Solar basics, by the numbers
Do solar panels need direct sunlight?
The short answer
No — solar panels do not require direct beam sunlight. Photovoltaic cells convert all daylight, including diffuse (scattered) light, into electricity. On fully overcast days they still produce roughly 10–25% of rated output, and on partly cloudy days 50–80%. For a mobile or manufactured home, the deciding factors are location, heat, shade, and — above all — your roof, not whether the sky is blue.
1. The physics: direct vs. diffuse light
When light strikes a photovoltaic cell, the U.S. Department of Energy explains, that light "may be reflected, absorbed, or pass right through" — and the absorbed photons knock electrons loose to create a current. Nothing in that mechanism requires the light to arrive as a straight, unscattered beam. Cells respond to photons across the visible spectrum whether they come as direct beam or diffuse sky light.
Three irradiance terms (all in W/m²) describe what actually reaches a panel:
- DNI (Direct Normal Irradiance) — the direct beam, measured perpendicular to the sun's rays. This is what casts sharp shadows.
- DHI (Diffuse Horizontal Irradiance) — scattered sky light, arriving from all directions after being diffused by water vapor, aerosols, and clouds.
- GHI (Global Horizontal Irradiance) — the total on a horizontal surface: GHI = DNI (adjusted for solar angle) + DHI + ground reflection.
A clear-sky maximum at sea level is about 1,000 W/m² (also the lab benchmark used for panel ratings), while any irradiance over roughly 300 W/m² is "good" operating light. On a clear day about 10–25% of energy arrives as diffuse; on a fully overcast day it can be nearly 100% diffuse — and the panel still works.
2. Cloudy-day performance
Multiple sources citing NREL/DOE irradiance data converge on the ranges below. A real-world field study at the University of Port Harcourt on a 250W monocrystalline module measured 76.2% of output retained under light cloud and 33.25% under heavy cloud.
Solar output by sky condition (% of rated capacity)
Panels keep producing in cloud. Output as a share of clear-sky rated capacity, from irradiance data attributed to NREL/DOE's National Solar Radiation Database (as compiled by installer sources) plus a field study.
| Sky condition | Output (% of rated capacity) |
|---|---|
| Clear / full sun | 100% (reference) |
| Light cloud / haze | 80–90% |
| Partly cloudy | 50–80% |
| Heavy / full overcast | 10–25% |
| Rain (thick cloud) | 10–20% |
| Field study — 250W mono, light cloud (Univ. of Port Harcourt) | 76.2% retained |
| Field study — 250W mono, heavy cloud | 33.25% retained |
Takeaway: Panels never stop working in cloud — even heavy overcast yields ~10–25% of rated output, and partly cloudy days often hold 50–80%.
Source: NREL/DOE National Solar Radiation Database · U.S. DOE — Solar PV Cell Basics · Vintage: Corroborated ranges, 2024–2026
⚠ Handle with care: These cloudy-day percentages are widely attributed to NREL/DOE irradiance data but appear mostly in secondary/installer sources, not one named NREL publication — treat as well-corroborated ranges, not exact constants. The field-study figures are from a single tropical-climate 250W module.
Germany is the definitive proof that cloudy climates work: despite averaging only ~3 peak sun hours per day, it reached 99.3 GW of installed solar across more than 4 million systems by the end of 2024 (Bundesnetzagentur). Note two nuances — the "edge-of-cloud effect" can briefly push output above nameplate when sunlight reflects off cloud edges, and monocrystalline panels capture roughly 15–20% more diffuse light than polycrystalline.
3. Heat matters more than clouds
Panels are rated at 25°C and lose power as they heat up — and cells typically run 25–35°C above ambient air. On a 35°C day, flush-mounted panels can hit 70°C. The metric to compare is the temperature coefficient of Pmax (%/°C):
Temperature coefficient of Pmax by cell technology
Panels are rated at 25°C and lose power as they heat up. Lower (closer to zero) is better. Bar length = %/°C lost above 25°C.
| Manufacturer / technology | Pmax temp. coefficient |
|---|---|
| REC Alpha Pure (HJT) | −0.24%/°C |
| Panasonic EverVolt (HJT) | −0.26%/°C |
| Jinko N-type (TOPCon) | −0.30%/°C |
| Qcells Q.PEAK DUO (PERC) | −0.34%/°C |
| Jinko Tiger (PERC) | −0.35 to −0.36%/°C |
| Industry PERC average | ≈−0.38%/°C |
Takeaway: At 65°C cell temperature a −0.36%/°C PERC panel loses ~14% of rated power vs ~10% for HJT. On a dark, low-vent manufactured-home roof, prioritize −0.30%/°C or better.
Source: Manufacturer datasheets (REC, Panasonic, Jinko, Qcells) · Vintage: Current model-year datasheets
⚠ Handle with care: Coefficients vary by exact model and production year — always check the datasheet for the specific panel you are quoted.
This is especially relevant for manufactured homes, which often have dark roofs, minimal attic ventilation, and low-profile flush mounting — all of which push cell temperatures higher. In a hot climate, a low coefficient (TOPCon or HJT, −0.30%/°C or better) and an air gap under the panels are worth prioritizing.
4. Shade is the real enemy
Because panels in a string are wired in series, shade has an outsized effect — and this, not cloud cover, is what most often disappoints owners.
How one shaded panel drags down the system
Panels in a string are wired in series, so shade on one module hurts the whole string — unless module-level electronics isolate it. Share of a 10-panel array's output lost when one panel is fully shaded.
| Scenario | Effect |
|---|---|
| Shade covering one cell (~1.5% of a 400W panel) | Cuts that panel's output 35–40% |
| One fully shaded panel — string inverter | Reduces the whole string 30–80% |
| One fully shaded panel — microinverters / optimizers | Loss isolated to that panel (~10% of a 10-panel array) |
| Bypass diodes (≈3 per module) | Route current around shaded cells; the bypassed section still makes 0 W |
Takeaway: For manufactured homes near trees, awnings, or neighboring units, module-level electronics (microinverters or DC optimizers) are the single highest-value design choice.
Source: EnergySage — inverters & shading · Vintage: Current
Bypass diodes (about three per module) route current around shaded cell groups, and half-cut-cell panels handle partial shade better. But the decisive fix is module-level electronics: microinverters or DC optimizers isolate each panel so shade on one doesn't drag down the rest. For a manufactured home sited near trees, awnings, or neighboring units in a park, this is the highest-value design choice you can make.
5. Where you live sets the ceiling
Using NREL's PVWatts v8 model, the same 1 kW of panels produces very different annual energy by location:
Annual solar production by location (per kW installed)
Same 1 kW of panels, very different yield by location — modeled in NREL's PVWatts v8 with typical-meteorological-year weather.
| Location | kWh/kW/year | Peak sun hours/day |
|---|---|---|
| Phoenix, AZ | ~1,800 | ~6.5 |
| U.S. average | ~1,500 | ~5.0 |
| Seattle, WA | ~1,100 | ~3.7–4.0 |
Takeaway: Location drives a ~57% gap — a 6 kW system makes ~10,400 kWh/yr in Phoenix vs ~6,600 in Seattle — yet cloudy Seattle is still very much worth doing.
Source: NREL PVWatts v8 · NREL Annual Technology Baseline · Vintage: PVWatts v8 (TMY modeling), ±10–15%
As NREL puts it, "flat-plate PV can use direct or indirect insolation, so PV modules need not directly face and track incident radiation" — which is exactly why rooftop solar has such broad geographic reach. Seasonal swing is real, though: summer months run ~150–180% of the annual monthly average, winter ~60–80%.
6. For manufactured homes, the roof is the constraint
You can absolutely go solar on a HUD-code home (built after June 15, 1976) — but the limiting factor is structural, not solar. Under HUD Code (24 CFR § 3280.305), roof live-load zones are 40 psf (North), 30 psf (Middle), and 20 psf (South), while a solar array plus racking adds only ~3–4 psf of dead load. The catch is that this weight lands as concentrated point loads at the rack attachments, and manufactured-home trusses are lightweight. Every HUD home has a metal data plate (often in a kitchen cabinet or closet) listing its roof-load zone and wind zone — installers and permit offices require it before issuing a solar permit.
That's why many installers recommend a ground-mounted array: it sidesteps roof-load and truss concerns, allows optimal tilt and orientation, and typically produces more energy per kW. Pre-1976 homes are rarely approved for roof mounts at all. For sizing, manufactured homes average 11,787 kWh/year of electricity use (ACEEE, using DOE RECS data) — higher than the site-built average, so efficiency upgrades before solar can meaningfully shrink the system you need.
7. What to actually do
- Confirm feasibility first. Find your HUD data plate, record the roof-load and wind zone, and pull 12 months of electric bills. If the home is pre-1976, plan for a ground mount from the start.
- Design for your site. If anything casts shade, specify microinverters or optimizers. In hot climates, choose panels rated −0.30%/°C or better. If you have the land, strongly consider a ground mount.
- Verify and permit. Get a structural assessment (typically $300–$1,500), use an installer experienced with HUD-code homes, run your address through NREL's free PVWatts calculator, and notify your insurer before the project is closed out.
Frequently asked questions
Do solar panels need direct sunlight to work?
No. Photovoltaic cells convert all daylight — including diffuse, scattered light — into electricity, so they still work on cloudy days and in indirect light. Direct sun maximizes output but is not required.
How much power do solar panels produce on a cloudy day?
Roughly 10–25% of rated output under heavy overcast and about 50–80% on partly cloudy days, based on NREL/DOE irradiance data. A field study on a 250W module measured 76% of output retained under light cloud and 33% under heavy cloud.
Do solar panels work in the shade?
Partially. Because panels in a string are wired in series, one fully shaded panel can cut a string inverter's output by 30–80%. Microinverters or DC optimizers isolate the shaded panel so the rest of the array keeps producing.
Can you put solar panels on a mobile or manufactured home?
Yes — HUD-code homes (built after June 15, 1976) can go solar. The main constraint is the roof, not the sky: HUD roofs have tight load margins (North 40 / Middle 30 / South 20 psf live load), so many installers recommend a ground mount, especially for pre-1976 homes.
Is rooftop or ground-mount solar better for a manufactured home?
Ground mounts are often the smartest choice: they sidestep roof-load and lightweight-truss concerns, allow optimal tilt and orientation, and usually produce more energy per kW installed. Roof mounts work when a structural assessment confirms the roof can carry ~3–4 psf plus concentrated point loads.
Caveats worth knowing
The cloudy-day percentages (10–25%, 50–80%) are widely attributed to NREL/DOE irradiance data but appear mostly in secondary and installer sources — treat them as well-corroborated ranges, not exact constants. Manufacturer temperature coefficients vary by model and year. PVWatts yields are modeled estimates (±10–15%). And note: multiple 2026 industry sources report the 30% Section 25D residential solar tax credit expired December 31, 2025 under the One Big Beautiful Bill Act — verify current federal, state, and utility incentives independently before relying on them.