Off-grid solar sizing with real PVGIS production data

The problem with flat sun-hour estimates

A "peak sun hours" shortcut multiplies your panel wattage by a single number (often 4-5) to estimate daily production. It's fast, but it ignores your actual latitude, local weather patterns, panel tilt, azimuth and shading — all of which shift real output by a wide margin. Two sites 200 miles apart, or the same site with two different roof pitches, can produce meaningfully different daily energy even at the same panel wattage.

Satellite-based production data, like the one behind PVGIS (the European Commission's Joint Research Centre solar irradiance database, which also covers the Americas), accounts for your actual coordinates, tilt and orientation — giving a monthly production estimate specific to your site instead of a national average.

Step 1 — Daily energy consumption

Before sizing anything, list what you actually run and for how long: lights, fridge, water pump, any appliances. Multiply each device's wattage by daily hours of use and add it up — that's your daily Wh/kWh target. Everything downstream (panels, battery, inverter) sizes off this number, not the other way around.

Step 2 — Panel array from real production data

Required array size (kWp) = daily consumption ÷ average daily production per kWp at your exact location. This is where the flat sun-hours shortcut breaks down — the same formula with a generic assumption instead of site-specific data can be off by 20-30% or more depending on where you are and how the array is mounted.

Step 3 — Battery bank: autonomy and depth of discharge

Battery capacity (kWh) = daily consumption × days of autonomy ÷ depth of discharge (DoD). LiFePO4 tolerates a deeper DoD (often 80-90%) than lead-acid (typically 50%) at the same usable life — the chemistry you pick changes how much raw capacity you need for the same usable energy.

Step 4 — Inverter and cable sizing

Size the inverter for the peak load your appliances draw simultaneously, not the average — a fridge or pump's startup surge can spike well above its rated wattage, so a 20-30% margin above the estimated peak is common practice. Cable gauge (AWG in the US, following NEC Table 310.16) depends on expected current, run length and an acceptable voltage drop — undersized wire wastes energy as heat over long DC runs at 12/24/48V.

Common mistakes