A uniform film of dust costs a typical home array 2-5% of its annual output — usually less than the cleaning costs. A bird dropping or a baked-on mud line along the bottom frame costs 20-30% of a panel. Same glass, two very different decisions. Here's how to tell them apart and run the math.
Uniform dust, pollen, and road film cost a typical residential array 2-5% of annual production across most of the United States, and up to about 7% in the driest, dustiest regions. That loss is real and it is measurable, but it is small. On a 7 kW system it works out to roughly $30 to $65 of electricity a year, which is less than anyone charges to clean a roof array. Localized deposits are a completely different problem. A bird dropping, a lichen patch, a spray of stucco overspray, or the mud band that builds along the bottom frame of a panel can take 20-30% off that module, and in a string system the bad panels drag their neighbors down with them. So the real question is never "are my panels dirty." It is "which kind of dirty." Look at the glass before you look at a price.
Instrumented soiling stations around the U.S. accumulate somewhere between 0.03% and 0.15% of loss per dry day. True desert sites run 0.2-0.5%. Rain resets the clock, but only real rain: below roughly 1 mm per hour almost nothing comes off, and it generally takes about 4-5 mm (0.2 inches) in a single event to meaningfully rinse a panel. A light sprinkle is worse than nothing, because it lifts dust into droplets that dry into hard rings. Boise averages around 12 inches of precipitation a year, but July averages about 0.2 inches and August about 0.3. The Treasure Valley's highest-production months are also its only stretch with no natural cleaning. Seventy-five straight dry days at 0.05-0.1% per day leaves an array 4-8% down by Labor Day. Add harvest dust, gravel and canal-road dust, and August wildfire ash, and the top of that range is realistic in Kuna, Star, Melba, and out past Middleton.
Cells inside a panel are wired in series, so current through the whole string is set by the worst-lit cell — the narrowest section of pipe governs the flow. A uniform film dims every cell by the same small amount and costs you that same small amount. An opaque spot is different: it shades a few cells completely, the bypass diode engages, and an entire substring drops offline, typically a third of the panel, even though 95% of the glass is clean. Portrait-mounted panels have a related failure. Water sheets down the module and stalls at the bottom frame lip, leaving a dirt band an inch or two deep across the lowest row of cells. That band crosses every cell string in the panel, so it throttles the entire module. It is the one soiling pattern that rain makes worse instead of better, and it is the one genuinely worth paying to remove.
Run the numbers before you book anything. A well-oriented 7 kW array in the Treasure Valley produces roughly 10,000-10,500 kWh a year. Value a kilowatt-hour somewhere between the export credit on your Idaho Power bill and the retail rate you avoid by consuming it yourself — call it 9 cents blended, then check your own statement, because Idaho export credits were cut sharply in 2025. That is about $950 of annual production. A 4% soiling loss is roughly $38. A 7% loss is about $65. No cleaning service in Idaho costs $38. Here is the part that costs us money to put in writing: if your panels carry an even dust film and nothing else, cleaning does not pay for itself on recovered energy. It pays when the loss runs 15% or higher, which means deposits rather than dust, or when the array is large enough that one fixed fee spreads across tens of thousands of kilowatt-hours.
Some losses are not dirt, and nobody should sell you a wash for them. Modern panels degrade about 0.5% per year by design, so a ten-year-old array running 5% under its first summer is behaving exactly as warrantied. Hard-water spotting from sprinkler overspray starts as removable mineral deposit, but left through enough Idaho summers it etches into the anti-reflective coating, and at that point the haze is inside the glass surface and permanent. Abrasive pads, scrub powders, and steel wool do that same damage in one afternoon, and that AR coating is worth 2-3% of output on its own. Wildfire smoke cuts production by dimming the sunlight itself; washing the glass does nothing about it. Delamination, browning encapsulant, snail trails, and microcracks are warranty conversations with your installer, not cleaning jobs. If someone quotes a wash for any of the above, get a second opinion.
Temperature is what people get wrong. A panel on a Meridian roof in July sits at 140-160°F. Hitting that with 55°F hose water is a hundred-degree swing across tempered glass and a real microcrack risk. Manufacturers commonly want wash water within about 20°C (36°F) of module temperature and nozzle pressure under 500 psi, with a few allowing 1,500 psi at distance. Never point a pressure washer at panels — the perimeter seal is what keeps moisture out of the laminate, and we will not do it even though pressure washing is a service we sell. Work before 9 a.m. or on an overcast day. If your array is ground-mounted or on a low single-story pitch you can reach with a water-fed pole from the ground, plain water and a soft brush at dawn is a perfectly good DIY job. A two-story steep pitch is not, and it may not be a window cleaner's job either — ask a solar O&M crew with fall protection.
Guide →
Guide →
Guide →
Guide →
Guide →
Guide →
Guide →
Guide →
Still wondering about something? Call or text us at (208) 453-6760.
Open your monitoring app and compare your best clear-day peak output this month against the same month in an earlier year, or against your first summer. More than 10% below is worth investigating. If you have microinverters or optimizers, read the per-panel data: one module sitting 20-30% under its neighbors is a deposit you can wash off. Every panel down 3% together is dust plus normal age, and not worth a service call.
Real rain does most of the work. It generally takes about 0.2 inches in a single event to meaningfully rinse a module, while drizzle under 1 mm per hour mostly redistributes dust and dries into rings. In the Treasure Valley, spring and late-fall storms handle it for free; July through September almost never do. Rain also never clears the dirt band along the bottom frame, and it will not touch droppings that have baked on through a heat wave.
No. Surfactant residue dries into a film that grabs dust faster than bare glass, and streaking changes how light enters the module. Plain water beats soapy water. Pure deionized water is better still, because it carries no dissolved minerals to leave behind and dries without spots — it is the same water we run on windows, with nothing in the tank. If you wash with Idaho tap water, expect hard-water spotting, which is the problem you were trying to solve.
Usually no. Boise-area snow slides off a tilted array within a day or two once sun hits the glass, and December and January are a small slice of annual production, so you would be risking a fall and scratched glass to recover a few dollars. Scraping is a common way AR coatings get ruined. Flat or very low-tilt arrays that hold snow for weeks are the exception, and even then use a soft-blade roof rake from the ground.
For most homes, once a year at most, timed for late June or early July so clean glass covers the high-production, no-rain stretch. Twice a year makes sense on gravel roads, beside ag fields, or under trees where bird traffic and sap are constant. Arrays inside sprinkler throw are the real exception — those need mineral deposits removed before they etch, which is a maintenance decision, not an energy one. If you want it looked at alongside window cleaning, call (208) 453-6760.
Scale and tilt change the math. One fixed cleaning fee spread over hundreds of thousands of kilowatt-hours breaks even at a far smaller percentage recovery than it does on a 7 kW roof. Commercial arrays are also frequently mounted at 5-10 degrees on flat roofs, where rain runs off without carrying much dirt with it, so soiling compounds instead of resetting. Low tilt plus large scale is where scheduled cleaning stops being cosmetic and starts being maintenance.
Free estimate in 60 seconds. Same-day and next-day slots available.
Get Free Estimate Now