Darbe Geriliminde Sıkıştırılmış Gazlarda Delinme

Abstract

Darbe geriliminde gazların delinmesi üzerine yapılan bu çalışmada düzgün olmayan alan dağılımına sahip elektrot sistemlerinden biri olan küre- düzlem elektrot sistemi üzerinde çalışılmıştır. Uygulamada ortaya çıkan elektrot yüzey pürüzlülüğünün delinme gerilimine etkisi, küre-düzlem elektrot sisteminin düzlem elektrodu üzerinde boyutları, biçimi ve konumu belirli bir pürüzün varlığı ile göz önüne alınmıştır. Çalışmada, pürüzlülüğün etkisinin araştırılması planlandığından incelemeler, makroskobik boyutlarda yan küresel iletken tek bir pürüz ile gerçekleştirilmiş ve pürüzün olması ve olmaması durumları değerlendirilmiştir. Bu çalışmada, darbe geriliminde hava, N2 ve SFe gazlan ile SF6+N2 gaz kanşımlannda, düzgün olmayan alanda, pürüzsüz ve pürüzlü elektrot sistemlerinde, elektrot açıklığının, gaz basıncının, gerilim kutbiyetinin ve pürüzün varlığuun delinme gerilimlerine etkileri incelenmiştir. Çalışmada ele alınan küre-düzlem ve küre-pürüzlü düzlem elektrot sistemlerine ilişkin elektrik alan dağılımları, hem analitik hem de sayısal bir yöntem olan sonlu elemanlar yöntemi ile kuramsal olarak incelenmiştir. Darbe gerüiminde, Ua %50 atlama gerilimleri, literatürde önerilen yöntemler geliştirilerek, en küçük kareler yöntemine göre birinci dereceden eğri (doğru) uydurma yolu ile belirlenmiştir. Gaz yalıtkanların, uygulamada farklı basmç ve elektrot açıklıklarında kullanılmaları göz önüne alınarak basınç ve elektrot açıklığı parametre olarak seçilmiş ve basmç 0-4 bar aralığında, elektrot açıklığı ise 5-25 mm aralığında değiştirilmiştir. Deneysel çalışmalar, ortam sıcaklığında yapılmıştır. Darbe geriliminin kullanılması ile hava, N2 ve SF6 ile SF6+N2 karışımları ile yalıtılmış, alan dağılımı düzgün olmayan pürüzlü ve pürüzsüz elektrot düzenlerinin, yıldırımdan kaynaklanan darbe karakterindeki aşın gerilimlere karşı davranışını ve bu davranışta aşın gerilimin kutbiyetinin etkisini görme olanağı elde edilmiş ve her iki kutbiyetin ayn ayn değerlendirilmesi gerektiği gösterilmiştir.

Particularly, recent demands for more compact apparatus are making it necessary to design the insulated equipment with higher electrical stresses than conventional equipment This makes it an important requirement to attain higher insulation strength and higher insulation reliability. The general criteria for judging the quality of gas insulation can be arranged according to their degree of importance: - high dielectric strength, - good behavior in inhomogeneous electric fields, - a low level of induced over voltages during electric discharge, - high relative impulse strength, - great chemical stability, reversibility after the electric arcs, - high vapor pressure, - low toxicity and - acceptable price. To date, no insulator has been found to have properties better than sulphur hexafluoride (SFg). Therefore, SF6 is the most commonly used insulating gas. SFe has a high electric strength, low toxicity, is chemically inert and has good heat transfer properties. Its relatively large cross section for attaching low-energy electrons under electrical stress inhibits the initiation and growth of electrical discharges. These properties allow a reduction in size and enhance the reliability of high voltage equipment. However, when compressed SFe was first introduced as an electrical insulant in high voltage power equipment, it was discovered that the insulation strength of the system was less than that predicted by theory. An explanation for this apperent reduction in the dielectric strength of compressed SF6 can be given by considering the perturbations of the macroscobic electric field produced either by the microscobic roughness of the elecrode surface or by conducting particles. In SFö insulated equipment, the dielectric stregth is often lowered by factors of 2 to 10 by unavoidable contaminating particles, especially of long slender shape and conducting material. That such perturbations of the electric field should play a major role in the dielectric VII strength of SFg is due to the effective coefficient of ionization ct being a strongly varying function of the electric field strength E, particularly for E=Eiim is the value for which â=0. On the other hand high cost and toxic byproducts can be remarked as the other major disadvantages of SFe. Therefore, owing to the stated deficiencies, there has been much effort to develop alternative electronegative gases and gas mixtures with improved insulating behavior at reduced costs. As is well known, SFg is a strongly electronegative gas, but it can only remove free electrons effectively with energies in the low-energy range by the electron attachment process. On the other hand, the electron attachment cross sections for some perfluorocarbons such as G^Fg are substantially higher than that for SFg in the higher energy range. Moreover, many perfluorocarbons have higher dielectric strength than SFö. Therefore, gas mixtures comprising SFg may have a higher dielectric strength than pure SF6. Sulphur-hexafluoride (SFö) and its mixtures with other less expensive gases such as air, N2, and CO2, etc. are being investigated in recent years. Breakdown characteristics of SFğ mixtures with N2, air, and CO2 are reasonably well understood for uniform and quasi-uniform field gaps. However, there is still insufficient information regarding the impulse breakdown of such mixtures in non-uniform field gaps. Impulse breakdown characteristics of negative rod-plane gaps have been investigated for SF6-N2, SFe-air, and SF6-H2 mixtures. Experimental investigations on to the effect of electrode surface roughness upon the dielectric strength of strongly electronegative gases have adopted two different approaches, viz. (a) the use of geometrically well- defined artificial protrusions mounted on plane electrodes and (b) the use of an electrode having a surface finish comparable with that related to production processes, but which in comparison to (a) is ill-defined. The latter approach represents, with respect to insulation strength, the normal operating situation and thus results obtained from any investigation into the dielectric behavior of a gas are directly applicable to practical situation. Owing to the complex structure of the electrode surfaces utilized, we are unable to derive any information about; the associated field perturbations, and hence a deeper analysis of the results cannot be undertaken. The former approach provides a well-defined perturbation of the macroscobie electric field, and thus lends itself to an analysis of, for example, onset conditions. An analysis can be used to determine the relevant dimensions of artificial protrusions. These dimensions are found to be comparable with those of the roughness associated with production processes. To-date however, the viii dimensions of the artificial protrusions employed have all been much greater. This suggests an inability to obtain protrusions of the relevant dimensions. Consequently, as available artificial protrusions are unrepresentative of artificial conditions, results from investigations using such protrusions are not necessarily applicable to practical conditions. In this thesis the strength of several gas and gas mixtures, under lightning overvoltages was experimentally investigated. The experiments were carried out with sphere-plane field configurations placed in a plexiglas pressure vessel. The sphere electrode, which had a radius of 2 mm was made of brass. The lower plane electrodes were made of rounded brass plates of 75 mm diameter and one of them was with a protrusion on the center. The electrode gap spacings in investigation were varied within a range of 5 mm to 25 mm. The gap spacings were set under pressure using special measuring scales from outside the vessel. The partial pressures as well as the total pressure were measured using mechanical gauges with an accuracy of 1%. The gas mixtures were left at least 2 h before the application of the tests in order to obtain uniform gaseous mixture. Prior to each experiment, the electrodes were treated with metal polish and cleaned carefully with CCI4. The vessel was then evacuated to a pressure less then 10"1 ton* for thirty minutes. And then the vessel was filled with gas to a certain pressure. In order to condition the electrode surfaces, 10 unrecorded impulses were applied at the beginning stage of each series of experiments. Sufficient time (~30 sec.) is allowed between each flashover test so that the gas or gas mixture could completely recover to its original state after flashover. The ambient temperature during the experiments were about 20±6 °C. The experimental pressures extended over a range from about 0 to 4 bars corresponding to the pressure ranges of practical equipments. The 50% breakdown values were obtained using the "up and down" method and curve fitting method of first order, based on the least mean squares. At each pressure, impulse breakdown voltages for both negative and positive polarities were measured. Variations of positive and negative impulse breakdown voltages with gap spacings for the sphere-plane electrode arrangement at 4 atm are shown in Fig. 1 and Fig. 2, respectively. And also the variations of positive and negative IX 200 UbpW) 180 160 140 120 100 SO 40 - 20 L *Air 0 5 10 15 20 25 d (mm) Fig. 1. Curves of Ub=f(d). (p=4 atm.) 200 ub(KV) 180 160 140 120 100 en 20 ? %60S% 0 5 10 15 20 25 d (mm) Fig. 2. Curves of Ub=f(d). (p=4 atm.) 200 ub(kV) 180 160 140 120 100 40 20 ? *Air 0 5 10 15 20 25 d (mm) Fig. 3. Curves of Ub=f(d). (p=4 atm.) 200 Ub(kV) 180 160 - 140 120 100 - 60 40 - 20 %40S% *%20SF6 «Nj ? Air 0 5 10 15 20 25 d (mm) Fig. 4. Curves of Ub=f(d). (p=4 atm.) XI impulse breakdown voltages with gap spacings for the sphere-plane electrode with protrusion arrangement at 4 atm are shown in Fig. 3 and Fig. 4, respectively. It is clear from Fig. 1 and 3 that positive impulse breakdown voltages of SFg-N2 mixtures increase with increasing SFg content, and the breakdown voltages of N2 and SFg being the lowest and the highest respectively. Experimental Set Up A brief description of the arrangement is as follows: Gas Supply System Gas supply system comprises a vacuum pump, a compressor and several valves and manometers. Pressure Vessel The pressure vessel is made up of cylindrical insulating perspex tube of 120 mm diameter and exhibits an inner volume of 5.5 liters. It can be evacuated to a pressure less than 0.8xl0"4 bar and pressurized up to 5 bar. The bottom base and the top lid are made of aluminum with edges rounded off to avoid the formation of corona discharges. Investigations can be carried out by using this transparent perspex vessel up to the crest value of 250 kV, which has an advantage of visual observation of the flashovers and breakdown from outside and requiring no bushing. Impulse Voltage Generator The impulse voltages were supplied from an 260 kV, 100 Ws (J) two stage impulse generator. The impulse voltage wave shape was L2/50 |is. Both negative and positive impulse voltages were used in the study. The measuring system included a 2x140 kV (1200 pF) capacitive voltage divider and a storage impulse voltmeter (MWB type DSTM).