Methodology
Every number the calculator shows, where it comes from, and the assumptions that could make it wrong. If you disagree with an assumption you should be able to find it here and judge the result yourself.
1. Locating you
A ZIP code is matched against the US Census ZCTA gazetteer to get a centroid. A street address goes to the US Census Geocoder. Both are free public services with no API key, so nothing about your location is sent to a commercial data broker.
Some ZIP codes serve only PO boxes and have no mapped area. Rather than fail, the calculator uses the nearest adjacent delivery area and tells you it has done so.
Your state is derived from the ZIP prefix, not from the weather station the production model returns. Near a state line the nearest station is often across the border, which would apply the wrong price band.
2. How much electricity the panels make
Production comes from NREL PVWatts v8, the National Renewable Energy Laboratory model used throughout the US solar industry. It combines typical-year weather for the station nearest you with panel orientation and system losses.
Defaults, all overridable in the model itself:
- Tilt 20 degrees, azimuth 180 degrees (due south)
- Standard silicon modules, fixed open rack
- System losses 14% - wiring, soiling, inverter, mismatch
PVWatts is queried for a 1 kW array and scaled. Output is linear in array size, so this gives an identical answer while allowing one cached result to serve every system size at your location.
3. What your electricity costs
From NREL Utility Rates v3, which maps your coordinates to the serving utility and its filed residential rate. This is the single largest driver of the result: your payback depends far more on what you currently pay than on how sunny it is.
When NREL has no verified rate for an area, the calculator refuses to estimate rather than substituting a national average. A plausible wrong rate produces a confident wrong answer, which is worse than no answer.
4. Sizing the system
Your monthly bill divided by your rate gives annual consumption. Divided by annual production per kW, that gives the array size needed to cover your usage.
The result is then checked against roof reality. A typical home fits about 10 kW; beyond roughly 15 kW is more than almost any house holds. When the maths demands more, the calculator says so rather than quietly pricing an array that cannot be built. Enter your real roof area and it uses that instead, at about 17 watts per square foot.
5. What it costs to install
This is the weakest figure on the site and it is labelled as such wherever it appears. Production is physics and your rate is a filed tariff, but installed price is a market number: it moves with the installer, the equipment, the complexity of your roof and how hard you negotiate.
The national range used is $2.35 to $3.50 per watt, from 2026 US residential market surveys - LBNL Tracking the Sun (checked 2026-08-20). A handful of states with clear published evidence get their own band; every other state falls back to the national midpoint and the result says so.
There is deliberately no fabricated 50-state price table. Inventing a precise number for a state with no supporting data would be exactly the false precision this site exists to argue against. If you have a real quote, enter its price per watt - it beats every average here.
6. The federal tax credit is zero
The Residential Clean Energy Credit (IRC Section 25D) - the familiar 30% - was terminated by the One Big Beautiful Bill Act (P.L. 119-21), signed 4 July 2025, for expenditures after 2025-12-31. A home solar purchase made today receives no federal credit.
This is the biggest single reason results here look worse than elsewhere. On a $26,000 system the credit was $7,800. Calculators that still apply it understate your cost by five figures and shorten the quoted payback by roughly three years. State and utility incentives may still exist where you live and are not included, so check those separately.
7. Projecting 25 years
Each year the model applies:
- Electricity prices rising 2.9% a year. US residential prices went from a little over 12 cents per kWh in 2013 to 16 cents in 2023, which is 2.9% a year - and the EIA found they rose less than 1% over that decade once inflation is accounted for. Electricity has tracked inflation, not outrun it. This figure was previously set at 3.5%, which overstated savings; on cheap-power locations that was the difference between solar paying back and never paying back at all.
- Panel output falling 0.5% a year, the standard manufacturer warranty figure.
- Exported surplus valued at zero. Only electricity you would otherwise have bought counts as a saving.
- Running costs of $30 per kW a year, deducted from the savings. NREL's Annual Technology Baseline benchmarks residential upkeep at this level, covering insurance, cleaning, monitoring and the reserve that pays for replacing the inverter - which typically needs doing once in the panels' life. On a 10 kW system that is about $300 a year. Most calculators, including this one until recently, charge nothing for it, which overstates the lifetime result by anywhere from 7% to 40% depending on location.
Payback is the point where cumulative savings pass the net install cost, interpolated within the year rather than rounded up. If that never happens inside 25 years, the calculator reports never instead of extending the horizon until the answer turns positive.
These are future dollars, not today's money
Savings are counted in the dollars of the year they happen and are not discounted back to the present. A dollar saved in year 25 is not worth a dollar spent today, so a lifetime total is a nominal sum rather than a present value. Every solar calculator works this way, which keeps comparisons fair - but since the EIA finds electricity has merely tracked inflation, treat the real gain in today's money as meaningfully smaller than the headline figure.
Why exports are worth nothing here
This is the assumption most likely to differ from your reality, so it is worth stating plainly. Full-retail net metering is being withdrawn across the US - California moved to a far lower export rate under NEM 3.0 and other states are following. Export rates now vary from full retail down to a few cents.
Assuming a generous export rate is how oversized systems get sold. Valuing exports at zero means the estimate is a floor: if your utility still pays well for exports, your real return is better than shown. We would rather be pleasantly wrong than optimistically wrong.
What is not modelled
- Shading from trees, chimneys or neighbouring buildings
- Roof condition, age, or the cost of replacing it under the panels
- An unusually complex install, or a roof that needs work first
- Electrical panel upgrades, permits and interconnection fees
- State, utility and local rebates, and SRECs
- Battery storage, and time-of-use tariffs
- Tiered rates. Many utilities charge more per unit as you use more, and solar displaces your most expensive units first. We use the filed average residential rate, so a household on a high tier may save rather more than shown here.
- Financing costs - all figures assume a cash purchase
- Any effect on the resale value of your home
Several of these are worth thousands. This is a screening tool: it tells you whether solar is worth pursuing where you live and what a fair quote looks like. It does not replace a site visit.
Last reviewed 2026-08-20. Data sources are re-checked when the underlying models or the law change.