The topic addressed by the authors is relevant to the journal in terms of wind turbines and noise

reduction strategies, but the work and the methodology has several strong short comings and lack

of information which prevent me from accepting this manuscript in its current form and requires

several major revisions. Please find attached the main general comments of concern, followed by

specific comments.

The paper addresses an interesting topic but I feel like it misses a validation of the baseline simulation first. This should be done before proceeding to testing mitigations in the simulation. 

comments:

-line 175: why is the outer volume a cylinder? I had expected a rectangle with a wall boundary condition to represent the ground/building on which the turbine is installed. 

- why is dynamic mesh used and not moving mesh? Is the mesh quality around the airfoil good during all time steps? Can you please include plots showing details of the mesh around the airfoil? In case moving mesh is used, one can control better the quality of the mesh around the airfoil

- what is the courant number you have with the selected time step?

- the near-wall region and outer-flow region were not refined/made coarser  to confirm mesh convergence. I think this is a necessary step to be sure about the convergence. Also a convergence study on the time step is necessary.

-the acoustic power plot doesn't have 3 similar periods. Do you have an explanation for that?

small comments:

- 1st sentence of the abstract: "Wind turbines are a promising solution for sustainable energy": I suggest to narrow 

it down to Small vertical-axis wind turbines. The noise reduction methods for horizontal wind turbines are different 

than what is described in the following sentences.

- line 33: these numbers don't match the typical formula used in IEC61400-11.

This manuscript presents an investigation of aerodynamic noise emission of a small vertical-axis wind turbine in different configurations designed to alter and reduce noise emissions.

For the most part, the article is clearly written. However, the study of aeroacoustics is conducted with tools that are not well suited for the refined analysis required for estimating noise emission reductions in details and consequent environmental impact. The paper also suffers a number of deficiencies that are addressed following the sections' order below.

The introduction starts with generalities about wind turbine noise. The literature review appears a bit random and disorganized.

In Section 2, the numerical models are presented. The flow solution is based on a commercial software solving the compressible Navier-Stokes equations, with a standard RANS approach for turbulence modelling. In this respect, the unsteady term is missing in Eq.(2). The acoustic part is based on an integral acoustic analogy derived from Curle analogy. The integration surfaces for the acoustic calculations are on the blade surfaces. Note here, that strictly speaking, Curle's analogy is only valid for non-moving surfaces. To the authors' credit, in the present case, it may be a valid approximation though. A "correlation area" is introduced as a multiplicative factor in Eq.(11), but it is never numerically defined. Since this factor is related to wall-pressure correlation lengths, its value has a direct and significant impact on the noise emission levels. The accuracy of the overall model can then be questioned if it is not properly defined based on some physical considerations. In addition, the overall methodology for acoustic predictions seriously lacks qualitative validations, or at least references to it.

Furthermore, as shown in the results, the model provides only an overall (frequency-integrated) sound power level. There is no indication on how this energy is distributed over frequencies, which is a crucial aspect for environmental impact of wind turbines (or any other industrial facility for that matter). Not even an indication on A-weighted noise power levels can be obtained.

To make things worse, the model as it is, can most probably not capture scattering effects, which should be the dominant noise source in this context (at least in the audible frequency range). It may be that the present model can capture vortex-blade interaction noise. However, the use of a URANS strategy will limit the size of the vortices to the very large ones, resulting in sound emissions that would probably be in the infrasound range. It can be expected that this has little relevance in the present context.

In Section 2.4, the numerical implementation and discretization of the CFD model is presented. It is classical for these type of calculations to provide the size of the cells at the surface (for assessing mesh refinement) in term of y+ (as mentioned by the author in the introduction). But, this is never done here. Only cell sizes in meter are provided in Table 1.

Section 3 starts with a mesh convergence analysis which appears satisfying. The remaining of the investigation concentrates on noise emissions which is lacking a number of information to make it valuable. The comparisons concentrates on the impact of the geometry for noise reduction on torque and acoustic power. This is indeed a suitable approach as indeed, acoustic reduction should affect overall aerodynamic performances to a minimum, which should be addressed in the design phase. Nevertheless, the lack of physics in the noise modelling (see earlier comment about Section 2) does not permit to draw firm conclusions on the impact of the different designs, as the noise reduction (or increase) may occur at frequencies that are not relevant for environmental purposes.

As an additional comment, the authors use a very unusual metric for displaying the noise levels, namely watts. Note values up to 10^5 W in Fig.3 which is of the order of the noise emission of a turbojet engine....

To conclude, the paper presents an analysis of VAWT noise emissions which is questionable regarding its relevance for assessing environmental impact. Nonetheless, if interested in very qualitative acoustic results, that may be satisfying, e.g. if one is mainly interested in aerodynamic performances of the acoustic reduction devices. In the reviewer's opinion, the acoustic model strategy should be improved/enhanced, and noise spectra should be provided as a result, before this article can be considered for publication.