How to Size an Off-Grid Solar System: Load → Panels → Battery → Days of Autonomy

Most off-grid solar quotes get built backward. Someone picks a battery they can afford, bolts on however many panels fit the roof or budget, and finds out during the first gray week in December whether they guessed right. The other backward version costs more: an installer pads every number for safety and sells a system 30–40% bigger — and $15,000–$25,000 pricier — than the site needed. Real off-grid sizing only runs one direction: the load comes first, in kilowatt-hours per day, and everything else — array kW, battery kWh, inverter size — is arithmetic performed on that number. This is the whole method, worked through one real example, with the US 120/240V vs. European 230V split and the worst-month sun-hour numbers most guides skip.

Step 1: Turn the Load List Into One Daily Number

Everything downstream depends on this number being honest. The shape of a load audit never changes: running watts times hours actually used, summed for a full day — except anything that cycles on its own (a fridge, a freezer), which gets listed at its spec-sheet daily energy figure instead, since "running watts × 24 hours" wildly overstates a compressor that's only drawing current a third of the time.

Load	Daily energy
Lighting, 10 LED fixtures (~4h/day)	0.35 kWh
Refrigerator (efficient, off-grid rated)	1.20 kWh
Chest freezer	0.90 kWh
Well pump, ½ HP (30 min/day)	0.40 kWh
Satellite internet + router (24/7)	1.40 kWh
Washing machine (averaged across the week)	0.35 kWh
Laptop, phones, misc. electronics	0.30 kWh
Kitchen small appliances	0.30 kWh
Miscellaneous / phantom loads (buffer)	0.60 kWh
Daily total	5.8 kWh/day

Step 2: Size the Array to Your Worst Month, Not the Annual Average

A peak sun hour (PSH) is a unit built for exactly this job: one PSH equals one hour of sunlight at the lab-standard 1,000 W/m² intensity a panel's rating is measured at. A site averaging 4 PSH a day is, for sizing purposes, getting the equivalent of 4 hours of full-rated output. Multiply a panel's rated watts by PSH and you get roughly what it produces that day.

The mistake almost every first-pass estimate makes is sizing off the annual average PSH — usually the only number a brochure or a quick calculator leads with. An off-grid system doesn't get to average across the year; it has to survive its single worst month, almost always December, with no grid to fall back on. Sizing to the annual average is exactly how a "properly installed" system still browns out every January.

Divide the daily load by a system efficiency factor to find how much the array actually has to produce — wiring, charge-controller conversion, inverter DC-to-AC conversion, and battery round-trip losses take a combined 20–22% out of every kWh, so most designers plan around 0.78 (78% system efficiency). Divide that figure by the worst-month PSH to get array size in kW DC:

Required array output (kWh/day) = Daily load ÷ 0.78
Array size (kW DC) = Required array output ÷ worst-month PSH

Applied to the cabin above: 5.8 ÷ 0.78 ≈ 7.4 kWh of array output needed. This site, at a northern New England latitude, averages roughly 2.6 peak sun hours in December — well below the ≈4.2 PSH the same roof sees averaged across the whole year. 7.4 ÷ 2.6 ≈ 2.86 kW DC, rounded up to the next whole panel count: eight 400W panels = 3.2 kW DC. That's the array.

No site-specific number yet? These are reasonable worst-month planning figures — pull your exact PSH from the calculator below or a PVWatts (US) / PVGIS (Europe) lookup before you buy anything.

Region	Typical worst-month (Dec.) PSH	What it means for sizing
US Sunbelt — AZ, NM, S. TX, S. FL, S. NV	4.0–4.7h	Smallest arrays per kWh of load anywhere in the US
US mid-latitude — Mid-Atlantic, Midwest, high-desert Rockies	2.4–3.2h	Array roughly 40–60% bigger than Sunbelt for the same load
US cloudy/northern — Pacific NW, New England, Great Lakes snowbelt	1.2–2.2h	Battery and generator typically do more of the winter work than the array
Southern Europe — S. Spain, Portugal, S. Italy, Greece	2.2–3.0h	Close to US mid-latitude numbers
Northern/Central Europe — Germany, Benelux, N. France, UK, Poland	0.6–1.4h	Among the lowest winter insolation of any developed off-grid market

Step 3: Size the Battery to Your Days of Autonomy

Days of autonomy is the number of consecutive bad-weather days the battery alone has to cover with zero help from the array. It's a design choice, not a fixed constant, and it hinges on one question: is there a backup generator? A site with one can plan for 1–2 days, because the generator picks up anything beyond that. A pure solar-plus-battery site with no fuel backup needs 3–5 days, especially anywhere the table above reads below 2 PSH.

Usable battery capacity (kWh) = Daily load × days of autonomy
Nameplate capacity = Usable capacity ÷ usable depth of discharge

Usable DoD is roughly 80% for LiFePO4 and roughly 50% for flooded or AGM lead-acid, since lead-acid's cycle life falls off a cliff past the halfway point.

Applied to the cabin: its backup generator handles anything past a two-day stretch, so 5.8 × 2 = 11.6 kWh usable. At 80% DoD, that's 11.6 ÷ 0.8 ≈ 14.5 kWh nameplate — in practice, a 48V rack built from 280Ah cells (48V × 280Ah ≈ 13.4kWh), or two smaller wall units stacked to roughly 15kWh. Skip the generator and autonomy doubles to 4 days, and the battery number doubles with it to roughly 29kWh — a trade worth running both ways before committing to either.

	LiFePO4	Flooded / AGM lead-acid
Usable DoD for sizing	≈80%	≈50%
Cycle life at that DoD	3,000–6,000+	300–800
Nameplate needed for 11.6 kWh usable	≈14.5 kWh	≈23.2 kWh
Rough 2026 cost	$280–420/kWh · €260–390/kWh	$120–180/kWh · €110–170/kWh

Step 4: Size the Inverter for the Surge — Then Match It to Your Voltage

The inverter is sized off a different number entirely: peak watts, not daily kilowatt-hours. Continuous rating is what it sustains as long as several loads run at once. Surge rating is what it survives for the second or two a motor takes to spin up — a well pump, a compressor, a washing machine motor can pull 2–4× its running watts for that instant, the same problem a whole-house standby generator has to survive on fuel instead of battery (see our generator sizing guide for that mirror-image case).

Applied to the cabin: baseline running loads — both compressors, lighting, internet, electronics — draw roughly 500W continuous. Add the well pump's 750W and you're at 1,250W; a 25% margin brings continuous sizing to about 1,600W. But the pump can pull 3× its running watts for the instant it starts — roughly 2,250W of surge — on top of the ~500W already running, pushing required surge past 2,750W. Quality inverters typically surge to around 2× their continuous rating for a second or two, which is why designers round up rather than sizing to the bare minimum: a 3.5kW-class inverter covers both numbers with real margin.

Voltage doesn't change any of the math above — a kWh is a kWh at 120V or 230V. It only changes which inverter you buy and how it's wired into the panel. In the US, most homes need 120/240V split-phase output, produced either by a purpose-built split-phase all-in-one (Sol-Ark, EG4, Schneider XW+) or by phase-locking a pair of 120V units. In Europe, single-phase 230V/50Hz — a Victron MultiPlus-II-class unit is the common reference — covers most homes; larger loads, or a local grid code requiring balanced phases, push some builds to 400V three-phase instead, at which point a single-phase inverter physically can't do the job.

Put It Together: One Cabin, Four Numbers

Chain all four steps and this is what a real spec looks like end to end:

Number	Value	Comes from
Daily load	5.8 kWh/day	Load audit — Step 1
Array	≈3.2 kW DC (8 × 400W panels)	Load ÷ efficiency ÷ worst-month PSH
Battery	≈15 kWh nameplate LiFePO4 (2-day autonomy)	Load × autonomy ÷ DoD
Inverter	3.5kW continuous, 120/240V split-phase	Peak concurrent watts + motor-start surge

That's the entire method — the arithmetic is genuinely this short. What takes time is getting an honest load list and a real worst-month PSH number for your exact site.

• Sizing the battery to a stated autonomy target, not a round number, is the only way to know what you're actually buying protection against.
• Sizing the inverter to the surge, not the running total, is what keeps a well pump from tripping the whole system on day one.

Running these four steps by hand is worth doing once, so you understand what's happening under the hood. After that, the Off-Grid Solar Sizing Calculator runs the same four steps against your exact loads and site in under five minutes, and outputs a spec sheet ready to hand an installer.

Open the Solar Sizing Calculator →

Once you've got your four numbers, matching them to hardware is the easy part. Here's what a system built to a spec like this costs to buy right now, where you are:

In the US, the 30% federal residential credit (Section 25D) expired for systems installed after December 31, 2025 — the figures above aren't tax-adjusted. Some states still run rebates, and several European countries run national solar incentives; check what's live in your area before you budget off someone else's number.