A rooftop solar suitability assessment answers one question: is this roof worth the cost of a site visit? It is a triage decision, not a design decision. The point is to spend the field day on roofs that survive scrutiny.
The checks below work the same way on a 40,000 square foot distribution warehouse and on a 1,900 square foot ranch house. What changes is the threshold. A warehouse loses to obstructions; a house loses to plane geometry and canopy.
1. Usable area, not gross area
The assessor gives you a building area. Imagery gives you a footprint. Neither is the number a designer can use. Usable area is what remains after setbacks, obstruction clearances, walkways and access paths are removed.
The gap is usually large, and it is larger than people expect. Fire-code perimeter setbacks alone take a band off every edge. On a small or irregular roof, that band can consume a serious share of the surface before a single module is placed.
- Subtract the perimeter setback your jurisdiction enforces, not a generic figure. Local fire-code adoption varies and the setback is the single largest deduction on small roofs.
- Subtract clearance around every rooftop unit, not just the unit footprint. Service access is what the inspector looks for.
- Subtract required ridge and pathway clearances on pitched residential roofs.
- Subtract any plane too small to carry a contiguous run of modules. A 200 square foot triangle is area on paper and nothing in a layout.
- Record gross and usable as two separate fields. Collapsing them into one number is how a bad candidate reaches a rep.
2. Obstruction load
Obstructions do not reduce a roof proportionally. They fragment it. Ten small penetrations scattered across a flat roof do more damage to a layout than one large unit parked in a corner.
Count what is visible and note where it sits. Mechanical units, condensers, skylights, roof hatches, plumbing vents, exhaust fans, satellite dishes, antenna masts, ladders, screens and existing arrays all read from overhead imagery at typical public resolution.
- Count the obstructions, then look at their distribution. Clustered is workable. Scattered is not.
- Flag rooftop equipment that casts its own shadow, particularly tall units on the south side of a flat roof.
- Note skylights separately. They are both an obstruction and a clue about interior use and roof age.
- Note an existing array and its extent. A partially covered roof may be an expansion candidate or already spoken for, and overhead imagery cannot tell you which.
- On residential stock, count vents, chimneys and dormers per plane. Two vents in the middle of the only south-facing plane can end the conversation.
3. Plane orientation, pitch and count
A flat commercial roof is one plane, and orientation is a racking decision rather than a constraint. A pitched roof is several planes, and each one has its own azimuth, pitch and area.
The screening question on pitched stock is whether any single plane is large enough and oriented well enough to carry a worthwhile array. A house with six small planes facing six directions is a harder job than its total square footage suggests.
Elevation data helps here. Public lidar coverage, where a county has been flown, supports plane extraction and pitch estimation. Where it has not been flown, you are working from coarser terrain data and from what the imagery shows, and the pitch figure should be labelled as an estimate.
4. Shading, and the two-season problem
Shading comes from three places: adjacent structures, mature tree canopy, and the roof's own equipment. The first two are what a screen can see; the third you already counted.
Here is the trap. Public aerial imagery programmes typically fly leaf-on, near midday, in the growing season. That capture shows canopy at its densest and shadows at their shortest. A winter capture of the same roof shows bare branches and long shadows, and those are different pictures of the same problem.
- A leaf-on summer image overstates canopy obstruction and understates low-sun-angle shading from buildings.
- A leaf-off winter image understates canopy and overstates the clear area available in July.
- Neither image shows production loss. Shading geometry is an input to a production model, not a substitute for one.
- Where two captures from different seasons exist, use both and note the disagreement rather than picking the flattering one.
5. Roof cover type
Cover type drives attachment method, and attachment method drives cost. Some counties publish a roof-cover or roof-structure field in the assessor record. Where that field exists and is populated, use it and cite it.
Where it does not exist, imagery gets you partway. Gravel ballast, standing-seam metal, and asphalt shingle are usually distinguishable from overhead. Single-ply membrane types are frequently not, and a bright white surface tells you about reflectance rather than about what is underneath it.
The honest treatment is to mark cover type as observed, recorded, or unknown, and to let the unknowns stay unknown. An inferred membrane type that turns out to be wrong costs more in a proposal than a blank field does in a screen.
6. Building age as a proxy for membrane life
Year built is in nearly every assessor record, and it is tempting to treat it as remaining roof life. It is a weak proxy on its own, because it dates the building rather than the roof.
A roof can be replaced several times over a building's life without the year-built field changing. Strengthen the proxy by pulling the county permit index for roofing permits filed against the address, and by looking for replacement scarring in imagery. Where a jurisdiction publishes nothing searchable, say so. An absent permit is not evidence that no work was done.
Run the checks in elimination order
Order matters because each check costs a different amount of analyst time. Run the cheap, decisive ones first.
- Gross footprint and parcel identity, from the assessor record.
- Obstruction count and distribution, from current imagery.
- Plane geometry and usable area after setbacks.
- Shading from adjacent structures and canopy, with capture dates noted.
- Cover type, recorded or observed or unknown.
- Roof-age evidence from permits and visible replacement work.
A property that clears all six is a property worth a truck roll. A property that fails one should carry the reason it failed, because the reason may expire. Canopy gets trimmed, equipment gets removed, and roofs get replaced.