Surge Arrester Leakage Current Analysis by Using Particle Swarm Optimization Technique SAEED Vahabi Mashak a , ZULKURNAIN Abdul-Malek b *, HADI Nabipour-Afrouzi c , C.L. Wooi d and AMIR HESAM Khavari e Institute of High Voltage and High Current (IVAT), Universiti Teknologi Malaysia, Malaysia vmsaeed3@live.utm.my a , zulkurnain@utm.my b *, nahadi4@live.utm.my c , wooi5195@gmail.com d , hesamkhavari@fkegraduate.utm.my e Keywords: Surge arrester; Thermography Analysis; Neural Network; Condition Monitoring; Age diagnostics; Abstract. Surge arresters, as protective equipment, are used to limit any overvoltage in a power system resulting from various sources. The surge arrester affects from degradation due to the continuous operating voltage system as well as due to repeated lightning current discharge. Therefore, condition monitoring and health diagnostics of surge arresters are a necessary issue. Hence, as a feasible solution, a condition monitoring based on leakage current measurement techniques was selected to tackle the problem of age diagnostics of surge arresters. The Particle Swarm Search Algorithm was introduced as a method for extracting the third harmonic resistive component, Ir 3rd , from the total leakage current. To employ this method for extracting Ir 3rd , codes were developed in Matlab. The setting of Particle Swarm Search Algorithm was configured for extracting Ir 3rd and accuracy of 94% was obtained. Introduction Surge arresters, as protective equipment, are used to limit any over voltage in the power system. Operationally, a surge arrester discharges the surge current and protects the system and the equipment [1]. An equivalent circuit of a metal-oxide surge arrester can be modelled using a linear resistor Rg, a nonlinear resistor Rgb and a capacitor Cg [2]. Based on the equivalent circuit, at a normal working voltage a small leakage current flows in the surge arrester. This is because of the highly nonlinear V–I characteristic of the ZnO block [3]. Apart from the normal operating voltage system other parameters such as the level of external pollution and partial discharge activities cause degradation of the surge arrester block over time [4]. Prior research has illustrated that the leakage current flowing in the ZnO surge arresters causes degradation in surge arresters under normal voltage operation [5]. Therefore, various methods have been developed to monitor the condition of surge arresters and to predict probable interruptions caused by their degradation. Although the total leakage current composed of the capacitive and resistive components can be used as an indicator of age level for surge arresters, previous studies confirm that the total leakage current cannot be a trusted indicator for age diagnostics [6, 7]. Instead, the third harmonic of the resistive component in the total leakage current signal is considered as a better age indicator of the surge arrester [8, 9]. A complete experimental performance verification of metal-oxide surge arresters is not usually feasible because such testing requires expensive laboratory facilities. Therefore, performance verification requires reproducing operating conditions and meeting the necessities of the testing criteria. Consequently, a replacement model is required to be prepared to assess the surge arrester performance by measuring only the thermal conditions of the metal oxide surge arresters. Thus, in the developmental model, electrical properties of the surge arrester must be studied in correlation with its corresponding thermal properties. Currently, diagnostic methods of surge arresters use leakage current measurements for failure prediction, that is, the leakage current measurement is included in most methods of condition monitoring of gapless surge arrester [10]. Applied Mechanics and Materials Online: 2014-06-02 ISSN: 1662-7482, Vol. 554, pp 608-612 doi:10.4028/www.scientific.net/AMM.554.608 © 2014 Trans Tech Publications, Switzerland All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications, www.ttp.net. (ID: 161.139.222.35, Universiti Teknologi Malaysia UTM, Johor Bahru, Johor, Malaysia-15/02/16,11:50:52)