Affiliation:
1. State Key Laboratory of Electrical Insulation and Power Equipment School of Electrical Engineering Xi'an Jiaotong University Xi'an China
2. State Grid Shaanxi Electric Power Research Institute Xi'an China
Abstract
AbstractIn the past three years, a number of oil‐immersed ultra‐high‐voltage shunt reactors have experienced discharge defects when being put into operation, resulting in an overload of acetylene. However, detecting and identifying these discharge defects caused by switching impulse voltage is challenging under steady‐state conditions. This poses unpredictable and difficult‐to‐assess safety risks for the longterm operation of the equipment and subsequent transient processes. Hence, comprehending the discharge behavior of oil‐pressboard (PB) insulation under switching impulse voltage and devising a method to identify defects becomes crucial. This study focuses on investigating the frequent arc reignition (FAR) pattern exhibited by typical defects under both standard and oscillation switching impulse voltages. The objective is to uncover the mechanism behind FAR and propose a defect recognition strategy suitable for transient processes. The study reveals the FAR process will occur at least once during the breakdown process; the FAR phenomenon is the weakest in the surface defect with a weak vertical electric field. The average recovery voltage percentage and the average discharge interval of the FAR process decrease with increasing impulse amplitude or oscillation frequency. Additionally, the average number of discharges decreases with higher oscillation frequency, while it initially increases and then decreases with increaseing amplitude. Based on the analysis of the number of FAR processes and their variation in terms of amplitude or discharge interval, a method for recognizing oil‐PB defects during switching transient processes is developed and successfully applied to a case study involving acetylene overload in a 1000 kV shunt reactor.
Publisher
Institution of Engineering and Technology (IET)