Cleaning solar panels costs money. Not cleaning them also costs money — in the form of soiling loss. The optimal cleaning frequency is the point where the marginal revenue from the next cleaning session equals the marginal cost. Get this right and you extract maximum value from your O&M budget. Get it wrong and you either overspend on unnecessary cleaning or leave significant revenue on the table.
How much does soiling actually cost?
Soiling loss — the reduction in panel output due to dust, dirt, pollen, bird droppings and other surface deposits — varies enormously by location:
| Environment | Monthly soiling rate (% output loss) | Annual loss (no cleaning) |
|---|---|---|
| Desert / arid (Middle East, Atacama) | 1–4% per month | 15–40% |
| Semi-arid (Central Asia, inland Spain) | 0.5–2% per month | 8–20% |
| Temperate / humid (Northern Europe, UK) | 0.1–0.5% per month | 1–5% |
| Tropical (SEA, coastal Brazil) | 0.3–1.5% per month | 5–15% |
| Industrial urban (heavy particulates) | 0.5–2.5% per month | 8–25% |
Bird droppings are a special case — a single concentrated droppings deposit over a bypass diode can trigger the diode and kill an entire third of a module, causing disproportionate loss (sometimes 5–10× more than the deposit area would suggest). Manual spot-cleaning of bird droppings is almost always cost-justified regardless of your overall cleaning frequency.
Calculating cleaning ROI
The cleaning ROI formula:
Net cleaning benefit ($/year) = (Recovered energy × Electricity value) – Cleaning cost
Example calculation for a 1 MWp plant in a semi-arid climate:
- Annual generation without cleaning: 1,500 MWh
- Average soiling loss: 1% per month (12% annually if never cleaned)
- Electricity value (retail avoided or FiT): $0.10/kWh
- Annual energy loss from soiling: 1,500 × 0.12 = 180 MWh = $18,000
- Cleaning cost (2 cleans/year, commercial contractor): $2,000/clean = $4,000/year
- Net benefit of 2 cleans/year: $18,000 – $4,000 = $14,000/year
Compare to 4 cleans/year: soiling loss drops to ~6%, recovering 90 MWh more ($9,000 extra revenue), but costs an additional $4,000. Net extra benefit: $5,000 — still positive. Now compare 6 cleans/year: soiling loss drops to ~3%, recovering 45 MWh more ($4,500 extra), at $4,000 additional cleaning cost. Marginal net benefit: $500 — barely worth it. The optimal frequency in this case is approximately 4 cleans/year.
Monitoring-based adaptive cleaning
Rather than scheduling cleaning by calendar, leading O&M operators use monitoring data to trigger cleaning when soiling reaches a threshold:
- Set a baseline PR (Performance Ratio) for the plant when clean — typically measured immediately after commissioning or after a confirmed cleaning session.
- Monitor daily PR relative to the baseline. Correct for temperature and irradiance variations using a temperature-corrected PR formula.
- When corrected PR drops more than 3–5% below baseline for 3+ consecutive days (after ruling out cloud cover and equipment faults), schedule cleaning.
This approach adapts to real-world soiling patterns — including seasonal pollen peaks, dust storms, and rainfall that can clean panels naturally — rather than cleaning on a fixed schedule regardless of actual soiling.
Cleaning methods and their impact on panels
| Method | Cost per clean (per MWp) | Panel risk | Best for |
|---|---|---|---|
| Dry robotic brushing | $150–300 | Low if brush is soft | Regular dust/fine particle removal |
| Water + soft brush (manual) | $400–800 | Very low | General commercial cleaning |
| Deionised water spray (no contact) | $300–600 | Negligible | Areas with hard water (prevents mineral deposits) |
| Pressure washing (high pressure) | $300–500 | HIGH — can crack cells | Avoid — not recommended |
| Dry blowing / air | $100–200 | Very low | Loose dust only; ineffective for stuck soiling |
Never use hard tap water for cleaning without a filter. Hard water leaves calcium deposits (white film) that can increase soiling loss after cleaning. Always use deionised or filtered water, or ensure panels are cleaned in the direction of water run-off so mineral deposits don't accumulate at the panel bottom edge.
Rainfall and natural cleaning
Rainfall cleans panels — but only light to moderate rain (1–5 mm). Heavy rain with runoff can deposit additional particulates. In climates with reliable 10–20 mm monthly rainfall, natural cleaning significantly reduces the required cleaning frequency. In true desert climates, rainfall is too infrequent and insufficient to clean panels.
A practical rule of thumb: if your site receives more than 15 mm of rainfall in a month, that month can often replace a scheduled cleaning session. Adjust your cleaning schedule to skip months with significant rainfall and add extra cleans in dry seasons.