Mitigating wind farm cluster wakes by vertical staggering
Abstract. The expansion of offshore wind energy results in more wake influences between wind farms and wind farm clusters, which should be mitigated to reduce negative impacts on the power production. In this study we assess whether vertical staggering of wind turbine heights within a wind farm could mitigate farm wake interaction between two neighbouring wind farms. The aim is to understand how wake recovery mechanisms and three-dimensional wake characteristics differ between vertically staggered and non-staggered wind farms. We performed large-eddy simulations of a pair of an upstream and a downstream wind farm in a conventionally neutral atmospheric boundary layer and compared several vertical staggering layouts at the upstream wind farm to the non-staggered reference. To assess differences, we analysed mean wake characteristics, the three-dimensional flow field, the wake recovery budget, and power production. We found that vertical staggering hindered the mean wake recovery of the horizontal wind speed at the hub height of the lower turbines by reducing momentum at higher altitudes. Investigating spatial wind fields revealed a more heterogeneous, narrower farm wake with streaks of increased wind speeds for vertically staggered wind farms. The wake differences were explained by the formation of larger-scale advective flow structures for vertically staggering cases and their larger contribution to the mean wake recovery compared to turbulent mixing. A circulation zone at the northern wake edge was present for staggered and non-staggered setups, which caused a streak of increased wind speeds in the lower half of the boundary layer. The wake heterogeneities persisted far downstream and impacted the wake and power production of the non-staggered, downstream wind farm, causing locally increased or decreased power. The total power production of the vertically staggered upstream wind farm was strongly increased due to higher wind speeds at the taller turbines, a lower rotor density per height, and a better wake recovery at lower hub heights inside the wind farm To conclude, we showed that vertical staggering has the potential of locally improving the farm wake recovery by creating or enhancing large-scale advective flow structures, influencing the wake and power of a downstream wind farm. The advection-driven recovery of the far farm wake is relevant for wind farms beyond vertically staggered setups, requiring future analyses to focus on production mechanisms of large-scale advective structures.
Review comments on wes-2026-140
This paper presents a numerical study of wake developments for a single wind farm and two wind farms in tandem, using large-eddy simulations (LES) of a conventionally neutral boundary layer (CNBL) over an array or two arrays of actuator disks. The key novelty of the study is that the effects of vertically staggering the turbines are investigated in detail, mainly focusing on the formation of large circulation zones/streaks induced by the vertically staggered turbines and how they impact the farm-scale wake recovery.
Overall, the numerical results presented are interesting and informative, helping us better understand the mechanisms of farm-scale wake development and recovery, and the manuscript has been written very carefully, although some descriptions of simulated flow fields are a little redundant. I would prefer more concise descriptions of the simulated flow fields (e.g., some descriptions of the details of wind profiles could be simplified when they are obvious from graphs) but I think most of the results presented in this paper are still worth publication in Wind Energy Science. I have a few comments and suggestions, which might help the authors improve the paper before publication.
1. In this study the turbines are “modelled with an actuator disk model with rotation” (Line 161) but the authors do not mention explicitly in which direction each disk rotates. Since there are some previous studies on the impact of turbine rotation direction on wake developments (e.g. Englberger et al. 2020, https://doi.org/10.5194/wes-5-1359-2020; Nouri et al. 2020, https://doi.org/10.1016/j.apenergy.2020.115511) I wonder if such turbine-scale physics may be relevant to the formation and strength of the large circulations discussed in the present paper. It would be good if the authors could (at least briefly) discuss this point.
2. In this study the authors make comparisons of farm-scale wakes for different farm configurations (i.e. vertical arrangements of turbines) but I wonder if these comparisons could be presented in a better manner. When we make a comparison of turbine wakes for different turbines, we almost always want to check the turbine’s thrust coefficient and axial induction factor first, because if the thrust coefficients and/or the induction factors are different (or unknown), it would be difficult to make a fair comparison of turbine wakes. Similarly, to make a fair comparison of farm-scale wakes, it would be important to first check, at least, the total thrust of each farm for different farm configurations. The power production data in Section 3.3 are useful, but I think the thrust is more important than the power here (for comparison of wake development and recovery). If possible, it would be useful to also check the farm induction factor (see e.g. Nishino and Smyth 2026, https://doi.org/10.1017/flo.2026.10047) by calculating the farm-average wind speed across 30m < z < 270m (baseline) and 30m < z < 510m (case S1), for example.
3. The authors “applied a low-pass filter along the streamwise direction” (Line 254) to remove the oscillations related to gravity waves. I understand that this filtering was helpful/necessary to distinguish the effects of vertically staggered turbine layouts from the effects of gravity waves, but I do wonder what the original flow fields (including the effects of gravity waves) looked like. Would it be possible to present one example of the original flow data (including gravity waves) such as the wake recovery budget (like Fig. 5) in this paper, perhaps as an appendix? This might be more useful than the current appendix A2 on TKE contours (about which no text has been provided).
Other minor comments:
Line 173: The grid spacing (20m x 20m x 10m) has already been given in Line 167.
Line 187: What do you mean by “meaningful statistics” here? A little more explanation would be helpful.
Line 204: I guess what you mean by “wind farm power “ here is the rated capacity rather than the actual power production.
Line 260: What do you mean by “moving average” here? A little more explanation would be helpful.
Line 480/481: “This results for the vertically staggered wind farms in absolute deficit differences by up to 30 % at the end of the domain.” - This sentence was confusing to me. I guess “for the vertically staggered wind farms” should be placed at the start of this sentence instead.
Line 604: The definition of the Rossby number should be given here.