IMPULSE BREAKDOWN OF SHORT ROD-PLANE GAPS WITH A ROD COVERED WITH DIFFERENT DIELECTRIC MATERIALS

Size: px

Start display at page:

Download "IMPULSE BREAKDOWN OF SHORT ROD-PLANE GAPS WITH A ROD COVERED WITH DIFFERENT DIELECTRIC MATERIALS"

Valentine Reynolds
6 years ago
Views:

1 16th International Symposium on High Voltage Engineering, Cape Town, South Africa, 2009, paper No. 271 IMPULSE BREAKDOWN OF SHORT ROD-PLANE GAPS WITH A ROD COVERED WITH DIFFERENT DIELECTRIC MATERIALS P. N. Mavroidis*, P. N. Mikropoulos, C. A. Stassinopoulos and M. Zinonos High Voltage Laboratory, School of Electrical & Computer Engineering, Faculty of Engineering, Aristotle University of Thessaloniki, Building D, Egnatia St , Greece * pmavr@auth.gr Abstract: The breakdown mechanism of short rod-plane gaps with a dielectric covered rod under positive standard lightning and switching impulse voltages is investigated. Two different dielectric covers, made of PTFE and epoxy resin, were used and the simple air gaps with a bare rod were regarded as reference. Breakdown probability distributions were obtained and the time and voltage to breakdown were measured. Under lightning impulses the increase of the dielectric strength due to the dielectric cover is little dependent upon the cover material and only slightly higher to that of a simple air gap including the cover length. Under switching impulses this increase is greater; the presence of a dielectric cover results in a reduced effect of the impulse waveshape on the dielectric strength. Most of the times, the spark channel develops along the dielectric cover under lightning impulses whereas, under switching impulses it develops either partly or solely in free air away from the cover surface. The increase in the dielectric strength of the gaps due to the dielectric cover is associated with the effects on the discharge development pattern of the coronas preceding breakdown. 1. INTRODUCTION In modern high voltage equipment, there are increased economic and environmental requirements to be met, in terms of improved efficiency, higher reliability and lesser environmental impact. One of the modern trends for improving the insulation performance of gas insulated systems is the use of composite gas/polymer insulation by employing either dielectric barriers or covered electrodes [1]. Composite insulation can improve the compaction of the design for equipment, especially where high current rating at reduced insulating distances is required. This is beneficial in switchgear technology [2], and may also result in a reduction of the use of SF 6 [3, 4], which is considered a gas contributing to global-warming. Several studies have shown the improved dielectric performance of air gaps employing dielectric covered electrodes. In a covered plane-parallel electrode system under dc or lightning impulses the increased dielectric strength of the gap was attributed to charge formation at the electrode surfaces as well as in the air gap [5]. An improved dielectric behaviour due to dielectric covers has also been observed in rod-plane configurations [6-15], where the high electric field values in the vicinity of the rod tip may approximate the enhanced localized field values commonly found in insulation systems. The dielectric behaviour of a covered rod/plane gap depends on the electric field distribution in the gap as modified by space and surface charges associated with prebreakdown phenomena. Therefore, it would depend on the combined effect of the electrode arrangement [6, 7, 9-15], cover material and geometry [6, 7, 9-11], applied voltage waveshape [14] and polarity [6-10, 12], and air pressure [9]. Charge accumulation on the dielectric cover surface deposited by coronas preceding breakdown, may affect the discharge development pattern at breakdown [6, 7, 12]. Actually, breakdown occurs through both surface and air discharges [6, 8, 10], however, cases where breakdown occurred in air alone through a spark channel developing away from the dielectric cover [13-15], or after puncture of the dielectric cover at the limiting boundaries of the upward development of surface discharges along the rod [6] have also been observed. This study investigates breakdown in short dielectric covered rod/plane air gaps by considering as influencing parameters the gap length, waveshape of the applied positive impulse voltage and the cover material. An increase in the dielectric strength has been observed due to the dielectric cover; this is discussed based on the discharge development pattern at breakdown as influenced by the preceding coronas. 2. EXPERIMENTAL ARRANGEMENT 2.1. General settings The electrode configuration consisted of a cylindrical brass rod, 1.2 cm in diameter, with a square-cut tip hanging over an earthed aluminum plane 100 x 200 cm placed 80 cm above the floor of the laboratory. The gap spacing varied between 2.5 and 15 cm. A two-stage Marx generator 280 kv/ 0.5 kj was used to produce standard lightning and switching impulse voltages, stressing the rod. The latter was used in the experiments either bare or with a dielectric cover, made either of PTFE or epoxy resin with dielectric constants 2.1 and 3.9 respectively; their thickness was 0.4 cm and their length 12.7 cm (Figure 1). The dielectric covered rod/plane gaps are hereafter called composite; the simple air gaps were regarded as reference.

probability (p.u.) 1.0 0.8 0.6 0.4 LI, Bare SI, Bare LI, PTFE SI, PTFE LI, Epoxy SI, epoxy 0.

2 probability (p.u.) LI, Bare SI, Bare LI, PTFE SI, PTFE LI, Epoxy SI, epoxy 0.2 Figure 1: Schematic diagram of the electrode arrangement; numerical values are given in mm, L designates gap length Applied voltage (kv) Figure 2: Typical breakdown probability distributions; 7.5 cm gap length, fitting curves are drawn according to normal distribution. A Meek and Collins probe [16], positioned at the centre of the earthed plane, and a 400 MHz digital oscilloscope enabled the monitoring of the electric field strength at the earthed plane. The probe, with a circular area of 1.2cm in diameter, was covered by a perspex dome to prevent corona streamers from terminating on its surface. All the voltages were corrected to standard air density according to IEC [17]. Absolute and relative humidity varied between g/m 3 and 60 90%, respectively Measurement procedures Breakdown probability distributions were obtained by using the multiple level tests method [17]. Each voltage level, differing less than 3% from the previous one, consisted of 20 impulses applied at time intervals of around 60 s. Typical breakdown probability distributions are shown in Figure 2. The probability distributions were found to be well approximated by the normal distribution; thus, from each distribution the mean value and the corresponding standard deviation were computed. Only a few cases deviating from the normal distribution were observed; most of them referred to composite gaps. Typical oscillograms of the electric field strength at the earthed plane obtained at applied voltages causing approximately 50 % breakdown are presented in Figure 3. Corona occurrence is indicated by the sharp rise in the probe voltage, due to the associated positive charge injected in the gap. Part of this charge may be neutralised at the earthed plane as corona streamers reach the plane; indicative of the latter are the sharp drops in the probe voltage. Breakdown is clearly indicated in the field oscillograms by the collapse of the electric field, designated as t b in Figure 3. At each impulse voltage application, from the oscillograms of the electric field at the plane, the time to breakdown was measured and the instantaneous breakdown voltage was calculated after cross referencing the times with the impulse waveform. For these characteristic discharge parameters the mean value and the corresponding standard deviation per applied Figure 3: Electric field strength at the earthed plane; 10 cm gap length, 50% breakdown probability, t b designates time to breakdown (300 kv/m/div). (a) Simple gap, lightning impulse, 93 kv (1μs/div) (b) Epoxy resin cover, lightning impulse, 163 kv (1μs/div) (c) Simple gap, switching impulse, 78 kv (50μs/div) (d) Epoxy resin cover, switching impulse 174 kv (50μs/div) voltage level were calculated. Salient characteristics of coronas have also been measured; relevant results will be presented elsewhere. 3. EXPERIMENTAL RESULTS 3.1. Breakdown characteristics Figure 4 shows the variation of the 50% breakdown voltage, U 50, with the gap length for all cases studied; the corresponding standard deviation, σ, is listed in Table 1. There is a significant increase in U 50 due to the dielectric cover, becoming greater with decreasing gap length and under switching than lightning impulses. The highest U 50 values among all configurations were obtained for the case of epoxy resin cover under switching impulses. These results are better demonstrated in Figure 5a where the per unit increase in U 50 is plotted as a function of gap length. Under lightning impulses this increase can be expressed by a factor in the range of 3 down to 1.5 decreasing with gap length. These factor values compared well with the values of and found in [7] and [11] respectively, and with the values of and derived from [9] and [10], respectively, which refer to

3 similar to the present electrode configurations under positive lightning impulses. Under switching impulses the relative increase of the dielectric strength of the gap due to the dielectric cover is greater, between ~3.5 down to ~1.9. Also, from Figure 5b, showing the variation of the ratio of the U 50 obtained under lighting impulses to the U 50 obtained under switching impulses with gap length, it is obvious that there is a reduced effect of the impulse waveshape on U 50 in composite gaps as the ratio of U 50 is smaller for composite than simple gaps, specifically around unity and 0.9 for the PTFE and epoxy resin covers, respectively, for all gap lengths. voltages causing approximately 50% and 100% breakdown. Under lightning impulses and with the exception of the shortest gap the insertion of a dielectric cover results to longer breakdown times (Figure 6) despite the fact that breakdown occurs at significantly higher voltages (Figures 7); this is more pronounced for the case of the PTFE cover. It is noteworthy that the spread in the values of the time to breakdown is generally higher for composite gaps, especially under switching impulses. Concerning the instantaneous breakdown voltages, similar results to those obtained for U 50 as regards to the gap length and impulse waveshape can be observed from Figure 7. Figure 4: 50% breakdown voltage as a function of gap length; vertical bars represent σ. Table 1: Standard deviation σ% of the breakdown probability distributions. Gap length (cm) Bare LI PTFE Epoxy Bare SI PTFE Epoxy Figure 5: The effect of the insulating cover on the dielectric strength of the gap with the impulse waveshape and cover material as parameters Concerning σ, from Table 1 it can be deduced that composite gaps display in general smaller values under lightning impulses, especially in the case of the epoxy resin cover; an opposite trend is apparent under switching impulses. Figures 6 and 7 show the time to breakdown and the associated instantaneous breakdown voltage, respectively, as a function of gap length, at applied Figure 6: Time to breakdown as a function of gap length; vertical bars represent σ Figure 7: Instantaneous breakdown voltage as a function of gap length; vertical bars represent σ 3.2. Electric field calculations Electric field calculations have been performed via the Comsol multiphysics software package. Table 2 shows the maximum geometric electric field values in the gap and Figure 8 is a typical plot of the geometric field strength along the axis of symmetry of the gap; the latter is not the maximum field line within the gap, however, it is indicative of the effect of the dielectric cover on the electric field distribution. The geometric electric field strength attains always its maximum values along the circumference of the rod tip or the dielectric cover for the simple and composite gaps, respectively. The dielectric cover reduces the maximum field in the gap by a factor of ~ 7.3 and ~3.8 for the case of the PTFE and epoxy resin covers, respectively (Table2). However, after a short distance from the rod tip, the geometric field strength, reducing significantly with distance from the rod tip, becomes slightly higher for composite gaps (Figure 8). Also, the maximum geometric field strength in the vicinity of the cover upper end was found roughly equal for all gap lengths ~48 and ~90(V/m)/V for the

Table 2: Calculated maximum geometric electric field strength (kv/m/v) Gap length (cm) 2.5 5 7.5 10 12.

Under switching impulses, the initial coronas emerging at the vicinity of the cover tip during the rising front of the impulse voltage being of lesser growth as they initiate at lower voltages result

sufficient to cause 50% breakdown. Consequently, higher voltages must be applied for secondary corona to initiate, thus also for breakdown to occur.

Typical photographs showing the various spark trajectories observed in this investigation are shown in Figure 10.

4 Table 2: Calculated maximum geometric electric field strength (kv/m/v) Gap length (cm) Bare PTFE Epoxy and also shown in [13-15]. Under switching impulses, the initial coronas emerging at the vicinity of the cover tip during the rising front of the impulse voltage being of lesser growth as they initiate at lower voltages result in lesser extent of development of the upward discharges; this is obvious when comparing Figures 9b and 9c, which show still photographs of coronas at withstand cases both at applied voltages sufficient to cause 50% breakdown. Consequently, higher voltages must be applied for secondary corona to initiate, thus also for breakdown to occur. Under switching impulses most of the times at breakdown the discharge develops either partly or solely in free air away from the cover surface. Typical photographs showing the various spark trajectories observed in this investigation are shown in Figure 10. Figure 8: Geometric field strength along the axis of symmetry PTFE and epoxy resin covers, respectively. 4. DISCUSSION In composite gaps the electric field controlling the discharge development at breakdown is determined, besides the geometry of the electrode arrangement, by space and surface charges associated with corona discharges. Actually, as was also discussed in [13-15], a first corona emerging in air at the vicinity of the cover tip is followed by coronas emerging along the cover surface, developing partly along the dielectric cover and in air alone (Figure 9). These subsequent coronas, termed hereafter upward discharges, increasing the electric field at the cover upper end, determine the inception conditions of a secondary corona. The latter emerging at the upper end of the cover may eventually result in breakdown, through a spark bridging the gap that includes the dielectric cover length. The extent of development of the upward discharges along the cover surface would depend upon the electric field at the cover surface, thus upon the applied voltage (Figure 9a and 9b), and upon the surface charge induced by the first and upward discharges. This charge is likely to be negative as both first and upward discharges are positive, as has also been shown in similar to the present electrode arrangement under positive lightning impulse [12]. At breakdown, under lightning impulses the first corona is of ample growth allowing upward discharges to develop along the full cover length; this results in breakdown to occur through a spark bridging the gap most of the times along the surface. This is supported by the fact that under lightning impulses the dielectric strength of a covered rod/plane gap is only slightly higher to that corresponding to a simple air gap which includes the cover length, as can be derived from Figure Figure 9: Still photographs of corona discharges at withstand cases; epoxy resin cover, 10 cm gap length. Figure 10: Typical spark trajectories of composite gaps. Charge accumulation on the surface of a dielectric cover [6, 7], [12] and on cylindrical insulating surfaces bridging a rod-plane gap [18-23] associated with coronas preceding breakdown has been argued to exist. In the present experimental configuration, accumulated positive charge at the upper cover end may explain cases where the spark channel develops away from the dielectric cover in free air. Apparently, there is a need for further investigations, by employing also a surface charge measuring technique, to better understand the breakdown mechanism of the dielectric covered rod/plane gaps. 5. CONCLUSIONS The dielectric strength of composite rod-plane gaps with dielectric covered rod increases significantly with respect to simple air gaps. This has been demonstrated for lightning and switching impulses and for two

5 dielectric cover materials, specifically PTFE and epoxy resin. This increase in the dielectric strength is associated with the effects on the discharge development pattern of the coronas preceding breakdown. Under lightning impulses the increase of the dielectric strength is little dependent upon cover material, and only slightly higher to that corresponding to a simple air gap which includes the cover length. Under switching impulses this increase is greater; the presence of a dielectric cover results in a reduced effect of the impulse waveshape on the dielectric strength. The spark trajectory is affected by space and surface charges associated with corona discharges preceding breakdown. Most of the times at breakdown the discharge develops along the dielectric cover under lightning impulses whereas either partly or solely in free air away from the cover surface under switching impulses. 6. ACKNOWLEDGMENTS Mr. P. N. Mavroidis wishes to thank the Greek State Scholarship s Foundation for the support afforded by a merit Scholarship. 7. REFERENCES [1] L. Ming, A. Jaksts, L. Rongsheng, F. Owmann, and M. Leijon: Recent trends within insulation systems for transmission and distribution apparatus, in Proc. of 6th International Conference on Properties and Applications of Dielectric Materials, Xi an, June 2000, Vol. 2, pp [2] T. Shioiri, J. Sato, T. Ozaki, O. Sakaguchi, T. Kamikawaji, M. Miyagawa, M. Homma and K. Suzuki: Insulation technology for medium voltage solid insulated switchgear, Annual Report Conference on Electrical Insulation and Dielectric Phenomena, Albuquerque, October 2003, pp [3] J. Sato, T. Shioru, M. Miyagawa, T. Yoshida and K. Yokokura: Composite insulation technology for new compact 72/84kV C-GIS, IEEE Transmission and Distribution Conference, New Orleans, April 1999, Vol. 2, pp [4] N. Masaki, T. Yoshida, K. and Kato: Compact 66/77kV C-GIS That Meets Space Saving Requirements, Toshiba review, Vol. 52, no 9, pp , [5] H.J.M. Blennow, M.L.-A. Sjoberg, M.A.S. Leijon and S.M. Gubanski, Electric field reduction due to charge accumulation in a dielectric-covered electrode system, IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 7, No. 3, pp , June [6] G. Baldo and G. Pesavento: Impulse performance of air gaps in series with thick insulating layers, in Proc. of 5th International Symposium on High Voltage Engineering, Braunschweig, August 1987 paper no [7] L. Ming, U. Fromm, M. Leijon, L. Walfridsson, and A. Vlastos: Insulation performance of covered rod/plane air-gap under lightning impulse voltage, in Proc. of 10th International Symposium on High Voltage Engineering, Montreal, August 1997, paper no [8] U. Fromm, L. Ming, L. Walfridsson and D. Windmar: Temporal breakdown development of a covered rod/plane arrangement, in Proc. of 11th International Symposium on High Voltage Engineering, London, August 1999, Vol. 3, pp [9] T. Mizuno, K. Morita and Y. Kurata: The electrical performance of air or nitrogen gas with solid insulation and the application for switchgears, Transmission and Distribution Conference and Exhibition 2002: Asia Pacific, IEEE/PES, Yokohama, October 2002, Vol. 3, pp [10] T. Mizuno, K. Morita, Y. Kurata, K. Nagatake., and H. Saitoh: Electrical insulation performance in air gap with covered electrode, in Proc. of 7th International Conference on Properties and Applications of Dielectric Materials, Nagoya, September 2003, Vol. 1, pp [11] S. Stangherlin, C. Rein, G. Salge, and F. Koenig: Effect of different dielectric coatings in divergent fields, in Proc. of X International Symposium on Gaseous Dielectrics, Athens, March 2004, pp [12] F. Mauseth, A. Nysveen, and E. Ilstad: Charging of dielectric barriers in rod-plane gaps, in Proc. of 8th International Conference on Solid Dielectrics, Toulouse, July 2004, Vol.1, pp [13] P.N. Mavroidis, P.N. Mikropoulos, C.A. Stassinopoulos, A. Dodos, and P. Zannias: Discharge characteristics in short rod-lane gaps with dielectric-covered rod under lightning impulse voltages, in Proc. of XVII International Conference on Gas Discharges and their Applications, Cardiff, September 2008, pp [14] P.N. Mavroidis, P.N. Mikropoulos, C.A. Stassinopoulos, P. Rafailidis, and G. Smaragdakis.: Impulse Breakdown of Short Rodplane Air gaps with a Dielectric Covered Rod, in Proc. of 43rd Int. Universities Power Engineering Conference, Padova, September 2008, paper no [15] P.N. Mavroidis, P.N. Mikropoulos, C.A. Stassinopoulos: Lightning impulse behaviour of short rod-plane gaps with a dielectric-covered rod, IET Science, Measurement & Technology, 2008, submitted. [16] J.M. Meek, and M.M.C. Collins: Measurement of electric fields at electrode surfaces, Electronics Letters, Vol. 1, p. 110, [17] IEC : High-voltage test techniques, Part 1: General definitions and requirements, 1989 [18] L. Gao, C. Gomes, V. Cooray, and F. Roman.: Comparison of long sparks in air and over an

6 insulator surface, in Proc. of 11th International Symposium on High Voltage Engineering, London, August 1999, Vol.4, pp [19] S.G.J. Ehnberg, and H.J.M. Blennow: Effects of surface charge accumulation on impulse flashover on silicone rubber surface, in Proc. of 13th International Symposium on High Voltage Engineering, Delft, August 2003, paper no. P [20] N.L. Allen, and D.C. Faircloth: Corona Propagation and Charge Deposition on a PTFE Surface, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 10, no. 2, pp , April [21] F. Wang, Y. Qiu, W. Pfeiffer, and E. Kuffel,: Insulator surface charge accumulation under impulse voltage, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 11, no. 5, pp , October [22] B.H. Tan, N.L. Allen, and H. Rodrigo: Progression of positive corona on cylindrical surfaces Part I: Influence of dielectric material, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 14, no. 1, pp , February [23] L.A. Lazaridis, P.N. Mikropoulos, A. Darras, and A. Theocharis: Flashover along cylindrical insulating surfaces under positive lightning impulse voltages, in Proc. of XVII International Conference on Gas Discharges and their Applications, Cardiff, September 2008, pp

HUMIDITY INFLUENCES ON THE BREAKDOWN OF ROD-PLANE GAPS UNDER POSITIVE IMPULSES IN ATMOSPHERIC AIR

HUMIDITY INFLUENCES ON THE BREAKDOWN OF ROD-PLANE GAPS UNDER POSITIVE IMPULSES IN ATMOSPHERIC AIR P. N. Mikropoulos and C. A. Stassinopoulos Department of Electrical and Computer Engineering, Aristotelian