A wind tower near an Oklahoma town let go at 23 minutes past 7 with no gust anywhere on the plain, and 8 months later the same model dropped its whole top section onto a German field
Nothing was blowing across northern Oklahoma that morning.
There was no storm on the radar and no advisory out, and the air over the county was doing very little at all.
Then something weighing hundreds of tons came down without warning, and people two miles off felt it through the ground and took it for an earthquake.
The rotor stops being able to shed load, and in a rising wind it accelerates past its rated speed until the structure meets forces it was never built to absorb.
One of 74 machines on a northern wind farm had folded. Nobody was hurt, and the wreckage lay across open farmland in several pieces.
What has to go wrong to fold a tower on a still morning
Machines this size manage themselves with a pitch system. Motors inside the hub twist each blade to spill lift when the rotor turns too fast, and every blade needs power to do that.
Cut the feed to those motors and the blades lock wherever they last held. The rotor stops being able to shed load, and in a rising wind it accelerates past its rated speed until the structure meets forces it was never built to absorb.
That is the classic sequence, and it explains a great many failures. It does not explain this one, because the wind was absent.
With no gust to drive an overspeed, the load came from somewhere else.
A structural failure in the shell, a fault at a flange, a problem in the foundation, or a control event that ran the rotor up while the weather record showed nothing.
Crews on site did report one operational detail worth keeping. Several turbines nearby had shut themselves down shortly before the tower came apart.
A machine bigger than a 50 story building
The farm holds 74 turbines rated at 4.8 megawatts each, giving it a combined capacity just under 352 megawatts.
They went up through 2020 and have run since the spring of 2021, barely four years into a design life measured in decades.
The rotor on this platform sweeps 489 feet across and the hub sits as high as 538 feet.
Put those together and the blade tip passes something close to 780 feet above the plain at the top of its arc.
The collapse happened east and north of a small town, scattering debris widely over neighboring ground. No injuries were reported anywhere on the site.
The local fire chief told reporters that residents had heard the noise early that morning and that the whole thing looked plainly bizarre, for the simple reason that it had not been windy.
What the two statements actually said
The manufacturer confirmed a turbine had been found collapsed on the ground and that only material damage resulted. Safety measures went in immediately, it said, and a team of its own staff and outside experts began work alongside the customer.
The operator said all employees were safe and accounted for and the site secure.
It added that the plant remained safe and operational, which is the sentence worth pausing on. The other 73 machines kept turning while the cause was unknown.
Neither company has published a root cause since. That is not unusual, and it is why every comparison drawn to older failures deserves checking.
The Welsh case does not transfer
The precedent people reach for is a collapse in south Wales in the winter of 2022.
A fault beginning inside an uninterruptible power supply cabinet serving one blade knocked out both the main and backup feed to the pitch system, leaving all three blades stuck in a fixed position.
Two details get dropped whenever that case is cited. The machine was a two and a half megawatt model roughly three hundred feet tall from a far older platform, not a relative of the Oklahoma turbine.
And the day was not calm. Two named storms were crossing Britain and the site saw 50 mph gusts, which is what the manufacturer pointed to when it blamed an overspeed fault.
So it is a genuine warning about single points of failure and a poor template for a windless February. Welsh sites have supplied the industry with other lessons in how it handles failure, and openness is rarely the strong suit.
The one comparison that does hold
In October of the same year the same model failed again, at a small three turbine site in western Germany.
The upper tower segment went to the ground carrying the nacelle, the hub and the blades with it. That project had been standing since 2023 on towers around 410 feet, and again nobody was hurt.
The manufacturer opened a second investigation and again reported material damage only, with the cause undetermined.
Two failures, one platform, two continents, and no public finding from either. If the answer turns out to be a bolted joint, a fatigue crack or a control fault, the reach extends well past a single wind tower to 73 sisters on the same ground.
Fleets have been grounded for less. A bonding flaw traced to one factory floor put dozens of blades under inspection, and it took months to surface.
Hugo is an engineer with strong technical expertise. Multilingual from an early age, his writing combines technical clarity with a strong interest in science and energy.
