Yield as a share of the best fixed orientation at this site. Click a cell to try it.
Reopen this plan:
SolarPlan estimates what a photovoltaic array will do on a roof or on open land: how many panels fit, how much electricity they make each month, what the project will need as it grows, and what it may save. It works through every hour of a typical year at the site.
SolarPlan is a free, independent project, developed by enginerds from the University of Southern California and the University of Pennsylvania.
The model follows PVWatts version 8, the National Laboratory of the Rockies’ calculator (formerly NREL’s), as its source code defines it:
As a check, on the same weather as PVWatts, the model matches it within 0.2% a year and 0.6% in any month at 28 test systems across the US: roofs and racks, flat to vertical, east to west, with row shading and trackers.
The module search covers 21,049 modules of 200 W and more from the California Energy Commission’s list, with the single-diode parameters the System Advisor Model fits to each (library edition 2026.7.3). A chosen module runs through the same model as the generic panels, with cell temperatures on PVWatts’ assumptions: 45 °C nominal on racks and 49 °C on roofs. Most modules listed since 2020 give no dimensions; their size is estimated from their area and cell layout, and comes within 2% of the stated size for nine in ten of the modules that do give one. The extra output of bifacial modules from light reaching their backs is not counted.
In the Americas, the typical meteorological year from the National Solar Radiation Database (GOES version 4, latest edition): 8,760 hours of sun, temperature, wind and ground reflectance on a 4 km grid, assembled from many years of satellite data. Elsewhere, PVGIS from the European Commission’s Joint Research Centre. PVWatts still uses the 2020 edition of the NSRDB year, so its results for the same place can differ by a few percent. A real year differs from a typical one by a few percent either way.
On low-slope roofs (4/12 or flatter), OSHA 1910.28(b)(13) sets three zones for maintenance work. Within 6 ft of an unprotected edge, workers always need guardrails, a safety net, travel restraint or fall arrest. From 6 to 15 ft, a marked designated area is allowed for occasional, brief work. Beyond 15 ft, a work rule is enough for such work. A parapet at least 39 in high can serve as the guardrail. The plan outlines the 6 ft and 15 ft zones and marks the panels inside them.
Each year’s output counts against your use at your rate; any surplus is paid at the export rate, or at your rate under net metering. Rates rise 2.5% a year, output falls 0.5% a year, and future savings are discounted at 6%.
Until you enter your own, the rate is your state’s average retail price for homes or for businesses (EIA, January–July 2026), and the installed cost is Berkeley Lab’s median for systems installed in 2024: by state where it publishes one, nationally otherwise. Homes use the residential median; businesses the small (up to 100 kW) or large non-residential median; and ground arrays of 5 MW AC and more the utility-scale median for their size. These are prices before incentives, for systems without batteries. Your bill and your quotes are better: clear a field to go back to the default.
The federal credit for homeowners (§25D) ended for systems installed after December 31, 2025. Leases and power purchase agreements still carry the business credit (§48E) through the system’s owner. Projects that began construction after July 4, 2026 can claim §48E only if in service by December 31, 2027, and at 1 MW AC and above the full 30% requires prevailing wages and registered apprentices.
Shading from trees and nearby buildings, snow, batteries, and time-of-use or demand charges. The list of what a project needs is a screening, not a ruling: the building department, the utility and an engineer make the actual calls.
SolarPlan counts visits with Vercel Web Analytics, without tracking cookies: it tells visits apart by a hash of the request that is discarded after 24 hours, and records the page, the site that linked to it, a rough location, and the kind of device and browser. Vercel Speed Insights records how fast pages load. Links carry your place and settings, but only the page’s path is reported. Your history and last settings stay in your browser.