The coupled wind-wave solver (Yang & Shen, 2011) is extended to offshore wind energy research, with the embedment of turbine model (Yang & Sotiropoulos, 2015). The composite solver consists of a high-order spectral method for surface waves, large-eddy simulation for offshore wind on wave-surface-fitted dynamic grid, and an actuator based model for the representation of blades and nacelles. The blades are discretized by line elements, and its bound motion with floating platform is considered. Some smoothed spectral operations are applied to alleviate Gibbs phenomenon brought by local jump of turbine force. By performing simulations for a turbine array on land and over ocean waves, we have analyzed the impacts of waves on the wind farm in terms of mean wind speed and mean kinetic energy budget . It is found that a fast-propagating swell at the sea surface leads to a higher mean velocity at turbine hub height, which affects the wind power extraction. The influence of floating platform motion is also investigated. It is found that pitch motion has stronger influence than other in-plane motions, i.e., surge and heave. The interaction between tip vortex sheet and nacelle wake meandering is also studied. The intersection of these two flow structures coincides with a change of growth rate of mean velocity deficit width. The results will be useful for the under-standing and control of turbine wakes in offshore wind farm.