Solar photovoltaic "soiling" is the buildup of dust, pollen, soot, and other airborne particles on the glass of solar modules. That layer absorbs, reflects, and scatters sunlight before it reaches the silicon cells, lowering output. Researchers measure it with the soiling ratio — the short-circuit current of a soiled panel divided by that of an identical clean one — typically sampled between noon and 1:00 PM at irradiance at or above 500 W/m² to remove shading and angle errors.
The numbers below are drawn from national-laboratory studies, a large University of California dataset, and manufacturer warranty documentation. We have kept the framing honest in both directions: in much of California, rain keeps panels clean enough that routine washing barely pays for itself, while in specific conditions — wildfire ash, heavy pollen, flat-mounted arrays, and biological growth — cleaning recovers meaningful energy and protects the asset.
01 · The mechanics
The physics of solar soiling
Dry deposition is driven mainly by airborne particulate matter. Across regional studies, particulate concentration correlates with soiling losses more strongly than wind direction or other weather variables. Satellite- and ground-measured particulate data track localized soiling closely: the correlation for coarse PM10 is strongest 30–50 km from a system (adjusted R² up to 69%), while fine PM2.5 stays stable across 30–250 km (adjusted R² of 63–70%) using declustered distance estimation. Source: regional environmental soiling analyses.
Predictive frameworks such as the Humboldt State University soiling model assign static settling velocities to each particle size: 0.0009 m/s for PM2.5 and 0.004 m/s for the heavier PM10 fraction. Modern arrays add a twist — running at system voltages up to 1500 V increases electrostatic attraction, drawing fine particles onto the glass even after sunset while the system is idle.
Under the microscope, the deposited layer mirrors the local soil. Mean particle size runs 8–21 µm; roughly 90% of particles are dust under 50 µm and about 10% are larger sand grains over 50 µm.
| Parameter | Modeled value | Operational impact |
|---|---|---|
| PM2.5 settling velocity | 0.0009 m/s | Slow, steady accumulation of fine atmospheric particulates |
| PM10 settling velocity | 0.004 m/s | Rapid coarse-dust buildup, especially near agriculture |
| PM10 correlation distance | 30–50 km (R² ≤ 69%) | Local coarse-dust sources dominate regional soiling |
| PM2.5 correlation distance | 30–250 km (R² ≤ 70%) | Fine particulates act as a wide-area soiling driver |
| High-voltage attraction | Up to 1500 V | Accelerates particle adhesion, active day and night |
| Particle size mix | 90% dust (<50 µm), 10% sand (>50 µm) | Defines soiling texture and cleaning difficulty |
02 · The payoff
Energy loss & cleaning recovery
Does cleaning increase production? Yes — but the size of the gain depends on geography, tilt angle, and contaminant type. In the United States, annual losses from untreated soiling typically range from 0–7%. Heavily soiled systems can drop 20–25%, and extreme desert installations can lose 30–50% of capacity factor.
Field measurements reported by California solar maintenance providers show post-cleaning gains of up to about 21% on residential arrays and up to 60% on flat commercial installations. The gap is driven by tilt: panels mounted at 10° or steeper get gravity-assisted rinsing in rain, while flat or low-tilt arrays (under 5°) trap standing water and debris and can see a 100% jump after a single wash.
The nuance that hides the loss
Two system effects can mask soiling. First, in systems with a high DC/AC ratio, inverter clipping caps peak output during the sunniest hours, so a moderately soiled panel may still hit the inverter's AC limit and show little net loss. Second, daily losses must be weighted by insolation: a 5% loss in long, sunny summer days costs far more kWh than the same 5% in short winter days.
Passive factors cut the other way. In Alamogordo, New Mexico, a system endured 22 dust-storm events yet lost only 2–3% — prevailing south-to-southwest winds hit the tilted modules head-on and resuspended particles, and light rain of just 2.2 mm/hr restored performance. By contrast, a 1.1 MW Louisiana plant in a dry spell reached a 7.5% summer soiling loss.
Biological and organic contaminants are the real problem because rain does not remove them. An NREL study of utility-scale plants in North Carolina found frequent spring rain failed to wash off sticky tree pollen (cypress, juniper, pine, oak); output fell up to 15% at peak pollen season, for roughly 10% annualized loss if left uncleaned. Mechanical wet-brush cleaning restored 5–11% immediately. Mold and fungal films from humidity and pooling cause persistent 2–10% losses that rain cannot fix.
On the prevention side, NREL and Pellucere Technologies tested a silica-based anti-reflective, anti-soiling coating (MoreSun + Talus). In chamber testing mimicking the humid desert Southwest (55% relative humidity), uncoated modules degraded up to 20% while coated modules held soiling loss near 1% per cycle.
| Contaminant / environment | Loss (uncleaned) | Recovery after cleaning | Mechanism |
|---|---|---|---|
| Standard dust (arid) | 7–25% | 5–21% (residential) | Removes dry mineral particulates |
| Flat-mount modules (0°) | 25–50% | 36–100% | Clears pooled mud, dust sheets, standing water |
| Tree pollen (spring) | up to 15% | 5–11% | Breaks the sticky pollen bond rain leaves behind |
| Fungal / biological | 2–10% | Full baseline restore | Removes films that resist rain rinsing |
| Wildfire ash & smoke | up to 30% | Immediate 15–30% | Prevents ash from chemically etching the glass |
03 · The honest part
The California data
The most rigorous California-specific study comes from UC San Diego, which analyzed temperature-corrected efficiency for 186 residential and commercial installations from the San Francisco Bay Area to the U.S.–Mexico border, using California Solar Initiative data. The window was the historic 145-day summer drought of 2010, comparing output during the dry spell against output right after rain events over 0.1 inches (2.54 mm), which were assumed to fully clean the panels.
Across all 186 sites the average degradation was just 0.051% per day; only 26% of sites exceeded 0.1%/day. Over the full 145-day drought, total loss averaged only 7.4%. For a typical 5 kW residential rooftop, washing halfway through the drought recovered about $20 of electricity before the winter rains arrived — leading the researchers to conclude that for most standard California residential systems, routine professional cleaning does not pay for itself.
The counterpoint: regional averages hide local risk
An earlier PowerLight Corporation (later SunPower) study of 46 Southwest installations found soiling building at a much higher 0.2% per day. The difference is the dataset: PowerLight focused on large, low-tilt commercial systems in active agricultural and industrial corridors. The lesson — a regional average can completely mask severe localized soiling near highways, farming, or construction.
Year-to-year variability matters too. NREL modeled 16 Southwest installations over 24 years, combining soiling rates with historical rainfall and irradiance. Adding interannual variability and rate uncertainty widened the 90% confidence interval (P5–P95) for annual energy estimates from a narrow −0.7% out to 10.1% — driven largely by whether a given year brings a long summer drought or early autumn rain.
The rainfall threshold that counts as a "cleaning event" is critical. NREL's PV Fleets project found a daily rainfall total of 2.5 mm (about 0.1 in) restores the soiling ratio to 99.5%. But the standard pvlib HSU model uses a heavier 6 mm / 24 hr cleaning threshold — because light rain washes loose dust but often leaves sticky residue that needs manual cleaning. The Kimber soiling model simply resets accumulated soiling to zero once rain crosses its threshold.
04 · The exceptions
When cleaning is worth it
The UC San Diego finding applies to the average rain-cleaned residential system. Cleaning becomes clearly worthwhile — sometimes mandatory — in these conditions, several of which are common in the Sacramento and Placer County foothills:
- Wildfire ash. California's fire season (roughly July–October) can cut efficiency up to 30% instantly, and ash is alkaline and corrosive.
- Heavy tree pollen. Sticky spring pollen resists rain and can drive 10% annualized loss until brushed off.
- Flat or low-tilt arrays. Panels under 5° do not self-clean and can double their output after a wash.
- Biological growth. Mold, fungus, and lichen films cause persistent 2–10% losses rain cannot touch.
- Proximity to dust sources. Agriculture, construction, and highways create localized soiling far above regional averages.
⚠ Wildfire ash is a 48–72 hour clock
If wildfire ash is left on panels, its corrosive alkaline compounds react with morning dew and humidity to chemically etch the glass and permanently ruin the anti-reflective coating. Post-wildfire guidelines call for thorough cleaning within 48 to 72 hours of ash exposure. Emergency cleaning runs $200–$400 — far less than replacing etched modules.
05 · Schedule & cost
Cleaning frequency by California region
Annual solar maintenance in California typically runs $300–$850, with professional cleaning visits at $150–$300 each. Optimal frequency is highly regional:
| Region & environment | Suggested frequency | Cost / visit | Net outcome |
|---|---|---|---|
| Central Valley — agricultural dust, harvest soot, heat | 3×/year | $150–$250 | Highly profitable: high soiling justifies frequent cleaning |
| Inland Empire — desert wind, blowing sand | 2–4×/year | $150–$300 | Cost-neutral to positive; protects against hotspots |
| Los Angeles Basin — smog, carbon soot, industry | 2×/year | $150–$300 | Bi-annual, end of winter and summer |
| Bay Area & San Diego — marine layer, salt spray | 1–2×/year | $150–$200 | Rain keeps panels mostly clean; clean for bird droppings |
| Any region — post-wildfire | Emergency (48–72 hr) | $200–$400 | Mandatory: prevents permanent glass etching |
Many homeowners use a hybrid approach: low-cost DIY upkeep (monthly visual checks for debris, occasional low-pressure ground rinsing) paired with an annual professional inspection at $100–$200 to test electrical connections, verify inverter software, check grounding, and document maintenance for warranty purposes.
06 · Do no harm
Mechanical risks & safe cleaning
Improper cleaning can permanently destroy a panel and void its warranty. The worst risk is micro-cracks in the silicon wafers. Modules are a laminated sandwich: the tempered glass handles uniform loads like snow, but not concentrated point loads. Walking, kneeling, or leaning a ladder on a module bends and snaps the brittle wafers underneath — cracks invisible to the eye.
Over months of thermal cycling, moisture seeps into those cracks and corrodes the silver grid lines, creating visible dark "snail trails." Cracked cells also bottleneck current, generating hotspots that can exceed 150°F (65°C), melt backsheets, crack the glass, and create a fire hazard. If you file a warranty claim, manufacturers run an electroluminescence scan — and a micro-crack pattern shaped like a footprint gets the claim denied.
Manufacturer-approved cleaning rules
- Power down first. Shut the system off before applying water to prevent ground faults and high-voltage hazards.
- Avoid thermal shock. Sun-heated panels reach 150°F; cold water on hot glass causes instant cracking. Clean in cool morning, evening, or overcast hours.
- Low pressure only. Pressure washers force water past gaskets and seals, causing intrusion and corrosion. Use a gentle hose spray.
- No harsh chemicals. Bleach, ammonia, abrasives, and laundry detergent corrode frames, degrade coatings, or leave sticky residue that attracts more dust.
- Pure water. Use deionized or RO water at or below 10 ppm TDS to avoid hard-water scale and leave a streak-free finish.
How we approach it
Cable cleans solar panels from the ground with water-fed poles and deionized water — low pressure, no walking on modules, cool hours only. It is the method that protects both the glass and the manufacturer's warranty.
07 · The fine print
Warranty & California regulations
Major manufacturers tie warranty coverage to proper maintenance, and several reimburse certified cleaning labor:
| Manufacturer | Product warranty | Yr-25 output | Labor / cleaning provisions |
|---|---|---|---|
| Q CELLS (Q.PEAK / Q.TRON) | 25 years | 86% | Up to $250 labor reimbursement via Q.PARTNER; excludes biological infestation and acid rain |
| REC (Alpha Pure) | 25 years | 92% | Full 25-year labor via ProTrust when installed and maintained by a certified pro |
| SunPower | 25 years | 92% | Recommends low-pressure hose; permits long-handled soft brush at a distance for multi-story arrays |
| Tesla (standard panels) | 10–25 years | 80–85% | Rinsing restricted to glass face; no spraying under panels or onto wiring |
California water & runoff rules
Although the emergency wasteful-water regulation expired December 21, 2023, permanent conservation mandates remain under Title 23 CCR §997.2–997.4. Washing any outdoor surface with a hose that lacks an automatic shut-off nozzle is prohibited; violations carry civil penalties up to $500 per day under Water Code §377. Commercial cleaners must also keep runoff out of storm drains under local MS4 permits.
In drought-restricted zones, operators can stay compliant by using recycled water (Title 22 standards) or low-volume deionized water systems — water-fed poles use minimal volume and produce essentially no runoff, sidestepping drought triggers entirely.
Finally, utilities like PG&E and Southern California Edison may require documented maintenance to keep solar rebate eligibility. Under the CPUC's Net Billing Tariff (NEM 3.0), reduced export rates make peak-hour efficiency more valuable than ever — so soiling during high-value afternoon export windows directly slows the system's financial payback.
08 · Common questions