On this page [ 5 ]
Data sources
| Input | Source | How we use it | ||
|---|---|---|---|---|
| Electricity price | {{cite:eia-epm-56b | EIA Electric Power Monthly}} (state, year to date) and {{cite:eia-861-2024 | EIA 2024 utility data}} | State pages use the state average; city pages use the serving utility's average for its bundled residential customers. |
| Household use | {{cite:eia-861-2024 | EIA 2024 sales and customer counts}} | Average monthly kWh per residential customer, by state and utility. | |
| Solar output | {{cite:pvwatts | NREL/NLR PVWatts}} | Output per kW for a south-facing, fixed roof array tilted at 20 degrees with standard losses, at each city's coordinates, using typical-year weather. States without city pages use the PVWatts result for their largest city. | |
| Installed price | {{cite:lbnl-2026 | Berkeley Lab, U.S. Distributed Solar and Storage 2026 Data Update}} | Median pre-incentive price per watt for host-owned residential systems installed in 2025, by state. It is an estimate of a typical price, not a quote. | |
| Installed price (states without a median) | {{cite:lbnl-2026 | Berkeley Lab}} | Berkeley Lab shows a state median only where it has at least 20 observations. Elsewhere we use the U.S. median and label it that way on the state page. | |
| Export credit | Utility tariffs and rate riders | The rate each utility pays for exported energy, or retail-rate crediting where net metering applies. Where we have not added every utility yet, net metering states are modelled at retail; elsewhere we use the published export credit of the state's largest utility (treated as retail where that utility nets exports over a month or year). If we have not found that rate yet, net billing states use 40% of the state average price (an assumption, labelled on the page) and states where utilities set their own rules use zero. | ||
| Incentives | Statutes, program pages and IRS guidance | Applied only where the rule can be modelled; others are listed on each state's incentive page. |
Formulas
- System size (kW) = yearly use × share to cover ÷ output per kW.
- Price before incentives = size × 1,000 × price per watt.
- Incentives: cash rebates are applied first; percentage tax credits then apply to the remaining cost, up to their caps.
- Output in year n = size × output per kW × (1 − degradation)^(n − 1).
- Savings in year n = energy used on site × retail price + exported energy × export credit − any solar-specific fixed charges. Under net metering, exports offset purchases at retail up to your yearly use; any surplus earns the export rate.
- Retail price in year n = today's price × (1 + yearly increase)^(n − 1).
- Payback = the point at which cumulative savings first equal the net cost, interpolated within the year.
- Loan payment uses the standard amortization formula on the net cost less any down payment.
Default assumptions
- Share of output used at home: 50%. It depends heavily on when you use power; change it in the calculator.
- Electricity price increase: 2.5% a year. A planning assumption; recent year-over-year changes in the U.S. average have been higher, and future rates may differ.
- Panel degradation: 0.5% a year. A planning assumption typical of current warranties; use your panel's warranty figure if you have it.
- Analysis period: 25 years.
- Maintenance: none by default. Add a yearly amount if you expect costs, for example a future inverter replacement.
- Exports in Houston and Dallas: zero by default, because retail providers set buyback terms. Enter your plan's rate.
- California Net Billing exports: modelled at a flat planning average because credits vary by hour. Your installer can give an hourly estimate for your usage.
What we leave out
The model does not include shading on your specific roof, roof direction other than south, demand charges, battery dispatch, insurance or home-value effects, and it does not model taxes on the incentive itself. It treats the median installed price as typical; real quotes vary widely between installers for the same system.
Keeping the data current
Each record in our database stores its source link, a short quote from the source in our research files, and the date we last checked it. Pages show the oldest verification date of any number on them. Anything older than ninety days is flagged for re-checking, and changes are recorded in each page's change log.