Constellation, le dépôt institutionnel de l'Université du Québec à Chicoutimi

Résumé

This paper presents a simple method to assess the severity of ice built-ups in EHV substations. The proposed method is based on the monitoring of the leakage current (LC) and the applied voltage of an EHV standard station insulator. By simply studying the LC evolution of several experimental tests, it was established that the evolution of a LC envelope is divided in two distinct periods where the transition between these two periods is clearly identified and is correlated with the bridging of the shed spacing by icicles. It was also observed that the duration of the first period is only dependent of the icing rate and is not influenced by the applied water conductivity. From these observations, it is possible to estimate the icing rate of an accumulation using specific tools like artificial neural networks (ANN). ANN is used to detect the onset of the ice accumulation with permits to calculate the duration of Period 1 and estimate the icing rate which is one of the most important characteristic of ice built-up severity.