the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jan 2025
How does turbulence affect wake development in floating wind turbines? A critical assessment
Abstract. Research is flourishing on how to model, mitigate, or even try to exploit the complex motions floating offshore wind turbines (FOWTs) are subjected to due to the combined loading from wind, waves, currents, and buoyancy effects. While preliminary studies made use of simplified inflows to focus attention on blade-flow interaction, recent evidence suggests that the impact of realistic inflows can be much larger than expected. The present study presents a critical analysis aimed at quantifying to what extent turbulence characteristics affect the wake structures of a floating turbine undergoing large motions. Numerical CFD simulations, using a Large Eddy Simulation (LES) approach coupled with an Actuator Line Method for the rotor, are benchmarked against wind tunnel experimental data from the first campaign of the NETTUNO project on a scaled rotor tested both in static conditions and when oscillating in pitch. A comparative analysis of the results at different turbulence levels first confirmed that, whenever idealized flows with no significant turbulence are considered, platform motion in FOWTs indeed leads to the creation of induced flow structures in the wake that dominate its development and the vortex breakdown in comparison to bottom-fixed cases. More interestingly, analyses show on the other hand that, whenever realistic turbulence comes to play, only small gains in terms of wake recovery are noticed in FOWTs in comparison to bottom-fixed turbines, suggesting the absence of superposition effects between inflow and platform motion, with inflow turbulence contributing significantly to dissipating the structures induced by turbine oscillation. Finally, as an ancillary outcome of the study, evidence provided by LES high-fidelity simulations were used to understand to what extent a less computationally-intensive URANS approach can be used to study the impact of realistic turbulence. In particular, an innovative URANS approach featuring improved inflow boundary conditions proved to yield consistent results if mean wake profiles are considered.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Wind Energy Science.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.- Preprint
(25252 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2024-169', Anonymous Referee #1, 04 Mar 2025
Overall, this is an excellent paper, and I would like to sincerely congratulate you on your work. I really enjoyed reading it and have just a few minor comments.
- Can you highlight how often/realistic the conditions are offshore that are being studied?
- Abbreviations are not always explained, for instance NETTUNO (line 14), URANS (line 22), BEM & CFD (line 56), ALM (line 126), HAWT (line 196)
- Line 40: The word ‘the’ missing before ‘system’
- Line 68: ‘Turbulence in (typo?) not explicitly solved’ - can you please explain what this means in an extra sentence?
- Line 94: ‘Share’ not ‘shares’
- Line 139: Please explain ‘law of the wall’
- Line 156: Please explain ‘bottom shear layer’
- Line 229: ‘LES_t’ indexing not explained until the next page
- Table 6: Please explain what the +/- values are (standard deviation?)
- Equation 2: Please explain all the variables
- Very small and partially covered coordinate system axis description in Figures: 9,10,12,15,16,19,20,B1
- Line 354: ‘But then’ grammar mistake
- Equation 3: Typo in U_0 numerator
- Figure 15: Please explain the dotted line at the top
- Equation 4: Please explain u’x
- Line 493: An Equation to show the velocity gradient tensor would be nice
- Line 514: The word ‘with’ is missing before ‘respect’
- Line 547: Please explain why the pitching case is the most effective test case to be considered
- Equation 5: Please explain N (I am assuming it is the number of sections the blade is discretized into?). Is U0 the same here as U_0 in Eq. 3?
- Line 627: Please add units for the frequency (Hz)
- Appendix A, line 814: Please explain in more detail why there is a peak at 1Hz.
Great work, I can’t wait to see the final version!
Best,
Jonah Hanak
Citation: https://doi.org/10.5194/wes-2024-169-RC1 -
RC2: 'Comment on wes-2024-169', Anonymous Referee #2, 11 Mar 2025
Review of “How does turbulence affect wake development in floating wind turbines? A critical assessment”. In this manuscript, the authors compare experimental results of wind turbine with platform motion to LES and URANS simulations. The numerical simulations are performed with and without inflow turbulence and the properties of the wake are analysed. The manuscript provides relevant results and unique conclusions on the topic of wakes of floating offshore wind turbines. The manuscript is well-written, however, the title does not fully represent the manuscript and there is lack of description of the methodology. Also, questions can be raised on the flow structures observed in the laminar simulations, as mentioned in the main comments below. Other comments are listed as specific comments. If all comments are addressed, I believe this manuscript will be a relevant contribution to the knowledge on wakes of moving turbines.
Main comments:
- The title of the manuscript promises more than the manuscript deliver. The experiments are performed with only 1 turbulence intensity and the numerical simulations are performed with only 2 turbulence intensities. In my opinion, the manuscript does not fully answer the question proposed in the title, it is only a piece of a larger picture that is still being discovered by the academic community. I strongly suggest a modification of the title.
- It is not clear that the flow structures observed on the vortices between 2D and 3D are physical mechanisms, in the laminar case in Fig. 10 (a) and (c). These flow structures look similar to flow structures created by numerical effects. Therefore, the mechanism of vortex breakdown in the laminar case may be due to numerical effects. Please improve the spatial discretization in the wake and show that the results of Fig. 10 are not dependent on the grid.
- In Appendix A, is the velocity sampled from an experiment with platform motion or without motion? If the velocity is sampled from a case with motion, then the simulations of the “fixed” turbine with turbulence will be contaminated by the frequency of motion. If this is the case, this is an important limitation of the study and should be clearly discussed in the results.
Specific comments:
- Please define all abbreviations when they first appear. Many of the abbreviations are not defined, some examples being: ALM, URANS, BEM, DOF (list not extensive). AMR was only defined in the appendix.
- In line 66-65. (Xu et al., 2024) is mentioned as a paper that suggested that the turbulence change some of the phenomena described so far in numerical studies. However, this work did not only suggest this, it studied this effect. Please rewrite it making clear the contribution of the cited article.
- In line 95. Apparently there is a missing reference (“[X]”).
- In section 2, please add the Reynolds number and temperature of the experiment.
- In Table 2, please add the tip speed ratio and indicate the rate of pitching frequency to rotor frequency, which affect the growth of instabilities according to Kleine et al, 2022.
- The solver is not mentioned in the section 3.1, only in section 3.2. Please include the name of the solver, references and a brief description in section 3.1.
- Please reorganize section 3.1, dividing the section in smaller sections (or subsections), to make clear the parameters of each simulation and numerical method. Please provide a brief description of each numerical method and include references.
- In line 122. The term “nose cone” is more specific than “nose”.
- The use of the term URANS-Hybrid is very confusing. I was only able to understand it after reading section 4.4, hence it should be better described in the methodology. More importantly, it is not a hybrid between URANS and LES, as it could be understood from context. It is a URANS simulation with a different strategy of imposing inflow turbulence, without using any of fundamentals of LES. I do not believe the name is adequate.
- Please define what is the AMR threshold and provide a reference for the AMR.
- Please indicate the distances from the boundaries to the rotor.
- In line 189. Please indicate the width of the Regularization Kernel clearly in the manuscript.
- In line 97. There is a mistake in equation 22 of (Dağ and Sørensen, 2020), as pointed out by (Meyer Forsting et al., 2019) and (Kleine et al., 2023). Please indicate clearly how the correction velocity was calculated. References: [1] Meyer Forsting, A. R., Pirrung, G. R., and Ramos-García, N. "A vortex-based tip/smearing correction for the actuator line." Wind Energy Science 4, no. 2 (2019): 369-383.; [2] Kleine, V. G., Hanifi, A., and Henningson, D.S. "Non-iterative vortex-based smearing correction for the actuator line method." Journal of Fluid Mechanics 961 (2023): A29.
- Section 3.3 is not clear. Please define mathematically the desired velocity perturbation, the exponential distribution function and the input velocity perturbations. Include more details on the generation zone.
- Please present the the measured properties/statistics that shows that the position of the generation zone of turbulence is adequate for both the LES and URANS (“hybrid”) simulations.
- In line 220 and appendix B. Some distances are indicated in meters, without indicating non-dimensional distances.
- In section 3.4 and figure 12. Please indicate the sampling window in number of cycles instead of seconds. Please also indicate the number of rotor revolutions.
- Please define Δ in tables 5 and 6.
- In line 292. What do you mean by left part of the wake? Positive or negative y?
- In figure 9, reference to iso-surfaces.
- In figures 9, 10, 15, 16, 19 and 20. Include the coordinates in every axis.
- In equation 3. Probably there is a typo. Is it non-dimensionalization by area? Which is the integrated area?
- Please define the differential wake deficit mentioned in the caption of Fig. 12.
- In Figure 13. Please indicate clearly the frequency of movement, the blade passing frequency and the frequency of the rotor.
- Please explain why the results from Fig. 13 are far from symmetric.
- Show that the PSD converged. Convergence of the mean values (discussed in section 3.4) does not imply convergence of all statistics.
- Do the results from Fig. 13 change as the position move downstream? Can the same peaks in frequency be observed downstream?
- Please show the PSD at 5D.
- In lines 395 to 397. The explanation for the presence of a peak in 1 Hz for the fixed turbine is not clear. Please provide a more detailed explanation.
- In section 4.4. Please show figures similar to 19 and 20.
- In Figure 29(b) and page 27. The definition of ΔWD is not clear.
- Figure A1. Caption is probably incorrect.
- Appendix A. Where was the velocity sampled?
Citation: https://doi.org/10.5194/wes-2024-169-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 179 | 56 | 7 | 242 | 8 | 7 |
- HTML: 179
- PDF: 56
- XML: 7
- Total: 242
- BibTeX: 8
- EndNote: 7
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1