Analyzing parameters that affect the reliability of low voltage cable grids and their applicability in asset management

The majority of interruptions in the network of a distribution system operator (DSO) occur at low voltage (LV) levels. Although the number of affected customers is less than that for medium voltage level interruptions, contribution to the annual customer minutes lost is considerable, and the yearly costs of solving the outages are high. Underground LV cables cannot be visually inspected, and monitoring systems are still experimental. In order to improve asset management (AM) for LV cables, it is beneficial for DSOs to be able to perform condition assessments using historical data, in combination with asset and environment data. In this paper, survival analysis is performed using the Cox proportional hazard model. The results of this analysis can be used to identify those variables that predict a relatively high failure probability and to estimate the relative risk of failure for cables. This enables the improvement of AM strategies, such as the preventative replacement of cables. The method presented in this paper shows promising results, allowing for greater insight into the causes of failures.