Solar panel cleaning is a service industry, and like most service industries it is better at describing the problem than at sizing it. The question is not whether dirty panels produce less. They do. The question is whether the electricity recovered is worth more than the cleaning cost.
That is arithmetic, and it can be done before anyone gets on a roof.
How much soiling actually costs
The US National Renewable Energy Laboratory puts typical soiling losses at around 5% of output, ranging from about 3% up to 25% in the worst locations. That range is enormous, and where you sit in it is decided almost entirely by rainfall.
The clearest measurement of the dry end comes from a study of 186 solar sites in California, which found a 7.4% loss after 145 days without rain or cleaning. Nearly five months of accumulation, in a dry and dusty climate, to reach 7.4%.
A British roof does not go 145 days without rain. Rainfall is the mechanism that removes accumulated dust, it arrives regularly, and panels are mounted at an angle that helps it run off. The soiling loss on a typical UK rooftop array sits at the low end of the NREL range, and it is already inside the 14% system loss assumption in a standard yield model.
The break-even, stated properly
A recovered kilowatt-hour is worth different amounts depending on what happens to it. If it displaces electricity you would have bought, it is worth the import price, currently capped at 26.32p per kWh for direct debit customers in Great Britain for the quarter from 1 October 2026. If it is exported instead, it is worth your export rate, which for a household on a legacy Smart Export Guarantee tariff is a small fraction of that.
Those two values bracket the answer.
| Cleaning cost | If it displaces import at 26.32p | If it is exported at 4.1p | As % of annual output, at import value |
|---|---|---|---|
| £60 | 228 kWh | 1,463 kWh | 6.1% |
| £100 | 380 kWh | 2,439 kWh | 10.1% |
| £150 | 570 kWh | 3,659 kWh | 15.2% |
| £250 | 950 kWh | 6,098 kWh | 25.3% |
Percentages are against the 3,761 kWh annual output of the reference 4 kWp array used throughout these guides. Import price is the Ofgem price cap electricity unit rate for 1 October to 31 December 2026. The export figure is the legacy export-only rate this site assumes when a household does not know its own. Cleaning prices are shown as a range to work against, not as a quoted price.
Read the third column. Even in the most favourable case, where every recovered unit displaces an imported one, a £100 clean has to find 10% of annual output. Soiling losses of that magnitude belong to arid and dusty sites, not to a pitched roof in a maritime climate.
And most households do not consume every recovered unit. Extra generation at midday, when the panels are at their best and the house is often empty, is disproportionately likely to be exported, which puts the real break-even somewhere between the two columns and closer to the wrong one.
The three cases where it does pay
1. Localised opaque soiling
Bird droppings, lichen and packed leaf litter are different from dust, because they are opaque and they sit over a few cells rather than spreading thinly over everything. That is exactly the condition that makes a bypass diode conduct, and a conducting diode takes out roughly a third of the module it is in. A single dropping can therefore cost far more than its area implies, and removing it is worth doing. The mechanism is set out in what shade does to an array.
2. Shallow tilt
Rain cleans panels because water runs off them and carries dirt with it. Below about 10° it runs off slowly, pools at the frame edge and leaves a dirt line along the bottom of each module. Flat and near-flat installations accumulate soiling that rain does not clear, and they need intervention that pitched roofs do not.
3. A genuinely dirty environment
Sites near arable fields during harvest, unsurfaced roads, quarries, cement works or heavy industry accumulate far faster than the NREL typical case. If there is visible film on the glass within weeks of rain, the local rate is not the typical rate and the arithmetic above should be rerun with a soiling figure measured rather than assumed.
Measuring instead of guessing
The cheap way to find out whether soiling is costing anything is to compare output against the same period in a previous year, or against a neighbouring system, and then look at what happens after a heavy rainfall. If output steps up after rain and drifts down between, soiling is measurable and its size is the size of that step. If nothing changes, there is nothing to recover.
Systems with per-module monitoring make this easier still, because a single dirty module shows up against its neighbours rather than being averaged into a string total.
One practical note. Cleaning means either working at height or paying someone who is insured to. Deionised water and a soft brush on a pole is the standard method; abrasives, detergents and pressure washers risk the anti-reflective coating and the frame seals, and a damaged laminate costs considerably more than the output it was meant to recover.
Soiling loss range from the US National Renewable Energy Laboratory. The 7.4% figure after 145 days without rain is from Mejia and Kleissl, Soiling losses for solar photovoltaic systems in California, Solar Energy 95, 2013, pp. 357–363, covering 186 sites. Checked 2026. Source