This paper presents a novel idea which aims to increase the AEP in a wind farm, by forcing the advection of the atmospheric wind into the wake to reenergise the incoming flow for the next wind power generator. This paper presents preliminary RANS simulations using actuator elements to represent the wind turbines.

 While I have doubts that this type of system would be largely deployed in the future (due to structural integrity, robustness, reliability and control), the concept is original and it's worth doing this thought experiment.

The paper is globally well written and relatively easy to follow, despite some convoluted turns of phrase. The simulation results are interesting and seem possible to be reproduced as the settings and the code are apparently available.



I have mostly one main concern and one point that troubled me.

1. My main concern is the possible large blockage of such a system. When I look at figure 1, I see quite a lot of projected surface compared with more conventional wind turbines. I would imagine that not all the flow would pass through the system and would rather deviate and go around the system. If the flow deflects, it means that there is a smaller mass flow rate through the system, so less energy can be extracted. Could the authors confirm that in their RANS model, the flow deflection is correctly modelled? For example there are assumptions to define u^ele_ls and u^ele_in. May this assumption impact the mass flow passing through the system?

2. The point that troubled me is the terminology with "upward-lifting" and "downward lifting". To me, it seems the upward wind in the wake is due to a downward lift (for example in figure 3, the suction side of the airfoil points downward, so the lift is directed downwards, but it would create an upward wind). This does not impact the results of the paper, but I was doubting if I understood the concept correctly. Could the authors confirm or correct my thoughts and better explain and define this concept in the paper?

I have a couple of other minor remarks:

3. Why did the authors choose this airfoil for the lifting devices?

4. A Turbulence Intensity of 8% seems fair, but it could be much higher in the reality. As it is mentioned the Turbulent Kinetic Energy plays a minor role, I would interested to know whether the conclusions still hold with a higher TI (such as 20%).

5. Similarly, the difference of results between the different turbulence models seem quite large. Could the authors precise what could be the reasons for such a large difference between the k-omega and k-epsilon models?

The article describes a multi-rotor wind energy system with static lifting devices, aimed to increase the momentum entrainment and mitigate wake losses.

The paper is structured well, and the methodology is mostly clearly presented. Perhaps the paper could be shortened by not spelling out every well known concept, for example the RANS equation system with k-omega turbulence model (eqs. 1-4).

Some general remarks below.

It is not clear how the MRSL appears in the modeling grid. The name indicates several rotors, but Figure 4 indicates one (square) rotor with the diameter D=300m, and 186m hub-height. 

Nothing is said about the system integrity and loads on the structure. How does such a system turn into the wind? It could also be assumed that it is fixed and suitable for uni-directional wind climate. In this is the case, please state. While this device appears entirely conceptual and will highly unlikely ever be used at scale, it is however attractive to find an engineering way to capture some of the potential to double the energy density in large wind farms (e.g. Table 4). Have you perhaps tried to add the 5th wing at the lower edge of the system? What about if the wings are installed separate from the multi-rotor structure?

There are many acronyms in the article, but downward-lifting and upward-lifting are for some reason spelled in full over 100 times. Suggest using DL and UL instead.