Design loads for wind towers & jackets?

[Vigardo]

TL;DR Summary

I need reliable design data (loads and geometry) for wind power towers, both on- and off-shore, tubular, lattice, etc... Any idea where would I look for?

Dear experts,

I'm looking for some approximate (but reliable) design loads and geometry for wind turbine towers to perform a comparative FEA study. I don't need very exact data. I just want some load levels for different wind towers, e.g. for 5 or 10 MW on- and off-shore wind turbines, and off-shore jackets. I'm interested in any type of them, e.g. tubular, lattice, concrete, hybrid, etc... Furthermore, it would be great having data about towers for electricity transport or telecommunications as well.

Please, do you know where would I obtain such design data? Any textbook or academic paper?

So far, I've checked several research papers and NREL documents without success. They usually partially omit the data I'm looking for. The only reliable data I found is for a simplified structural model of a 76 m height tower* .

*The material properties, dimensions, and design loads for this example are: F= 2489 kN (horizontal thrust at hub height) and M= 88606 kN·m (overturning moment at base), according to to load case G+1.5W and including tower self-weight (Eurocode 1-1-4). The material is structural steel: E= 200 GPa, σ= 355 MPa, v=0.3, and ρ=7850 kg/m3. [HISTWIN RFS-CT-2006-00031, page 9].

Thanks!

 

[berkeman]

Paging @cjl ...

 

[Vigardo]

Thanks Berkeman! Cjl's posts contain interesting info and ideas. Upon searching "cjl turbine" (cjl has nearly 2000 posts so far :-), I quoted the relevant info here:

Thrust load on a modern ~2.5-3MW wind turbine with a rotor diameter in the 120-130m range is on the order of 500kN.

Most modern turbines run at a power coefficient (mechanical power vs kinetic energy power of incoming airflow) of around 45-50% at rated or so, and they also have an electrical efficiency around 90%. They also tend to be rated at more like 13m/s than 15, so if you factor all of those in, I suspect you'll get up from your 340kN up to the 500kN or so I quoted.

Thus, I would plug the 61% of maximum theoretic efficiency achievable to extract mechanical energy from the wind kinetic energy (Betz limit), an approximate 90 % of electrical efficiency, and the rated turbine power (e.g. 2, 5, or 10 MW) into some basic energy conservation formulas to obtain thrust force at hub height. But this is just the maximum horizontal force at rated power operation...

Ideally, I would like to know the loads of the most critical load case, i.e. those that determine the design of the tower. I think they must include the weight on top (nacelle, rotor, blades, hub, gearbox, etc), wind loads on the tower, rotor momentum, rotor speed (to avoid rotor resonance frequencies P1 and P3), etc. I mean, some references that I can cite, e.g. books, research articles, manufacturer specifications, etc.

Alternatively, and as last resource, I would add some safety factor to the thrust force, use Eurocode (EN 1991-1-4) for wind loads on tower, find some approximate weights on top, and perform my comparative study.

Do you think I would contact Cjl or any other wind turbines expert in Physics Forums?

 

[DEvens]

At some point you should also be looking at the time variation of the force. It's a very different design problem to sustain a constant force as compared to a force that is changing harshly over time. The company lab did testing on a wind turbine's tower. This consisted of bolting it to the test floor, and pushing on it with a piston. This was designed to be similar to the force the tower was expected to experience in operation, including the time variation.

The tower stood up until the test had to be terminated because the cement floor under the tower cracked. And nobody was happy about that.

So the tower, and its base, have to be able to withstand not just the force, but the variation over time.

 

[anorlunda]

I would imagine many variables factor into the safety factor you must allow.

Maximum wind gusts before the propellers are feathered. Maximum gusts while feathered. Ice loading. Unbalanced blades. Trauma the structure may endure during its lifetime. Many more.

What safety factor do you plan, 3x? 4x?

 

[Vigardo]

Thanks for your replies!

DEvans, you're right, but the aim of my study is just compare the strength and stiffness (maximum displacements and natural frequencies) for different topological alternatives of lattice towers. Thus, for simplicity, I will not include fundation interaction with the tower.

Anorlunda, I don't know the exact safety factor yet. A 3x factor over the nominal thrust load and mass on top seems reasonable. However, I would like that somebody recomended me some critical load case example, perhaps from some book or paper. Wind turbine towers design codes are complicated and have many loadcases that should be considered for each especiffic location of the wind farm. I don't want to go so deep. I just want some critical approximate load case for my study.

 

[anorlunda]

There are a lot of numbers in this paper. Choose one. If your needs are so approximate, it doesn't matter which you choose.

https://www.nrel.gov/docs/fy99osti/25787.pdf

 

[Vigardo]

Great contribution Anorlunda! Thank you very much!

In the cited NREL paper, they calculate the primary load cases for a 17 m tall and 10 m rotor diameter wind turbine according to the IEC 61400-1 industry standard and focus on the loads for the most critical cases, i.e. parked wind turbine at the extreme wind speed and running wind turbine. They provide the forces and moments at yaw bearing that I'm looking for. There is also info about different statistic models to predict the variation with time of the loads.
They also say that for several components the design is limited by ultimate loading, which is what I want to use in my approximate study. At this moment I will not pay much attention to fatigue.

However, the turbine is relatively old and small, quite far from the modern 2 or 5 MW machines with towers over 100 m tall. Perhaps you know some other paper with data for more recent and tall wind power generators, do you? Thanks a lot!

 

[cjl]

Sorry for not responding here earlier - I was off skiing and away from internet for the past several days.

Wind turbine load assessment is actually rather complicated (unsurprisingly). There are a number of different design load cases defined in the IEC 61400-1 standard (which you'll want to try to get your hands on if possible, since it is really what we design to), ranging from normal operation to extreme gust and direction change cases or startup or shutdown with gusts, or similar. Safety factor is actually different for different scenarios depending on their likelihood of occurrence, and then for fatigue load assessment, you have to combine a certain number of hours of normal operation, a certain number of startups and shutdowns, an assumed mean wind speed, etc. For actual tower designs, we'll run a few thousand (or more) simulations of the turbine behavior in all of these different scenarios (including full aeroelastic behavior and controller behavior), and then design the tower based on those observed loads (plus safety factor).

You can replicate this with a tool such as NREL's FAST, though you'll be somewhat limited in the scope of simulations you can run unless you have a large amount of time on your hands or access to a cluster or supercomputer. I believe FAST even contains a 5MW reference turbine model (based on the one defined in this paper), which should be a decent one for you to use to see some representative loads for a modern turbine.

If you want to focus on ultimate loading, you'll probably want to look at gust cases (likely either an extreme operating gust or an extreme coherent gust with a direction change) rather than normal static loads, though which case will be your design driver and by how much does depend pretty heavily on both your turbine design and your controller design. That 5MW reference and some simulation in FAST is probably your best place to start though, and should at least get you going.

I'm happy to try to answer further questions if you have any, though I can't give you any of our actual turbine load numbers unfortunately because I'm not sure how much of that is public information (the mean thrust number mentioned above is pretty easy to calculate from basic principles, but more detailed information about how we respond to gusts and what our load fluctuations look like are not as simple to come by).

 

[Vigardo]

In NREL documents database, I've found two more recent references intended for larger turbines:

In Rinker 2018 - WindPACT Reference Wind Turbines, you can find a series of four baseline models for different rated powers (0.75 , 1.5, 3.0, and 5.0 MW). Each model was designed to its specified rated power using the same methodology to facilitate investigations about the scaling of loads and turbine cost with size, so this seems ideal for my study purposes. For example, there are FAST calculations for the rotor thrust and torque from 0 to 25 m/s wind speed (pages 13-14), Campbell diagrams (pages 15-16), and general geometries and weights. In Jonkman 2009 - Definition of a 5-MW Reference Wind Turbine for Offshore System Development, there is similar data but for a 5 MW wind turbine.

However, as Anorlunda noted, the most critical load cases will be far from the energy production case that these data would approximate. Conversely to the paper Madsen 1999 (the one cited by Anorlunda), there were not analysed at all any of the most critical load cases from the IEC 61400-1. For example, the extreme loads in energy production with gusts or parked under extreme wind speeds (much higher than the 25 m/s rated for maximum energy production).

How would I approximate the (static) wind loads acting on the tower top (Blades + Rotor + Nacelle) for at least one of the most critical load cases?

Do you think that I would use Eurocode to compute the wind loads on the tower for some extreme wind speed?

Thanks!

PS: Sorry Cjl, I've just seen that you've answered while I was posting. I will answer you properly tomorrow, now it's really late here :-) Thanks for understanding!

 

[cjl]

No problem - it looks like you stumbled on a lot of the same resources yourself while I was responding anyways.

To approximate the critical loads, I would try to run FAST if you can (and I'd use that 5MW reference design). You should be able to download it and run it locally, and that'll give you the best representation of what's going on. For the tower, as I said, my guess is that it'll be driven by either an extreme gust with a direction change (Design Load Case 1.4 from the IEC standard) or by an extreme operating gust (DLC 2.3 from the IEC standard), so those are where I'd look when simulating. Also pay attention to the tower load at various angles - the worst case load direction likely won't be directly in the thrust or in the side-side direction, but will probably be a combination of the two.

I'm not familiar with Eurocode, but maybe it could also give you something useful? What kind of code is it and how are you implementing it?

EDIT: Actually, I see you linked it above. It looks like it just applies wind loads to the tower though, which is pretty negligible compared to wind load on the rotor. You're probably better off with FAST or some similar BEM code, and I believe FAST even accounts for nacelle and tower drag anyways (I know our in-house sim code does).

Another option if you want something simpler but not as powerful as FAST is ASHES. I'd probably recommend against it for designing a turbine unless you really can't get FAST running, but it's a neat tool to play around with some parameters and see how the turbine behaves.

 

[berkeman]

Sorry for not responding here earlier - I was off skiing and away from internet for the past several days.

I've seen some pretty good excuses in the past, and this is near the top of the list!

 

[Vigardo]

Hi all and thank you very much for helping me!

I will follow Cjl's recommendations. I'll try to simulate the DLC 1.4 or 2.3 for the 1.5 and 5 MW wind turbines included in OpenFAST suite (test suite r-test). On this initial stage of my study, I will use FAST wind loads instead of Eurocode.

Upon downloading the last version of OpenFAST, that can be downloaded here (currently v2.2.0), I've failed on the first attempt to run the pre-compiled windows binaries. The system complains about missing libifcoremd.dll file that is not present in the downloaded openfast_v2.2.0_binaries directory.

Instead of bothering you with naive installation questions, I'll check the installation issues in FAST forum first, try to understand the https://openfast.readthedocs.io/en/master/, and prepare my workstation accordingly (if necessary). I really wish I don't have to compile everything from scratch... it seems a cumbersome task!

What platform is easier to run FAST as fast as possible? Windows or Linux? (I have some experience on both). EDIT: Or, perhaps I would start by using the pre-computed data available at r-test directories? For example, at r-test-master\glue-codes\openfast\5MW_Land_BD_DLL_WTurb\windows-intel

PD: I'm happy to hear that somebody is having fun instead of fighting with turbines modeling and FAST installation issues during the weekend ;-)

 

[cjl]

I've had best luck on Windows, but I haven't spent much time trying to run it on Linux, so my suggestion would probably be to just go with whichever platform you feel more comfortable debugging on. Have you had any success yet running FAST? I haven't run it in a couple of years, but I can download it and see if I have any issues tonight if you still haven't had success.

 

[Vigardo]

Hi Cjl, sorry for the delay and thanks for your very kind help! It is highly appreciated!

Let me tell you more about my project. In a first stage, I just want to define a simple (but representative) base line for loading and geometries in order to compare different lattice tower configurations in some meaningful scenario. Thus, I decided to deduce myself the thrust formula and collect the necessary data from the WindPACT reference turbines (Tables 1 and 2).

With this data I tested how accurate the thrust predicted by basic physics (see formula "c" in Table 1) is in comparison with WindPACT paper. As it can be seen, the calculated and FASTv8 results match fairly well. In addition, I could check that these turbines are working at nearly 85 % of the Betz limit efficiency.

I also examined some of the pre-computed simulations from the four reference wind turbines to better understand what I would do and decide what to do next. For example, I plotted Thrust vs. time (Figure 1) and Thrust vs. wind speed (Figure 2) for the 10 minutes simulation for the 5 MW case.

Now I realize how complex the aero-elastic interactions between all turbine components are and why many long simulations are required. One question about all these stuff, do you know how the steady state thrust in Rinker 2018 (pages 13-14 or column FASTv8 in Table 1) was calculated? In the pre-computed data for the reference turbines (link above) I haven't found any clue.

In the meantime, I think that weight loads, rotor thrust, and torque plus some safety factor would do the job, at least as a first approximation. In a second stage, it would be great running some FAST simulations with the most critical DLCs commented above. I'll let you know when I'm ready :-) (thanks again!)

An alternative to running FAST would be using some publicly available DLCs simulations, perhaps you or somebody else knows where... Thanks!