Updated October 2026 · 50 states
Enter an electricity bill and a state. The result is the number of modules, the system size in kilowatts, the roof area required, the installed cost and the year the system breaks even — using that state's own solar yield, retail rate and export tariff rather than a national average.
Three numbers decide it: annual consumption, the yield of a kilowatt of panels where the property is, and how much of that consumption the system is intended to cover.
Taken from the bill rather than from a national average. A monthly bill is not all energy — a fixed customer charge sits on top of the per-kilowatt-hour part — so roughly $12 a month is removed before dividing by the state's retail rate. Entering annual kilowatt-hours directly, from the bottom of a utility statement, gives a more accurate answer than entering a bill.
A kilowatt of panels produces between about 900 and 1,800 kilowatt-hours a year depending on where it is. That single figure varies more across the United States than module efficiency ever does between brands, which is why a national rule of thumb such as "1,400 kWh per kW" produces sizing that is wrong by a third at either end of the country. The figure used here comes from NREL PVWatts v8 for each state's largest metropolitan area, south-facing at 20 degrees.
Covering 100% of annual consumption is the usual default, but it is the right answer only where exported power earns close to the retail rate. Where exports earn a quarter of retail, a system sized to daytime consumption — around 60 to 80% of the annual total — returns more per dollar spent than a larger one, because the surplus is sold at a loss against what it cost to generate. The calculator flags this for the state selected.
Each state page models a fixed 6 kW system. This calculator sizes to the consumption entered, and self-consumption moves with that: an array covering 60% of a household's use is almost entirely consumed on site, while one covering 120% exports a large share of its output. Where exports earn close to the retail rate that makes little difference to the payback; where they do not, it makes a great deal. The two agree where a 6 kW array happens to match the household's consumption.
Roof area is a rough guide at about two square metres a module and takes no account of obstructions, setback requirements or shading. Installed cost is a state median; a quotation may land $0.40 a watt either side of it. Loan interest and dealer fees, batteries, roof replacement and any utility-specific programme are all excluded, as they are throughout this site. The method page sets out every assumption and omission in full.
Most US households use between 9,000 and 12,000 kilowatt-hours a year. At a 430 W module and a typical yield, that is usually 16 to 28 panels — but the same house needs materially fewer panels in Arizona than in Washington, which is why the calculator asks for a state.
Floor area predicts consumption poorly: heating fuel, climate, occupancy and whether there is an electric vehicle matter far more. A bill is a better input than a square footage, and an annual kilowatt-hour figure from a utility statement is better still.
Only where exports are credited at or near the retail rate. Where they are not, a smaller system consumed on site earns more per dollar than a larger one that exports its surplus cheaply.
No. Cost rises roughly in proportion to size, while the value of the extra output depends on the export rate. Above the point where generation exceeds on-site use, each additional panel is paid for at full price and compensated at the export rate.
This is an estimate for comparison, not a quotation or a system design. Roof orientation, shading, structural capacity and the serving utility's interconnection rules all affect what can actually be installed. At least three written quotations are advisable before purchase.