Sei sulla pagina 1di 7
4.3. App Pye Wl 42 (2008) gp, 568.1574 Part 1, No. 4A, Apri 2003 102008 The Japan Socily of Appi! Physics ‘A Monitoring Device of Leakage Currents Flowing through ZnO Surge Arresters Bok-Hee Lee, Sung-Man KANG, Ju-Hong BOM and Tatsuo KAWAMURA! Research Center for Nes-Genertion High Voltage and Poser Technol, Univesity, 282 Yolen donk. Na-K, tacon 402-751. Kove "Center for ncaa Enon, Shara nse of Ten, 14 Siar Moa Tey HAS, een {Recieve option Oe 25, 202) In this study, we developed a device for monitoring resistive leakage caren and technique forthe diagnosis of the deterioration of ine oxide (2n0) surge aes. The esisive leakage cute’ that increases with ine rel i a herally nstbl state hat may even ed toa serious problem, We developed the lag crent monitoring devise forthe ageing tet {nd durability evaluation for 240 blocks. Te hir-harmonic component ofthe leakage cent of 230 blacks may be sed a tm inicatr of whether ZnO surge arses are ina god sate or in bad sats, Two methods forthe diagnosis ofthe deterioration of surge arses were assed. Both the measurement instruments forthe resistive leakage caren with & compensation cteut andthe thiharmonic iment component were designed and abated. For higher precision and svete elit of cant detectors, the cutent ransucer wit erie coe, which asa very high relative pemeabilty, ‘as wed to improve detection seaiviy. Laboratory tess for 220 blocks were performed by meassring the resistive leakage Cont and thitsnrmonic current components, together ith the fst Fourier transform (FFT) analy. The thité-ermonic arent component was cleat separated fromthe leakage cues flowing though Zn0 sarge sresters for power Frequency se applied voltages. As ares it was known that the proposed monitoring systems fr both the eiiveleskage current and the thirsharmooie cument would be useful fo developing a devie for editing the defect nd performance deterioration of x0 surge acters in power system aplcations. (DOV: 101143J5AP.2.1568} KEYWORDS: tghiring surge, Zn0 surge aster, lagnod ts deterioration resistive lskage cre, hha akaoe current, leakage current monitoring device 1. Introduetion The primary function of surge aresters in electric power systems is to protect transmission lines and generation and Aisteibution equipments from overvoltages and to absorb a substantial amount of electric energy resulting from transient overvoltages, lightning and switching surges. The higbly nonlinear V-1 characteristics of ZnO blocks eliminate the protective isolation gaps of surge aresters. Thus ZnO surge aresters are continuously stressed by the system working voltage and the leakage cutents always flow through Za0 surge aresters at working voltage levels, which lead to their performance degradation. That is, ZnO surge arresters are degraded by the system working voltage and high lightning and switching surge currents Some degradation mechanisms for Zn0 blocks have been reported.!~® ‘The leakage currents of ZnO blocks due to power frequency ac voltages consist of the capacitive cureat and resistive leakage current. Under normal conditions, the leakage currents lowing through ZnO blocks are classified into two groups: (1) the currents due to the intrinsic resistance and (2) the currents due to the parasitic capaci- tance in ZnO Blocks. However, ZnO surge arresters are ‘known to exhibit an increase ofthe resistive leakage eurrent with time, and the rate of increase in resistive Jeakage current increases with increasing applied voltage and ambient temperature. Harmonic components of the leakage current are due to the intrinsic resistance of Zn0 blocks, and these may cause overheating and deterioration of Zn0 surge amesters. AS a consequence, an increase of the resistive leakage current with time can cause failure or permanent degradation of ZnO surge arresters.™"® Thus, itis very important o develop a technique for the diagnosis of the deterioration of ZnO surge arresers in order to maintain ‘normal operation conditions in electric power systems. The diagnosis ofthe deterioration of Zi0 surge arresters is achieved by the measurement of reference voltage, ischarge residual voltage and power at the maximum operating voltage, and harmonic content analysis for power frequency ac voltages. The total leakage current flowing through ZnO surge arresters contains low harmonic content when the lightning surge arresters are energized with a sinusoidal working voltage. Because the harmonic current components increase with time, the amplitude of harmonic ‘currents can be used as an’ indicator of performance deterioration of ZnO surge arresters. The fundamentals of ‘evaluating the physical properties and the lifetime expec- taney of lightning surge arresters have been studied for many years and some methods for monitoring the deterioration of ZnO surge arresters have been proposed.''“* However, there is a need for guidelines by which electrical engineers can estimate the residual lifetime of lightning surge arresters in service to improve the reliability of electric power supply in modem power systems. In evaluating the deterioration of ZaO surge arresters ‘employed in power lines, the detection of leakage current under power frequency ac voltages is essential and is more {important than that under dc voltages. It is noted that the ac leakage current is significantly lager than that of de and its temperature dependence is weaker." Thus, for effective diagnosis of the deterioration of ZnO surge arresters, it is ‘very important to discriminate the resistive leakage or third harmonic current component from the total leakage current. ‘Therefore, in this work, in order to develop devices for the precise monitoring of the deterioration of ZnO surge arresters, the resistive leakage current measuring system with a compensation circuit was designed and fubricated. In addition, the on-line monitoring device for the third- harmonic component of the leakage current was developed. ‘A method for detecting the resistive leakage current is unsuitable for site survey of lightning surge arresters since they have different initinsic leakage conductances and parasitic capacitances. However, the method is appropriate for analyzing physical properties such as the causes of ~35- 4. J. Appl Phys, Vol 42 2003) PL 1, No. 4A, Aeterioration and failure mechanisms of ZnO varstrs with time, Alo, the test method for detecting the third-harmonic component from leakage currents i less precise than that for dctccting the resistive leakage curent, BU its very simple and convenient for monitoring the state of surge aresters in fetual power systems. A ferrite core, which has a very high relative permeability, was used for improving detection response. sensitiv. Calibration experiments for the current measbring systems were caried out. Laboratory tests for Zn0 blocks were carried out by observing the resistive leakage current and the third-harmonic current, together with the FET analysis. The thir-harmonie current components were clearly” isolated from the leakage curents lowing through the ZnO blocks stressed by power frequency ac voltages. 2. Experiment 2.1. Measurement system for the resistive leakage currents ‘As described previously, it is known that Zn0 blocks are degraded by the resistive leakage current. In order to achieve an effective diagnosis method of the deterioration of ZnO ‘surge arresters, the capacitive current resulting from the parasitic capacitance must be excluded, and the resistive leakage current and/or third-harmonic current components rust be detected alone. The electrical equivalent circuit of ZnO blocks can be approximated by a simple RC parallel network, where R is defined as highly nonlinear with increasing voltage.'™'"” The leakage current flowing through Za0 surge arresters consists of the resistive and capacitive leakage currents. In order to detect the resistive leakage current only, the capacitive current has to be subtracted from the total leakage current by supplying a reference capacitive ‘current. The on-line monitoring device with 2 compensation circuit for discriminating the resistive leakage current and capacitive current was designed and fabricated. Figure 1 shows the measurement circuit diagram used to separate the resistive leakage current and capacitive current from the total leakage current flowing through Zn0 blocks under power frequency ac applied voltages. In Fig. 1, Cy is the reference capacitor, R, is the vatiable resistor, and Re is the fixed resistor. The potential difference between b and a points is expressed as Vso = Vie + Veo a IfXcq , which is the reference capacitive reactance in the ‘compensation circuit, is much greater than Ri, then Jc is assumed as a pure capacitive current. When the capacitive Ra x Ug tile) + ler * Ri Vs > euilet seit es ie & le eee, br Fig, 1. Mewuring scheme forthe rexsive leskage caren ith the ompemstion crt for ZnO blocks: The operation! ampli i FPA BH. Lex er al current passing through ZnO blocks is compensated by the reference capacitive current, the equations are as follows. Mex Rul = le x Re @ Ue x Ral = Vie eo Vag z o RB Here, fc x Ry is the magnitude ofthe voltage Voc and Uk x Ra] is the magnitude of the input voltage Vbe Of the differential operational amplifier. The value of the resistor Ris adjusted so as to ensure thatthe capacitive current lc is cancelled atthe 2ero crossing point of applied voltages. Thus the total leakage current /r, and the capacitive and resistive currents, fe and Iq, respectively are calculated from the corresponding measured potentials ata and b points, Vie and Vee, and the potential difference between b and a points, Via . by using a diferentil operational amplifier. 22. Rogowski coil current probe with ferrite core ‘Because the measurement of the leakage current flowing through ZnO surge arresters by using @ current-sensing resistor is not desirable, a noncontact current-sensing device is required. In general the Rogowski coil current probe is a specially designed induction coil that is widely used in the measurement of time-varying currents without electrical contacts. The principle of operation of the Rogowski coil ‘current probe is based on Faraday's law relevant to electromagnetic induction. Figure 2 shows the schematic and equivalent circuit of the self-integrating Rogowski coil current probe. ‘The self-integrating Rogowski coil current probe is ‘mainly employed for measuring fast high impulse currents or their derivatives. In practical use of electric power systems, it is necessary to extend the operating frequency bandwidth, particularly for the lower cutoff frequency in the use of the Rogowski coil current probe. The self-integrating Rogowski coil probe witha ferrite core was designed so that the lower cutoff frequency is much less than the power frequency (50 Hz in Europe and 60 Hz in North Ametica) by using a higher magnetic permeability, and the response (0) Schema gram z ig. 2 Schomaie diagram and equiva ccuit of the seliegrating Rogowak ei erent probe, -36- 3.3. App. Phys. Vol. 42 (2008) No. 4A sensitivity for power frequency ac currents and their harmonies is excellent. In Fig. 2, L is the selfinductance of the sensing coil, Mt the mutual’ inductance between the sensing coil and the current path, R, the residual resistance of the sensing coil, Ro the signal resistance and Zp the cable characteristic impe- dance Because the voltage signal e() induced atthe ends of the Rogowski coil is proportional to the time derivative of the current being measured i(), integration is required for ‘obtaining en actual current signal. This may be accom: plished by a suitably designed passive network consisting of a simple R-L circuit. The Rogowski coil current probe with 4 passive integrator, an R-L circuit, is termed “self integrating Rogowski coil curent probe”. The voltage e() induced atthe ends of the sensing coll i proportional to the time derivative of the current being ‘measured i() and is equal to the voltage drop of the sensing circuit, i, = MO MO sas RoH. Since the imtgraion is necessary to obtain a proportional signal from an Re crcit the resistance Ry ofthe detection coll most be aller than the characteristic impedance ofthe signal cable.' Above the lower cutoff frequency, wL >> (e+ Ra) and Z5 > Ro; in other words, the curren i) in the RCL cri is controlled by the reactance wl. Hence, tt) ey eet © “Therefore, the output voltage onthe signal resistance Rois Re M 19 =F fetnar= Fo, ” In operation, the Rogowski coil itself provides the integrating inductor Z. It is clear ftom eq. (7) that the ‘output voltage of the self-integrating Rogowski coil current probe is directly proportional to the current being measured, ‘To measure a very small resistive leakage current lowing through ZnO surge arresters for power frequency ac applied voltages under normal operation conditions, we fabricated a penetration-type current transformer (CT) with a secondary coil of 200 turns, and a ferrite core with a diameter of ‘32mm, thickness of 2.5mm and permeability of 40,000 was, used to achieve high response sensitivity. 2.3 Measurement system for the third-harmonic current ‘When evaluating the deterioration of ZnO blocks by using the total leakage current including the capacitive current, detection accuracy is significantly reduced, Thus, itis very effective for the diagnosis ofthe deterioration of ZnO surge arresters to detect the third-harmonic current component of the leakage current, Figure 3 shows the detailed detection circuit diagram of the third-harmonic current, The total leakage current flowing through ZnO surge arresters under ‘normal operation conditions is very small under power frequency ac applied voltages. ‘The different ier Ay was applied in order to minimize common-mode noises. The multistep circuits of low-pass filters Az and As were employed to analyze the LEB eto | spat cet ! Fig. 3. Circuit ogra of the device for menurng the third-harmonic ‘amet. Al operational ampifers are L741 change of leakage current. Also we designed and fabricated the detection circuit which is composed of an active band- pass filter and a passive RC filter $0 as to detect the th harmonic component. 3. Results and Discussion 3.1 Characteristics of the measurement system for the third-harmonie current ‘The frequency response characteristics of the developed selfitegrating Rogowski coil current probe with a ferte core were examined in terms of the sinusoidal input current. ‘The frequency bandwidth ranges from SHz 10 about 5KEIz a illustrated in Fig. 4. The response sensitivity is 23 V/A. ‘Thus the frequency bandwidth and response sensitivity of the developed Rogowski coil probe are very suitable for detecting power frequency leakage coments and their harmonic components. Figure 5 shows the frequency response characteristics of the third harmonic detection device, and its resonant fiequency fo is 180Hz, The passband is 10H and the sharpness Q is 18. The sharpness increases linearly asthe resistive leakage current increases. The proposed third- harmonic detection device would be greatly suitable for ‘monitoring the performance deterioration of Zn0 surge arresters in electric power systems. [Experiments were carried out on Zn0 blocks taken from the commercial 18kV rated surge arrester, and the specifications and electrical properties of the ‘specimens used in this work are shown in Table I. ‘The measured current amplitude, which is in the range of. me z 3 1.0] Fos g i 06 FC kan aa ° a 10 ci we it Frequeny (Ha Fig. 4. Frequency response characteristics of the fabricated cament ttasfonne. -37- Jpn J. Appl. Phys: Vol 2 2003) Pe. 1, No. 4A 20.1 [mal 05 may 2 101mAl 15 {mal Output voltage (V] ° 6 9 120 150 180 210 240 270 300 Frequency (H2] Fig. 5. Frequency sponse characterise ofthe thi-hamonie-cumen- ‘messing device. "ble L Specications and electrical properis ofthe 2n0 blocks in his werk Diameter 33 mm ‘Thickness 29 (am) Rated voltage 31nd Rated ducharge eurent 5 (RAs) ‘Mari continuous working volge 255 BV DC 50-539 RVI at I {mA} 9.8 (kV put] a5 (KA) Nominal conduction vokage : “ lepton ig. 6. Response sensitivity to the third-harmonic cunt leakage currents flowing through ZnO surge arresters at normal working voltage levels in electric power systems, ranges from 0.1 to 1.5mA. It is evident in Fig. 6 that the response sensitivity of the developed third-harmonic current ‘measuring device exhibits good linearity. 3.2. Resistive leakage currents of ZnO blocks ‘Waveforms ofthe applied voltage, resistive and capacitive leakage currents were simultaneously observed using @ digital storage oscilloscope and the voltage and current signals were transmitted from the oscilloscope to a personal ‘computer to store the data on a floppy disc. Figure 7 illustrates typical waveforms of the applied voltage, total current, and capacitive and resistive currents ‘observed for a ZnO block at 25°C by the use of the leakage current monitoring device with the compensation circuit ‘The leakage current under steady-state conditions actually consists of odd harmonics as well as the Fundamental power frequency. The amplitude of the capacitive component ofthe BAH. Lew er at vl freee) jose a v4) somata Het a Fig. 7. Examples ofthe applied volage and leakage euret waveforms “etced bythe proposed measuring device with the compensation cru Vs isthe applied voltage, Fy the tual leakage cutent, e the espa cet, and he resistive leakage caret. leakage current is much greater than that of the resistive ‘component, Also we can see easly that a large amount of the third-harmonic current is included in the resistive leakage ccurent, We can see clearly in Fig. 7 that the waveform of the ive leakage current is an odd periodic function. Thus the resistive leakage current can be described in terms ofthe fundamental frequency and odd harmonics. We have analyzed the data that were observed for the stress of (60 Hz commercial frequency voltage of 4.8 kV. Waveforms of the total leakage current and resistive leakage current components and their FFT results are displayed in Fig. 8 10 illustrate the peak value of each hharmonic component included in the leakage currents. It is evident in Fig. 9 that the average values of the total leakage current and resistive current components are zero and the capacitive current is completely excluded from the total Upper Lesage carat 10. Se awe HT ated Fig. § Wavelorms of the total leakage caren and resistive Teakage ‘ren and thee FET rests oe pa J. App. Phys Vl. 42 (2008) PL 1, NOAA, etd mie amen: een 10 wo 0 "0 Reisve caret eh Fig. 9. Harmonie components ofthe resistive leakage eure leakage current, That is, the waveforms of leakage currents: flowing through ZaO blocks are not dependent on the polarity of power frequency ac applied voltages and exhibit a symmetrical, bipolar shape. From the FFT results, it was clearly evident that the amplitude of the fundamental ‘component for the total leakage current is much higher than that forthe resistive leakage current because the capacitive current is included in the total leakage current. Also the amplitudes of the harmonic components for the resistive leakage current were similar to those for the total leakage ‘current, but not to that of the fundamental component, Figure 9 shows the calculated results of the harmonic ‘components based on the FRT results for the resistive leakage-current at 25°C. The peak value of the fundamental component was the largest in all experimental ranges, and the peak value of the higher order harmonics. gradually decreased. The peak value of the fundamental frequency current is linearly increased with the amplitude of the resistive leakage current. However, the peak value of the higher ofder harmonics is significantly inereased with the resistive leakage current. The ratios of the peak value of the third-harmonic, fifth harmonic or seventh harmonic compo- nent to the peak of the fundamental component are relatively small in the high conduction region. ‘An example of the rates of changes of the resistive leakage current and its harmonic components at the applied voltage of 4.8KV is shown in Fig, 10, The rates of changes of the fundamental component and the harmonic current i i : t Tempe Fig. 10, Rates of changes of the resistive leakage current and harmonics Steel of 4 RV a fnction of emeratre ol. Lae al ‘components signifieuntly inerease with inereasing tempera- ture of ZnO blocks. It was found that the temperature dependence of the rate of changes of the third-harmonic ccurrent is very significant. The large rate of changes of the {hird-harmonie current of ZaO blocks can be applied as an Indicator of the deterioration of ZnO surge arrester 3.3. Third-harmonic current components Oscillograms of the test voltage and current signals were observed by applying a 601z power frequency voltage of various peak levels to the ZnO blocks. Figure 11 illustrates the waveforms of the applied voltage, total leakage current, resistive leakage current and third-harmonic current compo- nents. A great part ofthe total Ieakage current is a capacitive current in the low conduction region. On the other hand, the (@) Vy:3210%) M8008 1: 30008 fy 2OUA) 1:30) = ee ae & : : woo) a sense E (6) ¥y=481nV)) — M.000s 1300128 fy: 3OO(K), fy: COAG) — js 00m = 7 ae aes anal (o ¥g:52hy,)— MS.0oms 1000 1600(a) 130068) Fig. 11. Waveforms ofthe applied voltage and lekage caren isthe ‘plied voltage, fy the ttl leakage current fy the eesitive leakage ‘ent, snd ste third-harmonic cureat ~39- 4p. J. Apel Phys. Vol. 42 (2003) 1, No. 4 0 Sd hamonic curent(uA 0? 10? 10! Resistive omen (Ay) Fie. 12. Response senskiviy of te third-harmonic curen- measuring eve evaliate as «function ofthe resstveleskage eure peak resistive leakage currents significantly increase when the applied voltage is increased. In addition, the peak points of the resistive leakage currents are consistent with those of the applied voltage. The electrical conduction in the positive half cycle is the same as that in the negative half cycle. The higher the applied voltage and the greater the distortion of the waveform of the leakage current, ic. at higher voltage levels, the more the resistive leakage current becomes significant and nonlinear. ‘Also we can easily see that the third-harmonic current component can effectively be extracted from the resistive Teakage current, as illustrated in Fig. 11 Measurements of the response sensitivity of the proposed third-harmonic detection device were performed for the various peak values ofthe 60 Hz power frequency ac applied voltage. Figure 12 illustrates the relationship ofthe input and ‘output currents measured as functions of the resistive leakage current for the third-harmonic-current-measuring device, where the data are the peak values of the input and output signals ‘The graph of the third-harmonic current output against the resistive leakage current input on the whole showed a linear shape in the low conduction region and its slope slightly increased in the resistive leakage current range of 0.5 to ImA, because of the intrinsic characteristics of the ZnO blocks. The data from the third-harmonic-current-measuring device were very similar to the results obtained by the FFT ‘method, as shown in Fig. 10 and §3.2. The proposed third- ‘harmonic-current-measuring device has a high precision and ‘200d quality compared with the resistive-current-measuring, devices which have been widely used up to now. Asa result, it can altematively be used for the diagnosis of the deterioration of ZnO surge arrestrs. 34 V-I characteristics of 2n0 blocks Current, not voltage, is the essential input variable in understanding transient overvollage protection. In general, the V-/ curves for ZnO varistors are determined from the ‘measurements made with brief pulse currents, such as an 8) 20s wave shape, in order to eliminate the effects of heating the ZnO varistors." In actual practice, in order to evaluate the degradation of ZnO surge arresters and to predict their lifetime, because ofthe seasonal variation of temperatures of ZnO surge arresters, itis essential to examine the tempera- BAHL Lee ta Refenee Ba — tai 7 Capcitve 2 ee - , 10 10 10° 10° et Fig. 13. Voltage vers capcitive and resistive leakage cuent charac teristics measured a function of temperate, ture dependence of the resistive leakage currenis ait their harmonics in the prebreakdown region, ‘According to the general convention for obtaining data on electronic devices, when the constant ac voltage was applied to the ZnO blocks, the resistive leakage current flowing through the blocks was measured as a parameter of temperature. The ZnO blocks were placed in a constant temperature furnace, After the temperature ofthe test 210 block reached the set valu, the resistive leakage current was measured for Imin, That is, in products where ambient temperatures may be much greater than 25°C, the effects of temperature on the V=I characteristics of Zn0 blocks were investigated in detail. The V-T curves were obtained in the temperature range of 25 to 100°C and the current range up 9 ‘5OmAye- Figure 13 shows the temperature dependence of the resistive leakage current of Zn blocks stressed by the power frequency ac voltage inthe prebreakdown region and plotted by obtaining the peak value of the applied voltage and the resistive and capacitive leakage current components. In Fig. 13, the dotted line indicates the nominal reference voltage. The nominal reference voltage isthe peak value of the power frequency voltage divided by V/2 which shall be applied to the Zn0 blocks to obtain the nominal reference current. The reference voltage’ of a multi-unit aster is the sum of the reference voltages of individual Za0 block units." As illustrated in Fig, 13, the V-F curves are divided into two characteristic regions: (1) a prebreakdown region and @) a breakdown region. In the breakdown region, the leakage currents increase steeply with the applied voltage and the V-I curves with highly nonlinear characteristics depend on the temperature slightly, On the other hand, in the prebreakdown region, the V-/ curves show a strong temperature dependence. The strong temperature depen- dence ofthe V-1 curves of ZnO varisors has been reported in the literature.” I is predicted from the measured results that inthe prebreakdown region the current flow is due tothe thermal excitaion of electrons over the potential barrier between ZnO grains, while in the breakdown region, the leakage current is dve to the tunneling effect of electrons through the barrier. At very small curents, particulary for less than ImAp, 20 blocks behave like linear resistor with @ large value of resistance. Actually the resistive -40- pe. J. Appl Phys: Vol 2 2008) Pt. No. 4A are mre ee etage BV) Fig. H. tand agnnst the applied voltage plots as parameter of Temperate leakage current increases with increasing temperature; the temperature dependence of the resistive leakage current is still stronger in the Tow conduction region below 1 mA and rately appears in the high conduction region above mA. ‘Also, because ZnO blocks have a negative temperature coefficient of resistance in the idle current region, the resistive leakage current increases and the clamping voltage decreases as the temperature of ZnO blocks increases. That is, a high temperature tends to reduce the effectiveness of nonlinear characteristics of Zn0 blocks. On the other band, the capacitive current is increased with the increase of ambient temperature and its rate of increase is relatively smal. This result is attributed to the increase of dielectric constant with the increase of temperature in ZnO blocks.'® Because it is very important for the evaluation of the degradation and residual lifetime of ZnO surge arresters 10 investigate whether they are in a good state or in a bad state, the temperature dependence of the capacitive current and resistive leakage current and their harmonic components passing through ZnO surge arresters must be examined in detail forthe design and development of monitoring systems of ZnO surge arresters. ‘The dielectric loss factor is estimated from . 1 wand = 5G ‘The dielectric loss factor was plotted against the applied voltage as a parameter of temperature as shown in Fig. 14, ‘The tané plots indicate the performance of the materials and appear to be the first data obtained for the breakdown region.!®2”) A very strong dependence of the dielectric loss factor on the applied voltage could have been influenced by the moisture intake under laboratory experimental condi- tions. The conductivity is significantly increased compared with the capacitance as the applied voltage and iemperature are increased. The tan against ‘voltage characteristic curves fare similar to the V-J curves in the breakdown region, but the temperature dependence of the tan 5-V curves is widely observed. ® 4. Conclusions, Both the eesistve-leakage-current-measuring system with the compensation circuit and the third-harmonic leakage current detection device have been proposed. Laboratory “A1- BH. Lee ea experiments were performed 10 evaluate the fundamental characteristies of the proposed current measurement devices in the low conduction region of ZnO blocks. The resistive leakage current flowing through ZnO blocks under a 60 Hz. power frequency ac voltage was clearly observed using the fesistive leakage current measurement system with the ‘compensation cirewit, and its harmonic components were analyzed by the FFT method. The results could be summarized as follows: (I) ‘The resistive leakege current increases with increasing temperature; the temperature dependence of the resi tive leakage current is still stronger in the low conduction region below I mA and rarely appears in the high conduction region above SmA. This result is closely related to the temperature dependence of the ‘conduction mechanism of ZnO blocks. ‘The frequency bandwidth and response sensitivity of the proposed self-integrating Rogowski coil current probe are suitable for monitoring the power frequency Teakage currents and their harmonics passing through ZnO surge arresters under power frequency ac volta- ges. (3) The data obtained by the third-harmonic-current- ‘measuring device were in good agreement with those ‘obtained by the FFT method, Its clear thatthe third- hharmonie-current-measuring device can be used for monitoring ZnO surge arresters in electric power systems. @ 1b §. Fit, K, Koumoto, H. Yangida, Y. Watanabe and H, Kawazo: Jpn J Appl Phys. 38 (19) 1534 2) Pr Pincet end M. Glanseton: IEEE Trans. PD-102 (1999) 393, 3) M. V- Lat IBEE Trans, PAS-102 (1983) 2665 4), S.Nhiwak, H. Kite and T Stoh: IEEE Trans. PAS-103 (1984) as, 5) ¥. Ouch, A. Kimura, T. Kavahar, J. Morimoto and T.Teyods: pa 3. App Phys. 40 (2001) 3606. 6) ¥- Okbochi, A, Kimura, T Kaaba, Y. Okamoto and J, Moris: pa. Appl Phys. 39 (2000) 2665. MV, Lat IEEE Trans, PAS-L04 (1985) 266s. ‘A Hod, P. Nays, YK. Tong, W. A. Maly, LA. Metwally, DM CGerrman end RT, Waters: TBE Proc. A182 (1995) 482. ‘Endo, H Kita, Kobayashi, K Kut, K, Goto sad M. Sai: IEEE Trans. PD (1988) 98 10) J Landis, 1. Stenstom, A, Schl and B. Hansen: IEEE Trans, PDS (1990) 1. 1) A. Haddad, H. 8. B. Elayyan, D. M, German and RT. Waters: IEE Poe, A-138 (1981) 265 12) M. Mizuno, M. Hayashi 2664 13) M, Oyama L Ohshima 1363. 14) S. Tominaga, ¥. Shiba, Y.Faiara, M iaak and. Nit: IEEE. ‘Trans, PAS99 (1980) 1548, 15) AJ Schwab: High-Voltage Measurement Technique (The MIT Press, Mesachusts, 1972), 16) RB. Suandler. Protection of Blcrone Circuits from Overvotages, (ohn Wiley Sons, New York, 198). foteationa! Eletrotehnical Comission: Inemaionl Standard 15050099-4 (1998) 23 18) N. Oba, ¥. Terd, T- Ohgaki, §. Tanaka and. Tsurumi: Jp. ‘Appl. Py. 38 1999) 5028 19) A Madd J. F. Rost, DM. German and RT, Wes TEE Pre. ‘137 1980) 29. IXUM Pat M.A. Ponce, J M. Rio Fe a Rosu a V. M. Castano TEE Proc. A+43 (1996) 291 es 14K. Mitais IEEE Tans, PAS-100 (1981) 14M Honda: EE Trans. PAS-101 (1982) ™ Ey

Potrebbero piacerti anche