【作者】 谭志红；

【导师】 孙才新；

【作者基本信息】 重庆大学 ， 电气工程， 2011， 博士

【摘要】 h气体绝缘组合电器（Gas Insulated Switchgear,简称GIS）因稳定性好、占地面积小、可靠性高等优点,在电力工业中得到了广泛的应用。然而,GIS设备因安装、检修和长期运行等原因不可避免会存在各种内部绝缘缺陷,这些绝缘缺陷在长期运行中,有的会逐渐发展扩大,以致最终可能导致GIS设备绝缘故障。大量的研究表明,不同程度的绝缘缺陷大都会激发局部放电（Partial Discharge,简称PD）,可用检测PD的程度和特征来判断设备的绝缘状况。传统检测局部放电方法有脉冲电流法、超声法以及超高频法等,他们都有各自的特点,得到了不同程度的应用。但共同存在难以摆脱现场强烈电磁干扰的问题,使得他们用于在线检测的准确度受到质疑。由于非电检测不受电磁干扰,近年来,国内外对因PD导致SF₆气体发生分解的研究非常关注,初步研究结果表明,不同绝缘缺陷产生的PD使SF₆的分解特性有较大差异,且SF₆分解组分与绝缘缺陷之间存在一定关系,可用来诊断GIS绝缘状况,对提高GIS设备的稳定性具有较高的理论价值和重要的工程实用意义。因此,用检测SF₆的分解特性来诊断GIS内部绝缘缺陷已成为目前研究的热点问题。本文在课题组已有的研究基础上,开展了PD对SF₆分解特性和用于在线检测微量组分气体的碳纳米管传感器研究,初步提出了适用于SF₆组分分析的特征气体,得到了碳纳米管对SF₆分解组分的气敏特性,并从理论上计算和模拟了SF₆分解组分的吸附过程。具体工作如下:①构建了四种典型绝缘缺陷模型,对其进行了电场仿真。同时,设计了PD分解气室,建立了能对SF₆分解组分进行有效分离的气相色谱定量检测系统,并基于此平台,进行了不同典型绝缘缺陷产生PD下的SF₆分解气体实验,测量了分解组分中SOF₂、SO₂F₂、CF₄和CO₂的含量。研究表明,四种典型缺陷模型在预定位置和一定电压下均能产生PD,其中高压导体突出物缺陷产生的PD最稳定,并且其放电能量大,PD下SF₆产气量和分解速率高;自由导电微粒缺陷单次PD放电强度高,SOF₂、SO₂F₂两种气体组分含量较高;绝缘子表面金属污染缺陷由于PD发生时涉及与固体绝缘材料产生反应,故其CF₄含量增加明显;绝缘子外气隙缺陷虽然PD强度较高,但产气量相对较小。②不同绝缘缺陷产生的PD使SF₆发生分解的特性和产气的规律存在明显差异,初步提出将SOF₂/SO₂F₂、CO₂/CF₄、（SOF₂+SO₂F₂）/（CO₂+CF₄）三组气体含量比值作为统计特征量,用于对四种典型绝缘缺陷类型进行分析判断,指出可用SOF₂/SO₂F₂来表征PD强度及放电发展趋势,CF₄组分含量变化一定程度上可以说明GIS内绝缘材料的劣化状况。③采用叉指电极为基底,制备了SWNT-OH及本征SWNT碳纳米微量气敏传感器,对其进行了SOF₂、SO₂F₂、SO₂、CF₄四种分解组分的气性敏实验,对比了两者的气敏特性,重点研究和分析了SWNT-OH传感器对分解组分的的气敏特性、稳定性和恢复特性。结果表明,对于四种分解组分,SWNT-OH对SO₂响应速度最快,灵敏度最高,电阻变化率与SO₂的气体浓度满足一定的线性关系,且在空气中可以自恢复,恢复时间短,可初步满足检测稳定性要求。④利用Materials Studio（MS）建立了SF₆分解组分（SOF₂、SO₂F₂、SO₂、CF₄）与单壁碳纳米管的微观吸附模型,并根据密度泛函理论（DFT）对吸附过程进行理论仿真和计算,分析了羟基修饰单壁碳纳米管（hydroxylation single-wall carbon nanobutes,SWNT-OH）的微观气敏机理。首先,根据计算结果对SWNT-OH与四种分子的吸附能力进行了判断;其次,根据分子前线轨道理论分析了SWNT-OH与四种分子发生吸附作用的难易程度;再次,根据能隙计算结果研究了气体分子吸附在SWNT-OH上对其表面载流子转移能力的影响,并结合态密度计算结果分析了SO₂与SWNT-OH吸附过程中的原子轨道杂化情况;最后,将SWNT-OH与本征SWNT的计算结果进行对比,结果表明,四种分子中,SWNT-OH最容易与SO₂发生吸附作用,吸附能力最强,且吸附SO₂后能隙减小,导致其电子转移能力提高;另外,研究发现两者原子的部分轨道在吸附过程中,趋于杂化形成的电子输运通道有利于两者之间的电子转移,SWNT-OH对四种分子的吸附能力高于本征SWNT的吸附能力。更多还原

【Abstract】 Gas Insulated Switchgear （GIS） is widely used in electrical industry for its stability and compactness. However, there might be insulation defects formed in the processes of manufacture, transportation, installation and long term of operation. Those defects will develop and may finally lead to serious insulation faults. It has been shown by lots of research that most insulation defects lead to partial discharge （PD）. Detecting the characteristics of PD can reflect the insulation condition. The traditional methods to detect PD include pulse current method, ultrasonic method, UHF method, and so on. Each of them has its own features, and has been applied at different degree. However, the traditional measurement method is difficult to get rid of the electromagnetic disturbance. Because of what, the online detection veracity is doubtable. Since the non-electric methods don’t affected by electromagnetic disturbance, SF₆ decomposition components detection has drawn lots of attention recently. Preliminary study has shown that SF₆ decomposition characteristics are different in various insulation defects. And the SF₆ decomposition components are related to the insulation defect types, which make it possible to detect the insulation condition by analyzing the components. Study on this method has great value in both theory and engineering. In this article, based on the reaches have been done, SF₆ decomposition characteristics were studied, and a carbon nanotube gas sensor for trace gas online detection is developed. In the study, the characteristic components of SF₆ decomposition were selected preliminary, and the micro-sensing mechanisms and macro-sensing characteristics of carbon nanotube gas sensor have been investigated through theoretical simulation and experiments. The specific work is as follows.1) Four kinds of physical models of typical insulation defects were constructed and the distribution of electric field of these models was analyzed by Finite Element Simulation. A PD decomposition chamber was built. A gas chromatography detection system was built to separate and analyze the components. In these four models the decomposition experiments were carried out, SOF₂, SO₂F₂, CF₄ and CO₂ from the decomposition components were detected. The result shows that PDs happen in all four models at the certain positions and expected voltage ranges. The N-type defect has large discharge energy and high values in decomposition quantities and speed. In the P-type defect, every PD pulse has a high altitude. And the value of SOF₂, SO₂F₂ is high. The M-type defect produces more and more CF₄ as experiment time goes on. The G-type defect shows a high PD level but produce little decomposition.2) Different defects make the decomposition characteristics quite different. It was suggested preliminary to use SOF₂/SO₂F₂, CO₂/CF₄ and （SOF₂+SO₂F₂）/（CO₂+CF₄） three values as the characteristic quantities to identify these four kinds of defects. SOF₂/SO₂F₂ was explained to represent the PD intensity. And the content of CF₄ can reflects the level of insulation ageing.3) Using interdigitated electrodes printed circuit board as the base SWNT-OH and SWNT sensors were made. SOF₂, SO₂F₂, CF₄ and SO₂ four components were experimented on the sensors. After comparing their performances, we paid special attention to SWNT-OH. Its sensitivity, stability and recovery characteristics were studied. The result shows SWNT-OH sensor has high reacting speed and high sensitivity to SO₂ in those four components. It satisfied certain linear relations between the resistance change rate of SWNT-OH and SO₂ gas concentration. And the SWNT-OH sensor can self-recovery in the air costing a short time. To a certain extent, the SWNT-OH sensor meets the stability requirements.4) Adsorption models, between four types of gas components （SOF₂, SO₂F₂, SO₂, CF₄） and SWNT, SWNT-OH, have been established by Materials Studio. Based on density functional theory （DFT） theoretical calculations were carried on. The micro-sensing mechanism of SWNT-OH was analyzed. First, the absorption capacity of SWNT-OH on the four types of gas molecules was judged according to the calculation results. Second, the difficulty levels of interaction between SWNT-OH and four types of molecules have been analyzed according to the molecular frontier orbital theory. Third, how gas molecules adsorbed on the SWNT-OH to influence electron transfer ability was investigated according to the results of energy gaps. Combined with the results of density of states, the hybridization of atomic orbitals was also investigated during the adsorption process between SWNT-OH and SO₂. Finally, the calculation results of SWNT-OH were compared with that of SWNT. Simulation results show that SO₂ was the most easily adsorbed on SWNT-OH, and the adsorption capacity of SWNT-OH on SO₂ was the largest among the molecules; the energy gap of SWNT-OH was reduced after adsorbing SO₂, which improved its ability of electron transfer; during the adsorption process, atom partial orbits tend to hybridization between SO₂ and SWNT-OH, which could help form electron transport channel, conducive to electron transfer between SWNT-OH and SO₂. Simulation results also indicated that the adsorption capacity of SWNT-OH on gas components was larger than that of SWNT.更多还原