Solar Panel Soiling Loss Calculator

How to use this calculator

Enter seven values and the calculator returns the soiling rate, baseline annual loss with only rain to clean, loss with your chosen paid-cleaning schedule on top, and the dollar ROI of the cleaning programme:

1. System size (kW) — total nameplate of your array.
2. Annual specific yield (kWh/kWp) — your local PVWatts annual production per kW. Phoenix ≈ 1,800, Boston ≈ 1,350, Seattle ≈ 1,150.
3. Annual PM2.5 (µg/m³) — EPA AirNow annual average. Phoenix 8, LA 12, Bakersfield 14, rural Northeast 6, Wildfire-prone Sierra 15.
4. Average days between rain events — NOAA Climate Data Online wet-day count divided by 365. Seattle 3, Phoenix 30, Las Vegas 60.
5. Cleanings per year (paid) — how many professional washes you contract on top of natural rain.
6. Cost per cleaning ($) — what a local solar-cleaning company charges per visit; Angi 2024 lists $0.15–$0.25/W or $150 minimum.
7. Electricity rate ($/kWh) — your current retail rate, used to value the recovered energy.

Why soiling matters more than most homeowners realise

The NREL 2024 PV System O&M Best Practices white paper estimates that soiling alone costs the US installed PV fleet about $5 billion in lost generation every year. For utility-scale sites in California’s Central Valley, soiling regularly exceeds 8% per year between winter rain seasons. For a typical 6 kW Phoenix homeowner, the loss is more modest — about 4–5% per year on average, worth $60–$120 at PG&E retail rates — but it’s also the most preventable performance loss in the system.

Snow, shading, and temperature derate are mostly fixed at install time; soiling is the only big loss you can change after the fact, by cleaning. The arithmetic is what decides whether to bother.

The Kimber–Mejia–NREL soiling model

The reference work is John Kimber’s 2007 dissertation at SunPower (extended in Mejia and Kleissl’s 2014 paper at UCSD) measuring daily soiling rates across 250 California PV plants. Three findings drive the model in this calculator:

1. Daily soiling rate scales roughly linearly with airborne PM2.5. Sites at 5 µg/m³ accumulate 0.02–0.03% per day; sites at 30 µg/m³ in the Imperial Valley accumulate 0.15–0.20% per day.
2. Rain above 1 mm acts as a near-full reset. A 5 mm event washes 80–95% of dust off; 20 mm cleans the panel essentially fully. Light drizzle (under 1 mm) can make things worse by leaving mud streaks.
3. Soiling loss within a dry cycle is triangular. Day 1 after rain you lose 0.05%; day 7 you lose 0.35%; the seven-day average is 0.20%. This is why arid desert sites with 60-day dry cycles lose so much: the average is half the peak, but the peak gets very high.

US soiling loss benchmarks by region

Compiled from NREL soiling-station data (Tucson, Albuquerque, Las Vegas, San Diego, Boulder) and EPRI’s 2024 utility-scale soiling survey:

For ground-mount arrays close to dirt roads or agricultural fields, add 1–3 percentage points. For rooftop arrays at 5+ degree tilt, the natural runoff during rain events recovers nearly all of the loss.

When paid cleaning pays back — and when it doesn’t

A simple break-even rule from the Kimber–Mejia data: pay for professional cleaning only when your soiling rate is above 0.10% per day AND your average dry cycle exceeds 14 days. Below either threshold, rain does the job for free.

For a typical Phoenix homeowner (6 kW, 1,800 kWh/kWp/yr, $0.171/kWh) the math runs roughly:

• Soiling rate ≈ 0.05% per day at PM2.5 10
• Avg dry cycle 35 days → baseline loss ≈ 0.05 × 35 / 2 ≈ 0.88% per year
• 6 × 1,800 × 0.88% ≈ 95 kWh/yr lost ≈ $16/yr
• Adding two cleanings/yr at $150 each: total resets jump from 10.4 (rain only) to 12.4, average cycle drops to 29.4 days, new loss ≈ 0.74%, saving ≈ 0.14% or 15 kWh/yr ≈ $2.50
• Net benefit = -$297. Don’t pay for cleaning.

For a desert ground-mount near a dirt road (PM2.5 25, 80-day dry cycle, 30 kW farm pump) two cleanings can recover thousands of kWh and the ROI is clearly positive. The calculator lets you check.

How to reduce soiling loss without paying for cleaning

Pick a tilt that lets rain run off

A 10° tilt holds water on the glass and leaves mud streaks; a 25° tilt drains cleanly. NREL Albuquerque test-array data shows tilts above 20° self-clean roughly twice as effectively as tilts at or below 10°. If you’re choosing a roof slope for a new install in a dry climate, prefer the steeper side.

Anti-soiling coatings

DSM Anti-Soiling Coating, Solar-Pur, and Nanotech S6 are factory-applied hydrophobic glass treatments that drop soiling rates 20–40% in NREL accelerated trials. The premium is $0.02–$0.05 per watt — worth it for utility-scale in desert climates, marginal for residential.

Snow rinses

In northern US states, the first heavy spring rain after winter typically removes the entire winter’s accumulated pollen and soot in one event. Don’t pay for an April cleaning unless you’ve gone six weeks without rain over 5 mm.

Periodic visual checks

A drone flyover or a quick rooftop check in mid-summer catches the rare bird-dropping cluster, pollen mat under a tree, or HVAC soot deposit that the rain isn’t washing. A targeted hose-down of the affected modules costs nothing and recovers the loss instantly.

What the calculator assumes

• Daily soiling rate r_d (% per day) ≈ max(0.02, 0.005 × PM2.5). Calibrated to Kimber 2007 (California, PM2.5 10–14, ~1.5%/yr) and Mejia 2014 (Imperial Valley, PM2.5 18–22, ~5%/yr).
• Rain above 1 mm is a full soiling reset (per Sandia 2018 indoor-outdoor correlation study).
• Average soiling within an L-day dry cycle is r_d × L / 2 (linear accumulation, triangular average).
• Cleanings stack with rain events to shorten the average cycle: total resets = 365 / dry_days + N cleanings.
• No correction for spring pollen blooms or wildfire soot events — both can spike loss for a 2–4 week window.
• No tilt-based runoff modifier — assumes a 15°+ tilt where rain cleans effectively.

These assumptions hold for the great majority of US residential and small-commercial installs. For utility-scale plants in the Imperial Valley or near Bakersfield, hourly soiling-station modelling (kWh Analytics, Atonometrics) gives more accurate numbers.

Common mistakes

• Paying for spring cleaning every year out of habit. Most US sites don’t need it; spring rain handles pollen. Run the numbers first.
• Trusting installer marketing claims of “self-cleaning panels.” All glass is hydrophilic — none is genuinely self-cleaning in arid climates. Mejia 2014 showed identical soiling on coated vs uncoated panels after 90 dry days.
• Cleaning with tap water in hard-water regions. Calcium streaks reduce transmission by 1–3%. Deionised water or distilled-water rinse only.
• Ignoring soiling alongside microinverter or per-string monitoring data. A 3% drop in a single string that doesn’t recover after rain is usually a soiled module, not a degraded one. Cleaning first, then RMA second.

Sources

• NREL — PV System Operations & Maintenance Best Practices (2024 update) — US residential soiling benchmarks 1–7%/yr
• Kimber 2007 — The Effect of Soiling on Large Grid-Connected Photovoltaic Systems in California and the Southwest US — daily soiling rate vs PM2.5 dataset
• Mejia & Kleissl 2013 — Soiling losses for solar photovoltaic systems in California — UCSD field measurements
• EPA AirNow — annual PM2.5 by US ZIP — input data for the calculator
• NOAA Climate Data Online — wet-day counts by station — input data for dry-cycle length
• Sandia National Labs PV Performance Modelling Collaborative — soiling indicator — IEC 61853-4 reference