The formula installers actually use, explained in plain English — from your annual kWh to panel count, roof area and tilt angle, with worked examples for 800, 1,200 and 2,000 kWh per month, plus the derating factors most online calculators quietly ignore.
"How many solar panels do I need?" is the first question almost everyone asks, and it's the one most quotes answer badly. Installers often size a system around a round number or whatever fits neatly on the roof, rather than around your actual electricity use.
The arithmetic isn't difficult. This guide walks through the exact formula professional designers use, explains each input so you can judge whether a quote is honest, and works through three real examples. If you'd rather skip straight to the answer, our solar system size calculator does all of this for you — but understanding what's behind it is what stops you being over-sold.
System size (kW) = Annual kWh usage ÷ (365 × Peak sun hours × Derate factor)
Number of panels = System size (kW) × 1,000 ÷ Panel wattage
Four inputs. Get each one right and your answer will be within a few percent of what a professional design software package produces.
Take it from your bills, not from a guess. Add up twelve months of kWh — most utility portals will show this directly. If you only have one bill, multiplying a single month by twelve is risky, because summer air conditioning or winter heating will skew the number badly in either direction.
Rough anchors, if you have nothing else: a small apartment might use 2,000–4,000 kWh a year, an average family home 4,000–9,000 kWh, and a large home with electric heating, a pool or an EV can exceed 15,000 kWh.
Sizing for the future, not the past. If you're planning an electric vehicle, a heat pump or air conditioning within the next few years, add that load now. An EV typically adds 2,000–4,000 kWh per year. Panels added during the original install cost far less than a second visit later, and your roof space and inverter capacity are both finite.
To translate kWh into money — which is what you actually care about — run your usage and tariff through the solar savings calculator.
This is the input people get most wrong, because it is not the number of daylight hours.
One peak sun hour means one hour of solar irradiance at 1,000 watts per square metre — the standard test condition panels are rated at. Your location might see fourteen hours of daylight in June, but only about five or six peak sun hours, because early morning and late evening sun is weak and oblique.
Region typeTypical annual daily averageNorthern Europe, coastal UK, northern Canada2.5–3.2Central Europe, northern US, southern Canada3.2–4.2Southern Europe, most of the US, northern India4.2–5.5Middle East, North Africa, Pakistan, Australia's interior5.5–6.5
These are annual averages, appropriate for grid-tied systems where the grid covers seasonal shortfalls. For off-grid design you must use worst-month figures instead — a completely different and much larger answer.
Your country hub has regional figures, or check United States, United Kingdom, Australia, India and Pakistan directly.
Panels never deliver their nameplate rating in the field. The derate factor bundles every real-world loss into one multiplier — and it's the number that separates an honest estimate from a sales projection.
Loss sourceTypical impactTemperature (cells run hotter than 25°C test conditions)5–12%Inverter conversion2–4%DC and AC cable losses1–3%Soiling and dust2–6%Module mismatch and tolerance1–3%Shading (site-specific)0–20%+Annual degradation (year one)1–2%
Use 0.75–0.80 for most installations. Use 0.70 or lower if you have meaningful shading, a hot climate, a dusty location, or a suboptimal roof orientation. Anyone quoting you a derate above 0.85 is being optimistic, and any estimate that ignores derating entirely will overstate your production by a quarter.
Soiling is the one loss you control directly — see our guide on keeping panels clean and diagnosing underperformance.
Residential modules in 2026 commonly sit between 400 W and 600 W. Higher-wattage panels mean fewer modules, less mounting hardware and less labour for the same output — genuinely useful when roof space is tight.
What matters alongside wattage:
Efficiency (%) — output per square metre. Critical on a small roof, largely irrelevant on a big one.
Temperature coefficient — how much output drops per degree above 25°C. In hot climates this affects real yield more than headline efficiency does.
Degradation warranty — guaranteed output at year 25. Modern N-type modules typically warrant around 87–92%.
Compare real specifications in the solar panel database, or see how leading manufacturers stack up in LONGi vs Jinko and Jinko vs Trina.
Three worked examples
Example A — Small home, 800 kWh/month, moderate sun
Annual use 9,600 kWh · 4.5 peak sun hours · derate 0.78 · 450 W panels
9,600 ÷ (365 × 4.5 × 0.78) = 7.5 kW
7,500 ÷ 450 = 17 panels · roughly 34 m² of roof
Example B — Family home, 1,200 kWh/month, high sun
Annual use 14,400 kWh · 5.8 peak sun hours · derate 0.76 · 550 W panels
14,400 ÷ (365 × 5.8 × 0.76) = 8.9 kW
8,900 ÷ 550 = 17 panels · roughly 44 m² of roof
Notice that stronger sunlight produced a similar panel count despite 50% more consumption. Location does as much work in this equation as usage does.
Example C — Large home with EV, 2,000 kWh/month, low sun
Annual use 24,000 kWh · 3.4 peak sun hours · derate 0.74 · 500 W panels
24,000 ÷ (365 × 3.4 × 0.74) = 26.1 kW
26,100 ÷ 500 = 53 panels · roughly 116 m² of roof
At this point roof area becomes the binding constraint, not budget. This is exactly the scenario where high-efficiency modules earn their premium — and where offsetting only part of your consumption may be the sensible answer.
Will it fit? Roof area and orientation
A rough rule: a modern residential panel occupies about 2–2.3 m² (22–25 sq ft), and you need roughly 20–25% extra for walkways, fire setbacks, edge clearances and obstructions like vents and chimneys.
Roof area needed ≈ Number of panels × 2.2 m² × 1.25
Orientation changes the answer significantly. Relative to a true equator-facing roof:
South-facing in the northern hemisphere (north-facing in the southern): 100% baseline
Southeast or southwest: roughly 90–95%
East or west: roughly 75–85%
North-facing in the northern hemisphere: 55–70% — usually only worth it as fill-in capacity
East–west split arrays are worth considering even though total output is lower. They spread generation across morning and evening, which often matches household consumption better than a midday peak — and self-consumed energy is worth more than exported energy under most tariffs.
Tilt angle: the quick answer
The classic rule of thumb is to set tilt equal to your latitude for the best annual yield. Adjust from there:
Latitude minus 10–15° favours summer production
Latitude plus 10–15° favours winter production — the right choice for off-grid systems and for homes with winter-heavy loads
Below 10° tilt, panels stop self-cleaning in rain and soiling losses climb noticeably
Practically, most rooftop systems are fixed to whatever pitch the roof already has. That's fine — being 10–15° off optimum typically costs only 2–5% of annual yield, rarely enough to justify tilt frames and their wind loading. Planning your layout properly matters more; see our video on rooftop layout mistakes that cost you energy.
Do you also need a battery?
Panel sizing and battery sizing are separate questions. Batteries don't generate anything — they shift energy from when you produce it to when you use it.
A battery is usually worth it if your export rate is much lower than your import rate, if you face frequent outages, or if you're on time-of-use tariffs with an expensive evening peak. It's usually not worth it if you have generous net metering, since the grid already acts as free storage. Our analysis of 2026 net metering changes covers how that calculation is shifting, and the battery size calculator handles the sizing.
Common sizing mistakes
Sizing from one month's bill. Seasonal swings make single-month extrapolation wildly inaccurate.
Confusing daylight hours with peak sun hours. This inflates estimated output by 50% or more.
Skipping the derate factor. Nameplate ratings are laboratory numbers, not field numbers.
Ignoring future load. An EV or heat pump added two years later leaves you undersized on a roof that's already full.
Chasing 100% offset regardless of cost. The last 10–15% of offset is often the least economic part of the system, especially where export is poorly compensated.
Not accounting for shading. A chimney shadow crossing one string for two hours a day can cost more than every other loss combined.
Turning panel count into a decision
Once you have a system size, three questions decide whether to proceed:
What will it cost? See Solar Panel Cost 2026 for current per-kW pricing.
What will it save, and when does it pay back? Use the payback calculator and ROI calculator.
How will you pay for it? Compare cash, loan and lease outcomes with the financing comparison tool.
Do the arithmetic before the sales visit. Walking into a quote already knowing your target system size is the single most effective thing you can do to get a fair price — and to recognise a bad design when you see one.
Related tools and guides
