A probabilistic study of the influencing factors on distribution cable failures using cox proportional hazard model

Analysis of cable failure data are often limited to calculations of the number of failure per one hundred kilometres in length per year with considerations in some instances of cable length, types and their voltage ratings. This paper proposes the use of Cox Proportional Hazard Model (Cox PHM) for analysis of cable failure data. A set of covariates which have significant effects on the cable failure are considered to demonstrate the appropriateness of the model. This paper analyses failure data related to medium voltage (MV, 10kV) cables collected from two metropolitan cities in China using correlation coefficient and PHM. The paper demonstrates that the proposed method is capable of determining the significant factors that are most relevant with failures using a hypothesis test of 5% of p value. The value of relative risk (Exp(β)) is used to evaluate the significance of influencing factors on cable failure. Results prove that the model is more robust than the Weibull distribution in that failure data does not have to be homogeneous and can help provide guidance for improving future practice in cable procurement, installations and maintenance.