Soiling losses are the accumulation of dust, pollen, salt, soot, agricultural residue and bird droppings on module glass, and the common assumption that rain handles it is only sometimes true. A heavy downpour on a steep array does clean well. A light drizzle can mobilize dust into a film and dry into a haze that is worse than what was there before. What determines the loss is the local dust, the rainfall pattern, the tilt and what sits upwind.
What actually accumulates and where
Soiling is not uniform across a module. Water running down the glass carries particles to the low edge, where the frame stops the flow and a band of concentrated grime builds up along the bottom row of cells. That band matters far more than its area suggests, because the cells it shades are in series with everything else in their sub-string. A narrow, dense strip of dirt along the bottom edge can cost more output than a much larger area of light haze elsewhere. This is also why frameless and frame-flush module designs and steeper tilts shed better: there is no ledge for the accumulation to build against. In agricultural areas, near unpaved roads, under flight paths, near industrial sources or along a coast with salt spray, the accumulation rate can be many times higher than in a suburban setting.
Why rainfall pattern matters more than rainfall total
Two sites with identical annual rainfall can have very different soiling losses. A climate with regular heavy rain events keeps arrays clean. A climate with a long dry season followed by a wet one lets soiling build for months, and the first rains of the season arrive as light showers that wet the dust without carrying it away. Dew does the same thing on a smaller scale every night in some climates, cementing dust into a film that a subsequent light rain will not lift. The practical implication is that soiling loss estimates are strongly regional, and a national average number in a production model may be far off for your site. Local installers with data from nearby systems are the best source.
When cleaning pays and when it does not
The calculation is simple in principle: compare the energy recovered against the cost and the risk. In a temperate suburban location with regular rain and a decently tilted array, the recoverable energy is often small enough that professional cleaning does not pay for itself. In a dusty, dry or agricultural setting, or where birds roost nearby, it clearly does. Monitoring gives you the answer for your own site. Watch production against a weather-corrected expectation through a dry spell and see how far it drifts, then see how much a cleaning recovers. That measurement, repeated a couple of times, tells you your real cleaning interval far better than any general recommendation.
How to clean without destroying anything
Module glass has an anti-reflective coating that abrasives and harsh chemicals damage permanently. Cleaning is done with plenty of clean water, a soft brush or squeegee, and patience, preferably early or late when the glass is cool. Spraying cold water onto hot glass risks thermal shock. Pressure washers can drive water past the junction box seal and the frame gasket. Standing on modules cracks cells invisibly. And climbing onto a roof to clean an array is, statistically, far more dangerous to the owner than the soiling is to the production. On most residential roofs the right answer is either to leave it to rain or to hire someone with fall protection who does this professionally.
The things that are not really soiling
Production drops get blamed on dirt when the cause is something else. A tree that has grown, a new rooftop unit casting an afternoon shadow, a failed module-level device, a degraded connector, an inverter derating on hot days, or a monitoring sensor that drifted all look like gradual decline. Before scheduling a cleaning, look at whether the loss is uniform across strings or concentrated, whether it appeared gradually or as a step, and whether it correlates with time of day. Per-module or per-string monitoring answers those questions quickly. Cleaning an array to fix a failed optimizer wastes money and leaves the real problem in place.
What to check before you act
- Look for a concentrated band of grime along the bottom module edge.
- Track production against a weather-corrected model through a dry spell.
- Rule out shading, hardware failure and derating before blaming dirt.
- Never use abrasives, harsh chemicals or a pressure washer on module glass.
Common questions
Do coatings that repel dirt work?
Results vary widely by product and environment, and some wear off within a season. Measure your own soiling rate before paying for a treatment that promises to reduce it.
Is it worth cleaning once a year on a schedule?
Only if your measured soiling loss justifies it. In many temperate locations the recovered energy does not cover the cost, while in dusty regions more frequent cleaning pays easily.
The practical takeaway
Measure your soiling loss instead of assuming it, and check that the loss is really dirt before you spend money on cleaning. If cleaning does pay, use water and a soft brush, never abrasives, and hire someone with proper fall protection rather than climbing up yourself.
References and further reading
- NREL PVWatts: production modeling with adjustable system loss assumptions including soiling
- NREL: photovoltaics research covering field performance and soiling studies
- NREL: a technical report from the laboratory’s photovoltaics publication library
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
