Pressure Washing Solar Panels Destroys the Coating That Makes Them Work

What the anti-reflective coating is, and why pressure takes it off

Solar glass is low-iron tempered glass, usually 3.2 mm, with something bare window glass does not have: a nanoporous silica layer on the sun side, roughly 100 to 130 nanometers thick. That is about one seven-hundredth the thickness of a human hair, and the thickness is not arbitrary. It is tuned to a quarter wavelength of visible light so that reflections cancel each other. It lifts transmission from about 92 percent to about 95 or 96 percent, which is worth 2 to 4 percent of everything the array will ever produce. It is also soft and porous on purpose, because density is the enemy of a low refractive index. A pressure washer does not dissolve that layer. It abrades it, and the abrasive is the dust already sitting on the panel the moment a 2,000 psi jet picks it up. The glass looks cleaner afterward. In raking light you will find shinier streaks where the coating used to be.

Put the pressure numbers next to the panel's actual ratings

A garden hose delivers 40 to 60 psi at the nozzle. An entry-level electric pressure washer runs 1,300 to 2,000 psi. A gas unit off the rental counter runs 2,800 to 4,000. That is a 30- to 70-fold jump in the one variable a module was never engineered around. Modules are qualified under IEC 61215 to a uniform front load of 2,400 pascals, or 5,400 where heavy snow is rated, which is 50 to 113 pounds per square foot pressing evenly across the whole laminate. A 3,000 psi jet is roughly 20 million pascals concentrated on a coin-sized spot, and it gets aimed straight at the frame edge seal, the junction box potting, and the MC4 connectors. Read the IP rating carefully too. IP67 means half an hour under a meter of still water. It certifies nothing about a pressurized jet, which is a separate test class. Sealed against standing water is not sealed against being shot at.

Thermal shock, microcracks, and why "mine was fine" proves nothing

On a Meridian roof in July, with the air at 98, module glass sits at 140 to 160 degrees Fahrenheit by early afternoon. Water out of a valley spigot is 55 to 70. Hitting hot glass with that is a 70- to 100-degree swing in about a second. The tempered glass usually survives it. The silicon underneath is the problem: modern cells are wafers 160 to 180 microns thick, thinner than two sheets of paper, soldered rigidly to ribbon. They crack. Those cracks are invisible from a ladder. The standard way to find them is electroluminescence imaging, which means a darkened array and a camera no homeowner owns. Production on day one reads normal. The cracks propagate through a few hundred more daily heat cycles, the array quietly gives up a percent or two a year, and hot spots follow. Nobody ever connects that to the wash eighteen months earlier, which is exactly why the anecdote is worthless.

What can be fixed, and what cannot

Be blunt about which column each of these lands in. Dirt: fixable. Hard-water spotting on the bare glass edges: usually fixable. A stripped anti-reflective coating: not fixable at any price. It is a functional optical layer applied at the factory, either sol-gel dipped and fired or vacuum deposited, and there is no field product that puts it back. Polishing compounds make it worse, because polishing is the process that removed it. Cell microcracks: not fixable. Delamination and cell corrosion after water gets past the edge seal: not fixable, and that is the path that eventually trips a ground fault and shuts the system down. The remedy for all three is a replacement module, generally $200 to $400 for the panel plus a truck roll, install, and electrical labor, so $400 to $800 landed on a residential roof. The warranty that should have covered it will not. Module product warranties exclude damage from abrasives, improper cleaning, and high-pressure water.

The method that actually works: pure water, soft bristle, hose pressure

Panels want what glass wants, only gentler. Deionized water through a water-fed pole at ordinary hose pressure, a soft bristle brush, no soap, worked from the ground wherever the array can be reached. Zero-TDS water carries nothing to leave behind, so the panel dries with no spotting and no film for the next round of dust to grab. Soap is worse on panels than on windows: the porous coating holds surfactant residue, and common exterior detergents run pH 10 to 12, where silica begins dissolving outright. That is the same chemistry that etches window glass, running against a layer a hundred nanometers thick. Time it before 10 a.m. or after sundown with the glass under about 85 degrees, so nothing shocks and nothing dries mid-pass. Keep every drop of spray off the back of the module, the junction box, and the connectors. That is the same pure-water, no-soap method Reflekt runs on windows daily, and it transfers to panels unchanged.

When paying anyone to clean your panels is not worth it

Run the arithmetic before you book. A 7 kW Treasure Valley array makes roughly 10,000 to 11,000 kWh a year. Soiling around Boise typically costs 2 to 5 percent of that, and Idaho Power's residential retail rate is around 11 cents per kWh, with exported kWh credited at less. So a full year of accumulated dust is worth $25 to $60 of electricity. A $200 cleaning does not pay for itself, and putting a pitched 25- to 30-degree array on an annual schedule, when winter and spring rain already rinse it, is mostly spending your money to feel organized. Buy the service when the loss is concentrated rather than film-thin: bird droppings, a lichen patch, wildfire ash after a smoke week, cement dust from the lot next door, or canal-water overspray from a misaimed irrigation head, because one opaque spot can pull an entire string down through its bypass diode. Otherwise every two or three years is plenty. Steep-pitch arrays belong to a solar O&M contractor, not to us.