Can Hail Damage Solar Panels?

What a hailstone actually delivers

A hailstone is a lump of ice arriving at terminal velocity, and the damage it can do is governed by kinetic energy: E = ½mv². Both terms work against the glass as stones grow. Mass rises with the cube of diameter, and larger stones also fall faster, so energy climbs steeply with size.

The arithmetic is worth doing once, because it explains why hail reports are dominated by a narrow band of stone sizes.

Doubling the diameter from 25 mm to 50 mm multiplies the energy by roughly fifteen. That is why a storm producing golf-ball hail is a different event from one producing pea hail, and why "it hailed" on its own says very little about the risk.

Drag sideways to see the rest

Enlarge

Why the glass usually wins

The front sheet of a module is not window glass. It is thermally tempered: heated and then cooled rapidly so the outer surfaces end up in compression and the core in tension. A crack has to overcome that built-in compression before it can start, which makes tempered glass several times stronger in bending than annealed glass of the same thickness.

It also fails differently. When tempered glass does break, the stored stress releases at once and the pane disintegrates into small blunt fragments rather than long shards — which is why a hail-broken module looks like a crazed sheet held together by the laminate rather than a hole.

Drag sideways to see the rest

Enlarge

Underneath sits the part that cannot be toughened. A silicon cell is brittle, wafer-thin — a slice cut from a grown or cast crystal — and bonded into a polymer encapsulant. The encapsulant cushions it, but a sharp enough impulse through intact glass can still crack the wafer.

What the certification test actually proves

Modules qualified to the IEC 61215 series undergo a hail impact test — IEC 61215-2, test MQT 17. In the standard class, a 25 mm ice ball is fired at about 23 m/s at eleven impact points across the module: the corners, points along the edges, over the gaps between cells, the centre, and above the junction box. Those locations are chosen because they are where the laminate is least forgiving. Larger ice-ball diameters exist as optional, harsher classes for regions where big hail is routine.

Drag sideways to see the rest

Enlarge

Two things follow, and both get forgotten. Passing the test proves the module survives that impact at those points — it is a qualification threshold, not a guarantee against any storm. And the test uses a standard ice ball on a bench: a real stone may be irregular, denser, driven by wind at an angle, and landing on glass already weathered for a decade.

Those harsher classes are specified before purchase rather than discovered afterwards.

Tilt helps, quietly

Only the component of the impact perpendicular to the glass does the work of breaking it. A stone falling vertically onto a panel tilted at 30° strikes at 30° from the surface normal, and the perpendicular component of its velocity is reduced by the cosine of that angle — with energy scaling as the square of velocity, the normal-direction energy falls by about a quarter.

Drag sideways to see the rest

Enlarge

Steeper arrays therefore fare better in hail, all else equal — though tilt is chosen mainly for yield, which is its own trade-off and the subject of choosing tilt angle and direction.

The damage you cannot see

Here is the part that costs money. A module can pass a visual inspection and still have lost cells.

A microcrack is a fracture in the silicon wafer beneath unbroken glass. The cell still conducts, so nothing looks wrong, but the crack interrupts the fine metal fingers that carry current across the cell face. Sections of the cell become electrically isolated, and because cells are wired in series, the weakened cell limits the string current — the same mismatch mechanism described in how partial shade affects solar panels, arriving from a different cause.

Drag sideways to see the rest

Enlarge

Cracks also tend to progress. Thermal cycling flexes the laminate daily, and a crack that isolated a sliver of cell after the storm can isolate a larger area two winters later. A module that lost 2% of its output in March may have lost considerably more by its third summer.

What to check after a storm

1. Take the production data first. Compare daily output against a comparable period before the storm, at similar irradiance. A step change on the day of the storm is the clearest evidence there is.
2. Compare strings. If two strings share an orientation and only one has dropped, the weather affected them equally and something else did not. Where the system has module-level electronics, the per-module data narrows it further, to individual panels.
3. Look at the glass, carefully. Crazing, chips at the edges, and a spidery pattern radiating from a point are all worth photographing.
4. Check the frames and mounts. Hail arrives with wind; bent frames and loosened clamps are ordinary findings and matter for the next storm.
5. Ask for electroluminescence imaging if output has dropped without visible damage. Passing a current through the module in darkness makes it emit infrared light, and cracked or isolated regions show up black. It is the only practical way to see a microcrack.

If output has fallen and none of this explains it, the systematic hunt in why a system produces less power than expected covers the non-weather causes — and a storm is a good moment to rule out the mundane ones, since soiling and cleaning affect output in the same direction.

When hail is genuinely a design problem

In most climates, hail belongs on the list of things that occasionally happen rather than things that drive design. Where large hail is a recurring seasonal event, it becomes a specification question — impact class, glass thickness, mounting tilt, and whether the array can be stowed at a steeper angle — and all of those are decided before anything is bolted down.

What does not work is improvisation during the storm warning. The useful preparation is a production baseline you can compare against afterwards, which costs nothing and is the difference between knowing and guessing.