the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 May 2024
Wind farm layout optimisation including meandering correction and momentum conserving superposition
Abstract. This study introduces a combined analytical approach, to simulate the wake flow in large-scale offshore wind farms and forecast their power output. The developed tool, Qwyn, integrates the Ishihara-Qian single-wake model, a momentum-conserving superposition method by Zong & Porté-Agel, and a wake meandering correction by Braunbehrens & Segalini. A validation against both measurement data and Large Eddy Simulation (LES) results at the Horns Rev 1 (HR1) wind farm is performed. This demonstrates the tool’s capability in capturing the wind farm flow and power output for different wind directions. Notably, the momentum-conserving superposition method significantly enhances the accuracy of power prediction for narrow directional wind bins, while the wake meandering correction improves precision for wider bins. Subsequently, the validated computational tool is used to optimise the layout of HR1 to enhance its annual power production (AEP). By introducing a convexly shaped layout a projected 0.12 % increase in AEP when optimising for a wind rose with 72 wind bins is found. Furthermore, this study shows, that omitting the meandering correction leads to an even more convexly shaped layout without substantial change in AEP improvement.
- Preprint
(16815 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2024-62', Anonymous Referee #1, 16 Jun 2024
The authors introduced a novel analytical framework (Qwyn) for the simulation and optimisation of wind farms. The proposed tool has been applied to the Horns Rev 1 wind farm, showing promising results. The effect of the selected wake superposition and wake meandering models on the predicted power output has also been outlined.
The reviewer believes that the topic and the activity are very interesting, innovative and worthy of investigation. The study is extensive and carried out with rigour and a consistent approach. Both methodology and results are presented in a clear and concise way. A few consideration to further improve the quality of the manuscript:
- Introduction: as the authors are introducing a new simulation framework, it would be interesting to add an overview of the tools currently available, highlighting their characteristics, advantages and drawbacks. A comparison table would be very useful in this sense;
- Introduction: please add a small paragraph at the end to briefly outline your work;
- Figure 1 seems redundant, I would remove it for conciseness;
- Conclusions are too discursive, the use of a few bullet points highlighting the main takeouts of the study would make them clearer for the reader;
Citation: https://doi.org/10.5194/wes-2024-62-RC1 -
AC1: 'Reply on RC1', Daniel Sukhman, 22 Jul 2024
Thank you very much for your positive review on our manuscript.
We value your feedback and ideas to improve our manuscript further:- We will consider adding such a table, as suggested. However, a comparison table would be most interesting, especially when comparing the performance and characteristics of available tools on the same hardware. Such a comparison would require a separate study. We are therefore inclined to add this to our list of future work, to ensure a sophisticated conclusion.
- We will adjust the introduction accordingly.
- Previous presentations have shown that a figure clarifying indices often helps reduce confusion when following the given equations. Therefore, we would like to retain this figure in our manuscript.
- Thank you very much for this feedback. We will amend the conclusions as you suggested.
Citation: https://doi.org/10.5194/wes-2024-62-AC1
-
RC2: 'Comment on wes-2024-62', Anonymous Referee #2, 17 Jun 2024
- AC3: 'Reply on RC2', Daniel Sukhman, 22 Jul 2024
-
RC3: 'Comment on wes-2024-62', Anonymous Referee #3, 18 Jun 2024
General comment
This manuscript presents a methodology to develop a tool that can be used for predicting the annual power production of a wind farm. The tool is based on the parameterization of the flow inside a wind farm as a function of the wake characteristics of individual wind turbines. The objective is to achieve a model that could be used to optimize the layout of a wind farm to increase the power production. In general, the manuscript is well written and structured.
- My main question regarding the analysis presented in the manuscript is the reason of performing this study in only two dimensions. On which basis the estimation of the thrust is performed by integrating only along the spanwise direction?
- Furthermore, how are the wake model parameterizations as a function of the thrust coefficient and the turbulent intensity derived (for example Eq.3 and 6)?
- Additionally, there is no information about the LES data used for validation, besides the text in line 355. Please add more information regarding the LES data the first time that it is mentioned in the text.
- Finally, the optimisation results show that a modification of the layout of the wind farm could result in an increase of 0.12% and 0.10%, depending on if the wake meandering is considered or not. What is the reference used here? Does the black dashed curve correspond to meaturements. Moreover, what is the estimated AEP using the proposed tool, the result of which is presented in Fig.11? I am asking this because the proposed tool is overestimating the normalised power output at the reference wind farm layout for most wind directions.
Specific comments
- Line 64: Which LES data was used to fit the wake expansion parameters of Equation 2? Please either provide a reference or present in detail the LES data.And how do the authors reach to two expressions of Eq. (3)?
- Lines 75.—76. How are the far and near wake regions defined here?
- Line 129. Why does the integration of Eq.(22) take place only over the spanwise axis?
- Lines 253-255. How is it assessed that those wind directions lead to worst case operating scenarios? Please either explain more (I guess the results from Figure 11 are used here) or add a reference.
- Lines 260-264. The wind farm data correspond to mean values over 10-minutes?
- Line 280, Table 2. How is the data uncertainty defined and how it is estimated?
- Lines 298 – 306. Why the impact of the wake meandering is more evident when a wider wind direction sector is selected? I guess when the wind direction sector is wider the wake centre is found in different locations along the rotor, but the wake meandering is the same.
- Lines 326 – 327. It is written that “the wake meandering correction performs worse when no averaging of wind bins is done during modelling”. I guess what is meant is that models of the normalised power presented in Fig. 11 correspond to estimated values using 1-degree bins. Have the authors tried to plot this figure using 5-degree bins?
Technical corrections
- Line 47. Replace “hubheight” with “hub-height”
- Lines 102-103. Shouldn’t the regions inside and outside the wake be defined by the absolute value of y?
- Line 234. The zeta parameter has values m/s. Please add them after the 0.001.
- Line 239. Rewrite “wind farms” to “wind farm’s”
Citation: https://doi.org/10.5194/wes-2024-62-RC3 - AC2: 'Reply on RC3', Daniel Sukhman, 22 Jul 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 405 | 83 | 30 | 518 | 23 | 29 |
- HTML: 405
- PDF: 83
- XML: 30
- Total: 518
- BibTeX: 23
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1