right-angle wind turbine gearbox

[Windward]

Advice on this topic is requested below. First a description of the arrangement.

The right-angle drive would be installed at the top of the tower. Power would be transferred to the base through a vertical drive shaft. This would reduce tower top weight and put the generator and electricals at ground level. The cost of erecting and maintaining the turbine would be substantially reduced. There would be no heavy power cables running from tower top to base.

The vertical drive shaft would also act as a torsional spring. It would absorb the shock from wind gusts that cause much of the gearbox damage in conventional turbines.

A right-angle gearbox would allow the rotor shaft to be straddle mounted on dedicated bearings, eliminating gearbox stresses caused by supporting some or all of the rotor weight and bending moments. The rotor shaft would be hollow to allow access to the hub for controls.

Direct drive has none of these advantages. The tower top weight of direct drive is just one of its insoluble problems. One estimate for the tower top weight of an 8 MW direct drive wind turbine is 800 tons. There are several other good reasons to abandon direct drive for large wind turbines.

The only question I cannot answer at this time about this arrangement is whether there would be unacceptable torsional vibrations in the vertical drive shaft that could not be counteracted. Any advice on this question would be appreciated.

 

[TVP]

Vibration would certainly be a concern, but I think the key point is can it be counteracted? Lots of options to address this from low weight, high stiffness composite materials, multiple-piece construction with periodic bearing support, etc.

 

[Windward]

Thanks, TVP. I hadn't thought of composites. Here is a statement from BAC technologies, which makes race car drive shafts:

"4. Torsional Damper: For the same fixed diameter and fixed length, it is possible to engineer a carbon shaft to have a relatively low modulus of rigidity resulting in a shaft with a high torsional flexibility compared to the same size metal shaft. This unique characteristic will "smooth out" and absorb torque spikes of the engine extending the transmission life."

I had been considering drill pipe, for known properties and wide availability. Composites would cost more but probably worth it. They would certainly be lighter.

 

[gearcutter]

Seems like a good idea but how would you counteract the torsional forces being transmitted through the vertical shaft via the nacelle?
A correction to the position of the nacelle, to keep the blades pointed into the wind, would require releasing of the brake. Under power; this would cause the nacelle to begin to rotate rapidly.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[Windward]

Excellent point, gearcutter. It does require a solution. I have seen counter-rotating shafts, or nested shafts turning in opposite directions if you see what I mean. That would do it, but it is a complication.

I think the best solution would be to reduce the torque acting on the vertical axis of the nacelle. This is done by speeding up the vertical drive shaft. I had already understood that this would be necessary in order to reduce the size of the shaft. I have been thinking of a ratio of 20:1 or more over the rotor speed. This would also get the vertical drive shaft up to the speed of a medium speed generator, so that no further speed increase would be needed.

This speed increase would reduce the torque on the vertical axis of the nacelle by the factor of the speed ratio. Although the full torque would still act on the nacelle, it would be on the rotor axis. But this is just the torque that all conventional nacelles are designed to resist. An additional torque of one-twentieth or less of the full torque, acting on the vertical axis, may not be too hard to handle.

 

[MikeHalloran]

Putting the speed increaser at the top restores the problem of a large mass at the tower top.
... and forces you to deal with critical speed issues for a very long shaft.

The next thing that came to mind was putting a crank on the turbine hub, and using a push/pull rod to run a second crank on the ground. Unfortunately, you need two cranks, quartered, to have a hope of not toggling into destruction.

Next thing that came to mind is a really long timing belt drive from turbine shaft to a ground shaft. Provide a tensioner with some stroke, and let the turbine yaw to follow the wind, twisting the belt a bit, but the center distance is so huge that twist may not be such a problem. You need intermediate rollers or slippers to keep the rising teeth from fouling the falling teeth. ... or you just yaw the whole tower including the base shaft, speed increaser, and generator.



Mike Halloran
Pembroke Pines, FL, USA

 

[Windward]

Can't agree, Mike. The right-angle drive would weigh less than the gearbox in a conventional arrangement. The generator, power cabinets etc. would be at the base of the tower. The tower top weight would be cut roughly in half.

I did consider the ideas you came up with, and others. None of them were practical, as you seem to have concluded for those you mentioned.

I know how to build the right-angle drive. My main concern is the shaft. If you are an expert in that field and believe it won't work, please say so and offer some evidence. Opinions are welcome also, and I believe that is all you have offered.

 

[MikeHalloran]

Opinions are all you get for what you are paying me.


Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

Windward-

Have you considered just how much torque the nacelle mounted right-angle gearbox and vertical driveshaft system would be subject to? A 3MW wind turbine rotor produces over 1,400,000 ft-lbs shaft torque at rated speed/power. The size/weight of the bevel gear mesh and vertical driveshaft system would be massive. What would make the (nacelle mounted) right-angle gearbox especially heavy is that all of the torque must be transmitted through a single gear mesh, as opposed to the conventional planetary gearbox designs used in wind turbines where the torque is distributed among multiple planet gears.

 

[Windward]

Let’s double the power. The bigger the turbine the more effective the arrangement, and 6MW machines are now in production. A rotor speed of 10 rpm would produce a torque of 4.2 million foot-pounds. With a 30:1 ratio for the right-angle drive, the torque on the vertical drive shaft would be 141,000 foot-pounds.

This would be the new torque, acting in the vertical axis, that the nacelle and tower must handle. To me it does not seem too high to be manageable. A harder problem is the behavior of the vertical shaft. It is a mystery to me whether any torsional vibrations would be too much to handle. I know there are experts who could solve this problem on paper. If it appears that the idea might work, I think it is worth a try.

 

[tbuelna]

Windward-

Let's start with the reasons you feel this concept would be a better approach than the industry standard multi-stage gearbox, variable frequency AC generator and power electronics. First of all, contrary to what you think, the weight of nacelle components (like gearboxes or generators) of existing commercial wind turbines is not a big issue. A 6MW gearbox or DFIG generator is no more difficult to lift than the rotor hub or blades. The total mass of the generator and gearbox are also not the driving factor with tower design. The tower structure is designed mostly based on the out-of-plane moments produced by the rotor.

As for your 30:1 ratio bevel gear drive at the rotor shaft output, how would it be any less sensitive than a conventional gearbox to the effects of deflections in the rotor shaft bearing system? As for weight, a conventional 6M 3-stage epicyclic gearbox might have a total ratio of around 150:1, and the input stage would have a ratio of far less than 30:1. Plus the epicyclic input stage of the conventional gearbox would split the massive rotor torque between at least 3 gear meshes. Your 30:1 ratio bevel gearbox would need to pass 100% of the rotor torque through a single mesh. Plus it would be very difficult to design a single bevel gear mesh that could achieve a 30:1 ratio. In fact, your 30:1 bevel gearbox would likely end up being much heavier and less efficient than a conventional 150:1 3-stage epicyclic gearbox.

Hope that helps.
terry

 

[Windward]

Terry, some of us know you are an ingenious fellow, particularly in mechanical drives. You might figure out a way to build a 30:1 ratio right-angle drive with none of the deficiencies you listed, and none that would make it impractical. If you did, would you publish the details before filing for a patent on it? I might, but not yet.

There would be no point in doing so if it can be shown that the vertical drive shaft would not work. That is the main question I am asking here, although it may be too complicated for this forum. I just wanted to see whether someone could answer it.

Granted that the greatest stress on the tower is the overturning moment created by the resistance of the rotor blades to the wind. It seems that you believe there would be no advantage in reducing tower top weight and moving most of the equipment to the base. If the vertical drive shaft works, a detailed analysis of the arrangement would show whether it is an improvement.

 

[tbuelna]

Windward- Thanks for the compliment, and I did not intend to rain on your parade. It's just that I have spent quite a bit of time looking at this particular issue, which is how to create a more efficient and economical commercial wind turbine drivetrain. And what I found is that the current systems employed by 90% of commercial wind turbines (ie. a gearbox and DFIG generator) are used for very sound reasons, the primary reason being simple market economics.

It's not that I think a pair of right angle drive transmissions and vertical driveshaft system could not be made to work, it's that based on my experience I don't see how such a system would be an improvement on the current designs. I spent over 4 years and a fair amount of my own money trying (unsuccessfully) to sell a variable speed transmission to the wind turbine companies. I had lots of analysis that showed how much better my variable speed transmission was in terms of cost, efficiency and reliability. And it could even be integrated to existing wind turbine designs without much trouble. But I was unable to get anyone interested in the idea.

All I can say based on my own experience is that if I could not generate any interest from the wind turbine industry with my variable speed transmission concept, which was far easier to integrate into existing turbine designs than what you're proposing, you are likely facing a very difficult task.

Good luck to you.
Terry

 

[Windward]

Here is a quote from an April 2014 Siemens brochure on their new 6 MW turbine:

“With a tower head mass of less than 60 tons per megawatt, the D6 wind turbine is genuinely lean. This new low-weight standard for offshore wind turbines offers significant cost benefits in terms of substructure requirements, shipping, and installation. All of this is made possible thanks to Siemens’ proven direct drive technology.”

http://www.energy.siemens.com/hq/po.../D6 Offshore brochure_English_Apr2014_WEB.pdf

American Superconductor is developing high temperature superconductors for direct drive generators which they believe will substantially reduce tower top weight:

http://web.mit.edu/windenergy/windweek/Presentations/P7 - McGahn.pdf

These are just two examples of a serious effort in the industry to reduce tower top weight. In their presentation, AMSC estimates that the tower top weight of a conventional 8MW direct drive turbine would be 800 tons, while their superconducting design would weigh 480 tons.

But in both of these cases, the decision to use direct drive has prevented them from achieving even greater TTW reduction. A geared turbine will always have a lower TTW than a direct drive machine. But those who advocate direct drive invariably claim that the elimination of the gearbox is necessary to increase reliability.

Gearboxes do have a bad reputation in the industry, but it is too soon to conclude that they will always be a problem. Recent research has discovered some likely causes for these failures. This will allow for improved designs. One discovery is that the gearbox is greatly overloaded in both directions during braking. Not the fault of the gearbox, just fix the braking system. Another discovery concerns bearings. They are failing sooner than expected, and the focus of research now is on “white etching area.” Fix that problem and the gearbox becomes even more reliable. Distortion of the gear case from external loads is also a problem, again not due to design of the gearbox, but to inferior design of the rotor shaft support.

Not that I want to start a new controversy here, but the planetary gearbox design must be abandoned if gearbox reliability is to achieve an acceptable level. The planetary design is inherently inferior to the multi-branch parallel shaft type. In the planetary design it is impossible to make the planets share the load equally. The planet carrier will distort under load, misaligning he gears. Gravity aggravates these effects, and I have not exhausted my objections. Here is a good study of the subject, although it avoids the conclusion that the planetary design itself is the main problem:

http://www.nrel.gov/docs/fy12osti/55207.pdf

There are many other important papers on gearboxes where that one is listed:

http://www.nrel.gov/wind/grc/publications.html

Further to the question of replacing gearboxes with direct drive, I cannot agree with the commonly heard claim that fewer parts necessarily mean greater reliability. What would you rather have in your car, a one-lunger or a V-8? Werner von Braun made this point as well as anyone (quoted in an earlier thread on this forum):

“ …when he was asked how something as complex as the moon rocket for the Apollo mission could possibly be reliable with so many parts involved. He replied that in his garage, he had a car and a lawn mower. The lawn mower was much simpler than the car, but the car started every time he pulled the starter, and the lawn mower didn't.”

In fact, power electronics fail twice as often as gearboxes, as shown in the following:

http://www.gl-garradhassan.com/asse...bility_-_Results_of_the_Reliawind_Project.pdf

Note that this study was conducted before any gearbox improvements indicated by recent research have been put into practice. If power electronics are the main cause of turbine downtime, how can direct drive advocates justify their claim that they are intent on increasing reliability? Direct drive cannot work without them. If they are worried about part count, maybe they have a point after it reaches a certain level. Power electronics have many trillions of moving parts.

One of the foremost advocates of direct drive, Sandy Butterfield of Boulder Power, formerly chief engineer for many years at the national wind research organization NREL, has finally succumbed to industry pressure by providing a generator for geared turbines:

http://boulderwindpower.com/home/products/medium-speed-generators/

Neither can direct drive achieve what DeWind has been doing with ever greater success, using a geared turbine to eliminate the power electronics. Their website explains it well.

Terry, the proposition that most wind turbine manufacturers have made the best possible choice by settling on a planetary geared turbine, whether low, medium or high speed, with a DFIG and power electronics, and everything mounted in the nacelle, reminds me of the attitude of the major American car companies when the Japanese introduced fuel efficient small cars. There was no end of ridicule from the majors. At first these small cars were not very good, but the Japanese kept at it. That is why Toyota is the biggest car company today. I regard the advocates of direct drive and conventional geared turbines as the modern version of General Motors circa 1972, before the first big oil shock.

With the arrangement I propose, tower top weight would be reduced to a minimum. Most of the equipment would be located at the base for easier erection and maintenance. It may even be possible to replace the power electronics in this arrangement. All of these changes would be highly advantageous. Direct drive cannot achieve any of them. The only question I have is the behavior of the vertical drive shaft. If it is controllable, I think the arrangement should be studied in detail.

Concerning your failure to interest any wind turbine manufacturers in your CVT design, which I believe would require at least some gearing, which again would be in opposition to direct drive, I am familiar with the problem. I first proposed a vertical drive shaft more than ten years ago to Vestas with no reply. For more than twenty years I have been working on an alternative to the planetary gearbox. At least two leaders in this field have found that my work is “unique and has technical merit” but have declined to develop it. There is a wise saying about this:

“Establishments like nothing so little as progress not established by themselves.”

Gearbox development is extremely expensive. Unless the inventor has millions of his own money to spend to demonstrate his concepts, he will almost certainly get nowhere. But no one who gives up can call himself an inventor, even when he goes to the grave a forgotten pauper while others profit greatly from his work. There is no better example of this than John Fitch, inventor of the steamboat. I cannot recommend a more interesting story than his. See

https://archive.org/stream/lifejohnfitchin02westgoog#page/n52/mode/2up

 

[AWloo]

The only thing I haven't seen mentioned here is the cost of having a very long shaft rotating in the inside of the wind turbine, have you accounted for this decreased efficiency? I'm more curious than judgmental. It is an interesting concept!

Cheesr

 

[Windward]

That is one of the unknowns, AWloo. There must be several spring bearings to keep the shaft from whipping, how many I don't know, but they would be a parasitic load. Although I doubt that this would be more than a minor inconvenience, nothing about the behavior or design of the vertical shaft can be known without a detailed study, including ways to avoid critical speeds.

Another point about this arrangement may be obvious, but the slewing of the nacelle to face the wind would not cause a problem with a central, vertical drive shaft.

 

[tbuelna]

Windward-

Regardless of what Siemens' marketing claims, the total mass of the components at the top of the tower is not a legitimate concern. Regarding nacelle components, the only concern with regards to weight is what is the maximum weight of a single piece of hardware that must be lifted into place atop the tower, and heavier components require bigger cranes that cost more to operate. Things like gearboxes and generators must be lifted in one piece, but even a 6MW gearbox or 6MW DFIG generator is still significantly lighter than a 6MW direct drive generator.

A while back, utility scale wind turbine gearboxes did have reliability problems. But over the past few years the reliability of gearbox designs has greatly improved. If you look at the industry reliability rates compiled over the last 6-8 years, you'll find that gearbox statistical failure rates are much lower than many other systems like power electronics. However, it is also important to consider the operational downtime caused by the failure of a component. Replacing a gearbox might require several weeks, while replacing failed PEs might only require a couple days.

The primary factor commercial wind companies use to decide which type of turbine to purchase is ultimately economics. Which turbine design provides the best balance of capital costs, reliability, power production, operating cost, etc. The overwhelming majority of utility scale wind turbines sold each year are still conventional designs using gearboxes and high-speed generators. For an outsider looking in, it might appear that the commercial wind turbine operators do not appreciate the technical sophistication of something like a 6MW turbine with a massive PM direct drive generator and solid state power electronics, but they do. They prefer to use conventional turbine designs because they are a known quantity that provides acceptable cost, reliability and efficiency.

If you think your wind turbine drivetrain concept is superior to existing designs, then all you need to do is convince commercial wind companies of this. Unfortunately, this won't be easy, but I wish you good luck all the same.

Best regards,
Terry

 

[Windward]

Terry, glad you agree with several of the points I made in my long post above. And I doubt that anyone would disagree with this statement of yours: “The primary factor commercial wind companies use to decide which type of turbine to purchase is ultimately economics.”

One thing about which we will probably never agree is the significance of tower top weight. I believe that reducing it is important and you don’t. Below are some quotes from the 2011 UpWind report:

http://www.upwind.eu/~/media/UpWind/Documents/Home/21895_UpWind_Report_low_web.ashx

More of the same can be found by searching the document for “weight.” Although the report concerns very large turbines which do not yet exist, their conclusions about tower top weight would apply to existing units:

“Key weaknesses of the extrapolated virtual 20 MW design are the weight on top of the tower, the corresponding loads on the entire structure and the aerodynamic rotor blade control. The future large-scale wind turbine system drawn up by the UpWind project, however, is smart, reliable, accessible, efficient and lightweight.”

“Reducing the loads and the nacelle weight enables the offshore substructure design to be optimized.”

“Size and weight do matter. We had to investigate the limits of current designs. UpWind provided the necessary tools for upscaling. Integral design tools will significantly increase the reliability of future drive trains, and avoid costly drive train failures. UpWind specified the future generator type, which is proportionally lighter than today’s designs.”

“Concentrating on upscaled designs, the weight on top of the mast is critical for transport, installation and design loading. The perfect generator would be reliable, efficient and light.”

Can you suggest any studies that contradict these results?

 

[gearcutter]

One area where weight of the nacelle components is critical is in off-shore systems. In areas like Europe; off-shore units will soon out number the ones on-shore. I can't see how a vertical shaft system can be implemented in an off-shore environment so this may significantly reduce the market size for your idea.

Back to your original enquiry; perhaps you need to talk to someone that designs deep hole drilling rigs, particularly rigs that are mounted to floating platforms.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[Windward]

That is right, gearcutter, the first thing I searched for was drill pipe torsional vibrations, and they can be severe. My only consolation is that a wind turbine drive shaft would be very short compared to a well string.

I am curious why you think that a vertical drive shaft would not work for offshore turbines. I was thinking that this is the best place for it. Erection and maintenance is harder there, so anything that makes those jobs easier should be considered. Especially if it also significantly reduces tower top weight.

I see that Alstom is mounting the big power cabinets at the base of their Haliade turbine. One step closer to the optimum.