Guide · off-grid power
Solar and battery sizing for remote equipment: size for winter.
Short version: most remote solar setups die in July, not January. They get sized on a sunny day or a yearly average, and the first stretch of grey winter days runs the battery flat. The method that survives: measure the load, size the battery for two or three cloudy days, and size the panel to recharge it in winter sun while still running the load. Here is the maths, worked on a small wireless link.
Last updated 2 July 2026 · by Alien IT Solutions
Winter is the design case
A remote power system has one job: keep the gear alive through the worst week of the year, and in Australia that is almost always a winter week. Short days, low sun, cloud hanging around, while a wireless link draws the same power at 3am in July as at noon in January.
Most of Australia sees 5 to 7 peak sun hours a day in summer; the south in winter sees roughly 2 to 4. Same panel, same load, less than half the energy coming in. Every number below gets worked at the winter figure.
Step 1: measure the real load
Nameplate ratings are worst-case maximums, not what the gear draws. A radio badged at 24W might pull 7W on the bench. Measure it: an inline DC power meter for 12V gear, a plug-in energy meter for anything on a power supply, running under normal traffic, because the draw moves.
Count everything hanging off the battery: radio, router or switch, any PoE injector, and the losses in whatever DC-DC converter sits in the chain. A radio at 6 to 7W plus a small router at 3 to 4W lands close to 10W all up. Round up, never down.
Step 2: daily energy
Watts times hours. A link runs around the clock, so: 10W × 24h = 240Wh per day, sun or no sun. Gear that runs all night is what makes remote power different from a garden light.
Step 3: size the battery for the cloudy days
The battery's job is autonomy: carrying the load through cloudy weather when the panel produces next to nothing. Two to three days is the minimum I would design for, longer for a link people depend on. Three days at 240Wh per day means 720Wh of usable storage.
Usable is the word that catches people. Lead-acid only gives you about half its rated capacity before you start shortening its life, so 720Wh usable means about 1,440Wh of battery, 120Ah at 12V. Lithium (LiFePO4) gives you 80 to 90 percent, so the same 720Wh needs roughly 800 to 900Wh, call it 67 to 75Ah at 12V; a 100Ah unit buys comfortable headroom.
Cold takes a further bite: capacity drops with temperature, and most lithium batteries will not accept charge below freezing unless they have built-in heating. If the site gets frosts, add margin and keep the battery out of the weather; more on placement below.
Step 4: size the panel to recharge while still running the load
Peak sun hours are the honest way to rate a day of sunshine: everything the sun delivers from dawn to dusk, squashed into the equivalent hours of full-strength sun. Southern Australia in winter runs roughly 2 to 4; check the solar data for your own area rather than borrowing someone else's number.
The panel maths: daily energy divided by winter peak sun hours. 240Wh ÷ 4 = 60W in a good winter spot; 240Wh ÷ 2 = 120W in a poor one. Ideal-world figure, not the answer yet.
Step 5: the multipliers everyone skips
Real panels never deliver the sticker rating. Dust on the glass, an off angle, voltage drop in the cable, controller losses and charging inefficiency all shave the harvest. Assume you keep about 70 percent and divide by 0.7: the 60 to 120W ideal becomes roughly 86 to 171W just to break even on a winter day.
Break-even is still not the target. After a cloudy stretch the panel has to run the load and refill a low battery at the same time, or the battery never sees full before the next front arrives. Add about 50 percent for recovery: roughly 130 to 260W for this 10W load. If it were my gear in southern Australia, I would fit 200W and stop thinking about it.
On the controller: a PWM unit drags the panel down to battery voltage and throws the difference away. An MPPT unit converts that surplus into charging current and typically harvests 10 to 30 percent more, with the biggest gains in cold, low-light winter days. On a small system the price gap is trivial. Fit MPPT.
The worked example in one place
A small wireless link, measured at 10W. Ranges, not product specs: your site and your winter move the numbers.
The load
10W measured: radio plus router plus conversion losses. 10W × 24h = 240Wh per day, cloud or not.
Battery, lead-acid
Three days of autonomy = 720Wh usable. At 50 percent usable that is about 1,440Wh of battery, roughly 120Ah at 12V.
Battery, lithium
The same 720Wh at 80 to 90 percent usable needs about 800 to 900Wh, roughly 67 to 75Ah at 12V. A 100Ah LiFePO4 gives headroom for cold and ageing.
Panel
240Wh ÷ 2 to 4 winter peak sun hours, then ÷ 0.7 for losses, is about 86 to 171W to break even. Add recovery margin: 130 to 260W. I would fit 200W.
Mounting for winter sun, and where the battery lives
Face the panel true north and tilt it steeper than you think, roughly your latitude plus 10 to 15 degrees. That points it at the low winter sun, and a steep panel sheds dust and rain-cleans itself. Watch shading harder than orientation: a shadow across even one row of cells craters the whole panel's output, and winter shadows are the longest of the year.
The battery wants the opposite: out of the sun and out of the frost. A ventilated, insulated box on the shaded side of the mount, up off the ground, does the job. Heat ages a battery fast in summer, cold cuts its capacity in winter, and a battery strapped to a sun-struck steel post gets the worst of both.
The three ways these systems die
Same three failures, over and over. All sizing problems, not equipment problems.
The panel that worked all summer
An undersized panel keeps the battery full through long summer days, then falls behind in winter. The battery drains a little further each grey day until it flattens. Nothing broke; the sums were done in January.
The battery that never gets full
If the panel only just covers the load, the battery sits half-charged for months. Lead-acid sulphates and dies early in exactly this pattern. The recovery margin in step 5 exists to prevent it.
Load creep
Sized for a 10W link, then a camera and a second radio landed on the same battery. Every extra watt is another 24Wh a day. Redo the sums whenever the load changes.
Who works this out for you
Long Range WiFi is a service of Alien IT Solutions, 18 years of networks and wireless in places without a power point in sight. Solar-powered relays are part of the day job: see off-grid & solar links, or the long-range wireless guide.
Questions people ask
What size solar panel do I need to run WiFi gear off grid?
Work from the load. Measure the real draw in watts, multiply by 24 for daily watt hours, divide by winter peak sun hours, then divide by 0.7 for losses. A 10W link lands around 90W to 170W to break even, so fit 150W to 250W for recharge headroom.
How many days of battery does a remote solar setup need?
Two to three days of autonomy minimum, more for a critical link. The battery carries the whole load through cloudy weather. Size on usable capacity, not the label: lead-acid gives about half its rating, lithium 80 to 90 percent.
Is lithium or lead-acid better for remote solar?
Lithium, in most cases. You can use 80 to 90 percent of its capacity and it tolerates the partial charging winter solar delivers. Lead-acid gives about half its rating and ages fast if it sits undercharged; if you fit it anyway, oversize it.
What are peak sun hours?
A day's total sunshine squashed into the equivalent hours of full-strength sun. Winter in southern Australia is roughly 2 to 4 peak sun hours, so a 100W panel yields roughly 200Wh to 400Wh a day on paper, less after losses. Check local solar data rather than guessing.
Why did my solar setup work all summer then die in winter?
Summer hides an undersized panel: long days keep the battery full. In winter the daily harvest drops below the daily load, the battery slips further behind every grey day, then goes flat. Size for winter and summer looks after itself.
Do I need an MPPT controller or is PWM fine?
Fit MPPT. PWM drags the panel down to battery voltage and wastes the difference; MPPT converts it and typically harvests 10 to 30 percent more, with the biggest gains in cold, low-light weather. The price difference on a small system is trivial.
Need a link where there is no power?
Tell us what has to run and where. We measure the load, size the system for your winter, and hand over a link that stays up in July.
Get a quote