Assessing Transformer Reliability using Post-Mortem Data and Health Indices

The reliability of aged transformers must be identified to ensure the continued secure and economic operation of the power system. However, the limited availability of ageing related failure data means that conventional methods of assessing transformer reliability are not feasible. This paper proposes a method for assessing transformer reliability based on degradation of a transformer’s paper insulation. The suggested method aims to identify the severity of degradation from a transformer’s assigned Health Index (HI). Post-mortem data obtained from a utility database of scrapped transformers is used to quantify the severity of insulation ageing at each HI score. Reliability is subsequently estimated using the Load-Capacity Interference Theory (LCIT). This method provides a convenient means for assessing transformer reliability based on assigned HI scores. From the results, two distinct regions of reliability are identified, highlighting the importance of conducting post-mortem investigations on transformers removed from service.