Autonomous localization of partial discharge sources within large transformer windings

Partial discharge (PD) condition monitoring inside a HV transformer generally and particularly along a transformer winding has become an important research area with the ultimate aim of providing asset health information that enables maintenance and replacement processes to be carried out effectively. As far as PD activity inside transformer windings is concerned, an electrical detection method has been developed based on the use of radio frequency current transducers and subsequent digital signal processing of captured measurement data. A localization approach based on the measurement of currents at the bushing tap point and neutral to earth connection has been developed, with the assumption that different PD source locations will generate unique signal profiles in terms of the distribution of measured current energies with respect to both frequency and time. Therefore the technique presented is based on analysis of measured current energies associated with different frequencies. Principal Component Analysis (PCA) is then applied to reduce the dimensionality of the data, whilst minimizing lost information in the original dataset. This non-linear analysis of captured current data is not practicable for the field but the process can be represented through the use of three finite impulse response filters that have the ability to perform PD source localization automatically and are straightforward to implement in monitoring hardware.