|---|---|---|
| 12 V | 167 A | Vans, boats, small systems under ~1 kW |
| 24 V | 83 A | Cabins, systems to ~3 kW |
| 48 V | 42 A | Anything larger — now the default for house systems |
167 A needs cable in the 50 mm² range for a short run. 42 A needs 10 mm². The copper saving alone often pays for the higher-voltage equipment.
Multiply each appliance by the hours it runs, then total. Be honest — underestimating the load is the most common cause of an off-grid system that disappoints.
Peak sun hours vary hugely by season and latitude. A system sized on the annual average will run short every winter.
Charge controller inefficiency, wiring drop, panel heat and soiling combine to 25–30% in a well-built system. Leaving this out undersizes the array by a third.
Autonomy is how many sunless days the bank must cover. Chemistry sets how much of the nameplate capacity you can actually use — 50% for lead-acid, 80% for LiFePO4.
Divide your daily watt-hours by peak sun hours and by system efficiency to get the array watts, then divide by the panel rating and round up. 3,000 Wh/day at 4 sun hours with 25% losses needs 3000 / (4 × 0.75) = 1,000 W, which is three 400 W panels.
The number of hours per day that solar irradiance would need to be at the standard 1,000 W/m² to deliver the same total energy as the actual day. It is not daylight hours. A location with 4 peak sun hours might have twelve hours of daylight, most of it at much lower intensity.
Panels are rated at 25 °C in laboratory conditions and never achieve that in the field. Real losses come from panel temperature (panels lose about 0.4% per °C above 25 °C), charge controller inefficiency, wiring voltage drop, dust and soiling, and module mismatch. Together these typically total 25–30%.
The fraction of a battery’s nameplate capacity you can regularly use without shortening its life. Lead-acid should not go below 50% state of charge for reasonable cycle life; LiFePO4 tolerates 80% discharge routinely. A 100 Ah lead-acid battery therefore provides 50 Ah of usable capacity, not 100 Ah.
Two to three days is typical for a system with a generator backup, and four to five for one without. More autonomy means a larger, more expensive bank that spends most of its life partially charged — which is itself bad for lead-acid. Beyond about five days, a generator is usually cheaper than more battery.
Usually, on cost per usable kWh over the bank’s life. LiFePO4 costs more per nameplate kWh but delivers 80% of it rather than 50%, tolerates several thousand cycles rather than several hundred, and does not degrade from partial-state-of-charge operation. The purchase price is higher; the cost per delivered kWh is generally lower.
Higher voltage means proportionally lower current for the same power, which cuts conductor size and resistive losses dramatically. 12 V suits small mobile systems under about 1 kW. 24 V works up to roughly 3 kW. Anything larger should be 48 V — at 5 kW, a 12 V bank would carry over 400 A.
Divide the installed array watts by the bank voltage to get the controller current, then add at least 25%. Panels can briefly exceed their rated output in cold, bright conditions, and a controller that current-limits is wasting the array you paid for.
Last reviewed .
Diameter, cross-section, resistance and current rating for any AWG size — with separate free-air and in-conduit ampacity, because they differ by a factor of two.
ΔVVoltage drop for DC, single-phase and three-phase runs in copper or aluminium, sized in mm² or AWG, checked against IEC and NEC limits.
WPower from any two of voltage, current and resistance — then energy in kWh and what it costs to run.
3ØkW, kVA and kVAr from line voltage and current, with star and delta phase values and what power factor correction would save you.