For most homes, no — routine solar panel cleaning does not pay for itself. The most complete study of the question, a 2014 Arizona State University master’s thesis summarised in the National Renewable Energy Laboratory’s photovoltaic operations-and-maintenance best-practice report (NREL/TP-7A40-73822), worked from measured soiling rates and measured cleaning costs and “concluded that cleaning is not cost-effective for residential, commercial, nor utility-scale plants.” In that study’s own simulation of a system in Mesa, Arizona, one cleaning a year cut annual soiling loss from 1.91% to 1.52% — it bought back 0.39 percentage points of a year’s output, roughly $4 of electricity for every $1,000 your array offsets. Cleaning is still worth doing in specific circumstances: bird droppings, construction dust, agricultural dust, pollen and industrial soot. Those are the exceptions, and the rest of this page is about telling them apart from the sales pitch.
Does cleaning solar panels actually pay for itself?
Usually not, and the finding is older and better evidenced than the industry that sells cleaning would suggest. The federally funded best-practice report for photovoltaic operations and maintenance — report number NREL/TP-7A40-73822, published December 2018 by the National Renewable Energy Laboratory with Sandia National Laboratories and the SunSpec Alliance — summarises the work this way:
“Naeem (2014) did a comprehensive study based on empirical data for both soiling rates and cleaning costs and concluded that cleaning is not cost-effective for residential, commercial, nor utility-scale plants. He found that cleaning costs varied from $0.25/m² for large systems to $1/m² for single residential systems, and water consumption was around 1 liter/m² of system area.”
Two things about that sentence matter more than the conclusion. First, the finding is not the laboratory’s own — it belongs to Mohammad Hussain Naeem’s 2014 master’s thesis at Arizona State University, and the report is relaying it. Second, residential systems are the worst case in his cost data: a single home array costs about four times as much per square metre to clean as a utility-scale plant, because the crew has to travel to your house either way.
The same report gives the benefit side, from the same study’s hourly simulation. This is the table the cleaning quotes never show you:
| Cleanings per year | Annual soiling loss remaining | Reduction in the loss | Output recovered vs. never cleaning |
|---|---|---|---|
| None | 1.91% | — | — |
| One | 1.52% | −20% | 0.39 percentage points |
| Two | 1.32% | −31% | 0.59 percentage points |
| Three | 1.20% | −37% | 0.71 percentage points |
Those are simulated results for one system in Mesa, Arizona — a dry, dusty, high-sun location where cleaning should look better than it does in most of the country. Read the last column slowly. The whole prize for one cleaning a year is 0.39% of a year’s production. The prize for tripling the effort is 0.71%.
So here is the arithmetic to run with your own numbers. Take what your array offset last year in dollars. One annual cleaning, on these figures, buys back about $4 for every $1,000 of that. Then get a written price for the cleaning and compare the two.
For scale only: the U.S. Energy Information Administration reports that in 2022, “the average annual amount of electricity sold to (purchased by) a U.S. residential electric-utility customer was 10,791 kilowatthours,” and its Electric Power Monthly puts the average residential price across 2025 at 17.30 cents per kilowatt-hour. Multiplying those two figures — our arithmetic, across two EIA series from two different years — gives roughly $1,867 of household electricity a year. An array sized to offset all of it would recover on the order of $7 a year from one annual cleaning. EIA itself cautions that purchase figures understate consumption in homes with solar, because “net-metered PV systems effectively reduce electricity purchases,” so treat $1,867 as a scale marker, not as your bill.
How fast do solar panels actually get dirty?
Slowly, and the honest unit is a daily rate rather than an annual one. You will find a “solar panels lose about six percent a year to dirt” claim on most pages about this topic. It comes from a single sentence in the same best-practice report — and that sentence is uncited, and it sits underneath a heading clause reading “However, where special conditions (listed below) occur.” The special conditions are agricultural dust, construction dust, pollen, bird populations, diesel soot and industrial sources. It is not a figure for an ordinary suburban roof, and this page does not use it.
The report explains the problem with annual figures in its own words: annual values “are confounded by the rain cycle, and it is more helpful to look at how soil accumulates daily, in between heavy rains.” So here are the daily accumulation rates it reports, which are fleet and field measurements rather than a rule of thumb:
| Setting | Output lost per day to soiling | Where the figure comes from |
|---|---|---|
| Desert with no farming, construction or industry nearby | 0.01% / day | Reported in the literature survey |
| General case — the number to plan on | about 0.05% / day | “Studies report about 0.05% reduction in output per day due to soiling” |
| Fleet measurement across 186 systems | 0.051% / day | Study recounted in the 2014 thesis |
| Another reported spread | 0.04% – 0.07% / day | Study recounted in the 2014 thesis |
| Heavy agricultural activity | 0.36% / day | Reported in the literature survey |
| Large bird populations | up to 0.5% / day | Reported in the literature survey |
| Dust-storm regions (reported in India) | around 1.5% / day | Reported in the literature survey |
The endpoints of that table are a factor of 150 apart. Put the two figures the report contrasts directly against each other and the point lands harder: at 0.01%/day, a roof in clean desert air accumulates in a month roughly what a roof beside a ploughed field accumulates in a day. That is the real answer to “how often should I clean my solar panels” — it depends on what is in the air where you live, not on a calendar.
One subtlety worth knowing if you watch your monitoring app closely. The report notes that soiling losses in “the morning and evening [are] about twice that in the middle of the day because of high incident angle — the shadow cast by each dirt particle grows longer with increasing incident angle.” A dirty array looks disproportionately bad at the shoulders of the day. Judge it on daily totals in kilowatt-hours, not on a bad-looking hour after breakfast.
Compare Solar Options for Your Home
Tell us a little about your home and compare options in about two minutes.
Rain already cleans your panels, and it does it better than you think
The default maintenance strategy for photovoltaic arrays is not a strategy at all. In the report’s words: “Most rely on rain to keep the array clean; no cleaning regimen is employed.” That is not neglect — it is what the cost-benefit arithmetic recommends.
Rain is not uniformly good at it, though, and the distinction is useful. “Heavy rains result in a nearly complete cleaning effect, whereas light rains clean much less effectively and can even increase soiling if dust then sticks to sparse water droplets.” A drizzle after a dry spell can leave your glass measurably worse than it was. If your production dips right after a light shower, that is the mechanism, and the next real downpour will fix it for nothing.
The report also supplies the best argument against putting cleaning on a schedule: an instrument “may indicate that the array is dirty, which would trigger a cleaning, but there may be a heavy rain the next day that would clean the array for free.” Its recommendation for evenly dirty arrays is not an interval at all — it is that “a local, site-specific cost-benefit analysis should be performed.” The parameters it lists are worth knowing, because a good contractor will discuss them and a bad one will not: the cost of cleaning (a fixed fee to mobilise a crew, plus a per-square-metre cost), how fast soil accumulates where you live, how good your solar resource is, what your electricity is worth per kilowatt-hour, and your module efficiency — a less efficient panel means more square metres to clean for the same benefit.
When cleaning solar panels is genuinely worth it
There is a real case for cleaning, and it is not about squeezing out a percentage point. It is about damage. The report is blunt: “Soiling reduces the energy output of the PV array and can lead to localized hot-spot failures if the soiling is uneven. Efforts should be taken to reduce uneven soiling, for example, from bird droppings.”
That reframes the whole question. A thin, even film of dust is an economic nuisance you can rationally ignore. A concentrated opaque patch — droppings, a leaf mat, a smear of construction overspray — is the uneven case the report singles out, and it is a hardware risk rather than a yield rounding error. Even soiling is optional to clean. Uneven soiling is not.
Beyond that, the report names six specific sources that justify a cleaning regimen, each with its own timing:
- Agricultural dust. “Cleaning can be scheduled following plowing.” This is the setting behind that 0.36%/day figure — if you live beside worked fields, you are in a different regime from everyone else reading this.
- Construction dust. “Cleaning can be scheduled after completion of nearby construction.” One cleaning at the end of the neighbour’s build, not a subscription for the duration of it.
- Pollen. “Schedule cleaning after the end of pollen season” — once, on the way out, rather than repeatedly into a headwind.
- Bird populations. The report treats this as a design problem before a cleaning problem, and lists the fixes: reduce open gaps between panels where birds nest, fit plastic bird slides to turn flat surfaces into steep ones, seal under the array with netting down to the roof, add spikes along the top edge, and time any rooftop work to the nesting season. Prevention beats a cleaning contract here, because “birds are creatures of habit, and their behaviors can be changed over time to avoid your roof.”
- Diesel soot. Present in cities and concentrated near depots; the report says it “may require frequent cleaning.”
- Industrial and cooking exhaust. Greasy soiling that rain does not shift. The report’s own example is a filter fitted to a fryer to cut the oil in a kitchen’s exhaust air — fix the source, not the symptom.
If you genuinely do not know which of these you have, there is a diagnostic that costs less than a year of cleaning: “a sample swabbed from the PV module surface can be taken to an analytical laboratory to ascertain its origin.” Some sources can be eliminated at source. The rest are the only ones a cleaning schedule actually solves.
How to clean solar panels without damaging them
- Read the module manufacturer’s instructions first. The report’s opening rule is “follow the PV module manufacturer’s recommendations with any array cleaning,” and it applies to the detergent as much as the method. Your installer’s handover pack has the datasheet; the manufacturer’s site has it if the pack does not.
- Use plain demineralised water, and a mild detergent only if the manufacturer names one. Verbatim: “Clean PV modules with plain demineralized water and mild detergent recommended by the manufacturer.” The specification is demineralised for a reason: you are cleaning to remove a deposit, and the rinse should not dry into a new one.
- Work from the ground, with a bucket, a strip cleaner and a squeegee. The report calls this “an economical method… using overlapping vertical strokes in the same way window glass is cleaned on commercial buildings.” A strip washer and squeegee on opposite ends of one long pole is the tool, and the pole is what keeps you off the roof — the next section explains why that matters most.
- Rinse rather than scrub, and let the glass dry itself. The panel’s front surface is engineered, not just transparent: the Department of Energy notes that “anti-reflection coatings and textured surfaces help decrease reflection,” which is why abrading it costs you output permanently rather than temporarily.
- Never use a pressure washer, a brush, a solvent or an abrasive. The report’s prohibition is one sentence, it names five things, and it gives no exception and no threshold: “Do not use high-pressure water, brushes, or any types of solvents, abrasives, or harsh detergents.” That rules out most of what a general-purpose exterior-cleaning crew arrives with.
The report’s framing of the whole exercise is worth keeping in mind: “Care must be taken with array cleaning to avoid damaging the components.” The downside of a bad cleaning is not a wasted afternoon — it is a scratched anti-reflective surface or a compromised seal on a component you intended to keep for decades. One thing to ignore while shopping: robotic cleaners. The report describes them, but scopes them to “large systems,” and adds that “many of these require that the design of the system accommodate the movement of the robotic cleaning system.” A residential roof was not designed to accommodate one.
The cost nobody puts in the quote: getting on the roof
The economics above are already unfavourable, and they assume nothing goes wrong. The physical risk is the part that turns a marginal decision into an easy one.
Look at what the federal workplace-safety rules require of the trades who do this professionally. Under 29 CFR §1926.501(b)(13), “each employee engaged in residential construction activities 6 feet (1.8 m) or more above lower levels shall be protected by guardrail systems, safety net system, or personal fall arrest system.” Roofing work on low-slope roofs triggers the same six-foot threshold under paragraph (b)(10), with guardrails, nets, arrest systems or a warning-line arrangement.
Those rules bind employers and their workers. They are not a law a homeowner breaks by climbing their own ladder. Take them as calibration instead: the standard of care the government requires of people who do roof work for a living is a harness. Weigh that against roughly $4 per $1,000 of annual production.
There is a second cost that outlives the afternoon. The same report warns that installing a photovoltaic system “can increase a roof’s potential for leaks and damage due to increased rooftop foot traffic and additional attachments to and through the roof membrane.” Foot traffic is named as its own mechanism. Roofing workmanship guarantees run “frequently between 5 and 10 years,” and the report notes those typically cover leakage or total failure — not wear and tear associated with the array.
Should you clear snow off solar panels?
Generally no, and here the guidance is a safety instruction rather than an economic one. The report states plainly that “snow removal is generally not recommended because it may damage the modules,” and that where it is genuinely necessary — to reduce snow weight on a roof, or to clear ice dams — “snow removal is by powerful turbo-fan, not shovel or other mechanical means.” A roof rake or a broom on cold glass is exactly what that sentence is warning against.
The production at stake is bigger than soiling and still smaller than most people assume. Snow accumulation “can reduce annual average performance by 0%–2% in southern states, 2%–4% in arid states such as Colorado, and 10%–16% in places with heavy snowfall such as Michigan, Wisconsin, and Maine.” If you are in the first two groups, the answer is simply to wait. If you are in the third, the lever is geometry rather than labour: “snow generally slides off steep arrays (e.g., a 30-degree tilt) but does not slide off low-sloped arrays” below about twenty degrees — a design decision made before installation, and a good question to put to bidders if you are still at the quoting stage.
Your production dropped. Here is what to rule out before you blame dirt
Soiling is the first explanation people reach for and rarely the right one, partly because it is the only one with a service you can buy on the spot. Work through the cheaper possibilities first.
- Shade that grew. This is the one nobody checks and the one with the largest effect. The report says a residential handover pack should carry “clear documentation that states the customer is responsible for maintaining original insolation/shade study results by completing routine bush and shrub trimming.” Your array was modelled against a shade study; trees do not respect it. Trimming is free and the recovery dwarfs anything cleaning can offer.
- The inverter. An inverter fault produces exactly the gradual, unexplained shortfall people attribute to dirt, and the report treats inverter service life as a planned event rather than a surprise. Our guide to solar inverters covers how each architecture fails and what the monitoring will show you.
- Normal ageing. Modules lose a small, specified fraction of output every year by design. Check the figure against your panel degradation rate before treating a slow decline as a problem.
- An inspection you were always going to have to do yourself. The report is explicit that for homes, “residential onsite inspections are the responsibility of the… homeowner.” Compare this month against the same month last year, not against last month.
One structural fact worth knowing before you buy a maintenance plan. When the report specifies what a residential operations-and-maintenance agreement should look like, it says performance guarantees should be written so that “module cleaning and snow removal (by turbofan) is not provided.” The professionals who write these contracts for a living deliberately exclude cleaning from the scope. That is the industry’s own revealed opinion on whether a residential array needs washing, and it is worth more than any brochure.
Explore Related Solar Topics
Keep going with these guides and tools:
Frequently Asked Questions
Do solar panels need to be cleaned?
Usually not on a schedule. The photovoltaic operations-and-maintenance best-practice report published as NREL/TP-7A40-73822 states that most owners “rely on rain to keep the array clean; no cleaning regimen is employed,” and that heavy rain produces “a nearly complete cleaning effect.” Cleaning is warranted where a specific source is at work — bird droppings, agricultural or construction dust, pollen, diesel soot or industrial exhaust — and especially where soiling is uneven, because that can cause localized hot-spot failures.
How much output do dirty solar panels lose?
About 0.05% of output per day of accumulation in the general case, per the studies summarised in NREL/TP-7A40-73822. A measurement across 186 systems found 0.051% per day; another study reported 0.04% to 0.07% per day. The extremes are far apart: 0.01% per day in a desert with no farming, construction or industry nearby, and 0.36% per day under heavy agricultural activity. Rain resets the accumulation, which is why daily rates are more meaningful than annual ones.
Is cleaning solar panels worth the money?
For a residential roof, generally no. A 2014 Arizona State University master's thesis by Mohammad Hussain Naeem, summarised in NREL/TP-7A40-73822, worked from measured soiling rates and measured cleaning costs and concluded that cleaning is not cost-effective for residential, commercial or utility-scale plants. In its simulation of a Mesa, Arizona system, one cleaning a year reduced annual soiling loss from 1.91% to 1.52% — recovering 0.39 percentage points of a year's output, or roughly $4 for every $1,000 your array offsets.
What is the safest way to clean solar panels?
From the ground, with a long pole, using plain demineralised water and only a mild detergent the module manufacturer recommends. NREL/TP-7A40-73822 describes the method as a bucket, strip cleaner and squeegee using overlapping vertical strokes, and prohibits the alternatives outright: “do not use high-pressure water, brushes, or any types of solvents, abrasives, or harsh detergents.” Workplace-safety rules require fall protection for workers six feet or more above a lower level — a fair measure of the risk of doing this from the roof.
Should I remove snow from my solar panels?
Generally no. NREL/TP-7A40-73822 states that snow removal “is generally not recommended because it may damage the modules,” and that where necessary it should be done by powerful turbo-fan rather than a shovel or other mechanical means. Snow reduces annual average performance by 0% to 2% in southern states and 2% to 4% in arid states such as Colorado, so in most of the country the right action is to wait.
Sources
- National Renewable Energy Laboratory, Sandia National Laboratories and the SunSpec Alliance — Best Practices for Operation and Maintenance of Photovoltaic and Energy Storage Systems; 3rd Edition (NREL/TP-7A40-73822, December 2018) — §5.7.1 Cleaning, §5.7.2 Snow Removal
- Naeem, Mohammad Hussain. 2014. “Soiling of Photovoltaic Modules: Modelling and Validation of Location-Specific Cleaning Frequency Optimization.” Master’s thesis, Arizona State University — the primary study behind the cost-effectiveness and daily-soiling figures on this page, as summarised in NREL/TP-7A40-73822 §5.7.1. The thesis itself is not available at a stable public URL we could verify.
- U.S. Occupational Safety and Health Administration — 29 CFR §1926.501, Duty to have fall protection (residential construction and low-slope roofing)
- U.S. Energy Information Administration — Electric Power Monthly, Table 5.3: Average Price of Electricity to Ultimate Customers by End-Use Sector
- U.S. Energy Information Administration — FAQ: How much electricity does an American home use?
- U.S. Department of Energy — Solar Photovoltaic Performance and Efficiency Basics
