PD Signal Propagation in GIS: Ultra-High Frequency Detection-Based Modeling

Abstract

Gas-insulated switchgears (GIS) have become essential parts of electrical power substations due to the associated merits of these capital assets. Although such resilient devices can rarely suffer from failure, partial discharge (PD) is responsible for around 85% of their recorded collapses. Ultra-high frequency (UHF) techniques have been widely used in the detection and localization of PD for a long time because of their immunity to noise and high sensitivity. Understanding electromagnetic (EM) wave behavior in GIS systems is significant for improving the utilization of UHF sensors in PD detection and for the optimal allocation of UHF antennas inside GIS systems. Thus, this paper is devoted to building a detailed 3D finite element (FE) model based on UHF detection techniques to understand the propagation behavior of EM waves inside GIS. A disk-type UHF sensor is used for acquiring EM waves inside the GIS. The sensitivity of the sensor has been obtained using a gigahertz transverse-electromagnetic (GTEM) test cell. The proposed model investigates the impact of multiple disconnecting parts including L-structure, relative angle between PD source and sensors, and disconnecting switches on the propagation of electromagnetic waves based on step 1 of the CIGRE recommendations. To validate the modeled GIS, a simple L-structured model is initially built, and a comparative analysis has been conducted between the built model and the experimental and analytical results from the literature.

A Novel Method for Pattern Recognition of GIS Partial Discharge via Multi-Information Ensemble Learning

“…The UHF detection method has certain disadvantages for the detection of free metal PD. The random distribution of amplitude and phase makes it easy to be overwhelmed by noise during the field detection process and cannot be effectively identified [ 16 ].…”

Numerical and Experimental Investigations on Influence of Internal Defect Parameters on Partial Discharge Induced UHF Signals in Gas Insulated Switchgear

“…Partial discharge (PD) detection using ultrahigh frequency (UHF) method has been widely investigated for condition monitoring of power apparatus due to its high sensitivity, immunity to external noise and ability to detect variety of PD activity [1][2][3][4][5], identify defect type [6,7], locate the source [2] and assess the insulation condition in high voltage (HV) equipment [8]. UHF PD sensing is gaining acceptance for condition monitoring of ageing gas insulated switchgear (GIS) [9][10][11][12]. GIS is a metal encapsulated power equipment consisting of several HV components.…”

“…GIS is a metal encapsulated power equipment consisting of several HV components. The changing dimensions of the GIS due to the internal components, electrical discontinuities, and branching result in complex electromagnetic (EM) wave behavior inside the GIS and influence the propagation path and characteristics of PD induced UHF signals [11,[13][14][15][16][17][18][19].…”

“…Numerical probes yield optimistic results as they have isotropic and broadband behaviour unlike practical internal UHF antenna. Darwish et al studied the influence of disconnecting switches on PD induced UHF signals inside a L-GIS with spacers and handholes using an artificial internal transmitter connected to a wideband signal generator to inject PD-like signals [11].…”

“…Existing studies on UHF signal propagation in GIS are mostly numerical investigations in varying GIS geometry [11,15,17,18,29,30], defect location [27,28,31,32] and UHF orientation [11,27,32]. A significant number of them are simulations and only a few provide insights on EM wave propagation [11,27,30].…”