Dear Authors,

please carefully consider the comments of all reviewers, and especially the ones from reviewer #3 who has found a number of important shortcomings in the paper. Your revised submission will be reviewed once more.

Thank you, and kind regards
Carlo Bottasso

Dear Authors,

please carefully consider the reviewer's comments. I am indicating "minor revisions", but please take into account that some of the reviewer's concerns are of major importance. Therefore, please make sure to fully address those concerns in the revised manuscript.

Best regards
Carlo Bottasso

Dear Authors,

thank you for your replies and the updated manuscript. However, the article still falls short of fully addressing the reviewers' comments. I recommend that the manuscript is further modified as follows:
- Concentrate the discussion on ultimate conditions, which is the novelty of your study. This will improve readability and should drastically simplify the discussion, which at the moment is confused and difficult to follow.
- Move the discussion on fatigue to a less prominent position (for example, in an appendix). Better highlight the limits of your results on fatigue (which might increase or decrease, depending on the site and specific location of a turbine in a farm).
- Use design as a way to preliminarly quantify the effects of changes in ultimate loads and tip deflection. Even if fatigue is approximate (see above), still the re-design exercise can have the role of indicating the impact of the changed ultimate states. Clearly indicate the qualitative nature of the re-design and its limits.
- Reflect the new focus also in the title, abstract and introduction.
- Avoid claiming that the study is general because you look at sensitivities. Clearly your results still depend on the specific turbine that you have selected.
- Eliminate statements to the effects that loads are overlooked (abstract), because clearly loads cannot be overlooked by industry given the obvious implications on safety.
- Please review all the discussion from line 128 to the end of Section 2. I cannot follow the logic, and I suggest to eliminate it altogether.
- The discussion on directional dependence of fatigue (from line 259 onwards) is rather obvious, please reduce or consider eliminating.
- "Downwind turbine" might be misunderstood, use downstream or reword.

Best regards
Carlo L. Bottasso

Dear Authors,

Thank you for submitting a revised version of your work.

The new version of the manuscript is certainly much improved with respect to the previous one. However, before publication the following points need to be addressed:
• Page 4 “Unfortunately, regulations, in their current status, do not consider yet the fact that a turbine may operate out of the design conditions according to a farm control.”. This is not entirely accurate, and you might want to consider this position paper: https://www.windfarmcontrol.info/-/media/Sites/WindfarmControl/Publications/FarmConners-WP2-D2-1-Position-paper-on-WFC-certification_revised.ashx?la=da&hash=E5801009B447D914DD7A566C9FCA9426B4483C9D. In addition to considering wind farm control, the paper also explains well how in-farm inflow conditions are currently addressed by the standards. Material from this paper can help improve the introduction and methodology sections.
• Page 5, “1D geometrically exact …”: I understand what you mean, but the sentence might be misunderstood. The model is 3D, and I believe that the term “beam” already implies what you are referring to.
• Page 7: is there any possible failure mode of the wind farm control laws that should be considered? If not, please explain why.
• Page 9 (but also elsewhere, wherever you refer to AEP): AEP is not a meaningful KPI in this context, as a turbine will never work with the same misalignment angle for a full year. Power losses for a steering turbine are typically quantified through the cosine law exponent, not AEP (unless you are looking at a specific site, wind rose, farm layout, etc.). The power loss exponent however has a large scatter, and very different values have been reported in the literature. It is doubtful that your (and most other) BEM models would be able to accurately predict this effect, unless high yaw misalignment corrections are implemented and calibrated. I suggest that you simply refer to the literature on this point, which is also not central to the present discussion.
• Page 9 (but also elsewhere, wherever you refer to ADC): the previous comment applies also to ADC: even if you use Weibull weighting, it is not reasonable to assume operation for a full year at a given misalignment angle. Additionally, you refer to pitch ADC, but one should first have a discussion on yaw ADC (which is the main concern in practice, since it grows significantly, while pitch ADC will tend to decrease).
• Page 11, end of sect. 4.1: this is also the reason for one-sided wake steering (in addition to the much reduced effects on yaw ADC), which has been consistently used in most of the field tests conducted so far.
• Page 14: please note the strange apparent inconsistency: here you look at the ADC of the driving input (pitch), while for wake steering you look at the ADC of pitch, although the driving input in this case is yaw. As noted above, pitch ADC will decrease, but yaw ADC will very much increase.
In addition, here again the WF controller will not be operating all the time, so this KPI does not seem to be very useful. As the work focuses on ultimate loads, I suggest that you eliminate references to AEP and ADC: as for DELs, any analysis of these KPIs requires creating first a reasonable scenario, because WFC is not used all the times. In my opinion, as for DELs in the previous version of the manuscript, references to these aspects unnecessarily complicate the discussion and distract from the main point.
• Page 16: you decided to focus only on 2 deg amplitude. But is it enough to achieve a useful increase in wake mixing? Please justify this choice. What happens if we really need 4 deg to achieve some power boost?
• Page 17: “… have an impact on ultimate loads, especially on the maximum tip deflection, …”. I understand what you mean, but tip deflection is not a load. Please rephrase.
• Page 17, lines 389-392: I am not sure I understand the motivation not to consider wake steering in the redesign exercise, since previously you said that it causes a significant increase in tip deflection. When you say “The design process of the baseline generated an optimal solution compliant to all optimization constraints, with a structure mildly different …” are you talking of the redesign accounting for wake steering? Please rephrase, this is unclear.
• Many sentences are long and a bit convoluted. Please do not rely only on the language editor to improve the language during production, and try to shorten and simplify the text wherever possible.

Best regards
Carlo Bottasso