الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 176 |  |  |
| | | from | 176 | | |
Abstract
Gas-insulated systems (GISs) have been widely used in the electric power grid with substantial growth. It is a complete enclosure of all energized parts in a grounded metallic encapsulation insulated with compressed gas. The use of GIS in the power system network has acquired considerable importance because of its compactness, maintenance free, safe and reliable operation. However, the reliability of GIS is adversely affected by the presence of contaminating conducting particles in the insulation structure. The existence of a particle in GIS initiates partial discharge (PD) when the electric field at its tip exceeds the limiting dielectric strength of the gas and may in time initiates the process of breakdown. Therefore, the detection of partial discharge at early stages and the identification of geometry, size and location of a particle contaminant causing the discharge play an important role in evaluating the integrity of equipment insulation and allow the user to predict which equipment of GIS is in need of maintenance.
This thesis presents a method for detecting and identifying the geometry of a fixed contaminating metallic particle in GIS. A fixed particle–initiated negative corona in air insulated co-axial cylindrical configuration is investigated at a voltage slightly above the corona onset level. Then, the onset voltage of negative corona is calculated based on the criterion developed for the formation of repetitive negative corona, Trichel pulses. This calls at first for an accurate calculation of the electric field in the vicinity of the particle where avalanches grow. The investigated gap in presence of a particle is a three-dimensional field problem due to the asymmetrical position of a particle inside the gap. The three-dimensional electric field is calculated using the charge simulation technique with a new charge distribution. The characteristics of the initiated Trichel pulses are
calculated by mathematical modeling the processes taking place during the discharge of negative corona. The Trichel pulse characteristics are calculated for different particle sizes and shapes at an applied voltage slightly above corona onset voltage level. These data of pulse characteristics and onset voltage level are used as a bias for designing and training the artificial neural network technique (ANN).
An experimental set-up has been built up to measure the onset voltage of negative corona initiated by a particle in air insulated co-axial configuration to check the accuracy of the present calculation. The effect of varying field nonuniformity, particle shape, size and position on onset voltage of negative corona is investigated. The calculated onset voltage values agreed well with those measured experimentally. Also, by the same experimental set-up, the Trichel pulse characteristics are measured to check the accuracy of the present calculation. The calculated values of pulse amplitudes and repetition rates agreed well with those measured experimentally.
Hence, for identification of particle contamination geometry in GIS, the Trichel pulse characteristics, peak current and repetition rate are measured. These measured values and the applied voltage are given as an input data to the trained ANN to test for the presence of a contaminating particle and reveal its geometry and location.