How Many Solar Panels Do I Need? Sizing a System From Your Bill

Published 2026-04-21 · By the CalculatorHive editorial team

Solar quotes vary wildly for the same house, and most of the variation comes from assumptions you can check yourself in about ten minutes. The sizing math is genuinely simple; what makes it feel opaque is that installers bundle four separate estimates — consumption, sunlight, losses, and panel selection — into a single number on a proposal. Here is each one, separately.

Step 1: get your real kWh from the bill

Do not use the dollar amount. Rates change, tiered rates distort the picture, and delivery charges are not proportional to usage. You want kilowatt-hours, which appear on every utility bill, usually as a monthly usage figure and often as a 12-month bar chart.

Add twelve months. One month × 12 will mislead you badly in either direction: a home with electric heat peaks in January, a home with central air peaks in August, and either extreme extrapolated across the year is off by 30% or more. If you only have a few bills, most utilities show 12 or 13 months of history in the online account.

Also decide whether you are sizing for today's usage or tomorrow's. An electric vehicle adds roughly 3,000–4,000 kWh a year for average driving; a heat pump replacing gas heat adds more. It is far cheaper to size for that at installation than to add panels later, since the inverter, permitting, and labour do not scale down. Our appliance energy calculator is useful for estimating a specific new load before you commit.

Step 2: find your peak sun hours

A "peak sun hour" is one hour of irradiance at 1,000 watts per square metre — the standard test condition panels are rated at. It is not daylight hours. A location with 4.5 peak sun hours may have 14 hours of daylight in June; the metric compresses a whole day's varying intensity into an equivalent number of full-strength hours, averaged across the year.

These are approximations for a south-facing array at a reasonable tilt. For an actual number at your address, NREL's PVWatts tool models irradiance from long-term weather station data at a specific latitude, tilt, and azimuth, and it is free. Use it before you accept anyone's production estimate.

Step 3: apply the derate

Panels never produce their nameplate rating in the field. The losses stack multiplicatively and every one of them is real:

• Inverter efficiency — about 3–4%. DC from the panels has to become usable AC.
• Temperature — 5–12%. Panel output falls as cell temperature rises above 25 °C, which on a dark roof in July is most of the time.
• Soiling — 2% typical, more in dusty or agricultural areas and more where it snows.
• Wiring and connections — 2–3%.
• Module mismatch and nameplate tolerance — 2–3%.
• Shading — 0% to catastrophic, depending on your site.
• Degradation — roughly 0.5% per year, so a 25-year panel produces around 85–88% of its year-one output at end of warranty.

Stacked, PVWatts defaults to about 14% total system losses for a clean, unshaded, well-oriented array. In practice, once you include the fact that most roofs are not perfectly south-facing at the optimal tilt, planning at 15–25% is honest. Use 0.80 as a derate factor for a typical installation and 0.75 if you have east-west split arrays or partial shading.

Step 4: the worked example

A household in the Midwest uses 11,000 kWh per year, has 4.5 peak sun hours, and a 20% derate (factor 0.80). Panels are 400 W.

Check it backwards: 8.37 kW × 1,314 = 11,000 kWh. The solar panel calculator runs this both ways, so you can also start from a roof area or a panel count and see what production it supports.

Location changes this dramatically. The same 11,000 kWh home needs about 9.9 kW (25 panels) at 3.8 peak sun hours in the Northeast, and about 6.5 kW (17 panels) at 5.8 hours in Arizona. Same house, same appliances, a 47% difference in system size.

(All at 4.5 peak sun hours and a 0.80 derate.)

What the roof actually allows

Module area is not roof area. A residential 400 W panel is roughly 74 × 41 inches, about 21 square feet. On top of that you need:

• Fire-code setbacks. The International Fire Code and most local amendments require clear pathways — commonly a 3-foot path at the ridge and along at least one edge — so firefighters can vent a roof. This can remove 20–30% of a small roof plane.
• Obstruction clearance around vents, chimneys, skylights, and satellite mounts.
• Usable orientation. A north-facing plane in the northern hemisphere produces so little that it is rarely worth the racking. East and west planes produce roughly 15–20% less than south, which is often still worth doing, particularly where time-of-use rates reward afternoon generation from a west-facing array.

Practical planning figure: 25–30 sq ft of roof per panel once you allow for setbacks and rails.

The caveats worth taking seriously

Sizing math gives you a target, not a guarantee. Four things routinely make real results differ from the spreadsheet:

• Shading is non-linear. A tree shading one corner of one panel can drag down a whole string on a system with a central inverter. Microinverters or DC optimisers largely fix this and are worth specifying on a shaded site.
• Seasonality. Annual sizing means you overproduce in June and underproduce in December. Whether that balances out depends entirely on your utility.
• Net metering rules vary and change. Some utilities credit exported energy at the full retail rate, some at a lower wholesale or "avoided cost" rate, some only within a monthly window, and some have moved to net billing that makes exports worth far less than self-consumed energy. This is the single biggest determinant of whether over-sizing pays. Read your utility's current tariff before deciding to build a system that exports heavily.
• Incentives change. Federal, state, and utility incentives are revised regularly and have eligibility conditions and deadlines. Check the current rules rather than a figure quoted in an old article, and confirm with a tax professional how any credit applies to your situation.

Safety and permitting. A grid-tied PV system involves high-voltage DC, roof penetrations, and a service-panel interconnection. It requires permits, an inspection, and utility approval to operate essentially everywhere, and the interconnection work must be done by a licensed electrician. This is not a weekend project, and unpermitted work commonly voids homeowners insurance.

Common questions

Should I size for 100% of my usage?

Only if your utility credits exports at a rate that makes the surplus worth building. Where net metering is full-retail, sizing at or slightly above 100% is usually optimal. Where exports are credited at wholesale rates, the economics shift toward sizing closer to your daytime self-consumption — often 60–80% of annual use — sometimes paired with storage. Ask the utility, not the salesperson, what the current export rate is.

Do more efficient panels mean fewer panels?

Yes, but the relevant comparison is watts per square foot, not the efficiency percentage in isolation. A 22%-efficient 430 W panel and a 20%-efficient 400 W panel are close to the same physical size, so the higher-efficiency module buys you roughly 7% more production from the same roof. That matters only if roof area is your binding constraint; if you have plenty of roof, cheaper panels usually win on cost per watt.

What happens to my system in a power outage?

A standard grid-tied system shuts down. This is a safety requirement — anti-islanding prevents your array from energising lines that utility crews believe are dead. To keep power during an outage you need a battery with an islanding-capable inverter, or a transfer-switch arrangement. If outage resilience is the goal, say so before the design is finalised, because it changes the equipment selection.

How much does the tilt and direction really matter?

Less than most people fear. Relative to an optimally tilted south-facing array, a typical roof pitch facing southeast or southwest loses only a few percent; due east or due west loses roughly 15–20%. Flat roofs are usually racked at a shallow tilt to balance production against wind load and self-shading between rows. Run your specific azimuth and tilt through PVWatts rather than assuming a non-south roof rules solar out.