Customer Says Panels Look Dirty But Production Fine Clean Now vs Defer Decision Tree

Why this matters

A customer who can see dust on their panels often assumes dirty equals lost money and asks for a cleaning. But uniform light soiling on a tilted array usually costs very little production, rain clears most of it, and rooftop cleaning carries real fall and electrical risk. The decision is not "clean because it looks dirty," it is "does the measured production loss, the soiling type, and the rinse outlook justify the risk and effort of cleaning now, or defer to the next rain or a scheduled visit?" Cleaning the wrong array wastes a truck roll and exposes a tech to roof risk for a fraction of a percent; deferring the wrong one leaves a real, localized loss (bird droppings, lichen, ash) clipping a string for months.

Symptom presentation

• Uniform thin dust film, production tracking the model: cosmetic soiling, typically small loss, rain-clearable.
• Localized hard soiling (bird droppings, sap, lichen, agricultural/industrial deposit, wildfire ash) on specific modules: can cause hotspots and disproportionate loss because one shaded cell drags a substring via bypass diodes.
• Low-tilt or flat arrays with pooling/edge crud: rain does not self-clean these well; soiling accumulates at frame edges.
• A production droop that recovers after rain: confirms soiling was the cause and rain is an effective cleaner here.

Quick checks

1. Compare actual production to the model/expected specific yield over the recent clear days. If production matches expectation, "dirty" is cosmetic.
2. Inspect the soiling type and distribution. Uniform film differs entirely from concentrated droppings or lichen.
3. Check the array tilt and the rain outlook. Steeper tilt plus regular rain means deferral is usually fine; low tilt plus a dry season favors cleaning hard deposits.
4. Look for hotspot risk: any module with concentrated opaque soiling over cells can run hot and is worth removing even if array-level production looks fine.

Isolation tree

Step 1: Is measured production materially below expectation? No (production matches the model within normal scatter): the array is not losing meaningful energy. Go to Step 2 to decide whether any non-energy reason to clean exists. Yes (a real droop): go to Step 3.

Step 2: Is the soiling localized and capable of causing harm even at low array-level loss? If there are concentrated droppings, sap, or lichen on specific modules, those create hotspot/PID-adjacent risk and uneven aging even when the array total looks fine; targeted cleaning of those modules is justified. If it is only a uniform cosmetic film with good production, defer: the energy case does not support a risky roof visit, and rain will handle it. Set expectations with the customer.

Step 3 (real droop): Will rain clear it? Assess soiling type. Uniform dust on a tilted array that recovers after rain: defer to the next rain unless a long dry spell is forecast, in which case schedule a cleaning. Hard/sticky/biological deposits that rain will not remove (lichen, sap, baked-on dropping, ash crust): clean now, because rain will not fix it and the loss persists.

Step 4: Weigh the access risk against the recoverable energy. Even with a real, rain-resistant loss, factor roof pitch, height, surface condition, and weather. If access is hazardous, schedule the cleaning into a planned visit with proper fall protection rather than an ad hoc climb, and quantify the recoverable energy so the customer understands the trade.

Confirming diagnosis

• Cosmetic, defer confirmed when production matches the model, soiling is a uniform film, the array is tilted, and rain is in the outlook.
• Clean now (energy) confirmed when a measured droop correlates with soiling and the deposit type will not rinse off naturally.
• Clean now (harm) confirmed when localized opaque deposits create hotspot risk regardless of array-level production.
• Rain will fix it confirmed when a prior post-rain interval shows production recovered to model after a similar film.

Remediation

• Defer: document that production is within model, advise the customer that uniform dust is normal and rain-cleared, and set a re-check at the next scheduled visit.
• Clean (uniform, dry-season): use water and a soft soft-bristle method, deionized or low-mineral water to avoid spotting, no abrasives or harsh chemicals that void module warranties; clean in cool hours to avoid thermal shock.
• Clean (localized/biological): targeted removal of the deposit, inspect underlying cells for hotspot damage, and verify the affected substring recovers.
• Recurring soiling: recommend a maintenance cadence matched to the site (agriculture, construction, wildfire-prone, low-tilt) rather than reactive calls.

References

• IEC 62446-1, Photovoltaic (PV) systems - Requirements for testing, documentation and maintenance
• NEC (NFPA 70) Article 690.4, General Requirements for PV Systems
• UL 1741, Inverters, Converters, Controllers and Interconnection System Equipment
• NEC (NFPA 70) Article 690.12, Rapid Shutdown of PV Systems on Buildings