Posters

Abstract

The wind energy market leads to develop wind farms in more and more complex terrain affected by extreme wind conditions. This tendency has developed much faster than the accuracy of the modelling capabilities in the wind industry leading to higher risk on the wind resource assessment but also on load assessment and consequently on turbine integrity. Risk assessment is difficult in such configuration and problems are often discovered during turbine operation with consequent financial losses. The presentation will illustrate an operating wind farm located in Europe, in complex terrain and coastal area. In 2017, a first wind resource assessment based on RANS simulations and industrial standards at the time (IEC ed. 3) has shown no exceedances. In 2022, 3 years after the wind farm commissioning, a turbine blade failed under normal operation conditions. The root cause analysis could not find any negligence in design, manufacturing, transport, installation or controlling. Therefore, the complex terrain and wind conditions were further analyzed using both standard and cutting edge tools and processes. The whole wind farm is located South of a large mountain (1.3 km height difference and about 20 km horizontal width). Measured wind data at the wind farm shows the highest frequency and wind speed magnitude for the North sector, despite the wake of the nearby mountain. In opposition to the measured wind conditions, a microscale modelling (RANS and LES) shows very low wind speed magnitude for the North sector due to the wake of the nearby mountain. Unlike RANS and despite the low wind speed magnitude from the North sector, LES detected exceedances in several IEC extreme wind conditions at several turbine locations, including the turbine location with observed blade failure. Analysis of weather situations highlighted the recurrent occurrence of Low Level Jets from the North sectors in combination with a relatively shallow maritime-origin boundary layer. By applying the mesoscale model WRF we demonstrate, that the interaction of the low boundary layer height with the complex terrain led to the high wind speeds at the wind farm location. An LES with low inversion layer has confirmed the speed-up at the wind farm location on the leeside of the mountain, in better agreement with mast measured data. The detection of the IEC extreme wind conditions has shown that most of the exceedances occur at the turbine location with observed blade failure. The findings of this case study unveil both the limits of the industrial standard approach and the potential of innovative tools and processes such as mesoscale to microscale coupling to understand complex flow behavior that can occur at wind farms in complex terrain.