Monte Carlo modeling of tornado hazard to wind turbines in Germany.

Using a Monte Carlo hazard modeling approach, this study assesses the damage and losses to wind turbines caused by tornadoes over Germany. The model combines observed climatology of tornadoes, exposure map of wind turbines and their capacities, vulnerability curve (first developed in this study), and costs of wind turbines to estimate the likelihood of financial losses of different magnitudes. The losses are presented in terms of aggregated and highest annual losses as a function of the return period. Deterministic modeling of the vulnerability function produced larger variability of the most severe losses (rare, but high-intensity events) compared to the pseudorandom sampling of the damage from the vulnerability curve. Doubling the number of current wind turbines results in smaller expected losses than installing fewer but more powerful units (repowerment). Tornadoes are most frequent in the summer months when the capacity factors of wind farms are the lowest and the electricity consumption is also lower than in the winter months. By repeatedly placing a tornado track of a fixed size over the same wind farm, we further investigated the sensitivity of our model to different parameters. The probability of tornadoes of different intensities hitting a wind turbine over Germany is also calculated.