Solar Site Survey Checklist

Six categories, and most published lists cover four of them. The electrical service, which means the busbar rating, the main breaker rating, available breaker spaces and their position, and the meter. The roof and structure, meaning framing measured in the attic, the covering and its age, each plane measured with pitch and azimuth, and obstructions located. Shading, measured on site rather than modelled remotely. Routing and equipment locations, walked rather than drawn. Access and logistics, so every later visitor inherits the gate codes rather than rediscovering them. And the account data, meaning the utility bill, because the name on it decides who can sign the interconnection application. The last two are the ones most often missing.

Because the proposal was built from aerial imagery, a utility bill and a production model, and the survey is built from a tape measure, a camera and an open electrical panel. When those disagree, the survey wins. The three recurring contradictions are an electrical service that cannot accept the backfeed the design assumed, a roof plane that cannot carry the module count that was sold, and shading that the remote model under-read. Each of those changes something the homeowner already agreed to: the production estimate, the price, or both.

The panel label, showing the busbar rating. The busbar rating and the main breaker rating are different numbers and are frequently assumed to be identical when they are not, and together they decide how much inverter output the existing panel can accept before a derate, a supply-side connection or a service upgrade becomes necessary. It cannot be read from aerial imagery or from a utility bill, and it is the input most likely to cut a sold system down at design.

Yes, and it is usually treated as a sales artefact rather than as survey data, which is why it is so often missing. The bill carries three things nothing else does: the account holder's name as the utility records it, the service address in the utility's own format, and the account number. The name is the one that matters most, because the utility contracts with its account holder rather than with the contractor, so a mismatch between the person signing your agreement and the person on the account stalls the interconnection filing while the permitting track proceeds as though nothing is wrong.

Somebody competent to open an electrical panel safely and to work in an attic, which in practice means a trained surveyor or a qualified technician rather than a sales representative with a phone. The distinction matters because the highest-value items on this list, the busbar and main breaker ratings and the framing measurements, are exactly the ones that require access most people should not be taking. A survey that skips those items is a photography exercise, and it produces a design built on the same assumptions the proposal used.

Usage varies between organisations, so the useful question is what the visit is for rather than what it is called. A survey in this sense is the pre-construction visit that converts a sales proposal into a buildable design: it gathers the measurements and conditions an engineer and a plan examiner need. A later visit for a different purpose, such as verifying installed work or diagnosing a performance complaint, is a different exercise against a different standard. Confirm which one is meant before assuming a checklist applies to it.

Almost every avoidable revision traces back to a design built on an assumption the site later contradicted: a busbar rating assumed rather than photographed, framing taken from a typical rather than measured, a meter location taken from imagery. Each of those produces a design that is internally consistent, gets approved, and then meets a roof. The change that follows has to go back through the plan review queue, so its real cost is the wait it re-enters rather than the redraw. Capturing the constraint at survey is the same work done at a hundredth of the cost.