An Improved Weibull Model with Consideration of Thermal Stress for Analysis of Cable Joint Failures

Abstract

The Weibull model as a statistical method has been widely used for analysis of cable failure data and for prediction of future failure rate on an annual basis. However, previously published approaches cannot be relied on to model the phenomenon that higher number of failures of cable joint has been observed during summer time by Scottish Power Energy Networks (SPEN). Seasonal temperature higher than $20^{\circ}\mathrm{C}$ resulted in increasing failure rate of cable joints, especially for the joints which connect PILC cables. When the ambient temperature was lower than $15^{\circ}\mathrm{C}$, the failure rate of cable joint became smaller. This paper aims to improve the standard Weibull model to allow predictions of number of failures with consideration of seasonal temperature. The thermal stress caused due to seasonal temperature is treated as a variable to correct the failure rate obtained via standard Weibull analysis, as the number of failures has been heavily related to seasonal temperature. With the improved model, failure prediction can forecast higher number of failures during the summer season.