Uncertainty quantification of aerodynamic characteristics of wind turbine blade airfoils

• Published in: Renewable energy Vol. 248; p. 123151
• Main Authors: Hu, Weifei, Wang, Shengjun, Zhang, Tongzhou, Zhang, Yiming, Shi, Wei, Li, Qingyi, Zhang, Fanghong
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.08.2025
• Subjects: aerodynamics; Dynamic aerodynamic characteristics; normal distribution; Probabilistic design; Reliability; renewable energy sources; Safety factor; uncertainty; Uncertainty quantification; wind; wind tunnels; Wind turbine blade airfoil; wind turbines; Dynamic aerodynamic characteristics; Uncertainty quantification; Safety factor; Wind turbine blade airfoil; Reliability; Probabilistic design
• ISSN: 0960-1481
• DOI: 10.1016/j.renene.2025.123151

Wind turbines are subject to multiple sources of uncertainties that need to be meticulously considered during the design phase. Among these, the aerodynamic characteristics uncertainty of wind turbine blade airfoils directly impacts the aerodynamic load on the blades. Moreover, the dynamic stall phenomenon exacerbates the variation of aerodynamic characteristics. This paper addresses the impact of airfoil aerodynamic uncertainties on wind turbine blade design, particularly considering the dynamic stall phenomenon. The uncertainty in static aerodynamic characteristics is firstly quantified by parameterizing lift and drag coefficients from wind tunnel test data and treating them as random variables. A statistical Beddoes-Leishman model is then used to incorporate dynamic stall effects, considering its empirical parameters as random variables to account for epistemic uncertainties. Model errors relative to wind tunnel measurements are fitted using a Gaussian process. The uncertainty is used in a reliability analysis of wind turbines using the first-order reliability method to calibrate partial safety factors. Validation with National Renewable Energy Laboratory (NREL) data shows that the predicted confidence intervals align closely with test data, and the method can predict the uncertainty of new airfoils using dynamic stall wind tunnel data from existing airfoils. Additionally, applying the aerodynamic uncertainty to the NREL 5 MW wind turbine enables a 3 % reduction in load safety factors while maintaining reliability, potentially lowering design costs.