Dynamic behavior of wind turbines influenced by aerodynamic damping and earthquake intensity

Abstract

In the present paper the effects of aerodynamic damping and earthquake loads on the dynamic response of flexible-based wind turbines are studied. A numerical analysis framework (NAF) is developed and applied. NAF is based on a user-compiled module that is developed for the purposes of the present paper and is fully coupled with an open source tool. The accuracy of the developed NAF is validated through comparisons with predictions that are calculated with the use of different numerical analysis methods and tools. The results indicate that the presence of the aerodynamic loads due to the reduction of the maximum displacement of the tower attributed to the dissipation of earthquake excitation energy in fore-aft direction. Emergency shutdown triggered by strong earthquakes results to a rapid change of aerodynamic damping, resulting to short-term instability of the wind turbine. After shutdown of the wind turbine, enhanced dynamic response is observed. For the case where the wind turbine is parked, the maxima displacement and acceleration of tower-top increase linearly with the peak ground acceleration. With the use of the least-square method a dimensionless slope of tower-top displacements is presented representing the seismic response coefficient of tower that can be used to estimate the tower-top acceleration demand. Moreover, on the basis of the seismic response coefficient, an improved model for the evaluation of load design demand is proposed. This model can provide accurate predictions. KEYWORDS aerodynamic damping, dynamic behavior, earthquake intensity, seismic demand, wind turbines | INTRODUCTIONWind energy technology has a leading role in the renewable energy sector. 23.4 GW of new installed capacity of wind turbines that was added in China 1 for 2016; this value corresponds to the 42.8% of the overall added capacity globally for the same year. New installed wind farms are located along the northwest and southeast coasts of China. These wind farms are located in earthquake prone areas, close to the Eurasian and Pacific seismic belts. Wind farms that are located in the aforementioned areas are susceptible to damage on the event of an earthquake.Over the past decades it has been well accepted that earthquakes have a significant effect on the structural dynamic response of wind turbines. 2 Response spectrum method and finite element method (FEM) have been commonly used for the estimation of the seismic load demand of wind turbines. 3 With the use of the response spectrum method the rotor and nacelle are modeled as point lumped masses at tower-top. The seismic load demand of wind turbines can be calculated on the basis of the natural periods, mode shapes, and mass distribution of the tower. 4,5 Usually the aerodynamic damping and higher-order modes are neglected because of the overestimation of the seismic load demand. On the basis of experiments, 6 it is concluded that the seismic excitation increases significantly the lateral displacement of tower-top. On the other hand FEM is applied for the analysis of win...

Investigation on the sensitivity of flexible foundation models of an offshore wind turbine under earthquake loadings

“…It can be argued that the SSI effect is better addressed with the use of the AF model than the CS model. maximum bending moments decrease along with the support structure height, this result is consistent with the analogous results of the same wind turbine calculated using the boundary element method [10]. The maximum bending moment of the rigid model at the mudline is close to that of the DS model, while a significant difference is observed for the AF and CS models.…”

“…Generally, the aerodynamic loads increase exponentially with the rotor diameter for large-scale wind turbines. The resulting aerodynamic effects have been determined to be unneglectable from a comparative study on operational and parked states [10]. Therefore, over-simplification of aerodynamic loads is never precise, thereby undermining the accuracy of results in the seismic analysis of large-scale wind turbines.…”

Analysis of seismic behaviour of an offshore wind turbine with a flexible foundation

“…In the parked state, the peaks of the towertop fore-aft displacement in the fixed and flexible conditions are 1.77 m and 2.25 m, respectively while the values in the operational state are 1.63 m and 2.10 m. This indicates that the aerodynamic damping has a positive effect in mitigating the vibration amplitude because of energy dissipation resulting from aerodynamic damping during earthquake event. This finding is consistent with the conclusion for a land based wind turbine under earthquake events[59].In the case of emergency shutdown condition induced by the earthquake, the generator is turned off at 808.468 s and 807.336 s for the fixed and flexible cases, respectively. The variations of the results in the emergency shutdown state are significantly different from those in the operational state, especially for the fore-aft responses.…”

Seismic Analysis of 10-MW Off Shore Wind Turbine with Large-Diameter Monopile in Consideration of Seabed Liquefaction

“…2 Wind power and earthquake distribution in the South China Sea: (a) the average wind power density (Wan et al 2018) and (b) seismic epicenter distribution (BC1767-2018) (Li et al 2020) The existing research on WTs subjected to seismic loadings mainly includes three aspects. The first is the dynamic responses of onshore WTs under wind and earthquake loadings or offshore WTs under the wind, wave, and earthquake loading in consideration of aerodynamic damping, seismic loading conditions and types (Alati et al 2015;Huang et al 2021;Kjørlaug and Kaynia 2015;Patil et al 2016;Santangelo et al 2018;Santangelo et al 2016;Sigurðsson et al 2020;Yang et al 2018;Zafeirakos and Gerolymos 2013). Then, the second is the fragility analysis and collapse evaluation employing the nonlinear incremental dynamic analysis approach to assess the exceeding probability for various damage states (Asareh et al 2016;Fan et al 2019;Kim et al 2014; Martín del Campo and Pozos-Estrada 2020; Quilligan et al 2012).…”