Solar shading analysis finds how much sun a roof loses to shade and what that costs in yearly energy. Check far shade with a horizon profile and near shade with a 3D model or an on-site tool. Run a full year. Read the electrical loss, not just the shaded area. Then confirm on site.
SurgePV publishes this guide and sells one of the tools it mentions. For a price table of shading tools, read our shading analysis tools comparison. This page covers the method: what shade does, how to measure it and how to work out the loss.
In short
Shade costs more than its share of the roof area, because cells in a string share one current. Use the right method for the kind of shade. Simulate every hour of the year. Check the model against the real roof.
What is solar shading analysis?
Solar shading analysis measures how much sunlight shadows block from a solar array over a year. The result is a loss figure, in percent of yearly energy, by month and by roof face. You use it to move modules, change the wiring, or decide not to build.
Do it before you sell the system. If you are new to the subject, start with how solar panels work. A shading estimate that is too low leads to a yield shortfall after install. The customer notices, and you pay for it.
Types of shade
PVsyst, the yield software, splits shade into two kinds in its shadings documentation.
| Type | What causes it | How to model it |
|---|---|---|
| Far shade | Hills, mountains, a distant skyline. Objects roughly ten times the size of the PV field away or more | A horizon profile: the height of the skyline in each compass direction |
| Near shade | Chimneys, parapets, dormers, nearby buildings, trees | A 3D model of the system and what is around it |
| Self-shade | The array shading itself, such as one row of modules shading the next | Row spacing and tilt in the layout |
With far shade, the sun is either visible or hidden for the whole field at any moment. With near shade, a shadow covers part of the array. PVsyst calls near-shade treatment “much more complex” than far shade, and it must be computed at each time step.
Shade also moves. A shadow that is absent at noon in June can cover part of the array on a December morning. Tall trees and new buildings change the shade from year to year.
How shade cuts output
Cells in a panel are wired in series. The weakest cell sets the current for the whole string. PVsyst states it the same way: in a string, the total current is set by the current in the weakest cell.
A bypass diode gives current a path around a shaded group of cells. PVEducation explains that diodes usually sit across groups of cells, not single cells. NREL notes that a typical 60-cell module has three bypass diodes, so each one covers a third of the module (NREL shading testbed report).
The diode protects the cells, but the bypassed group stops producing. So a small shadow on one cell can cost a third of the module’s power.
PVsyst reports two kinds of loss for near shade:
- Irradiance loss is the sunlight the shadow blocks.
- Electrical loss is the extra power lost because the cells in a string do not all make the same current.
A tool that reports only the shaded area leaves out the electrical part. PVsyst offers a rough “according to strings” estimate, which it calls an upper limit, and a detailed calculation from the exact module positions.
Module-level power electronics (MLPE) such as optimizers and microinverters change how the loss spreads. See the glossary entries on MLPE, bypass diodes and maximum power point. For more on panel behavior, read how shading affects solar panels.
How much energy does shade cost?
It depends on the site. We found no single figure that fits every roof, so use your own model.
One real data point comes from NREL. Its 2012 report used a survey of 66 California homes from a large residential installer. Shade there was the share of yearly sunlight lost to shade. The median home lost 7.6%. NREL then picked three real homes as test cases: 7% (light), 19% (moderate) and 25.5% (heavy) shade (NREL/TP-5200-54876).
These are shade levels, not energy losses. They do not say how much energy a home lost.
The same NREL test compared a string inverter with a microinverter on two identical 8 kW arrays. The microinverter made 3.7% more energy in the light case, 7.8% in the moderate case and 12.3% in the heavy case. This was one trial with one product pair, so treat it as an example, not a rule.
Shading analysis methods
Pick the method by the shade you need to find.
| Method | What it finds | Good for | Watch out for |
|---|---|---|---|
| Horizon profile (for example PVGIS) | Shade from hills and mountains | A first check of distant shade | It does not model a chimney, a building or a tree on your roof |
| Satellite or aerial imagery | Roof outline and visible obstructions | Screening leads before a visit | Imagery goes out of date. It can miss new buildings and tree growth |
| On-site tool (Solar Pathfinder, SunEye-210) | The shade at one spot, by day and month | Roofs that are hard to model | You measure one point. The result goes into design software for yield |
| 3D model with hourly simulation | Shade from each modeled object, hour by hour | Design and proposals | The result is only as good as the site data |
| Drone or laser (LiDAR) survey | Accurate roof and tree geometry | Complex roofs and large sites | It gives the geometry. You still run the simulation |
We do not list an accuracy figure for each method. Vendors publish their own figures, and they are not comparable. Read each vendor’s documents for what it claims, and see the tools comparison for published prices.
A quick visual check of where shadows fall is called a qualitative shading analysis. It is fast, but it does not give an energy loss. Many teams stack methods. They screen with imagery, build a 3D model, then confirm key obstructions on site. Solar panel design software can run the 3D step. For a drone survey, see our drone site survey guide.
How to do a solar shading analysis, step by step
- Collect the site geometry. Record the roof outline, roof pitch, and the position and height of every obstruction. Include chimneys, vents, parapets, nearby buildings and trees.
- Pick the method. Use a horizon profile for distant terrain, a 3D model for nearby objects, and an on-site tool where the roof is hard to model.
- Run a full-year simulation. One day is not enough. Winter and summer shadows both count.
- Read the loss by month and roof face. Look at each string, not only the yearly total.
- Change the layout or the wiring. Move modules off the worst areas. Change string groups. Consider MLPE where shade cannot be avoided.
- Check on site. Confirm that the obstructions in the model still exist and that the dimensions match.
Tree growth matters. Model trees at the height they will reach, and check whether a tree is protected before you count on pruning.
How to calculate shade loss: worked examples
The numbers below are round numbers chosen to show the arithmetic. They are not measured data.
Example 1: irradiance loss
Shade loss is the sunlight each shaded period blocks, added up over the year:
loss = sum of (share of yearly sunlight in the period x share of the array shaded in that period)
Take a roof with a chimney. Say 30% of the year’s sunlight arrives in the morning, 40% around midday and 30% in the afternoon. The chimney shades 20% of the array in the morning, none at midday and 5% in the afternoon.
- Morning: 0.30 x 20% = 6.0%
- Midday: 0.40 x 0% = 0%
- Afternoon: 0.30 x 5% = 1.5%
- Total: 6.0% + 0% + 1.5% = 7.5% of yearly sunlight lost
A real tool does this for all 8,760 hours of the year. It also treats direct, diffuse and ground-reflected light differently, as PVsyst describes. This simple example treats all light the same.
Example 2: the bypass diode effect
Take a 60-cell module with three bypass diodes. Each group is 20 cells, or one third of the module. Put 10 identical modules in a string. Fully shade one cell in one module, enough to switch the diode on.
- Shaded area: 1 cell of 600 in the string = 0.17%
- Power lost: the bypassed 20 cells, 20 of 600 = 1/30 = 3.3% of string power
The loss is about 20 times the shaded area share (3.3% / 0.17%). Without the diode, the shaded cell would limit the current of the whole string, and the loss would be far larger.
Now say this shade only happens in the morning, which carries 30% of the year’s sunlight. The yearly loss for that string is 3.3% x 30% = 1.0%. This is a simplified, per-string estimate. A real tool also checks the voltage window of the inverter and the effect on the other strings.
How to read a shading report
A good report shows:
- The skyline or horizon profile and a picture of the 3D scene
- The shade loss by month and for the year
- The loss for each string or roof face
- The method used, the date of the site data and the obstructions it left out
Ask who made the model and from what data. Use the loss by string to decide where MLPE might pay off. For more, read how to read a solar shade report.
Check Shade on a Roof You Know
Bring a roof with real obstructions to a guided demo. SurgePV’s shading analysis is part of its design workflow, so you can compare it with your own survey.
Book a Free DemoNo commitment required
See shadow analysis for what SurgePV covers. The result depends on the site model and the assumptions you set.
Common mistakes
- Leaving the default “no shading” setting. Build the obstructions into the scene before you trust a yield number.
- Using one tool for every kind of shade. A horizon file does not describe a chimney.
- Checking one day. Run the full year.
- Reading only the shaded-area share. Ask for the electrical loss too.
- Skipping the site check. Imagery goes out of date.
- Counting on pruning. A promise to prune is not a contract, and some trees are protected.
How to buy solar shading analysis services
Say what the result is for: screening, proposal design, detailed engineering or a yield review. Give the site coordinates, the survey or imagery date, the roof or terrain geometry, the obstruction sizes, your tree assumptions, the proposed layout and any module-level electronics.
The report should show the data source and date, the horizon profile or 3D scene, the near- and far-shading method, the layout and orientation, the time step, the loss result, the obstructions left out and the assumptions. Ask for screenshots or geometry that let another qualified reviewer repeat the result. Compare doubtful results with site measurements, current imagery or an independent model.
Remote imagery may be old. It may not show tree growth, planned building, parapet height or fine roof features. A shading report helps with yield and layout decisions. It is not a production guarantee and does not replace a field check. See the PVsyst report guide for the yield-model checks that come next.
Disclosure: SurgePV is promoting Heaven Designs through a commercial relationship. Apply the same source-data, reproducibility, field-verification, and qualified QA checks to every provider.
Review the Heaven Designs pre-design and shading service, visit the Heaven Designs website, and request a project-matched sample before seeking a quote.
Sources
- PVsyst documentation, Shadings
- PVEducation, Bypass diodes
- NREL, Deline, Meydbray, Donovan and Forrest, Photovoltaic Shading Testbed for Module-Level Power Electronics (NREL/TP-5200-54876, May 2012)
- SurgePV, shadow analysis
Frequently Asked Questions
How do you do a shading analysis for solar?
Collect the roof outline and the position and height of every obstruction. Pick a method: a horizon profile for distant terrain, a 3D model for nearby objects, or an on-site tool where the roof is hard to model. Run a full-year simulation. Read the loss by month and by roof face. Change the layout or the electrical design where the loss is high. Then check the model against the real site.
How much does shade reduce solar output?
It depends on the site. An NREL test report cites a survey of 66 California homes where the median system lost 7.6% of its yearly sunlight to shade. Shade can cost more than its share of the area because cells in a string share one current. A small shadow on one cell can cost a whole third of a module’s power if a bypass diode switches on.
What is the difference between far and near shading?
PVsyst defines far shadings as objects far enough away that the sun is either fully visible or fully hidden for the whole array at a given time, usually more than about ten times the size of the PV field. Near shadings come from nearby objects such as chimneys, parapets and trees. They cast shadows on part of the array and need a 3D model.
Do solar panels work in partial shade?
Yes, but output falls by more than the shaded area. Cells in a string share one current. A bypass diode lets current skip a shaded group of cells, but that group stops producing. Module-level power electronics such as optimizers and microinverters change how the loss spreads through the system.
What is the difference between irradiance loss and electrical loss?
Irradiance loss is the sunlight that the shadow blocks. Electrical loss is the extra power lost because series-connected cells and modules do not all produce the same current. PVsyst reports the two separately. A tool that shows only the shaded area leaves out the electrical part.
Do I need a site visit if I use shading software?
Yes. The model is only as good as the site data you give it. Imagery goes out of date, trees grow and neighbors build. Check that the obstructions in the model still exist and that the dimensions match before you finalize the layout.
Where this fits
This article is part of SurgePV's Solar Design hub, which works through the topic from first principles to the decisions a project team actually has to make.
