Failure rate estimation is an important tool for planning and operating decision making in asset management of the power system. Moreover, the knowledge of how different explanatory variables impact the failure rate of the power system equipment is crucial for substation design. This study investigates 2191 work orders of 1626 non-current breaking disconnectors with 344 major failures. In particular, this paper analyses the disconnector failure data regarding recurrent failure data. Since the original PHM cannot handle recurrent event data, different extensions were developed such as the Andersen-Gill (AG), Prentice, Williams and Peterson (PWP), and the Wei, Lin, and Weissfeld (WLW) model. These models are applied to the disconnector dataset with 140 recurrent time-to-failure processes. The explanatory variables age at admission, remote control, preventive maintenance, and voltage level are assessed. The results show that preventive maintenance has a significant and positive impact on the recurrences with all tested methods. Also remote control, voltage level, and age are significant covariates. Compared to the single failure study previously conducted, where age had no significance, age is significant when assessing the recurrent failure which is the most critical difference to the analysis without recurrences.