The design of wind turbines installed in various regions of the world where earthquakes are likely must take into account loads imposed on the turbine due to ground shaking. Currently, design standards such as the International Electrotechnical Commission’s standard, IEC 61400-1, do not provide detailed guidelines for assessing loads on wind turbines due to seismic input excitation. In regions of high seismic hazard, it is extremely important to perform a thorough seismic analysis. Various simplified and full-system wind turbine models have been published and used for seismic analysis of turbine loads in recent years. Among these models, the open-source software, FAST, allows for full-system simulation of the response of wind turbines subjected to earthquake ground motion along with other sources of loading such as from the mean wind field and turbulence. This study employs this open-source software to simulate seismic loads and presents statistical and spectral summaries resulting from extensive analyses undertaken by simulating turbine response to various input motions from Western U.S. earthquakes. A total of 150 different earthquake ground motion records with varying magnitude and distance from fault rupture are selected and normalized/scaled to selected target levels prior to response simulation using a utility-scale 5-MW wind turbine model. The records selected are divided into six groups of 25 records each; the groups consist of different magnitude and distance-to-rupture values. The records in each bin are scaled to have similar demand levels as the average of the demand of the unscaled records in that bin. Two different normalization options are considered—in one, the scaling is at the rotor rotation rate (or the once-per-rev or 1P frequency); in the other, the scaling is done at the tower fore-aft first mode frequency. A study of various turbine load measures is conducted. It is found that turbine tower loads, in particular, are especially influenced by the earthquake excitation.
A statistical analysis of the extreme loads for the 5-MW NREL baseline wind turbine over the range of operational mean wind speeds at hub height i.e., from the cut-in wind speed of 3 m/s to the cut-out wind speed of 25 82 m/s was carried out for 150 ground acceleration motion records belonging to six different earthquake bins based on the magnitude of the earthquake and the distance of the recording station from the point of rupture. This study attempted to explain the dynamics of the wind turbine under combined aerodynamic and seismic loads using statistical summaries of extreme tower
and blade loads. Some important conclusions drawn from this study follow.
• Extremes of the tower base bending moment are significantly affected by the seismic ground acceleration input. The distribution of the extremes of the tower base fore-aft as well as side-to-side bending moments suggest that when seismic ground acceleration is included in turbine loads analyses during normal operation, tower base bending moments are significantly higher than those during normal operation with only turbulent
wind input.
• The maximum tower loads for a given ground motion bin are observed to occur at a mean wind speed at hub height close to the rated wind speed. Extreme tower loads at a given wind speed increase with an increase in the magnitude of the earthquake to which the ground acceleration record is associated. Also, these extremes decrease with an increase in distance from the fault rupture.
• The orientation of the ground acceleration records with respect to the wind turbine plays an important role in determining extreme tower loads. This study suggests that the maximum tower loads generally occur when, among the two horizontal components of ground acceleration, the one with the higher intensity of shaking (or higher peak ground acceleration) is aligned with the longitudinal wind direction (i.e., with the non-zero 83 mean wind direction). This is due to the mean component of the tower base fore-aft moment that results from the aerodynamic loading. In all other orientations of the horizontal ground acceleration components, the tower base bending loads are generally lower. The present study focused on tower loads during the worst-case orientation scenario.
Tower deformation
• Unlike tower loads, blade root bending moments do not show a significant influence of the seismic input. The distribution of the extremes of the blade root in-plane and out-of-plane bending moment suggest little to no dependence on the ground acceleration input.
• The variation in the transient wind-only period before any ground acceleration does not affect the extreme values of turbine tower loads. The extremes of the tower loads depend less on the turbulent wind input but are greatly influenced by the ground acceleration input.
• Time series and power spectral density functions for the tower base response parameters show contributions from frequencies associated with seismic input as well as low-frequency energy from the turbulent wind field; resonant frequencies associated with tower bending are clearly influenced by the level of seismic input.
Posts
0 Comments :
Post a Comment