The storms arrived overnight, and they arrived three times.
By dawn the fields were flattened, the grass was shredded and the ground was white with stones.
Between them stood rank after rank of solar modules, tipped up almost on edge, facing nothing in particular.
They had been standing that way since the evening.
Almost every one of them came through it.
The exceptions were on one row.
Why the angle matters more than the glass
A hailstone falling on a flat panel hits it square. All of the impact energy goes straight into the surface, and solar glass is only certified against a fairly modest strike.
Tip the same panel steeply and the geometry changes before the stone arrives.
The stone now hits at a shallow angle, only part of its energy goes into the glass, and the rest carries it sideways across the surface and off the edge.
That is the whole idea. Not a stronger panel, just a worse target.
The second effect is area. A panel standing near vertical presents a fraction of its surface to something falling out of the sky, so most of the stones that would have hit it now miss it entirely.
Steepness also sheds water and slush instead of letting it pool, which keeps the accumulated weight off the frame.
What a hail stow actually involves
Trackers exist to follow the sun, and their motors can just as easily point the rows somewhere useless.
A stow protocol is a subscription to a weather feed, a trigger radius drawn around the site, and an instruction that overrides normal tracking when a hail threat crosses that line.
Rows drive to a steep fixed angle and hold there until the cell has passed.
The cost is obvious. A farm in stow produces almost nothing, so the operator is deliberately shutting down hours before anything happens.
That is a bet against a forecast, and the false alarms are pure lost revenue.
It is also entirely a software and maintenance question. The hardware to do it is already installed on every tracking site in the country.
What the three sites recorded
Three storms crossed Fort Bend County, Texas, inside twelve hours on the 15th and 16th of March in 2024.
Hail reached three to four inches at the eastern end of that path and one and a half to three inches across the sites to the west.
All three exceeded what engineers classify as a 500 year hail event.
Two of the farms went to a 52 degree stow and reported no hail damage at all.
The third, a 270 megawatt site, went to the same angle and came through the same way, apart from roughly 40 modules that never got there because a tracker motor had failed on their row.
Those 40 stayed flat. Those 40 broke.
What one county in one storm does not settle
This is three solar farms and one weather system, not an experiment.
The engineering firm that published it inspects and certifies solar projects, so it has a professional interest in the finding, which does not make the finding wrong but is worth stating.
Stow angle is also not the only variable. Module thickness, frame design, glass thickness and the tracker’s own stiffness all sit inside the same result and cannot be separated from it here.
And a stow only helps with enough lead time. A fast forming cell that appears inside the trigger radius leaves no room to drive the rows anywhere.
The most persuasive detail is the accidental control group, because the failed motor produced exactly the comparison a designed test would have asked for.
The president of the firm put the conclusion narrowly, that operationalized stow protocols are effective at preventing module glass breakage during a severe hail event, which is a smaller claim than most of the coverage made.
What this changes for anyone insuring a field of glass
Hail is a small share of solar insurance claims and more than half of the total money paid out.
That imbalance is what drove insurers to write hail sublimits into policies, commonly in the range of 15 to 20 million dollars, which leaves an owner carrying the rest of a bad day.
A documented stow protocol, with maintained motors and a logged trigger, moves a site from that exposure toward almost none of it.
Which turns tracker motor maintenance from a housekeeping task into an insurance line item, because the row that cannot move is the row that pays.
The counterexample sat a few miles east. A neighboring plant took the largest stones in the same twelve hours and lost modules in the thousands, and the public record of what it had done that night is thin.
Glass is the part of a solar farm that was engineered hardest, and it still breaks when it is pointed wrong.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
