Daily load to panel watts
Solar Panel Sizing Calculator
Turn daily energy use, sun hours, and system losses into a practical solar panel target.
Array sizing is a division problem wrapped around two numbers most people get wrong. The first is peak sun hours, which is not hours of daylight — it is total daily insolation restated as equivalent hours at 1,000 W/m², and it varies enormously with season, latitude, tilt, and orientation. The second is system loss, which is not one number but a chain of them.
The most common sizing failure is using an annual-average sun-hour figure for an off-grid system. Off-grid systems do not get to average across the year. If the array cannot refill the battery in December, the system fails in December, however comfortable July looks.
Size for the worst month you intend to operate through, not the mean.
The method
Nothing here is hidden. These are the exact expressions the calculator evaluates, in the order it evaluates them.
system_efficiency = 1 − (system_loss% / 100)array_W = daily_Wh / peak_sun_hours / system_efficiencybuffered_W = array_W × 1.2The 1.2 is a deliberate 20% margin, not part of the physics. It covers a worse-than-expected month, panel ageing, and the load creep that happens on every real system. The unbuffered figure is shown alongside it so you can see exactly what the margin costs you.
panels = ceil(buffered_W / panel_W)total_loss = 1 − [(1 − L₁) × (1 − L₂) × … × (1 − Lₙ)]If you are deriving your own loss figure from individual factors, they combine multiplicatively, not by addition. Adding the percentages overstates total loss and oversizes the array.
What each input means
| Input | What it actually is | Typical range |
|---|---|---|
| Daily energy use in watt-hours | The output of a load audit: every device's watts multiplied by its hours per day, summed. Use the worst season's figure if loads vary. | 200–3,000 Wh for small systems |
| Peak sun hours | Daily insolation in kWh/m² for your location, tilt, and orientation, in the month you are sizing for. Run PVWatts for your own site rather than using a generic figure. | 1.5–3.0 worst month, 4–6 annual average |
| System loss percentage | Everything lost between the panel's nameplate and energy stored in the battery: soiling, mismatch, wiring, controller conversion, and battery round-trip loss. Enter the loss, not the efficiency — 25 here means the system delivers 75% of nameplate. | 20–35% loss |
| Panel wattage | The nameplate rating of the modules you intend to buy, used only to turn the array target into a module count. | 100–450 W |
Worked example
A cabin using 850 Wh per day, sized for a worst month with 2.4 peak sun hours, entering 30% system loss, planning on 200 W modules.
| Step | Calculation |
|---|---|
| System efficiency | 1 − 30/100 = 0.70 |
| Energy needed at the array | 850 Wh / 0.70 = 1,214 Wh |
| Minimum array | 1,214 Wh / 2.4 h = 505.95 W, shown as 506 W |
| With 20% buffer | 505.95 W × 1.2 = 607.14 W, shown as 608 W |
| Panel count | ceil(608 W / 200 W) = 4 panels |
Result: 506 W minimum, 608 W with buffer, so four 200 W modules (800 W).
Compare this against sizing on a 5.0 peak-sun-hour annual average, which gives 243 W — less than half. An array sized on the annual average leaves this cabin dark all winter. This one input choice matters more than anything else on the page.
Assumptions baked into the result
Every calculator makes assumptions. Most do not tell you what they are. These are ours, and if one of them does not describe your system, the answer will be wrong in a direction you can now predict.
- Panels are unshaded during the productive part of the day. Partial shading on a series string costs far more than its proportional area.
- Panels sit at a fixed tilt appropriate to the sizing month.
- The charge controller can accept the array's output. An oversized array on an undersized MPPT is clipped, not harvested.
- The battery can accept the charge rate. A flooded bank limited to C/10 cannot absorb a large midday harvest however much array you install.
- Nameplate wattage is at standard test conditions of 25 °C cell temperature. Panels in sunshine run much hotter and produce measurably less.
What this tool will not tell you
- This calculates an average daily energy balance. It does not model consecutive cloudy days — that is what the battery's days of autonomy are for.
- It assumes a fixed tilt. Seasonal adjustment changes the sun-hour figure and therefore the answer.
- It does not account for snow cover, which can take an array to zero for days regardless of irradiance.
- Panel temperature derating is folded into the system efficiency figure rather than modelled separately. In hot climates, use a lower efficiency figure.
Common questions
Where do I get an accurate peak sun hours figure?
Run NREL's PVWatts calculator for your exact location with your intended tilt and azimuth, then read the monthly output rather than the annual total. Generic sun-hour maps rarely state what tilt they assume, which makes them close to useless for sizing.
What system efficiency should I use?
For a small off-grid system with an MPPT controller and a lithium battery, 70–75% is a reasonable planning range once you include array losses, controller conversion, and battery round-trip. Lead-acid and PWM push it lower.
Is it better to oversize the array or the battery?
For most small systems, the array. A larger array recovers the battery faster after bad weather and improves worst-month performance, and panels are usually cheaper per watt than batteries are per watt-hour. The limits are your controller's rating and your battery's charge acceptance.
