Economic evaluation of degradation diagnosis for cables

Degradation diagnosis is aimed at preventing unexpected failure of electric power apparatus. It may be an effective method to prolong the service time of apparatuses. For the total economic evaluation parameters other than accuracy such as cost, interval and criterion of the diagnosis must be considered. In this paper a power cable with water-tree degradation, for which available data were relatively rich, was studied as an example of power apparatuses to discuss the economically optimum parameters of diagnosis. From the relation between accelerating-degradation period and maximum water-tree length, the probability distribution of the maximum water-tree length for the duration for which the cables have been in use was computed. The relation between maximum water-tree length and failure probability of the cable was computed by using the Weibull distribution plot of the breakdown voltage for the cables with different maximum water-tree length. From the failure probability for the cables in use and the replacement rate by diagnosis, which were computed from the probability distribution and the relation between maximum water-tree length and failure probability, we evaluated the effect of diagnosis and discussed the economic optimization of diagnostic parameters (e.g. diagnosis interval, criterion).