Joint wind and ice hazard for transmission lines in mountainous terrain

Abstract

This paper focuses on analyzing the ice and wind conditions for a high voltage transmission line system going over mountainous terrain in British Columbia, Canada. After analyzing sixteen surrounding weather stations, an Artificial Neural Network (ANN) is performed for wind speed predictions, the Inverse Distance Weighted Interpolation method (IDW) is used for temperature estimations, and the K Nearest Neighbor Imputation (KNNI) is performed for precipitation rate on the unsampled site of interest. In-cloud icing is estimated using a simple empirical equation proposed by Makkonen & Ahti (1995), in addition to the ice accretion model proposed by Makkonen (2000). The weather data from the described procedure is compared with data from a Weather Research and Forecast model (WRF) reported in previous studies. The wind speed and ice accretion data are fitted to the Weibull and Generalized Pareto distributions, respectively, and five hundred years of data pairs are simulated in addition to the historical data based on those distributions. Finally, hazard contours are drawn, and the results are compared with the combined wind and ice values proposed in the National Standard of Canada (CSA-C22.3) for different reliability levels.