WP4: The main objective of WP4 is to develop and validate mathematical models for structural reliability, which will enable probabilistic design in practice in line with SO3 while maintaining consistency with the design load case approach currently used in international design standards IEC61400-1 and IEC61400-3. This capability will also allow the demonstration of the effect of the uncertainty reduction efforts carried out in WP2 and WP3, on the design reliability. The results are compared against the current design practise, which is the partial safety factor-based design approach, and thereby the reduction in uncertainty and conservatism is quantified.
As a part of this WP, the methodological challenges in computing reliability for wind energy applications are considered: connecting models of phenomena on different scales and fidelity, the highly nonlinear turbine response which is a challenge for efficient uncertainty propagation, and the curse of dimensionality, meaning that model performance deteriorates significantly when the number of variables is increased. These are addressed by implementation of novel techniques optimally combining machine-learning-based dimensionality reduction with advanced surrogate models from UQ7 . The reliability assessment procedure needs to take a very large number of design scenarios into account, considering both Ultimate Limit State (ULS) and Fatigue Limit State (FLS).
To improve the load variability coverage while keeping feasibility in terms of time requirements for industrial design, WP4 develops simpler, straightforward procedures for reliability assessment. For ULS, the reliability is computed without having to aggregate information from multiple load scenarios, by determining the short-term wind and wave parameters associated to critical loading conditions with given exceedance probability, through an optimization problem formulation8 . An adaptive design of experiments for FLS9 is implemented, which allows computing the fatigue reliability of bottom-fixed and floating wind turbines with a minimum amount of calculation ensuring safe design. To evaluate the effect of wind turbine dynamics on the reliability, the safety margins in wind turbine blades, drivetrain, and tower are evaluated as function of their position in the wind farm, while using the uncertainty assessments from WPs 2-3.