Fatigue limit state design check based on S-N curve and Palmgren-Miner rule approach is required by design standards for floating wind turbines since floating wind turbines are subjected cyclically varying environmental loads. Fatigue damage over a given period could be considered as a summation of short-term fatigue damage of each stationary environmental condition multiplying by the number of occurrence of the environmental condition during the given period. In reality, environmental condition varies continuously. While fatigue damage due to transient load processes that are induced by startup and shutdown operations cannot be captured by this conventional approach. Large data bases of real-time measurements show shutdown and startup operations may appear in any operational conditions and the number of the operations could be considerable. Analysis for startup and shutdown operations induced fatigue damage is required by standards for offshore wind turbines. However, relevant publications addressing this issue are very limited in particular for floating wind turbines. In contrast to bottom fixed wind turbines, floating wind turbines have rigid-body motions in 6 d.o.f.s, while the floating wind turbine hulls are moored by mooring lines rather than directly mounted on sea bed or land.
This paper focuses on shedding light on the importance of startup and shutdown induced transient load processes on fatigue damage in the tower of two MW-level horizontal axis semi-submersible wind turbines by comparing short-term fatigue damage in several environmental conditions with and without transient load processes induced by startup and shutdown of the wind turbines. In some situations, 4,600 seconds short-term fatigue damage may be dominated by the transient load process induced fatigue damage which may make the fatigue damage be increased by up to 300%, while in many cases, the fatigue damage may be increased by 10% to 100%. The importance of the transient load processes on long-term fatigue damage is related to occurrence frequency of startup and shutdown events and needs more analysis in future.