A cranberry bog in early November, flooded shallow and red, with the vines under the water.
Lying across it is something long and white that does not belong in a field.
It is the better part of a hundred feet, split open at one end, mud pushed up along one side.
A few hundred feet away a tower stands with two blades on it instead of three.
Nobody saw it come down.
Several people heard it.
What holds a blade on and what lets go
A turbine blade is not bolted to a hub at its edges. It is anchored through a ring of studs set into the thick root end, which is the only place the composite is strong enough to take that load.
That ring bolts into a pitch bearing, a large slewing ring that lets the blade rotate about its own axis to catch more or less wind.
Everything the blade does passes through that one circle of steel. Weight, thrust, bending, and the twisting load of pitching in gusts.
It is also the most heavily instrumented joint on the machine, because the control system must know the pitch angle of each blade at all times.
So when a blade leaves, the imbalance is immediate and enormous, and the controller senses it within a fraction of a rotation.
The machine stopping itself is the design working, not a lucky outcome.
Four machines, one of them now short
The site sits on Head of the Bay Road in south Plymouth, near the Bourne town line, and it carries four turbines rather than a single one.
The machines are about 300 feet tall, rated at 2 megawatts each, and they came online roughly a decade ago.
The blade that came off belongs to one of them. The other three were stopped as a precaution and the whole project was taken out of service.
That decision is worth pausing on, because nothing was known to be wrong with the other three.
Machines installed together tend to share a build date, a component batch and a service history, so an unexplained failure in one is a question about all of them.
Shutting a whole site is cheap caution until a cause exists.
The afternoon, and who reported what
The call came in at about 1:52 in the afternoon on a Friday, from a neighbor who looked up and counted two blades where there had always been three.
Firefighters found the blade several hundred feet from the base of the tower, out in open bog, and reported no injuries and no fire on the ground.
Two residents described the sound rather than the sight. One heard a boom like an explosion, another felt a rumble inside the house and compared it to a small earthquake.
Those are useful and they are not measurements.
The operator said it had shut the project down and begun a root cause analysis with the company that built the machines.
State environmental regulators are monitoring the removal rather than investigating the failure.
The mark on the tower and what it is worth
One detail from the scene has carried most of the speculation since, and it deserves handling carefully.
A television crew reported that the nacelle at the top of the tower appeared charred. That is the housing behind the rotor, not the hub the blades bolt into.
Other outlets covering the same scene the same day reported no fire, no scorching and no burn marks, and the fire department account mentions none.
So what exists is one visual impression of one component, from the ground, at distance, on a gray afternoon.
Even taken at face value it settles nothing about sequence. Heat at the top of a tower can follow a violent mechanical separation as easily as it can precede one.
Blades also come off for reasons with no heat in them at all, as a parked machine off Yorkshire showed when it lost 351 feet of blade in a storm.
The charred appearance was reported by a Boston station and has not been confirmed as a finding.
A scorch seen from a road is an observation and not a cause.
Getting it out of the bog
The wreck is now an environmental job before it is an engineering one, because the ground it is lying in is a working agricultural wetland.
A licensed site professional is drawing up a removal plan, and the bog surface will not carry ordinary heavy equipment.
Then comes the part with no good answer. Blade shells of this generation are glass fiber in cured thermoset resin, which does not melt back down and resists conventional recycling.
Cutting and burying has been the default, and the alternatives are only now reaching real volume, including an Iowa plant that shreds blades without heat or solvents.
The site details, the ownership and the removal process are set out by the local paper.
The machine behaved correctly on the way down, and the question left standing is why the blade did what it did.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
