Large scale wind farms consisting with floating offshore wind turbines (FOWTs) will be a solution for offshore wind energy industry to access more wind resources. However, wake structures and wake interactions of FOWTs subject to motions are still not yet been fully understand, especially when they are under turbulent inflow conditions with realistic turbulent intensities. These will be critical for designing floating offshore wind farms. Note that the majority of previous research are conducted with single rotor using models having relatively low fidelities and/or focusing on laminar inflow conditions. To advance the knowledge about wake and wake interactions of FOWTs, numerical studies about rotors of FOWT with prescribed harmonic surging motions are conducted in this thesis project with high fidelity CFD models, namely large eddy simulation (LES) coupled with actuator line model (ALM). Cases with single rotor without controller, dual rotors in tandem without controller, and dual rotors in tandem with controller are simulated with various of settings, including different surging settings and different inflow turbulence intensities.

For the cases with single rotor without controller, it is found that the differences of wake structures between fixed and surging rotors are pronounced when under laminar inflow conditions, where the periodic structures related to the harmonic surging motions can be detected straightforwardly; while the differences are much less significant when under inflow conditions with realistic turbulence intensities, and the periodic structures are clearly revealed only after phase-locked averaging. Moreover, surging cases with laminar inflow conditions have wake recovery rates which are significantly higher than the fixed case with laminar inflow; however, with turbulent inflow, wake recovery rates for surging cases are only slightly higher than the fixed case.

For the cases with dual rotors without controller, it is found that the wake interaction modes between the two rotors are significantly affected by the surging settings for the laminar cases, while the turbulent cases are insensitive. However, the power performances of the downstream rotors will be increased slightly with surging upstream rotors for the turbulent cases.

For the cases with dual rotors with controller, it is found that the implemented simple controller cannot improve the performances of the rotors as designed due to the large rotational inertia, and thus the modes of wake interactions are not altered a lot. However, the downstream rotors' operational parameters were successfully changed to more desirable values by the controller, demonstrating the controller's potential for numerical analysis of wake interactions between wind turbine rotors.

The findings of this effort demonstrates that the wake structures due to the surging motions of FOWT rotors will be smeared out by the ambient turbulence; and to achieve better power performances, more advanced controlling strategies may have to be implemented for FOWTs subject to surging motions.

Measurement campaigns and CFD simulations have recently identified a large-scale flow phenomenon called wind-farm flow blockage. This is found to bear a significant and far-reaching reduction in wind speed upstream of a wind farm. The wind farm blockage is attributed to the cumulative induction effects of multiple wind turbines placed in series. Wind-farm flow blockage has important consequences on energy production because it reduces the available kinetic energy in the incoming wind flow. In turn, this causes leading wind turbines in a wind farm to produce less energy than they each would in isolation. To date, the physics of this global blockage effect is not entirely understood, and they are therefore an active research topic. Due to the increasing demand for wind energy, reducing annual energy production (AEP) uncertainties and power production bias seems to be a challenge for wind energy researchers. Understanding wind farm blockage in complex terrain becomes crucial to account for uncertainties and power production bias.
This thesis set out to perform Reynolds-Averaged Navier-Stokes (RANS) simulations to assess the impact of wind-farm flow blockage in complex terrain using the open-source software OpenFOAM. A laterally infinite row of turbines is simulated on top of a 2-D hill defined by the mathematical curve ’Witch of Agnesi’. The set of simulations is performed for varying atmospheric conditions: truly neutral and stable free atmospheric conditions. Thermal stratification imposed under stable conditions is of particular interest due to the excitation of atmospheric gravity waves (AGWs) by the turbine array and the topology. The velocity fields due to the presence of the turbine array on top of the hill are compared to the ones without. The resulting flow reduction is then compared to the cases without the hill in order to assess the impact of complex terrain on wind farm blockage. A series of sensitivity analyses are performed for varying inter-array spacing and hill size variations in order to further the understanding of wind farm blockage.
The results obtained in this study show that the magnitude of wind farm blockage is amplified due to the presence of the hill. Additionally, the excitation of AGWs is seen to have a major impact on the wind farm blockage due to alterations caused to the pressure field. The impact of blockage is seen to be dominant up to at least 10-15 turbine diameters upstream of the turbine array under truly neutral conditions. While the effects are more pronounced and much more dominant under stable free atmosphere conditions. All the stable free atmosphere cases simulated show a reduction ranging from 1-4% at different upstream locations while neutral cases show slightly lower yet non-negligible reduction due to blockage.
This study ultimately concludes that the existing ’wakes-only’ approach for estimating energy losses still has a significant power production bias. Therefore accounting for the blockage effects in the farm upstream is also equally important and must be analysed before commissioning a wind farm.

Computational Fluid Dynamics based on RANS models remain the standard but suffer from high errors in complex flows. In particular, turbulent kinetic energy is over-produced in high strain rate regions, such as the near wake of wind turbine flows. Data-driven turbulence modelling methods aim to derive novel turbulence models with lower uncertainties, which generalize well to a certain class of flows. These state-of-the-art constitutes to first derive model-form corrections of a selected baseline model from high fidelity reference data, followed by regressing the corrections in terms of RANS-known flow features. For data-driven wind turbine wake modelling, industrial-scale wind turbines and non-neutral atmospheric boundary layers have yet to be considered. In this thesis, the first steps are made to address this research gap. First, Large-Eddy Simulation data is generated and validated against literature. The considered cases are under neutral, convective and stable atmospheric conditions. The frozen-RANS methodology, a technique used to derive turbulence model corrections given the high fidelity data, is then extended to for non-neutral conditions. The new framework now provides corrections to both the Boussinesq eddy viscosity hypothesis for the Reynolds stress and the gradient-diffusion hypothesis for the turbulent heat flux. By injecting the obtained corrections into dynamic RANS simulation, the baseline turbulence model deficiencies are corrected. In particular, high rate-of-strain regions now no longer show an overproduction of mechanical turbulence. Similarly, the lack of buoyant turbulence production in the free-stream atmosphere under convective conditions is solved. In the stable case, buoyant destruction is too large in the free-stream but not large enough in the wake. For the neutral and stable case, the corrected models produce wake velocity profiles that show excellent agreement with the large-eddy simulation reference data. Issues in the wall stress solution of the convective large-eddy simulation propagate to issues in the corrected RANS solutions, proving the necessity of high-quality data. Furthermore, it is shown that for most cases a single scalar correction to the turbulent heat flux, as opposed to the full vector correction, is sufficient for improving the error introduced by the gradient-diffusion hypothesis. This result is considerable since the simpler correction would be much easier to regress in terms of mean RANS-known quantities. The computational cost of the corrected RANS models is around the same as that of baseline RANS models; only 2\%-5\% of the large-eddy simulation computational cost.