Solar Panel Calculator

Where the missing 15 percent goes

A panel rated 400 watts produces 400 watts under standard test conditions: 1,000 watts per square metre of irradiance, a cell temperature of 25 °C and a specific spectrum. No roof ever meets all three simultaneously.

Soiling costs roughly two percent in most climates — dust, pollen and bird droppings between rains. It is higher in agricultural and arid areas where months pass without washing rain.

Shading takes another three percent even on a roof described as unshaded, because a vent stack, a chimney or a neighbouring tree clips the array for part of the day. Partial shading on a string can cost far more than the shaded fraction suggests, which is why module-level electronics exist.

Mismatch and wiring take two percent each. No two panels are identical, and a string performs to its weakest member. Wiring losses are ordinary resistance between the array and the inverter.

Light-induced degradation removes about 1.5 percent permanently in the first hours of exposure. Nameplate tolerance accounts for manufacturing spread. Availability — three percent — covers the hours the system is offline for maintenance, grid outages and inverter restarts.

Multiply those together and about 86 percent survives. Then the inverter converts DC to AC at roughly 96 percent efficiency, leaving 82 or 83. That figure, not the panel rating, is what determines how many panels you need.

Why 100 percent offset is not always the goal

Sizing to cover annual consumption is the default assumption, and in a full net metering market it is correct: every exported kilowatt-hour banks at retail value and offsets one you later import.

Many markets no longer work that way. Under net billing or avoided-cost export rates, power you send back earns a fraction of what you pay to buy it — sometimes a quarter. Every kilowatt-hour of overproduction is then sold cheap and bought back dear.

In those markets the economics favour sizing to daytime consumption rather than annual total, which typically lands at 60 to 80 percent of usage. The array costs less, the payback is shorter, and the electricity you do produce is worth full retail because you consume it directly.

Battery storage changes the calculation again by letting you time-shift production into evening use, which is why storage attach rates rise sharply in markets that have moved away from net metering.

Check your utility’s current export rate before deciding on offset. It is the single input with the largest effect on whether a system pays back, and it is the one most likely to have changed since you last looked.

What this estimate cannot see

Peak sun hours is a single number standing in for a great deal of local variation, and three factors it hides can move production substantially.

Orientation and tilt come first. The figures people quote assume a reasonably south-facing array near latitude tilt. East or west facing typically produces 10 to 20 percent less, and a shallow tilt in a northern latitude gives up more in winter than it gains in summer. A split array facing both east and west produces a flatter curve, which is sometimes worth more than peak output under time-of-use rates.

Temperature is the one people find counterintuitive. Panels lose roughly 0.3 to 0.4 percent of output per degree Celsius above 25, so a hot roof in July produces less per panel than a cold clear day in April. Cool sunny climates outperform hot ones at the same irradiance.

Snow and seasonal variation matter in northern latitudes, where a month of coverage removes production entirely rather than reducing it.

NREL PVWatts models all of this from your actual coordinates and is free. Use this calculator to get to a shortlist; use PVWatts before signing anything.