Abstract:Free metallic particles are the main factor inducing flashover incidents and threatening the safety of gas-insulated switchgear (GIS) and gas-insulated transmission line (GIL) devices. Acoustic diagnosis is widely used to detect free metal particles due to its high sensitivity, but it generally fails under multiple particle conditions because of the overlap of acoustic signals. This paper proposes a new method of acoustically recognizing multiple free metallic particles. The pattern of Δt-Normalized N that can be used to distinguish single/multi-particles is determined first. Second, each particle's neighboring collision signals are extracted from raw data by the established physical relation between adjacent collision signals. Wire, lamellar, and spherical free metal double-particles' mass values are calculated as per the numerical connection between adjacent collision signals. The geometry of the particles is assumed to be known, so the density of the spherical particles can be effectively estimated. The proposed method is validated in a scale concentric hemisphere platform. The results of this work may provide a workable reference for GIS free metal multi-particle monitoring.

Acoustic Method for Multiple Free Metallic Particle Recognition in GIS/GIL