Separation of Partial Discharge Sources Measured in the High-Frequency Range with HFCT Sensors Using PRPD- teff Patterns

Ricardo Albarracín-Sánchez; Fernando Álvarez-Gómez; Carlos A Vera-Romero; Johnatan M Rodríguez-Serna

doi:10.3390/s20020382

Separation of Partial Discharge Sources Measured in the High-Frequency Range with HFCT Sensors Using PRPD- t_eff Patterns

Sensors (Basel). 2020 Jan 9;20(2):382. doi: 10.3390/s20020382.

Authors

Ricardo Albarracín-Sánchez¹, Fernando Álvarez-Gómez¹, Carlos A Vera-Romero¹, Johnatan M Rodríguez-Serna¹

Affiliation

¹ Department of Electrical and Electronic Engineering, Automatic Control and Applied Physics, School of Industrial Design and Engineering (ETSIDI), Universidad Politécnica de Madrid (UPM), Ronda de Valencia 3, 28012 Madrid, Spain.

Abstract

During the last two decades, on-line partial discharge (PD) measurements have been proven as a very efficient test to evaluate the insulation condition of high-voltage (HV) installations in service. Among the different PD-measuring techniques, the non-conventional electromagnetic methods are the most used due to their effectiveness and versatility. However, there are two main difficulties to overcome in on-line PD measurements when these methods are applied: the ambient electric noise and the simultaneous presence of various types of PD or pulse-shaped signals in the HV facility to be evaluated. A practical and effective method is presented to separate and identify PD sources acting simultaneously in HV systems under test. This method enables testers to carry out a first accurate diagnosis of the installation while performing the measurements in situ with non-invasive high-frequency current transformers (HFCT) used as sensors. The data acquisition in real-time reduces the time of postprocessing by an expert. This method was implemented in a Matlab application named PRPD-time tool, which consists of the analysis of the Phase-Resolved Partial Discharge (PRPD) pattern in combination with two types of interactive graphic representations. These graphical depictions are obtained including a feature parameter, effective time (t_eff), related to the duration of single measured pulses as a third axis incorporated in a classical PRPD representation, named the PRPD-t_eff pattern. The resulting interactive diagrams are complementary and allow the pulse source separation of pulses and clustering. The effectiveness of the proposed method and the developed Matlab application for separating PD sources is demonstrated with a practical laboratory experiment where various PD sources and pulse-type noise interferences were simultaneously measured.

Keywords: PRPD-teff pattern; PRPD-time tool; effective time; insulation condition; noise rejection; on-line PD measurements; partial discharges; pulse source separation; wideband PD measurements.