the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Aug 2025
Wind-tunnel analysis of wake-steering control strategies on a multi-column model wind farm
Abstract. Wake-steering control has the potential of improving the power production of wind farms by deflecting the wakes of upstream turbines away from the downstream ones, thereby increasing the velocity impinging on the latter by sacrificing the performance of the former. In this work, a wide range of wake-steering control strategies are systematically applied to a 3 × 3 wind farm in a series of wind-tunnel experiments. When each streamwise column is operated identically to the others, the maximum measured power gain is approximately 5.3 %. It is observed that the columns respond differently to a given yaw configuration, with the central one generally improving to a smaller degree than the lateral ones. Nevertheless, our data indicate that tuning each column independently of the others does not result in further power improvements. Furthermore, we show that increasing the free-stream velocity enhances the baseline power production of the model farm but reduces the scope for improvement achievable with wake-steering control.
- Preprint
(1722 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2025-135', Anonymous Referee #1, 21 Sep 2025
The manuscript investigates yaw configurations in a 3×3 turbine array exposed to a sheared, turbulent inflow in a wind tunnel. The dataset is extensive and carefully obtained, with clear attention to ensuring realistic inflow conditions. In general, the manuscript is well-written, interesting, mostly easy to follow, and well-situated within the existing literature.However, the following issues, two of them major, require clarification before publication.Major concerns1. Representation of an infinite wind farm through wall proximityLines 194–195: I disagree with the claim that the presented setup resembles an infinite wind farm. The lateral turbine columns are positioned very close to the solid wind tunnel walls. This proximity likely constrains the natural wake expansion of an uncontrolled wake and artificially limits wake deflection under yaw. This raises the following questions and concerns about the validity of several conclusions:1.1 What is the blockage ratio of the setup?1.2 If the setup were representative of an infinite wind farm, why do the left and right columns react differently to yawing of the first row turbine (lines 291 ff.)? In an infinite wind farm, all columns should behave equally.1.3 Section 3.1.3 (particularly lines 331–333) contradicts the infinite-farm hypothesis. The authors state that lateral columns benefit from lateral entrainment, yet by definition in an infinite wind farm lateral entrainment does not exist, because each column is neighbored by two other column.1.4 In Fig. 18, why does the relative performance of the columns change so strongly with wind speed (e.g., the center column is the worst at 7 m/s but the best at 8 m/s)? Could this be an artifact of blockage and local acceleration effects from insufficient spacing to the tunnel walls?The conclusions would be significantly strengthened if the influence of the side walls were assessed, for example, by repeating experiments with only a single turbine column placed at left, center, and right positions. This would clarify the wall’s impact, especially when wakes are steered toward it.2. Asymmetry in yaw steering performanceThe asymmetry in power generation between positive and negative yaw angles (Fig. 7) is surprising and highly relevant for ongoing research and recent efforts in commercial deployment of wake steering. If wake steering benefits only from one yaw direction, the concept itself requires reevaluation. This point needs more careful discussion:2.1 Lines 277–282: The authors attribute the asymmetry to lower tip speed ratio compared to real turbines, causing enhanced wake swirl. To what extent do these results translate to utility-scale turbines? Should we expect directional dependence of yaw steering benefits in practice?2.2 How sensitive is this asymmetry to inflow conditions? Would a differently sheared inflow produce the same result?2.3 Fig. 11c: The proposed explanation of the linear trend due to vertical wake displacement is not convincingly demonstrated. Can this be substantiated with wake measurements or LES?Further concerns3. Graphical presentation3.1 Several figures use the label “config Nr” on the x-axis. While interesting trends are presented, this label does not convey the underlying physical parameter being varied, making interpretation difficult. Replacing “config Nr” with a meaningful physical variable would improve clarity.3.2 Many results are shown as scatter plots with both marker color and shape encoding additional variables. This creates very dense visualizations where trends are difficult to see. Moreover, continuous colormaps for discrete yaw angles are confusing. Alternative formats, such as contour plots with the row-wise yaw angles on x- and y-axes, and performance metrics (Cp, thrust, etc.) as a colormap, may more clearly highlight the trends. Additionally, varied parameters (e.g., γ3) could be separated into dedicated subplots.Minor commentsLine 3: “wide range of wake steering control strategies” is vague; consider “wide range of yaw steering configurations in a wind farm.”Line 40: The abbreviation “LES” should be defined.Fig. 6: Why is Cp,opt=0.325 so much lower than the Betz limit of 0.593?Line 147: The use of gradient-based optimization for each yaw setting makes sense scientifically, but how realistic is this for actual wind farm operation? Could comparable methods be applied in practice?Fig. 14: Appears redundant with Fig. 13, given the correlation between power and thrust.This study addresses a relevant topic, and the dataset has significant value. However, before acceptance, the concerns outlined above—particularly regarding wind tunnel wall effects and yaw asymmetry—must be resolved.Citation: https://doi.org/
10.5194/wes-2025-135-RC1 -
RC2: 'Comment on wes-2025-135', Anonymous Referee #2, 23 Nov 2025
The submitted manuscript presents the results of an experimental campaign where a 3x3 cluster of turbines are yawed at various angles. The experimental setup is described in detail and no context or information seems to be missing in the text. The results are mostly presented well and show interesting findings. Although the present manuscript makes it difficult to draw concrete conclusions for turbine arrays, the yaw-asymmetric findings for turbine columns are interesting and might warrant further investigation. It is difficult however to draw conclusions for turbine arrays, because of the high irregularities between the lateral columns.
I think it highly likely that these irregularities are caused by unintended wall-interactions, due to the lateral turbines being placed only 1.33D from the wall. Depending on how the wall-boundary layer evolves, the steered wakes probably heavily interact with this layer. The assumption of neighbouring specular farms on line 137 regarding this is rather questionable. An argument might be made if the neighbouring farm was mirrored perfectly, but this falls apart when the yaw-asymmetry presented here is taken into account, since this makes mirroring impossible.
Could you please provide further explanation on how the wall-nearness would not cause measurement problems?
Furthermore, the authors show many scatter plots that could perhaps be better visualised as heatmaps for easier interpretation.
Other comments and suggestions:
1. The literature review in the introduction might benefit from a table summarizing the previous works. Sec.4 could then refer back to this as well.
2. Fig.1: for clarity, perhaps color/shade/texture the different components and add a legend or define them in the caption. For this and all figures, place them closer to where they are referenced.
3. Please refer to Micheletto et al. (2023b) in the caption of Fig. 2 as well and make it clear in the caption that the data presented there is not from this campaign.
4. Fig. 3: The data points should extend to the lateral rotor tips instead of the hub. Perhaps add another subplot of the same size of a x-z slice with a turbine sketch and horizontal lines labelled a/b/c/d/e
5. How did you calibrate the zero position of the yawing stepper motor? Additionally, how did you ensure that the zero position aligned with a step position of the stepper? In other words, how did you ensure that your target of 0deg resulted in 0deg and not something on the interval [-0.9, 0.9]? You do mention this briefly in Fig. 11.
6. Fig. 5 shows a tower with a very high roughness. Can you comment on how this could show up in your measurements?
7. Lines 240-246: It would have been interesting to replace C2 with a tuft at hub height to see if the local wind direction changed visibly.
8. For all scatter plots, please use a higher aspect ratio to space the markers out more: I'd say use up to 80% of the text width. For Fig. 7 I would suggest swapping the axes too.
9. In Fig. 7, could you include the LLS fit of each column separately too?
10. At the end of line 343 you state: "yawed in the opposite direction of the turbines upstream." However, in EXP2 there are Type 2 configurations according to table 1, meaning that C3 is yawed in the same direction as C2?
11. Sec. 3.2: For consistency, it makes more sense to compare Ct values than thrust values. Since you use a constant hub height wind velocity, Ct and T, and Cp and P are just scaled pairs. Therefore, I would advice to either present Ct and Cp results, or T and P results, but not mixed.
12. Sec. 3.3: Consider defining an EXP3 collection and including a row of Uinf values in table 1.
13. Is there a reason you only tested increased free stream velocities, but not decreased?
14. h and l are defined in the captions of figures, but not in text.
15. Typesetting of h on line 104.
16. Table 1. bottom right cell is formatted differently, I suggest -5.4:5.4:5.4 for consistency.
17. Fig 13d is not labeledCitation: https://doi.org/10.5194/wes-2025-135-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 613 | 49 | 29 | 691 | 40 | 49 |
- HTML: 613
- PDF: 49
- XML: 29
- Total: 691
- BibTeX: 40
- EndNote: 49
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1