A COMPARISON OF SELECTED MODELS FOR ESTIMATING CABLE ICING

In many cold climate countries, it is becoming increasingly important to monitor transmission line icing. Indeed, by knowing in advance of l ocalized danger for icing overloads, electric utilities can take measu res in time to prevent generalized failure of the power transmission n etwork. Recently in Canada, a study was made to compare the estimation of a few icing models working from meteorological data in estimating ice loads for freezing rain events. The models tested were using only standard meteorological parameters, i.e. wind speed and direction, tem perature and precipitation rate. This study has shown that standard me teorological parameters can only achieve very limited accuracy, especi ally for longer icing events. However, with the help of an additional instrument monitoring the icing rate intensity, a significant improvem ent in model prediction might be achieved. The icing rate meter (IRM) which counts icing and de-icing cycles per unit time on a standard pro be can be used to estimate the icing intensity. A cable icing estimati on is then made by taking into consideration the accretion size, tempe rature, wind speed and direction, and precipitation rate. In this pape r, a comparison is made between the predictions of two previously test ed models (one obtained and the other reconstructed from their descrip tion in the public literature) and of a model based on the icing rate meter readings. The models are tested against nineteen events recorded on an icing test line at Mt. Valin, Canada, during the winter season 1991-92. These events are mostly rime resulting from in-cloud icing. H owever, freezing rain and wet snow events were also recorded. Results indicate that a significant improvement in the estimation is attained by using the icing rate meter data together with the other standard me teorological parameters.