Global monitoring approach in a large autotransformer through PD, DDF and DGA analysis: PD source location and maintenance action planning

Feasibility and effectiveness of a Global Monitoring approach on a 250 MVA autotransformer are illustrated in this paper. The on-line Global Monitoring System (GMS) is constituted by a Partial Discharge Monitoring (PDM) system, a bushings Dielectric Dissipation Factor Monitoring (DDFM) system and a Dissolved Gas Analysis (DGA) on-line monitoring unit. Combining the results obtained from the three systems, it was possible to detect, identify and locate two different PD sources producing a critical level of hydrogen. The first PD activity, ascribable to presence of gas bubbles in the oil, disappeared after oil treatment. The second one, due to wrong constructional arrangement at the winding connections to the bushings, did not extinguish after the oil treatment and started again to produce a large and critical amount of hydrogen. The trend of the PD activity was analyzed for six months, resorting to statistical indexes, such as the amplitude and repetition rate, in order to help the maintenance manager to plan proper actions. This successful example highlights how predictive monitoring systems can reduce the maintenance costs in critical assets as large power transformers, stressing the importance of analyzing various diagnostics marker together in a global monitoring approach.