This manuscript shows the results of a layout optimization exercise for three floating wind turbine sites. The mooring system and initial power cable designs were developed in previous work, such that the present work only addresses the layout optimization. The assumptions are clear and the results are believable given the assumptions and approach, but the scientific contribution of the work is not very clear to me. What research questions does this work address? The novelty of the procedure only seems to be the need for buffer zones which would not generally be considered for a bottom-fixed layout. The procedure for cable routing is also already described in previous work, so it is not very clear to me what is new here.

The premise of the optimization is that a uniform grid layout shall be obtained for a uniform water depth for each location. The procedure assumes that the mooring and cable design are already selected, while only layout optimization is performed here. It is unclear to me how this division can be included in a practical design situation, as the water depth in general will vary throughout the farm and require modifications to the mooring systems. Furthermore, in Fig. 1, it seems that the cable sizes are input to step 4 separately from the cable design in step 3 – aren’t these taken from step 3?  

There are also some notable shortcomings in the calculation of both the objective and constraints in the layout optimization. For the objective function, the AEP calculation is very sensitive to the number of wind directions considered, as the authors recognize. It does not appear that the sensitivity to this selection has been assessed for either the optimization process or the final designs. For the fatigue and extreme response criteria, the frequency-domain tool RAFT has been applied. Has the applicability of a linearized tool for 500-year response conditions been assessed? The conditions for linearization are not likely to be met in extreme conditions. Furthermore, the choice of extreme weather conditions, particularly for the GoM/GoA site which is expected to be dominated by hurricanes, deserves a bit more attention. (I have not thoroughly reviewed the reference which provides the background for these choices, but the Hs, Tp, and Uw for the SLC seem a bit low compared to what I might expect in a hurricane-prone area).

The optimization algorithm is described in limited detail. A PSO approach was used, but the number of particles, constraint handling, and convergence criteria are not described.

The writing is generally pretty clear, though it could be a bit less colloquial in some instances (“pulled directly from…”) and tense can vary in some parts of the manuscript. There can also be some confusion between what is meant by “we chose” vs. “the optimizer chose”.

Dear authors,

first of all, I would like to apologize for the late review, which was due to a problem with system, as indeed I revised the paper quite a long time ago. Please accept my apologies.

The paper is interesting and will contribute to the research in floating wind, especially in the current context of partial loss of momentum from industry due to the modified economic framework.

Please find below a few comments for your consideration:

1) I see that the paper is registered as a “research article”. While this has no great drawbacks, I invite you to consider switching it to a “Data description” paper since I think this fits better the nature of the study and would put it in a better spotlight. Maybe this can be a topic of discussion with the Editors.

2) Is there any additional info on the real bathymetry in the site and the type of seabed available? You clearly stated that these are not accounted for at this stage, but adding such info (e.g., in the GitHub repository) could be useful to other researchers willing to further optimize the layouts in the future

3) Another piece of information that could be very useful to maximize the future exploitation of this study is represented by more detailed metocean conditions, including waves and wind-wave misalignment. The selected metocean conditions would deserve more attention also in the paper, as they are not addressed in a fully convincing way. If not fully measured, you could reconstruct them using for example the approach proposed in DOI: 10.1088/1742-6596/2385/1/012117

4) Selected cost should be put in context, ideally providing more references and realistic “ranges” of variation of such costs for the time being