Story by Eric Worrall –
![wind_turbine_bearings[1]](https://wattsupwiththat.files.wordpress.com/2014/08/wind_turbine_bearings1.jpg?quality=83)
A few years ago, I used to know a senior wind turbine engineer. One evening, over a few beers, he told me the dirty secret of his profession:
“The problem is the bearings. If we make the bearings bigger, the bearings last longer, but making the bearings larger increases friction, which kills turbine efficiency. But we can’t keep using the current bearings – replacing them is sending us broke. What we need is a quantum leap in bearing technology – bearing materials which are at least ten times tougher than current materials.”
At the time there was very little corroborating online material available to support this intriguing comment – but evidence seems to be accumulating that bearings are a serious problem for the wind industry.
Siemens citing bearing failures as part of the reason for a substantial fall in profit;
http://www.offshorewind.biz/2014/05/07/siemens-energy-division-profit-down-54-pct/
In the announcement of the opening of a new Siemens research facility;
http://www.greenoptimistic.com/2013/03/19/siemens-wind-turbine-research/
“… The Brande test center would evaluate the main parts of their wind turbines such as main bearings …”
http://www.geartechnology.com/newsletter/0112/drives.htm (an attempt to make direct drive turbines, to reduce bearing wear)
“… More accurately, it is typically the bearings within the gearbox that fail, in turn gumming up the gearbox, but that’s a story for another time. …”
http://en.wikipedia.org/wiki/Burbo_Bank_Offshore_Wind_Farm
“… During summer 2010 Siemens decided to change the blade bearings on all 25 turbines as a pre-emptive measure after corrosion was found in blade bearings found on other sites. …”
Of course, there is the occasional video of catastrophic turbine failure;
Suggestions the industry is trying to conceal the scale of the turbine fire problem;
All of which creates an interesting question – just how much of our money is the government prepared to waste, to keep their wind dream afloat? If the costs are far greater than the industry admits, how long is the wind industry going to carry that additional hidden cost, before they try to push the costs onto taxpayers, or abandon wind technology altogether?
http://www.dailymail.co.uk/news/article-2116877/Is-future-Britains-wind-rush.html
My experience with bearings comes from kilowatt sized machinery and power tools. My first reaction to this post is ” Where are the bearings manufactured?”
As an old engineer I think that the bearing problems will be found to be harmonic in nature. They need to do some serious thinking. My thoughts would be abandon the useless things and forget the engineering problems.
Power plant turbines are routinely operated for 40 and 60 years continuously.
Large foundations in a very large building that is NOT moving wildly 300 feet up in the air.
Large bearings aligned to 1 mil radially and axially with vibration aligned out down the entire shaft by laser and tight-wire tools before and after every outage. Wind turbines are slung around moving several feet sideways, and are pitched and yawed every second, with irregular thrusts hitting the bearing from 3 cantilevered 100 meter propellers every rotation. High thrust (forces) while on top in relatively undisturbed air, low and irregular gusts creating low and unpredictable thrusts and vibrations down low, and a quick pulse every time they cross the wind turbine tower.
Regular outages fully opening and inspecting and correcting problems every 3 years, and short inspections every 18 months. Wind turbines are “looked at” every 18 months, but not torn apart and refurbished. Ever.
Power plant bearings are sized and lubricated to fit around 24 and 36 inch diameter shafts rotating at 3600 rpm for 50 and 60 years. Wind turbines are built to generate profits by getting sold to investers who HAVE TO BUY them to meet “greed” (er, green) energy specifications and to receive greed energy rebates and tax credits and loan guarantees. Wind turbines are designed to get erected. Not to run.
Useless as tits on a bull…
@RACookPE1978
August 26, 2014 at 8:13 pm
From your description of the problems with wind turbines, it would seem then that there is no valid long term solution if the present design philosphy is continued.
Sounds like a good reason to cut losses.
ralfellis
August 26, 2014 at 12:10 pm
Rolls are now the second largest aero-engine manufacturer in the world.
Yes, the carbon blades were a failure, but the backup titanium blades were the company’s salvation, because they proved to be a world-beating design and are now (semi) copied be most other companies. Interestingly, these huge blades are made flat, and then heated and blown up like a balloon. Its a very secret technique.
ralfellis,
Super Plastic Forming (SPF) and Diffusion Bonding (DB) of titanium alloys is not ‘super secret’. SPF and DB of Ti-6Al-4V alloy sheet has been in existence since the the 60’s and used in fighter aircraft since the first F-15 flew.back in 1972. Certain materials can be roll processsed to have a uniformly fine and equiaxed alpha phase grain size with small amounts of transformed beta phase at the grain boundarys in sheet product form. This sheet product can then be pressure ‘blow formed’ or ‘super plastic formed’ into high temperature dies at 1650F, using inert argon as the pressure and shielding gas.
Diffusion bonding can also be achieved with titanium at 1650F, provided the surfaces of the titanium components are held in intimate contact in an inert argon gas atmosphere at high temperatures. Titanium is one of a few elements that reduce their own surface oxides at temperatures like 1650F, making metallurgical diffusion bonding achievable.
The process parameters for SPF of titanium are so similar to DB that the two processes are ‘a marriage made in argon pressurized hot presses’. SPF/DB is routinely performed in 2 sheet, 3 sheet, and 4 sheet ‘packs’.
Here’s a backgrounder on turbine blade fabrication:
http://www.sciencedirect.com/science/article/pii/S0924013699003775
Applications of superplastic forming and diffusion bonding to hollow engine blades
Mac the Knife August 26, 2014 at 9:31 pm
Super Plastic Forming (SPF) and Diffusion Bonding (DB) of titanium alloys is not ‘super secret’.
_______________________
Did I say it was?
The secret component is Roll’s ‘three sheet’ diffusion bonding and inflation process, which saves a great of weight. Unless you can show me otherwise, Rolls is the only manufacturer to use this advanced ‘three sheet’ process.
And regards the bearings, it is much easier for an aero-engine designer, who deals with predictable mono-directional airflows. A wind turbine has all kinds of forces to deal with, from turbulence, to tower vibration, to tip vortices, to blade vibration, to tower interference, to gyroscopic forces.
Designing a viable rotor bearing for such a structure must be a nightmare. And even aero-engine makers get it wrong sometimes. A few years ago there was a general recall of all main fan bearings in some CMF jet engines. But it was all done out of the public eye, and nobody noticed.
R
You seem to make this sound like rocket science. It’s not. All those are merely forces to design for. I flat out guarantee you that aero-engine design is a LOT closer to rocket science than these simple things. All that stuff you listed is fairly easy to deal with from a designer’s POV. (I have 40 years under my belt, using bearings of all kinds in all kinds of harsh industrial environments and heavy shock loading, so this thing sounds like a big baby in a cradle.) I see people crying about the awful shock form the wind coming around the tower and causing some chatter. At like 22 rpm max, and with a 34 mph wind, like, wow, that honestly can’t be diddly squat. It’s not like the wind is totally blocked, either.
Seriously, all I hear is crying, “Oh, isn’t it HARD?” I’ve designed a number of double-width 144″ industrial airfoil fans with steel rotors probably as heavy as these wind turbines, going at 600 rpm and moving MASSIVE amounts of air. This with 22 rpms sounds like a piece of cake.
They seem to have been over their heads about bearing applications and how to design for protection against the elements. If that bearing in the image above is what they have been using, then they don’t know much.
As to the loads on the bearings, the load is the load, and you design for it. It is not that difficult.
Nah, the Achilles heel is that the dominating design can only work most efficiently if the wind comes from one direction. Why not a wheel design that to states on its side that can take wind from any direction?
You do not seem to be aware that they turn with the wind…
Use fluid bearing. It’s low friction. It’s a simple problem it’s not nuclear fusion.
I was thinking of that. I was not sure they would be acceptable because of the desire/need for low friction. I used to design very large industrial fans (up to 16 feet diameter, and used “sleeve-oil bearings”. No rollers at all, no balls at all. We’d have grooves in them and often positive lube pumping, periodically. In theory, no metal ever touches metal, which one almost always tries to avoid. The oil gap is designed so that the shaft rides on a film of oil. It’s a very old technology, actually. One end should be a fixed housing bearing, and the other should float, so that the two bearings are not fighting each other.
For high loading, plain bearings have proven to be far superior than roller bearings. They have many times the bearing surface and the force is much more evenly spread out. You can’t find roller bearings in an engine any more. Really surprising to hear they are using roller bearings.
Actually, that is not true at all. Roller bearings are SPECIFICALLY designed for high loads. Ball bearings intrinsically have less load capacity, because of the higher unit stress (Look it up) on the point contact of balls versos rollers. Rollers have more surface area, so they can take more load.
As to roller bearings, your car’s wheels have them – tapered roller bearings – and they are used BECAUSE of the high impact loads that wheel bearings are subjected to. Such as potholes and bumpy roads.
The bearing problem is due to an intrinsic problem with horizontal-shaft wind turbines. Each time a fan blade passes in front of the pylon the airflow around it changes abruptly, causing a large asymmetric force on the shaft. So for every revolution the shaft (and the bearings) get three big sideways kicks. This beats the hell out of the bearings.
Note that large jet engines have restrictions on how fast the aircraft can change direction (e. g. when rotating on the take-off run) to protect the main bearings. And that doesn’t happen every few seconds.
Sorry. That sounds like normal wear and tear on equipment. Nearly every piece of equipment out in the world has some difficulty with SOMETHING going in – overhung load, vibrations, heat, high surges. It is the engineer’s job to anticipate and properly design for those environmental factors.
What we are hearing in this is that these engineers were not up to snuff. Hey, that happens. Not every designer know how to deal with every condition, and then they try something and it doesn’t work. That is the difference between good design engineers and bad ones. My professional guess is that these guys came out of college with one CAD class and a sheepskin that said they were engineers. And because they were going into the saving the world world of renewables, they were ll starry-eyed and eager. Well, that sheepskin and 20 years – under someone who really knows WTF he is doing – will make them a REAL senior engineer (which I am). In the meantime, they cost their bosses a lot of grief.
Everything I am reading about this screams that the were over their head and overlooked stuff that any experienced designer would have known to do.
Hahaha – This is the kind of hard-headed world that design engineers live in, and it is unforgiving. And when they screw it up, and the world sees the problems, someone will come and rub their nose in it. Right now, I guess I am that guy.
At 8:54 AM on 29 August, Steve Garcia had observed:
On the face of it, that doesn’t seem possible.
The information presented in this article and the comments appended thereunto appear to indicate that the phenomena under discussion are pervasive in the wind power generation industry, and show up in these Brobdingnagian bat-manglers bloody everywhere they’re boondoggle’d into installation. Is this impression correct?
If so, then – ceteris paribus – experienced, ingenious, skillful engineers have been dragged into the design process all along, and all of them have been fretting over this issue for decades, along with metallurgists and other strength-of-materials experts, uniformly staggering into frustration and failure. It’s not for want of proper brainpower that this problem plagues a whole industry, or that it hasn’t been recognized, even if recognition is being thought-blocked by the government thugs pushing it, the K Street commandos greasing their palms, and the Democrat Party Audiovisual Club making up the legacy media.
It may be – most probably it is – simply that wind power generation is intrinsically and utterly unsuited to the needs of an industrial civilization, and there are no engineering breakthroughs within the reach of established or even anticipated materials technologies that can make it either reliable or cost-efficient.
Ever take a good look at the wheel bearings on a big truck? Specifically the ones on the “steer axle”? Go take a good look someday. They operate in an oil-bath environment, and last for quite a long time that way. The driver gives a visual inspection— either by looking through the sight-glass to see the oil level, or if necessary he pulls the rubber fill-plug and looks that way– probably doesn’t add any oil or if he does he doesn’t add much because they don’t use much–and then drives without worrying about the bearings. Several hundred thousand miles can be had out of a set of front-wheel bearings operating in an oil-bath, and think of the loads imposed by front wheels on a 40-ton rig on those bearings.
Yeah, lubrication is kind of important, and those wind-turbine bearings are the “lubed for life” kind that you just install and forget. They’re not going to last long, they simply can’t.
Yeah, you are basically totally right. Except lube is not “kind of important” – it is all important. NO bearing is going to last without proper levels of lube and a proper system to freshen up the lube. These should NOT be lubed-for-life bearings. They should have a positive lube system built in for the bearings, and it should be designed so that new lube pumped in pushes old lube out, in such a way as to carry out grit. And it should do that pumping on a weekly or monthly schedule.
And look at the bearings shown above – no seals, no shields. Akkkkkkk!!!! They CANNOT possibly be using open bearings like that!!!
Xcel Energy said it cost $360,000 per turbine in 2012 to replace the gear boxes on their GE 1.5 MW turbines at their Grand Meadow wind facility. The failure rate was 10% in the first 3 years of operation. Xcel said in public filings with the Minnesota Public Utilities Commission that they anticipated this would continue since the other folks they talked with said this is “normal”. This was after the underground cables burned up in the first year of operation.
F-ing keerist, these things sound like they were designed by the Marx Brothers or Three Stooges. That is my professional opinion as a Sr Mech Design Engineer (retired). This all seems like one big screw-up. Only three years? Heck, the darned things don’t even run full time.
And WTF are they doing, taking someone else’s word for something being “normal”. Ay yi yi.
Nowhere is this actual vs. claimed cost difference more monumental than in Germany with $1.4 trillion on the line for wind power and needed transmission lines. In addition, servicing a bad turbine part offshore is probably much higher than land-based. The solar sector has similar issues with inefficiencies rated based on ideal temperatures, not rooftops and summer temps. These respective cost adjustments are not so hard for competitive cost investment selection processes but subsidies and political policy objectives undermine clear headed thinking to come up with totally different answers sometimes. See Solyndra and various solar CSP projects as examples.
Part of our end of summer drive was along US 50.
In Garden City, Kansas, south side of the road, there is an immense distribution center (?plant) for wind turbines. The inventory was impressive. Along side the road was a mile long train of rail cars each holding a turbine blade. Business looks good.
Google Earth coordinates: 37°57’27″N 100°49’32″W
Flicker image from 2012.
Yowch! I hadn’t thought about it before, but having bolt-on blades sounds like a REALLY bad idea. Dynamic balancing of big rotors is VITAL. In addition, the attachment point being bolted presents challenges for the proper deign, to make sure the entire rotor assembly works as an integrated whole. MUCH better to weld them (it is more complicated than just, though) in the factory and ship them as complete rotor assemblies. If that makes for shipping problems, helicopter them to the site. If that is too expensive, how expensive is it to replace a failed windmill?
From an engineering standpoint, this all really, REALLY does not sound like a problem that should have come up, if the engineers followed good engineering practice.
Balance problems— why didn’t the engineers think of that?
I sometimes run model power boats, and even in these incredibly small sizes properly balancing the propellers can be — nope make that is– a big issue. An out-of-balance propeller will shake the mountings loose, cause set-screws to loosen and could cause the prop itself to come off of the shaft when the retaining collar loosens and is lost. Now imagine that instead of a prop that is an inch and a quarter in diameter we’re talking a turbine that has blades of ninety feet each– it doesn’t take much of an out-of-balance condition to cause big trouble, and the bearings will feel it first. Next, the mountings will feel it, then the rest of the structure.
This is what happens when the world’s largest tunnel boring machine has bearing troubles:
http://seattletimes.com/html/localnews/2022992719_berthasealsxml.html
For the interested:
http://www.tunneltalk.com/Seattle-13Feb2014-Bearing-seal-damage-root-of-mega-TBM-troubles.php
The real issue is one of scale. Lots of devices with moving parts scattered all over the place. Even if the bearing issue is addressed in some way, its still a scale issue. With traditional power plants, there are a handful of turbines to produce power. Everything will need overhaul at some point. With wind there are thousands of turbines to produce power. Your maintenance overhead per megawatt is much higher.
So there’s a ubiquitous mechanical problem. Sounds like someone’s opportunity to make a better bearing.
@Bruce Cobb August 26, 2014 at 3:52 am
“In an honest free market industry, one would hope that ball bearing failure would have been foreseen by those engineering the systems…”
I have designed industrial equipment for 40 years. Of course you are right. The ONE main thing engineers are required to have is foresight. And mostly we do. I have NO idea why they would put these things out there without a PM (preventive maintenance).
Can the bearings be designed in so that they last longer? Absofreakinglutely.
The one thing that kills bearings is heat. You deal with heat by preventing its buildup. Yes, better bearings is good. But also by properly SIZING them in the first place. The LAST thing you do is play games, riding on thin ice with weak, undersized bearings, or going cheapo cheapo. Not when the damned things are 300 feet in the air in the middle of nowhere.
There are ways of positively cooling the bearings. I mean, every minute they are working there is power available to pump lubricants. Also the lubricants can – and SHOULD BE – be COOLED as they cycle around through the pump.
Another thing is to use the best synthetic lubricants available. I honestly recommend one called Amsoil. I’ve used it in applications going up to 500°F and the stuff is great. Best for cars, too. If you can find it. Normal organic lubricants are not up to snuff.
My guess is that the manufacturers balanced costs profits and they went the cheap route, even if only a little bit. And that includes the costs of maintaining the bearings – swapping them out before they go out. THAT is the main part of the PM plan. But it COSTS money. That has to be figured in.
Another thing they could have done is to design the dynamo packages so that they could be lowered easily, making preventive maintenance that much easier.
Last point: I may be dumb on this one, but I see the normal loading on the bearings as being pretty balanced. The rotors balance it (mostly) because they are balanced angularly. The overhung load (the rotors hanging on one end) is a negative, but not terribly. You just have to deign for it.
Perhaps the best thing they could do is mount rotors on BOTH ends of the dynamo/generator shaft. That way the weight load is balanced. They might also get away with shorter blades then. I think it would get a little complex with the way the wind loads hit the front or back rotors, but good engineers should be able to deal with that, with perhaps a little it of R&D. Perhaps the front (windward) blades would have a slightly different warp to them. Or have an automated pitch adjustment, based on feedback from the blade force.
All of these things are possible, but they cost money. They would have to decided which is more important – up front profits or long term profits and respect.
All this means nothings, of course. The things are monstrosities and have no place on our farmland. If they could take over all of our energy needs, fine; I’d shut up about them. But they can’t. Not enough overall energy in the winds. Not enough energy density. Especially when the old wind just ain’t blowin’.
Steve,
I’ve little doubt that the improvements you recommend would extend the bearing life, but every one of your suggestions will reduce the rated output of the mill by diverting power to maintenance tasks.
I’m guessing that the “rated output” is an absolutely central metric to the politicians and regulators who are already freaked out by the uneconomic nature of these beasts. Lower output, even more expensive designs would be unsaleable, even to them.
TYoke- Anybody who thinks that they are going to get by WITHOUT maintenance is a village idiot. It’s like hoping to invent a perpetual motion machine – bozo thinking.
There are trade-offs that sometimes must be accepted. If the cost of PM or better bearings brings the ROI down too far, then screw the whole thing. I am sure I know a bearing salesman who would want to let his engineers at it. Ultra-low friction bearings DO exist, and if necessary manufacturers would be willing to go the extra mile.
At the same time, others here bring up seals, an it is my guess that it is the seals that are the big drag elements. at the same time, I am sure that in my day (I am retired now) I would have been able to give them seals considerably lower in friction. But I think that even with infinitely small bearings the windmills are never going to deliver. And I LOATHE the bloody monstrosities, so I am not about to help them.
To minimize asymmetric loading on blades, instead of the typical 3-blade design, use 4 blades forming a cross of two straight lines. The opposite blades act like a lever and the shaft is the fulcrum. This will balance the force on the fulcrum and reduce stress on the bearing. Make the blades flexible like an aircraft wing so they can bend. This will reduce the force transmitted to the bearing.
The most efficient blade configuration for generating power is one. And some wind turbines have been built that way: A single heavy counterbalance is put on a short blade on the other side of the single long blade. (I-5, north of Vallejo CA for example.) Those models have not been repeated since the late 1980’s, but I do not know what the specific failure mode was. Regardless, as stated, single prop’s have not been built in other that test situations.
Dual blades can work, and many units are dual bladed. They are more efficient at converting the available wind energy into electric power than three-blade prop’s, but, the industry worldwide has found that 3 blades – each slightly shorter than a two-blade prop, are overall a better configuration.
Tests show a four-blade prop as you describe offers the worst of 2-blade in terms of vibration, loss of efficiency 9compared to three blades), heavier weights on the nacelle and bearings, and increased vibrational nodes (resonances) as the blades cross the line of the tower. 3-blade props also make installation and maintenance – what little maintenance is actually done that is – slightly easier and safer.
Wind turbine still kill more people than other energy system. But the greed-industry and their politicians don’t want to talk about that. They’d rather focus on the fear of non-existent deaths in the nuclear industry.
Are we talking bearing made in America or are we talking bearing outsourced to China here? You get what you pay for.
You sound like back in the 1960s and talking about “Made in Japan” a being a sign of poor quality then.
You evidently think it needs to say “MADE IN USA” to be good quality. Actually, made in the USA is as often erratic quality. As a former business owner of a small manufacturing company, I can tell you that American quality is fairly often the worst on the block. Oh, you take your chances in China, too. But we had spur gears made in both places and by FAR the worst were American. At the same time, one Chinese manufacturer got just about everything wrong on some shoulder screws, on a trial run. Like I said, you take your chances. But I assure you that MADE IN USA is no promise of good quality.
I’m completely against what currently exists as the wind energy “industry” (though I reject the bird issue since it would necessitate getting rid of cars, buildings, bridges, trees and everything else birds and bats continuously slam into). But if the bearings need to be 10 times stronger than current materials, if they discover a material that’s economical and 10 or more times stronger, would it make wind practical without subsidies? (Assuming one has either a market for intermittent power supply or an economical storage method.)
Snake Oil Baron asks @August 28, 2014 at 6:50 am;
In other words, are the bearings the only hangup? No; bearings are one of several hangups, each of which is individually a potential deal-breaker, show-stopper.
In the case of bearings (as will also apply with some of the other wind turbine adversities), hypothetical solutions that would improve the competitiveness of wind, will act in similar ways to improve the competitiveness of its competitors, such as petroleum and gas.
If we did create magic new bearings, everyone would stand to benefit. The conventional oil industry, and coal mining, leading the list.
And wind would still have several other albatrosses hanging from its neck.
good luck to the wind turbine crowd on developing a bearing material “ten times more durable” ( in the words of the “senior wind turbine engineer” quoted on the post). This is science fiction.
Agreed. Professionally speaking, I don’t think the guy is well-versed on bearings, “turbine engineer” or not. I am a retired senior mechanical deign engineer myself, and to be honest, to me this would be a cool project, but nothing very much “out there”. I think the engineers were simply not adequate for the job – him included. He might know lots about blades and power and all, but only years of dealing with bearings in equipment will teach someone how to deal with the forces and the elements .
At 7:42 am ON 28 August, mpainter wrote:
Nope. It’s fantasy.
David L. Hagen said @August 26, 2014 at 7:27 am;
The cited article does not mention bearings at all. How the necessary durability of the new bypass fan gearbox was achieved, is not discussed here.
A fancy gearbox needs good bearings, sure … but generally what limits ‘extreme’ gearbox-applications is the gears themselves, rather than the bearings.
Having ultra-reliable bearings is not a problem. There are MANY out there, available off-the-shelf. Bearings are rated based on millions and millions of cycles/revs. The thing is to size them right and protect them from the elements and heat. Thus, seals, shields, and automatic lubrication systems.
I am HIGHLY doubtful that hat spokesman really knows what he is talking about, with his “ten times stronger” line. There is no one-size-fits-all bearing or bearing type. They have to be properly selected. I mean, the RPMs on those things is so low (15?), the total revs per year can’t be that much. MOST bearings are run at speeds of 1800 RPMs, 900 RPMs, even many, many at 3600 RPMs. It is an “overhung load” application, so the front bearings are seeing most of the load. Proper design would be to minimize the loading and “unit stress”. The real experts on that are the bearing manufacturer’s in-house engineers. And they will work with anyone to get the very best.
But NO bearing is any better than its lubrication system. If they put them up without providing “in-shaft” lube ports and without heat sensors and vibration sensors, they don’t know what they are doing.
industrial wind turbines are an appliance, they use power same as a dish washer when counted over one year.
Sorry, dude. With all due respect, you do not know what you are talking about. You do not plug the wind turbine in to get it to go. They do not USE power; they produce power (even if not enough for the world).