Case for Including Atmospheric Thermodynamic Variables in Wind Turbine Fatigue Loading Parameter Identification

This paper makes the case for establishing efficient predictor variables for atmospheric thermodynamics that can be used to statistically correlate the fatigue accumulation seen on wind turbines. Recently, two approaches to this issue have been reported. One uses multiple linear-regression analysis to establish the relative casuality between a number of predictors related to the turbulent inflowand turbine loads. The other approach, using many of the same predictors, applies the technique of principal component analysis. An examination of the ensemble of predictor variables revealed that they were all kinematic in nature; i.e., they were only related to the description of the velocity field. Boundary-layer turbulence dynamics depends upon a description of the thermal field and itsinteraction with the velocity distribution. We used a series of measurements taken within a multi-row wind farm to demonstrate the need to include atmospheric thermodynamic variables as well as velocity-related ones in the search for efficient turbulence loading predictors in various turbine-operating environments. Our results show that a combination of vertical stability and hub-height meanshearing stress variables meet this need over a period of 10 minutes.