Investigators in Yeongdeok found a 21 year old wind turbine folded across a public road after a 16 mph breeze, and they say the blade that started the collapse should never have torn

The road runs along a ridge on the east coast, past a closed park, carrying almost nobody.

Across it now lies a steel tube 262 feet long, folded partway up like a drinking straw.

The nacelle sits in the grass at the end of it.

They are a shell of composite skin over internal webs, laid up in layers and bonded, light enough to keep tip stresses manageable across two decades of rotation.

Composite fragments are scattered across the asphalt for a hundred yards.

A driver had pulled onto the shoulder to make a phone call and filmed the whole thing.

The wind that afternoon was a light breeze.

Why one blade takes down the whole machine

A three bladed rotor is a balanced object spinning at the top of a very slender column.

Balance is the entire design assumption. Take one blade away and the remaining two throw a violent asymmetric force outward on every rotation.

At operating speed that force arrives as a lateral impulse hundreds of feet up, on a structure engineered to resist steady push from one direction rather than a hammer swinging in a circle.

The tower cannot absorb it. Nothing in that column is built to.

So the sequence is short and it is always the same. Blade separates, rotor unbalances, the top section is struck before it can flex enough to recover.

Which is why the interesting question is never the tower. It is what tore first.

What was standing on that ridge

The wind turbine was a 1.65 megawatt unit, one of 24 on a coastal ridge in North Gyeongsang Province.

Its tower stood 262 feet. Each blade ran 134 feet.

Those blades are not solid. They are a shell of composite skin over internal webs, laid up in layers and bonded, light enough to keep tip stresses manageable across two decades of rotation.

The strength lives in the bond lines, not the material.

A laminate can carry a flaw between layers, or a weak seam where skin meets spar, for years without a hint of it appearing on the outside.

Composites also accumulate fatigue differently from steel. Repeated flex opens microscopic gaps, and a gap grows slowly until the day it runs the full width and the part lets go.

What the record shows about that afternoon

It happened at about 4:40 in the afternoon on the second of February in 2026, at Changpo wind farm in Yeongdeok County.

Unit 21 had been turning since 2005.

Reported wind speeds differ between accounts, from about 14 miles an hour to roughly 28. The machine runs between 7 and 45.

Nothing about the weather was remarkable, which is exactly what makes the failure hard to read.

County officials reviewed video with the construction company and confirmed the order of events, that a blade rotating at high speed tore off, the rotor lost balance, and the tower was struck.

Then came the line worth keeping. Given the blade’s material, an official said, it should not have torn.

What the joint investigation has and has not settled

The county pulled in the national energy ministry and outside experts, and a review by the electrical safety body reported at the end of February.

It established that the wind was ordinary, around 6.4 meters a second against a working range of 3 to 20, and that blade damage came first.

It also flagged a suspected fault in an alarm sensor, and marked that as unconfirmed.

What it did not produce was a cause. Forty five days on there was still no definitive finding.

Age is not the obvious answer either. Korea has lost turbines four years into service, another at two, and one in a typhoon at ten, and this one at twenty one.

Blades come off elsewhere without a clean answer too, including a farm that lost three blades nobody has explained, and the county is inspecting the remaining 23 with the ministry.

What blade age means for a fleet built in the mid 2000s

Twenty years is the design life these machines were sold on, and Unit 21 was past it.

It is not alone. The build out that put turbines across South Korea, Japan and much of northern Europe in the early 2000s is now reaching the same milestone, still earning revenue and still passing periodic checks.

Unit 21 had passed its most recent safety inspection despite being past its design life, and that is the part that should worry people, because the checks were written for failure modes engineers expected.

Ultrasonic and thermographic scanning can see inside a laminate. Both cost money and time, and neither is required anywhere as a baseline for machines past twenty years.

The commercial answer is usually to replace the whole machine rather than interrogate the old one, which is why repowering keeps winning on paper.

What happened on that ridge is a physical argument for looking inside the material instead.

Nobody was hurt on that road, and that part was purely luck.