变压器油
电介质
离子
材料科学
铜
分析化学(期刊)
介电强度
击穿电压
介电常数
饱和(图论)
化学
变压器
电压
冶金
光电子学
电气工程
有机化学
工程类
数学
色谱法
组合数学
作者
Kaibin Liang,Feng Wang,Lipeng Zhong,She Chen,Qiuqin Sun,Changhao Hu,Xinyu Duan,Tao Peng
出处
期刊:IEEE Transactions on Dielectrics and Electrical Insulation
[Institute of Electrical and Electronics Engineers]
日期:2023-12-22
卷期号:31 (3): 1305-1313
被引量:1
标识
DOI:10.1109/tdei.2023.3345851
摘要
Corrosion and electrical erosion of metals within transformers lead to the presence of metal ions in the oil. Unlike metal particles, these soluble ions diffuse readily in the oil and are challenging to filter out, posing a significant threat to the transformer's reliable operation. This paper employs experimental evaluations to elucidate the effects of copper ions on the dielectric properties, breakdown performance, and partial discharge characteristics of insulating oil, coupled with a theoretical analysis. Dielectric spectroscopy test reveals that soluble copper ions alter the dielectric properties of the oil, specifically causing a reduction in its relative permittivity due to ion-induced endogenous electric fields (about 60% decline in the low frequencies). Additionally, the ions result in increased dielectric losses at low frequencies and notably enhanced conductivity at high frequencies, potentially due to electron tunneling and thermally activated polarization. AC Breakdown tests indicate that the breakdown voltage of the oil decreases as copper ion concentration increases. The relatively lower ionization potential and charge saturation effects are identified as contributing factors to the decreased breakdown voltage. Partial discharge experiments confirm that copper ions lead to reduced inception voltages (with a maximum decline of 17.5% in the test), and the frequency of discharges between needle-plate gaps in the insulating oil increases with ion concentration. The findings offer critical insights into the influence of metal ions on the electrical properties and discharge behavior of insulating oils.
科研通智能强力驱动
Strongly Powered by AbleSci AI