Autonomy Days for Solar Street Lights: Sizing Battery and Panel with PVGIS Data
Tender specifications for solar street lights almost always ask for a number of “autonomy days”. It is an easy number to write down and an easy number to compare between offers. It is also easy to misread. This article explains what the number means, what it does not guarantee, and how to size the battery and the panel together, with worked numbers from the European Commission’s PVGIS simulations.
If you want the result for your own site straight away, calculate it with our solar sizing tool: it uses your time zone, your device location or a city you choose, and shows for every model how many hours per night and how many backup nights it reaches there.
What “autonomy days” means
Autonomy is the number of nights the battery can run the light with no solar charging at all. The battery size follows directly:
Battery (Wh) = nightly energy × autonomy days ÷ (usable depth of discharge × temperature derating)
The nightly energy is the luminaire power times the burn hours times the average dimming level. Take a 40 W luminaire that burns 12 hours a night at an average of 60 % of full power:
- Nightly energy: 40 W × 12 h × 0.60 = 288 Wh
- With an 80 % usable depth of discharge (typical for LiFePO4) and a 0.9 derating for cold:
| Autonomy | Battery (Wh) | At 12.8 V (Ah) |
|---|---|---|
| 2 days | 800 Wh | 62.5 Ah |
| 3 days | 1,200 Wh | 93.8 Ah |
| 5 days | 2,000 Wh | 156.3 Ah |
Each extra day adds the same 400 Wh. Batteries are a large part of the cost and weight of a solar street light, so the number in the tender matters.
What the standard says
There is no single number that fits every road.
- IEEE 1562 is the IEEE Recommended Practice for Sizing Stand-Alone Photovoltaic Systems. The current edition, IEEE 1562-2021, covers sizing the array and the battery of systems where PV is the only power source and a battery stores the energy. It covers stand-alone PV in general, not street lighting specifically, and by its own scope lead-acid batteries only; our lights use LiFePO4. It is a paid standard, so its day counts are not quoted here.
In practice, 2–3 days is common for lit urban roads with a reliable sunny season; critical roads, remote sites and long cloudy seasons justify 3 or more.
The battery is only half of the answer
Autonomy tells you how long the light survives a dark spell. It says nothing about whether the battery is full when the dark spell starts. That is the panel’s job.
The panel is sized on the worst month’s peak sun hours (PSH) at the site:
Panel (Wp) = nightly energy × margin ÷ (worst-month PSH × 0.75)
The 0.75 covers panel temperature, dirt, wiring and charging losses. The margin is the part most quotations leave out. With a margin of 1.0 the panel refills exactly one night’s energy on an average worst-month day: nothing is left to recover the battery after a cloudy run.
We checked this with PVGIS 5.3, the European Commission Joint Research Centre’s free solar tool. Its off-grid calculation runs the system hour by hour over many years of satellite or reanalysis radiation data: surplus PV charges the battery, a shortfall is drawn from it, and the result is the share of days on which the battery reached full or reached its cut-off. We simulated the 40 W, 12 h, 60 % light above with a 3-day battery (1,200 Wh), panel tilted at the optimum angle, and raised the panel until the battery ran empty on no more than 5 % of worst-month days.
| City | Worst month | PSH | Panel at margin 1.0 | Empty days, worst month | Margin needed | Panel needed | Empty days, worst month |
|---|---|---|---|---|---|---|---|
| Ankara | Jan | 2.67 | 144 Wp | 38.5 % | 1.95 | 280 Wp | 4.7 % |
| Erzurum | Dec | 2.64 | 145 Wp | 44.7 % | 2.70 | 393 Wp | 3.9 % |
| Istanbul | Dec | 2.37 | 162 Wp | 28.4 % | 1.75 | 284 Wp | 4.1 % |
| Riyadh | Jan | 5.59 | 69 Wp | 37.1 % | 1.30 | 89 Wp | 4.7 % |
Three things stand out.
- The textbook panel leaves the battery empty on more than a quarter of worst-month days in every city, even with three days of battery. The battery does not save a system whose panel cannot refill it.
- Peak sun hours alone do not decide the panel. Ankara and Erzurum have almost the same worst-month PSH (2.67 and 2.64), yet Erzurum needs a 40 % bigger panel. Its cloudy days come in longer runs, so the battery is emptied more deeply and needs more surplus to recover. Only a simulation over real weather years shows this.
- Sunny sites still need a margin. Riyadh has more than twice Ankara’s sun, but at margin 1.0 the battery still ran empty on 37 % of January days.
How long does a full recharge take?
After three dark nights the battery has given 3 × 288 = 864 Wh. On an average worst-month day the panel in the table produces panel × PSH × 0.75 and the light uses 288 Wh of it the following night; the rest recharges the battery.
- Ankara, 280 Wp: 280 × 2.67 × 0.75 ≈ 561 Wh, a surplus of about 274 Wh a day, so about 3.2 average days to refill.
- Erzurum, 393 Wp: a surplus of about 490 Wh a day, about 1.8 days.
- Istanbul, 284 Wp: a surplus of about 216 Wh a day, about 4 days.
This is the real meaning of the margin: it sets how fast the system recovers before the next dark spell. A larger battery without a larger panel only stretches the recovery time.
Choosing your number
- Start from the tender. If the specification states the autonomy days, use that number.
- Otherwise, match the risk. Two days suits lit urban roads with a reliable sunny season. Choose three or more for critical roads, remote sites, or climates with long cloudy spells.
- Size the panel on the worst month with a margin, not on the yearly average and not at margin 1.0. Where you can, check the margin with an hourly simulation such as PVGIS off-grid.
- Reduce the energy before adding battery. Dimming after midnight cuts the nightly energy, and the battery and panel shrink with it.
- Ask suppliers for the simulation, not only the autonomy figure: the share of worst-month days with an empty battery is the number that tells you whether the light will stay on.
Our solar sizing tool applies these formulas with PVGIS data for your location, lets you set the autonomy days (its default is 2), and lists every LuxLuminaire model with the hours and backup nights it reaches at your site. For a full lighting layout and a quotation, contact us.
Sources
- IEEE 1562-2021, Recommended Practice for Sizing Stand-Alone Photovoltaic Systems: standards.ieee.org
- European Commission JRC, PVGIS user manual (off-grid calculation): joint-research-centre.ec.europa.eu
- PVGIS 5.3 off-grid simulations for the cities above: LuxLuminaire, November 2026 (load 40 W, 12 h, 60 % average dimming, 3-day battery, 80 % depth of discharge, optimum tilt).