Hourly wind data for aeolian vibration analysis of overhead transmission line conductors

Fretting fatigue of conductors due to aeolian vibrations is an important phenomenon for the design and maintenance of electric transmission networks. The evaluation of exposure and susceptibility of a given span to aeolian vibrations requires information on the distribution of hourly wind speed and direction. Interpolation has been used to obtain wind data at non-instrumented locations; however, their accuracy is not adequate at locations far from stations and at locations in mountainous regions. Numerical Weather Prediction data sets provide wind information on a regular grid covering equally all regions; however, the spatial resolution is not adequate. A procedure is proposed using a mass-conserving diagnostic model (WindNinja) that incorporates a detailed map of local topography to improve predictions of wind direction. A second correction is proposed to account for local surface roughness as a function of wind direction and average canopy height from high-resolution LiDAR data. The proposed procedure is applied to nine meteorological stations and two experiment sites over southern Quebec, Canada. WindNinja was found to provide good estimates for the effects of local topography in mountainous regions. Wind speed after correction provides good agreement with hourly observations as well as with Weibull distributions fitted to the actual observations.