When considering offshore wind farm projects, it is essential to understand the wind conditions in the target sea area. However, in the wind resource assessment at the initial stage, wind farm developers will conduct only the assessment with a numerical model due to no observation data. In this study, in order to evaluate the prediction error of the numerical model, we compared the 0.1km-resolution WRF simulation data with observation data at Happo-Noshiro offshore site, Akita prefecture. We confirmed that WRF can generally well reproduce the annual mean wind speed and wind rose, having good correlations with the observations. However, it was also found that WRF predicted more smoother wind speed fluctuations than the observations, which appeared as characteristic errors in the frequency distribution of wind speed for a year. These errors, combined with the nonlinearity of a wind turbine power curve, was found to cause an underestimation of the annual energy production.

View full abstract

Seismic loads on a 3MW offshore wind turbine supported by a monopile are investigated by a onedimensional Winkler model (1D model) with multi-soil springs and dashpots and are compared with those by a three-dimensional thin layer method (3D TLM). Firstly, the Winkler springs by Francis’s formula and dashpots by the simplified method are evaluated and compared with those obtained by the TLM. It is found that the predicted stiffness is close to that by the TLM, while the predicted damping is considerably smaller than that by the TLM. The seismic loads by the 1D models with the evaluated springs and dashpots are then compared with those by the 3D TLM. It is found that the seismic loads evaluated by the 1D model are close to those by the 3D TLM when the springs and dashpots are evaluated by the TLM, while the 1D model with the Winkler springs and dashpots obtained from Francis’s formula and the simplified method overestimates the seismic loads on the tower and monopile.

View full abstract