**5. Conclusions**

In this work, a CFD model based on the MRF approach was developed to assess wind turbine far wake characteristics according to operating conditions typically experienced in commercial wind farms. The influence of the TSR and free-stream wind speed on wake characteristics such as velocity deficit and TI was discussed and compared with the existing literature on this topic.

This paper reviewed most of the wind turbine wakes studies and wind farm CFD techniques from the literature. Overall, we found that the existing literature studies use different turbulence modeling techniques, as well as CFD solvers with different assumptions and boundary conditions. The wake results vary according to the approach adopted in each work. A gap was identified in the literature review of this work, showing that there is a need for more development of CFD models capable of simulating a whole wind farm. The vast majority of the CFD studies simulate single turbines, and only a few of them simulate more than one rotor. The computational resources may be a limiting factor for that, representing one of the biggest challenges on wind farm computational modeling: The expensive computational resources required to simulate several rows in a wind farm. In order to address this need, this work presented a novel methodology to analyze wind turbine wakes interaction with relatively reduced computational resources. The technique had never been applied before in the

context of wind farm numerical modeling. Even though multiple simulations are required for studying the interaction effect between upstream and downstream rows in a wind farm, this work successfully achieved a reduction in computational capabilities (processors) required to perform wake interaction simulations. This represents an advance for wind farm modeling, and many researchers could benefit using such techniques to improve wind energy CFD models.

The model presented in this work was previously validated in part I [56] with regards to near wake data. In part II, a verification of the model against other studies in the literature showed consistency in the wake results within acceptable levels. In regards to the velocity deficit and TI assessment, the values found in this work were similar to other CFD wake studies in the literature. This demonstrates the ability of the proposed CFD model in predicting wake characteristics, and this way the model is ready to be applied for determining the optimal spacing between turbines in a wind farm. The capability of the proposed CFD model showed to be consistent when compared with field data, kinematical models, and CFD results from the literature, showing similar ranges of wake deficit.

Further improvement of the model will include a transient approach modeling to determine wake characteristics according to variable rotor operating conditions. This will extend the capabilities of the proposed model by adding a more realistic modeling approach to derive the aerodynamic behavior of the turbine rows. Moreover, a FSI (fluid solid interaction) model would be relevant to determine how the structural behavior of the blades is affected by variable wind conditions. Although the deformation of the blades will have an impact on the blade fatigue lifetime, no study has previously shown that far wake aerodynamics is significantly impacted by the level of blade deformation. Furthermore, there is still room for improvement of the mesh layout in order to reduce even more the computational resources required for simulating wakes interaction effects.

**Author Contributions:** Conceptualization, R.V.R. and C.L.; methodology, R.V.R. and C.L.; software, R.V.R.; validation, R.V.R.; formal analysis, R.V.R.; investigation, R.V.R. and C.L.; writing—original draft preparation, R.V.R.; writing—review and editing, R.V.R. and C.L.; funding acquisition, R.V.R.

**Funding:** This research was funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), gran<sup>t</sup> number 249258/2013-7.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.
