How often should commercial solar panels be cleaned?

The most useful published work here is IEA-PVPS Task 13’s report Soiling Losses – Impact on the Performance of Photovoltaic Power Plants, released in January 2023. It estimates that soiling causes a loss of annual PV energy production of 3–5%, an economic loss the report puts on the order of 3 to 5 billion euros.

The second finding matters more. The report describes PV soiling as highly heterogeneous, both at the module and plant level, and as requiring multi-sensor networks to assess accurately. Soiling varies not only between sites but across a single array. That is the honest answer to “how often”: it depends, and strongly enough that a national rule of thumb carries little meaning.

Be clear about what the figure is not. It is a global average across all climates: the report notes soiling is a factor in equatorial regions with high densities of suspended atmospheric particles and in high-latitude regions because of snow cover. A rain-washed roof in the West Midlands is not the same problem as a plant in the Gulf.

Claims circulate in the UK market that cleaning lifts output by 25 to 40 per cent. Figures of that order do occur, but they describe arrays that were visibly and heavily soiled: thick bird fouling, an agricultural dust layer, moss creeping in from the frame edges. Restoring an array from that condition produces a large percentage change, because the starting point was very poor.

That is not the condition of a typical UK commercial roof receiving regular rainfall. Applying an uplift measured on a badly soiled array to a moderately dusty one overstates the return, often by a wide margin. Such numbers get repeated in good faith and harden into received wisdom, which is reason enough to ask where a quoted uplift came from.

For a normally weathered UK array, expect cleaning to recover a modest single-digit percentage of annual generation, and more on the modules worst affected. Whether that is worth a visit is arithmetic, answerable from your generation data and the rate you pay per kilowatt-hour.

Tilt is the most underrated factor, because rain only cleans when water runs off. Most large flat-roof arrays are mounted at a shallow angle to reduce wind loading and row spacing, so water moves slowly, gathers at the lower edge of each module and dries there, leaving a band of concentrated soiling along the bottom. Since cells are wired in series, that band holds back more than its share of the module’s output. Steeply pitched arrays shed water and dirt far better.

Rainfall pattern matters more than volume. Frequent light rain keeps glass reasonably clear; a long dry spell in spring or summer lets soiling build through the months when generation is highest. A wet winter cleans the array when there is least to gain from it.

Surroundings are the reliable predictor: farmland with its ploughing, harvest dust and slurry; quarries, cement works and food processing; a motorway or main road; a coastal salt environment; a flight path. Birds sit in their own category, particularly gulls near landfill and pigeons roosting beneath the array. Bird fouling is not dust — it is opaque, adheres hard, is not removed by rain, and shades cells completely rather than dimming them.

A monitored array gives a better answer than any schedule, and three comparisons do most of the work.

Measured output against expected yield. Monitoring platforms model what the system should have produced given the irradiance and temperature it actually experienced. A persistent gap between modelled and measured output, when nothing has faulted and nothing new is shading the roof, points at soiling.

Performance ratio. PR expresses actual output as a proportion of what the installed capacity should have delivered under the conditions measured, which strips the weather out of the comparison. A slow decline in PR through a dry spell that recovers after heavy rain is close to a signature for soiling.

The same month, year on year. Comparing this June against the last two is crude, needs no modelling and works: if output is down and conditions were comparable, something has changed — soiling, vegetation grown into the shading path, or a developing fault. The rule that falls out of all three: clean when the loss is worth more than the visit. Set a review interval, not a cleaning interval.

Almost all commercial array cleaning is work at height, which changes who can sensibly do it. The Work at Height Regulations 2005 require the work to be properly planned, supervised and carried out by competent people, following a hierarchy: avoid work at height where reasonably practicable; prevent falls using a place of work that is already safe or the right equipment; and otherwise minimise the distance and consequences of a fall.

HSE guidance on roof work is direct. All roofs should be treated as fragile until a competent person has confirmed they are not, and the main causes of death and injury are falls from roof edges or openings, through fragile roofs and through fragile rooflights. Older industrial roofs often have rooflights close to the array. This is why cleaning is not a task for site staff with a ladder and a hosepipe. Where an array can be reached from ground level with a water-fed pole system much of that falls away, which is worth establishing when a provider surveys the site.

The standard method is purified or deionised water with a soft brush, usually on a fed pole. Purified water matters because ordinary mains water leaves mineral deposits as it dries — cosmetic at first, a soiling problem later.

What to avoid is well established. Abrasive pads scratch the anti-reflective coating and permanently reduce light transmission. Detergents and solvents leave residues that attract dirt and may attack frame seals. High-pressure washing can force water past a frame seal or into a junction box. Walking on modules causes microcracks that do not announce themselves until output quietly drops. Cleaning outside the manufacturer’s stated method is a common way to compromise a module warranty, and it is that guidance which settles the point, not a contractor’s habit.

For UK practice, Solar Energy UK’s Rooftop O&M Guidelines 3.0, published in December 2025, is the current reference. It gives practical advice for the safe maintenance, cleaning and monitoring of domestic and commercial PV systems while emphasising the need to follow relevant legislation and industry best practice, covering safety, training, working at height, inspection, monitoring and warranty matters.

Monitor continuously. Review performance against expected yield and against the same month last year. Inspect the array visually at the annual maintenance visit, when someone is already safely on the roof. Then clean when the data or the inspection justifies it. Sites with a persistent soiling source — farmland, a quarry, a gull colony — are the exception, and a fixed frequency agreed with a provider who has seen the site is reasonable there. Where a system is funded through a power purchase agreement, operations and maintenance sit with the funder, who owns the array and carries that cost.

PPA Funding does not clean panels and does not maintain a standing list of cleaning firms. Tell us the array size, the roof type and pitch, and how the roof is accessed; we source and vet a specialist suited to the site and pass your enquiry on. The provider surveys, quotes and contracts with you directly.