多场耦合作用下GIS/GIL气固界面绝缘性能研究评述

Review on Insulation Performance at Gas-solid Interface in GIS/GIL with Multi-field Coupling

气体绝缘组合电器(GIS)和气体绝缘金属封闭输电线路(GIL)广泛应用于长距离、大容量输电线路中。但随着电压等级和输电容量的提升,气固界面性能面临着电荷积聚与固体绝缘子老化等挑战。为了降低绝缘故障的发生,亟需开展GIS/GIL气固界面绝缘机理、测量诊断、寿命评估等相关研究。文中首先总结了气固界面电荷积聚、消散机理以及影响沿面闪络的因素,结果表明温湿度、自由基等因素通过改变绝缘子电导率对沿面闪络造成影响。并针对在役运行绝缘子,着重分析了长期运行中多物理场耦合作用下GIS/GIL中气固界面绝缘失效与老化过程。论述了传统的电学、力学等诊断方法以及新兴的光学检测法、超声检测法以及分子模拟等技术,并分析了多种寿命评估模型的原理与适用范围。最后,提出气固界面绝缘性能评估需重点关注多物理场耦合下的绝缘性能退化过程,通过多种诊断方法提取性能退化的宏观性能与微观结构参数,利用机器学习等手段建立评估物理模型和寿命预测方法等建议。

Gas insulated switchgear(GIS)and gas insulated metal-enclosed transmission line(GIL)are widely used in long distance and large capacity transmission lines.However,with the increase of voltage level and transmission capacity,the performance of the gas-solid interface faces the challenge of charge accumulation and aging of solid insulators.In order to reduce the occurrence of faults of insulation,it is urgent to carry out such related research as the insulation mechanism,measurement diagnosis and life evaluation of GIS/GIL gas-solid interface.In this paper,the charge accumulation and dissipation mechanism at the gas-solid interface as well as the factors affecting the surface flash-over are summarized firstly.The result shows that such factors as the temperature,humidity and free radicals will affect the surface flash-over by changing the conductivity of insulators.For in-service insulators,the insulation failure and aging process of gas-solid interface in GIS/GIL under multi-physical field coupling are analyzed.Such traditional diagnostic methods as electricity and mechanics,as well as such emerging technologies as optical detection,ultrasonic detection and molecular simulation are described,and the principle and application range of various life evaluation models are analyzed.Finally,it is proposed that the evaluation of gas-solid interface insulation performance should focus on the insulation performance degradation process under multi-physical field coupling,extract the macro performance and microstructural parameters of the performance degradation through a variety of diagnostic methods,and set up the physical model and life prediction method of evaluation by using machine learning.