Implementation of statistical principles to reflect the long term performance of composite insulators

Abstract

There are numerous advantages of polymeric composite insulators over the traditional ceramic ones. However composite insulators, being comprised of organic elements, age more rapidly over time. Environmental conditions control both the manner in which composite insulators age, and the effective stresses that insulator strings are exposed to in service. The current work presents a statistical methodology to reflect the reliability of a network consisting of composite insulators. This accounts for the working environment and its impact on both the ageing of the insulator and risk of flashover. This is achieved by combining previously established statistical dimensioning principles according to the site's pollution severity and a four-state Markov ageing model. The risk of flashover is estimated according to the site's pollution severity using statistical principles. Then by choosing appropriate transition probabilities between the states of Markov's ageing process the population of insulators at risk of flashover because of ageing is estimated. The combination of the risk of flashover due to the pollution severity and due to ageing of the insulators results in an estimation of risk of flashover for a given number of insulators in a given point in time. The resulting forecast of flashover performance could be used to assist asset management decisions as well as optimising condition monitoring of composite insulators.