Probabilistic design of support structures for offshore wind turbines by means of non-Gaussian spectral analysis

authored by: Sebastian Kelma
supervised by: Peter Schaumann, Michael Beer
Abstract: Offshore wind energy is of special importance in order to meet the ambitious goals
to produce climate-neutral energy. Therefore, an accelerated installation of offshore
wind turbines is required. The design is to be achieved with respect to standards and
guidelines. Especially probabilistic design methods allow an accurate and economic
structural design.
Not only the environmental conditions vary during the lifetime, but the short-term
loads are also subject of random scattering. For the design of offshore wind turbines,
the required load simulations are usually carried out in time domain. In comparison, it is less time-consuming to obtain loads by means of frequency-domain analysis. This is very beneficial for the probabilistic design which requires significantly more load simulations in time domain. However, non-linearities and time-variant behaviour of offshore wind turbines cannot be represented well during the load simulation in frequency domain.
Extreme loads and fatigue loads can be calculated by means of frequency-domain analysis. The determination of the distribution functions of extreme values is well established on a theoretical background. As for the fatigue design, different empirical models exist which describe the distribution function of fatigue loads on the basis of frequency-domain analysis. In this thesis, a new model is introduced which leads to more accurate results.
Since frequency-domain analysis is not always suitable, the transformation of signals
given in frequency domain is required to generate respective random time series. As for the design of offshore wind turbines, only limited recommendations are stated in standards on how to carry out this transformation. Detailed analysis shows that accurate results with respect to wave-induced loads are also obtained for coarser discretisation of spectra. The resulting loads and their statistical properties are still accurate, while the numerical effort can be reduced in comparison to the stated recommendations.
On the basis of theoretical findings, time series from load simulations of offshore wind turbines are analysed regarding their spectral properties. Investigations are carried out to evaluate the agreement between the extreme load and fatigue loads which are either simulated or calculated on the basis of the spectral properties. It is also shown that currents within sea states lead to increased fatigue loads.
Organisation(s): Institute of Steel Construction
Type: Doctoral thesis
No. of pages: 199
Publication date: 2024
Publication status: Published
Sustainable Development Goals: SDG 13 - Climate Action
Electronic version(s): https://doi.org/10.15488/15784 (Access: Open)