Solar proposal shade loss: require evidence, not one unexplained percentage

Start with a safe, dated observation record

From ground level, photograph mature trees, neighboring structures, chimneys, dormers, roof peaks and other visible obstructions. Label the viewing direction, approximate time and date, and the roof plane affected. Do not climb the roof, trespass or promise tree work. One photograph is not a full annual shade study because the sun path and vegetation can change by season; its value is to anchor the questions a survey must answer. DOE’s homeowner guide specifically notes that excessive tree shade can make a rooftop less suitable. [2]

Ask which method produced the number

The proposal should name whether shade came from remote imagery, a site-measurement tool, a design platform, PVWatts input, another model or a generic default. Ask for the study date, the person or system that produced it and the output by roof plane when available. PVWatts exposes shading among its system-loss categories, but that field is a user input; the calculator cannot prove the entered number describes this property. [1] Replace vague labels such as “minimal shade” with the evidence and assumption actually used.

Trace shade through every modeling layer

Write down where shade is represented. A proposal platform may calculate obstruction effects separately and then send an adjusted value into PVWatts or another production model. If a second shading percentage is also included in a general loss field, the same effect may be counted twice. If remote imagery informed panel placement but no shade effect reaches the energy model, it may be counted nowhere. Ask the designer for a short bridge: observed source → roof plane → model field → production result.

Compare design alternatives, not just percentages

If shade affects only part of the roof, compare the layout consequence. Request a modeled option that removes the most affected modules, moves them to a different plane or retains them with the stated electronics and assumptions. Hold the remaining inputs constant so the tradeoff is visible. A larger module count does not automatically produce the stronger design, and a lower shade percentage does not prove a better model. Review annual energy, equipment count, roof use, installation scope and cost together.

Separate current canopy from promised vegetation work

List every tree-pruning or removal assumption beside the party responsible, required authorization and timing. A model based on work that is optional, prohibited or controlled by a neighbor is not the same as a model of current conditions. Preserve trees or obstructions expected to remain and identify future growth as uncertainty rather than pretending it can be forecast precisely. The proposal should state whether its annual kWh result uses current shade, a documented post-trim condition or two scenarios.

Freeze the post-survey baseline

After the qualified site survey, save the final shade output, roof-plane layout, production-model inputs, approved vegetation assumptions and proposal revision together. If the layout changes, require the shade treatment and annual kWh estimate to change with it. Keep the dated record for later service questions, but do not use it as an energy guarantee. Its purpose is narrower: a future reviewer can see what the designer observed, what the model assumed and which decision the homeowner approved.

Primary sources

1. National Renewable Energy Laboratory: PVWatts Calculator

   Primary source reviewed October 10, 2026. Utility rules, equipment availability, project design and contract terms remain specific to the property and proposal.

2. U.S. Department of Energy: Homeowner’s Guide to Solar

   Primary source reviewed October 10, 2026. Utility rules, equipment availability, project design and contract terms remain specific to the property and proposal.

Planning examples and questions are Origin’s editorial guidance, not a property-specific diagnosis, engineering design, tax determination or promise of savings. Manufacturer and utility references do not imply an affiliation with Origin.

Related guidance

• Request a free proposal review
• Check the complete system-loss stack
• Reconcile the array layout by roof plane