A Numerical Investigation of the Impact of Waves on a Large Floating Wind Farm in Northwest Atlantic

Abstract This study investigated the combined effects of waves and atmospheric turbulence on the power output of a deep-water floating offshore wind farm in the Northwest Atlantic. Using a scale-adaptive large-eddy simulation (LES) framework, we modelled a floating wind farm comprising an array of 15 MW turbines with 240-meter rotor diameters, staggered with 5 MW turbines featuring 126-meter rotor diameters. The marine atmospheric boundary layer (MABL) was simulated using wave drag parameterization, Monin-Obukhov similarity theory, and stochastic turbulence forcing. The simulations incorporated rated wind speeds with varying wind directions and wave amplitudes ranging 0.2 to 2 meters, resolving turbulence structures using 107 million grid points and a vortex stretching-based subgrid model. Turbine wakes were modeled with a Gaussian actuator disk approach, including wave effects. Data analysis validated the LES framework against met-ocean environments and identified dominant frequency ranges influenced by waves using Proper Orthogonal Decomposition (POD) and wavelet transforms. Results highlight that atmospheric turbulence is the primary driver of power fluctuations, enhancing overall power output through vertical flux entrainment. Large-amplitude waves were found to modulate turbulence at higher frequency ranges, leading to improved turbine performance under certain conditions. Proper Orthogonal Decomposition (POD) and wavelet-based analysis revealed dominant flow structures influenced by wave-induced wind stress. These findings emphasize the critical role of pitch control strategies that account for wind-wave misalignment and Ekman spiral effects, particularly for spar-based floating turbine platforms. This research underscores the need for integrating wind-wave interactions into the design and operation of large floating wind farms to optimize energy production and structural resilience in complex marine environments.