The three roof factors that separate a great solar site from a marginal one ??and why a site-specific shade analysis beats an average production estimate every time.
Direct answer
South-facing roof at 15??0° tilt produces maximum annual output. East or west-facing roofs produce 10??0% less annually but can be preferred in TOU rate structures where morning or afternoon production is more valuable. Shading is the most impactful variable: a single shaded panel can reduce the output of an entire string inverter circuit by 50??0% during shade hours. Roofs needing replacement within 5?? years should typically be replaced before solar installation to avoid panel removal costs.
Key takeaways
- A system sized for full sun on a shaded roof is not a good deal. Ask specifically how much of the roof area is shaded between 9am and 3pm in summer months, and whether the design uses panel-level electronics to handle it.
- Ask the installer: what shade analysis tool did you use?
- Ask: did the analysis include winter sun angles, or only summer peak production?
Evidence snapshot
This article was reviewed by Solar Payback Map Editorial against public solar payback sources and the Solar Payback Map editorial policy.
Roof orientation: south is best, east and west are viable
A south-facing roof at 30° tilt produces the most annual energy at most US latitudes. PVWatts models this as the reference case. East and west-facing roofs produce 10??0% less annually, but they spread production more evenly across morning and afternoon hours rather than concentrating it at solar noon.
In utilities with time-of-use pricing, a west-facing array can actually be more valuable per kWh than a south-facing one: it generates in late afternoon when peak TOU rates apply. If you are on a flat rate plan, south is still optimal.
| Orientation | Annual production vs south-facing |
|---|---|
| South (180°) | 100% ??reference case |
| Southwest (225°) | 96??8% |
| Southeast (135°) | 95??7% |
| West (270°) | 82??0% |
| East (90°) | 81??8% |
| North (flat roofs only, tilted) | 60??5% |
Shading: the biggest production variable
Shade from trees, chimneys, dormers, or neighboring structures affects solar production in two ways. First, shaded panels produce less power directly. Second, with string inverter systems, a shaded panel drags down the output of every panel connected to the same string ??sometimes reducing whole-string output by 50??0% during the shade event.
Microinverters and power optimizers (panel-level electronics) largely eliminate the string effect: each panel operates independently, so a shaded panel affects only its own output.
How shade analysis should be done
Professional shade analysis uses either a device like the Solmetric SunEye (physically measures shade on the roof) or shading software like Aurora or Helioscope, which models shade from satellite and LiDAR data.
Satellite-based shade models are faster and less expensive, but can miss seasonal shade from deciduous trees, new construction, or obstructions that do not appear clearly in imagery. An in-person roof assessment is more accurate for shade-sensitive sites.
- Ask the installer: what shade analysis tool did you use?
- Ask: did the analysis include winter sun angles, or only summer peak production?
- Ask: are any panels placed under known shade paths? If yes, does the design use panel-level electronics?
- Self-check with PVWatts: enter your address and check the 'shading' input ??it lets you manually model shade fractions by month.
Roof age and structural condition
Solar panels are warrantied for 25 years. If your roof is 15+ years old and made of asphalt shingles with an expected lifespan of 20??5 years, the roof may need replacement while panels are installed ??triggering a $5,000??15,000 panel removal and reinstallation cost.
Reroofing after solar installation is not impossible, but it adds cost and complexity. Most solar installers recommend replacing an aging roof before installation, particularly if the roof is within 7 years of expected end-of-life.
When roof constraints make solar not viable
North-facing roofs in the northern US (above 35° latitude) are rarely economically viable ??production is too low. Heavy year-round shading from mature trees or nearby buildings can reduce production enough that payback periods exceed 20 years on a realistic basis.
Flat commercial-style roofs on residential buildings are workable with tilted racking, but add cost and sometimes require structural assessment. Very complex roofs with many small planes are harder to design efficiently.
FAQ
- Can I put solar on an east-west roof with no south exposure?
- Yes, many systems use both east and west faces of a ridge roof. Each face gets half the system and produces at different times of day. The combined annual output is roughly 85??0% of a same-size south-facing system. In TOU rate markets, the morning east-array production and afternoon west-array production can align well with peak rates.
- How much does shading actually reduce solar production?
- It depends on the system design. With string inverters, a single panel at 50% shade during peak hours can reduce the output of 4?? connected panels proportionally during that window. With panel-level electronics (microinverters or power optimizers), the same shading affects only the single panel. In heavily shaded sites without panel-level electronics, annual production losses of 20??0% are not uncommon.
Next step
Sources and further reading
Editorial review
- Reviewed against public sources listed above, not installer lead-generation data.
- Written for homeowner decision quality, with conservative assumptions favored over sales optimism.
- Updated and checked for policy, rate, source, and quote-risk context.
Read the Solar Payback Map editorial policy and Solar Payback Map Editorial profile for source, correction, advertising, authorship, and review standards.
This article is general information, not financial, tax, legal, or engineering advice. Verify current incentives, utility tariffs, and quote-specific assumptions before relying on any estimate.