On Saturday, July 13, a newly installed turbine at the Vineyard Wind farm started undergoing routine testing before it could send power to the grid. It had been operating for six hours when an alarm sounded at 4:21 p.m., signaling a “high vibration” on blade 1, and prompting a turbine shutdown.
About one hour later, a helicopter team got eyes on the cause of that vibration: one of the 350-foot blades had snapped, an incident that would send fragments of foam and fiberglass into the ocean and onto beaches for weeks to come.
The blade failure — described with these new details based on documents The Light obtained through the Freedom of Information Act — was not the first broken blade for this blade model. Nor was it the last broken blade for the Haliade-X, manufactured by GE Vernova. Several weeks earlier, at Dogger Bank Wind Farm off England in the North Sea, a blade had failed due to an “installation error.” Weeks later, another blade failed at Dogger Bank. All three breaks were unrelated, the company has said.
As of 2023, 13,000 offshore wind turbines — about 39,000 blades — were installed or operating worldwide for more than 300 projects, and that does not include land-based turbines, according to the latest report from the National Renewable Energy Laboratory. Flaws in a turbine blade can be introduced during the design, manufacturing, transit, or installation stage. They can cause minimal issues, like a reduction in performance, or catastrophic failure, like the Vineyard Wind incident.
Experts contacted by The Light said they were not aware of a global or regional database that tracks how many turbine blades have failed. That means Massachusetts residents who have asked public officials how common blade failure occurs will likely be dissatisfied with available answers. Experts say blade failure is an unusual and rare event, though literature points to reliability becoming a greater issue with blades.
DNV, an international certifier of blade design, in a 2023 paper said the scale of the problem has grown as blades have become larger, and that it “perceives blade durability as a major challenge the wind industry must address.”
The 13-megawatt Haliade-X blade is being used in only two projects — Vineyard Wind and Dogger Bank — and its future is unclear. As of this fall, the Haliade-X, once billed as the world’s most powerful turbine, is not contracted for any other offshore wind projects.
The failed blade has contributed to significant delays for the Vineyard Wind project, which executives initially planned to have finished this year and sending power to Massachusetts homes and businesses. It has been a black eye on a new industry in the U.S. that green energy proponents are relying on to shift the nation from fossil fuels to renewable energy sources amid climate change — a shift that also requires trust in the infrastructure.
The Light looked at the failed blade, how blades — among the largest composite structures in the world — are manufactured, and what can cause them to fail.
History of the blade
GE Vernova is one of three major offshore turbine manufacturers. The other two are Vestas and Siemens Gamesa. Two New England offshore wind farms are using Siemens turbines: the recently completed South Fork Wind and the under-construction Revolution Wind, both developed by Orsted. Siemens turbine components are visible from I-95 in the Port of New London, Connecticut, a staging area for both Orsted projects.
No turbine manufacturer has escaped hiccups, with blade issues in Norway, Sweden and the UK this year alone.
GE introduced the large Haliade-X to the world in early 2018. Its selling point was its significant power advantage, in part due to its size. Its 351-foot length produced 45% more energy than other market models at the time. Graphics showcased its enormity. A complete turbine, with three Haliade-X blades rotating atop a tower, stands more than 800 feet tall. A graphic shows it overshadowed by the Empire State Building, but standing above the Washington Monument and just shy of the Eiffel Tower.
Because of this power capacity, Vineyard Wind could use fewer turbines to satisfy utility contracts: 62 turbines instead of a potential 100.
GE tested the blade’s design in Massachusetts in 2019 at the state’s wind technology testing center, which sits along the Mystic River in Boston. Due to its size, engineers had to remove the tip of the blade, as the 351-foot model was too long to fit in the 295-foot testing facility. Offshore wind critics have pointed to this to claim or infer that the model was not properly tested. Experts, however, disagree.
Two years ago, GE was embroiled in a patent lawsuit with Siemens Gamesa over intellectual property regarding the Haliade-X design. A judge initially ruled that only two projects could proceed with the model — Vineyard Wind and Ocean Wind (which has since been scrapped). But last year, the blade behemoths reached an “amicable settlement,” allowing each to use the contested patents.
LM Wind, a subsidiary of GE, has manufactured the Haliade-X blades for the Vineyard Wind projects in two locations: Gaspé, Quebec, and Cherbourg, France. The Gaspé site had been supplying most of the blades for the project until the blade failure. Since then, new blades have been shipped to New Bedford from Cherbourg.
Managers at the LM Wind plant may have falsified quality testing data, favoring quantity over quality, according to an October report from Canadian outlet Radio Gaspésie. The company terminated nine managers and suspended 11 floor workers as a result of the blade failure, The Light confirmed last month. Reuters, quoting unnamed sources, reported in November that “there were corners cut” at the plant.
Though factories manufacture the major turbine components, not all the work is automated. For blades, many “touches” are involved that can introduce error and create the need for precise quality testing.
What’s in a blade?
A wind turbine blade is composed of fiberglass, balsa wood, polyethylene foam, carbon fiber, gel coating, epoxy paint, resins, and adhesive pastes — materials that are found in other major industries, like boatbuilding and aerospace.
Blades contain only a bit of steel (mostly in the form of bolts and plates at the blade root), so it’s the adhesive and resins that keep everything bonded during strong winds, rain, and up to 200 mph revolutions at the blade’s tip. GE Vernova said preliminary analysis showed the Vineyard Wind blade failure was caused by a manufacturing defect of “insufficient bonding.”
A lot of blade manufacture is accomplished manually by dozens of workers, wearing personal protective equipment, moving in and around two molds, also called “clamshells.”
There are “a lot of touches” on these handmade items, said Josh Paquette, researcher at the federal government’s Sandia National Lab in New Mexico, in an interview this fall. The lab researches reliability and holds workshops to discuss the issues facing manufacturers and operators.
Defects can occur at this hands-on manufacturing stage, such as air bubbles in the resins or bonding gaps. Paquette said bonding issues don’t usually lead to catastrophic failure.
Workers lay dry fabrics of glass composite material in the molds. More layers go near the root of the blade, which experiences the highest load, and fewer layers go at the tip. A tacky coat keeps them in place before they’re cured with heat.
The workers then lay down the “core,” which can be balsa wood or polyethylene foam (or both), with panels individually marked and fitting together like puzzle pieces. They sandwich the core with more fiberglass. Workers then lay out a mesh film, preparing the materials to evenly receive a “molasses-like” resin, which a machine will pump down the length of the massive component.
“They line the individual sheets and wood inside special molds honed and polished to produce the right aerodynamic shape,” reads GE’s website. “They cover the layers with foil, pump the air out to create a vacuum and inject a special resin to fuse them together. The vacuum helps the resin seep into the tiniest nooks and create a solid shell.”
It can take up to 2,000 labor hours (the time of all involved workers) to produce one blade, per a GE Vernova article.
Workers cover the molds with blankets and turn on the heat source. The mold acts like a sort of hot plate, reaching more than 150 degrees Fahrenheit.
After everything is set, GE describes its composite result for the Haliade-X blade as “squished phyllo dough” — an ingredient that often appears in multi-layered Greek pastries.
Workers then install the internal support structures (formally called “shear webs”) before placing glue on the edges and joining the halves. This adhesive is also called “bond paste.”
It has to be mixed correctly. And once it is, workers must apply it before its properties change. Some manufacturers use two contrasting colors for the mixed materials, so that, when they are combined, workers have a visual cue that helps check whether the paste contains the correct ratio. The “pot-life” of the mixture — the time in which it must be applied — can vary, be it less than 45 minutes or more than two hours.
After more curing, the molds are removed, revealing a nearly complete blade that needs surface finishing, paint, and some aerodynamic accessories. Depending on the size, it can weigh 55 tons, or the weight of about 10 elephants.
Why do blades fail?
Blade failure can be superficial, functional or catastrophic. Some issues can be corrected with standard maintenance, while others (like the blade break) require total replacement. Failure includes cracks, breaks and unwanted bends.
Sometimes, the issue doesn’t affect the functionality, but can decrease performance. Outside of manufacturing, transport and installation, causes of failure include turbulent winds, out-of-control rotation, and lightning strikes.
Per the U.S. Department of Energy, failures in many cases can be “traced back to the manufacturing floor.” A 2022 Danish study concluded that the strength and durability of wind turbine blades is “controlled to a large degree by the strength of adhesive joints” in the blade.
Defects can come about when not enough paste is applied, which can lead to voids or air bubbles. Even a few millimeters of deviation can compromise the blade’s longevity, according to a lead GE Vernova engineer.
Turbine size can also affect failure rate, and experts anticipate that reliability could become more of an issue as blade size grows faster than manufacturing techniques can evolve and improve.
“As offshore wind blades increase in size the primary failure mechanisms can change or may be exacerbated,” reads a 2024 paper by Paquette and several scientists at Sandia and the National Renewable Energy Laboratory. “The time to market and speed with which wind blades are changing size mean that there is no historical data to effectively analyze failure rates and failure modes.”
The issue is compounded by proprietary data: “There is little to no data sharing of the blade design, manufacturing, or likely failure modes from the original equipment manufacturers … with the owner/operators,” the paper reads.
According to the Sandia Lab’s research on reliability, blades are one of the top sources of turbine downtime, which is industry-speak for the time the system is unavailable to generate power due to a component issue that needs maintenance or repair. Problems with the gear box (which is in the nacelle, or turbine generator) are another leading cause of downtime.
Paquette stressed that only experts can diagnose the reason for an individual blade failure.
“Generally, complete blade failure is very uncommon,” he said, noting the largest source of blade failure, per the lab’s data, has been lightning strikes.
Paquette said blade manufacturers have moved toward some automation or “pre-processing,” particularly for the more critical elements of the blade, in the form of pre-made or pre-cured parts that are of higher quality and easier to inspect.
Several defects are traceable in a factory setting by using non-destructive inspection techniques, he explained. This includes conducting ultrasonic testing (much like an ultrasound by a physician), taking thousands of scans, and dispatching remotely operated robots equipped with cameras into the blades. These crawlers can reach the tight areas near the tip in which humans can’t fit.
GE this year announced its use of artificial intelligence and algorithms as another tool to “scour each blade’s interior, looking for deviations.”
There is a bond error, however, that is hard to spot; colloquially, it’s called the “kissing bond.” That’s when the adhesive is touching the composite material, but is not adhered to it, Paquette explained. It’s hard for current inspection methods to detect because there is no actual space that would be reflected in a scan.
In response to an interview request to discuss blade manufacturing and bond issues, a GE Vernova spokesperson provided previous statements on the company’s blade recovery effort. GE’s preliminary analysis determined a bonding issue caused the blade failure at Vineyard Wind, but the company has not specified further.
GE Vernova has been re-inspecting more than 100 blades in New Bedford, in Canada, and at sea. This has resulted in the shipment of some to France, and the at-sea removal of others.
Blades can be repaired — as GE says it is doing to rectify the issue. If an area is missing adhesive, it can be added. If there’s a bad bond, workers can remove the adhesive and reapply it, or reinforce the bond.
However, repair costs time and money. And each day the Vineyard Wind project is not operating delays its return on investment. DNV, the Norwegian certifier, said in its paper that the significance of blade failures is increasing due to the cost of repairs and replacement of components. The implications become greater for turbines with increased power ratings (i.e., larger turbines).
As winter settles in, construction has slowed at Vineyard Wind. Sets of blades lie in stacks in bright blue cages at the New Bedford Marine Commerce Terminal, many with their roots still wrapped in plastic. Some of them, presumably, cleared inspections. Others had to be sent away; GE has not stated if or when they’ll return.
Email Anastasia E. Lennon at alennon@newbedfordlight.org.
The test procedure for the 107-meter (351 feet) LM Wind Company blades owned by GE is flawed. The prototype test blade used to manufacture 150 blades in Canada was cut in two parts at the Massachusetts Clean Energy Center test site because it did not fit. The engineers extrapolated figures to pass the test. The blades were never tested in an ocean environment for a year. There was a rush to get renewable energy credits by the end 2024. The blades are hybrid blades made with far less carbon fiber. The blades should be replaced with shorter blades made with the more expensive carbon fiber….
Look at the number of ocean based turbines, and multiply by three foe the total number of blades. Divide by how many blades have failed over a ten year period. The number is far lower than one percent.
The real reason why K Harris lost the election it was because this country is plagued with people like this Frank Haggerty that always complaint and see only the wrong on everything, he probably voted for Trump because everything that any administration is doing they think and find them wrong. He probably in a couple of years of the new administration he will be complaining again for everything. Good answer from Brad Abernethy showing the reality of new innovations. When Henry Ford built his first car didn’t have automatic transmission nor air conditioning. Time and try outs made things better.
The answer as to why these blades are falling apart is quite simple: corporate greed! They skimp on quality control to produce as many blades as possible in the lease amount of time. Time is money, and the more one can produce in the least amount of time equals greater profit. This is the story of corporate America.
And the story of the individual American.
They Joy of Capitalism.
I believe they g to nought the inferior blades we oils last long enough too squeeze all the subsidies money out of the unsuspecting consumer. That would be you and me!
Time to stop the “Fake it and just Take It in Massachusetts “failed wind energy program permanently.
I don’t see it ?
This is a very well written investigation on the subject matter. Thank you.
Thank you for the thorough article. It really helped me to see the challenges inherent in blade manufacturing. Some of these types of issues have appeared in the manufacturing of the new Dreamliner aircraft with the use of carbon resins. It’ll take time to resolve, but this industry is in its infancy.
Great article with credible references. Thank you.
Yes, a good article helping in lay understanding
Thank you for a very informative article.
Thank you for a very informative article.