Lifetime estimation regarding leading-edge erosion of wind turbine blades involves a series of mathematical and statistical procedures, including regression fitting, tolerance interval formulation, and weathering modeling approaches. This work investigates the intermediate steps in the fatigue assessment process and the associated sources of variability, using data from a whirling arm rain erosion test (RET) machine, time-series liquid precipitation and wind data from two sites, and the IEA reference 15 MW wind turbine. Results show that extrapolation into the turbine operation domain exerts a profound effect on lifetime estimates, with divergence increasing toward lower impact velocities. Regression choice significantly affects the predicted lifetimes, leading to variations of 7%–255% across the sites. Furthermore, advancing from constant to independent, dependent, and time-series weathering models resulted in lifetime differences ranging from 29%–240% and 38%–210% for the two sites, respectively. The findings highlight that methodological choices within the survivability framework have a significant influence on the final lifetime estimate.
