Partial discharge detection by extracting UHF signal from inner grading electrode of insulating spacer in GIS
2017; Institution of Engineering and Technology; Volume: 12; Issue: 1 Linguagem: Inglês
10.1049/iet-smt.2016.0419
ISSN1751-8830
AutoresXian‐Jun Shao, Wen‐Lin He, Jia‐Long Xu, Ming‐Xiao Zhu, Guanjun Zhang,
Tópico(s)Power Transformer Diagnostics and Insulation
ResumoIET Science, Measurement & TechnologyVolume 12, Issue 1 p. 90-97 Research ArticleFree Access Partial discharge detection by extracting UHF signal from inner grading electrode of insulating spacer in GIS Xian-Jun Shao, Xian-Jun Shao State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of China Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorWen-Lin He, Wen-Lin He Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorJia-Long Xu, Jia-Long Xu Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorMing-Xiao Zhu, Ming-Xiao Zhu State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of ChinaSearch for more papers by this authorGuan-Jun Zhang, Corresponding Author Guan-Jun Zhang gjzhang@mail.xjtu.edu.com State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of ChinaSearch for more papers by this author Xian-Jun Shao, Xian-Jun Shao State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of China Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorWen-Lin He, Wen-Lin He Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorJia-Long Xu, Jia-Long Xu Research Institute of State Grid Zhejiang Electric Power Company, Hangzhou, Zhejiang, People's Republic of ChinaSearch for more papers by this authorMing-Xiao Zhu, Ming-Xiao Zhu State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of ChinaSearch for more papers by this authorGuan-Jun Zhang, Corresponding Author Guan-Jun Zhang gjzhang@mail.xjtu.edu.com State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, People's Republic of ChinaSearch for more papers by this author First published: 01 January 2018 https://doi.org/10.1049/iet-smt.2016.0419Citations: 8AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract In this study, a novel ultra-high frequency (UHF) sensor using an inner grading electrode inside an insulting spacer is employed to detect gas-insulated switchgear (GIS) partial discharge (PD). The simulation of the waveguide modes of the 252 kV GIS and the frequency response characteristics of this UHF sensor are performed by using the finite difference time domain (FDTD) method. Three kinds of typical defects are employed to investigate the PD detection characteristics of this UHF sensor. It is shown that the resonance frequencies of this sensor are extremely abundant below 3 GHz, and the major resonance frequencies around 0.5–1.2 and 2.2–3 GHz. The phase-resolved PD (PRPD) images accumulated by using this UHF sensor present the typical PRPD images of defects. The main frequency spectrum of the metal protrusion on a high-voltage conductor and floating potential defects are distributed below 1.5 GHz, which is consistent with waveguide mode simulation. A resonance frequency of a grading electrode UHF sensor appears at about 115 MHz, which is close to the simulation value of a resonance frequency at 140 MHz. The PD detection sensitivity of this sensor is much higher than that of an external sensor; however, it is a little lower than that of an internal sensor. 1 Introduction The gas-insulated switchgear (GIS) is widely used as the main power equipment in the electric power system due to its high reliability and compactness [1-3]. The insulation breakdown of the GIS usually starts from partial discharge (PD), which is caused by the aging of insulating materials or the defects caused during the manufacturing process and operation. Therefore, in order to avoid the insulation failure, it is necessary to detect PD in the GIS at the early stage. Based on the discharge mechanism and secondary physical phenomena accompanying PD in the GIS, many kinds of methods for detecting PD have been developed in the past half century [4-8]. In recent years, the ultra-high frequency (UHF) method using the internal or external UHF sensors to detect the electromagnetic (EM) waves radiated from PD has been presented as a promising method to diagnose the insulation system of the GIS [9-12]. Compared with the internal UHF sensors, the detection sensitivity and anti-interference of the external UHF sensors are much lower, which cannot detect the PD signal of some defects even if the defect is critical to breakdown. However, it is usually difficult to install the internal UHF sensors in the GIS that has already been operated in power systems, and the electric field of the GIS will be varied by the installation of the internal UHF sensors, which will greatly affect the safety operation of the GIS. Generally, the internal UHF sensors are not installed in the GIS with the voltage grade <550 kV due to the cost and price. Several non-conventional UHF sensors are summarised by C. Neumann et al., and a new type of sensor by employing the inner grading electrode insider the insulating spacer is proposed by them [13]. The advantages of this UHF sensor are convenience and extensibility with low cost and high detection sensitivity, which shows the promising application in GIS PD detection. However, the detection characteristics of these sensors are not revealed deeply, including the frequency response characteristics of the UHF sensors and the differences of the detection characteristics caused by different insulation defects. In this study, the inner grading electrode inside the insulating spacer is employed as a UHF sensor for GIS PD detection. The finite difference time domain (FDTD) method is employed to simulate the waveguide modes and frequency response characteristics of this UHF sensor in a 252 kV GIS model with three extracting ports. The PD characteristics of the metal protrusion on a high-voltage conductor, free moving particles and floating potential defects are diagnosed experimentally by employing this novel UHF sensor, and the development features of PD under different applied voltage amplitudes are discussed. Finally, the sensitivity of this novel UHF sensor, internal and external sensors is also contrasted in this study. 2 Simulation of EM waveguide modes in GIS 2.1 Waveguide simulation model The GIS can be considered as an EM waveguide mode system due to its coaxial structure and the waveguide modes are determined by the structure size of the GIS. Therefore, it is necessary to calculate the frequency bands of the main waveguide modes propagating in the GIS when the grading electrode is used as the UHF PD sensor, and the resonance frequency of the UHF sensor should match the frequency band of the main waveguide modes in order to achieve high detection sensitivity. A 252 kV GIS model employed in this study is shown in Fig. 1a, which is composed of the bushing, potential transformer (PT), defect chamber, disconnector etc. with a single phase bus bar. The EM waveguide mode simulation model is simplified as a coaxial structure as shown in Fig. 1b. Fig 1Open in figure viewerPowerPoint Waveguide simulation of 252 kV GIS (a) 252 kV experimental GIS model, (b) Simulation model of 252 kV GIS waveguide modes, (c) Distribution of GIS waveguide main modes and corresponding cut-off frequencies The PD source is placed on the high-voltage conductor to excite the EM waveguide modes in the GIS. M. D. Judd's experiments showed that the PD current pulse width is very narrow (<0.5 ns) [14], therefore, a Gaussian pulse with 0.4 ns in width with 4 mA in amplitude is set as a PD current pulse to excite the EM modes. The numerical computation method is based on the FDTD method [6, 15]. 2.2 Waveguide simulation results The distributions of the main waveguide modes in this GIS model are illustrated in Fig. 1c. It can be seen that the waveguide modes of the EM waves excited by the PD pulse include the transverse electric (TE) and transverse magnetic (TM) waves. Besides the transverse EM, there are also high-order EM waves such as TE11, TE21, TE31, TM01, TM11, TE21 and so on. The high-order EM modes with low amplitude are not shown in Fig. 1c. The m and n in subscript of TE and TM represent the half wave order number along with the circumferential and radial directions at each standing wave mode, respectively. The cut-off frequency of each mode is also described in Fig. 1c, which means the EM waves can propagate in the waveguide system only when their frequency is higher than the cut-off frequency. The cut-off frequencies of the main high order waveguide modes in this GIS model are mainly around 250 MHz–1.5 GHz. 3 Frequency response of inner grading electrode inside insulating spacer 3.1 Structure of inner grading electrode A grading electrode is usually placed inside the edge of the insulating spacer during the casting process, which wraps the whole insulating spacer. The grading electrode is used to create a uniform electric field and to avoid the possible interface gap between flanges and insulating spacers. The grading electrode is connected with flanges by several bolts, and thus the grading electrode is actually electrically grounded. In this GIS model, the grading electrode with three connecting bolts A, B and C is shown in Fig. 2a. When one of the bolts is used to extract the PD signal, the grading electrode is still electrically grounded and formed as a loop antenna, which has no effect on the insulation condition of the GIS. Fig 2Open in figure viewerPowerPoint Frequency response simulation of inner grading electrode (a) Sketch of grading electrode configuration, (b) Waveform and frequency spectrum of voltage source, (c) S11 of grading electrode with A, B and C extracting bolts, respectively 3.2 Numerical computation of frequency response The computation model is the same as Fig. 2a. The frequency response is computed by employing the grading electrode as the transmitting antenna due to the reciprocity between the transmitting and receiving antennas [16]. Since the PD UHF detection frequency band is below 3 GHz, the cell width is set to 5 mm in the numerical model with the FDTD method [17]. The feeding point is set at the bolt with a 50 Ω matching impedance and the excitation source is the Gaussian voltage pulse source, as shown in Fig. 2b. Since the three bolts are not distributed uniformly as shown in Fig. 2a, the frequency response characteristics are different when the UHF signal is exacted from the bolts A, B and C, respectively. Therefore, the frequency response characteristics are computed when the bolts A, B and C are used as feeding points, respectively. S11 is the input port voltage reflection coefficient, which means the power reflected from the antenna. S11 is one of the most common parameters to the radiation efficiency of the antenna at the working frequency band. If S11 = 0 dB, then all of the power is reflected from the antenna and nothing will radiate. Therefore, the S11 is employed in this study to reflect the frequency response characteristics of the UHF sensor. The S11 curves at the feeding points of A, B and C, respectively, are shown in Fig. 2c. It can be seen that the resonance frequencies of the grading electrode UHF sensor are very extremely abundant below 3 GHz, which can be used as the wide-band antenna to detect a wide-band UHF PD signal in our GIS model combined with the main EM waveguide simulation results in Section 2.2. A maximum resonance frequency appears at about 140 MHz when the feeding points are bolts A, B and C, respectively. The S11 of bolt C is about 2.5 dBa lower than bolts A and B around 1.5–2.0 GHz, indicating that the bolt C is more suitable to be selected when the GIS main propagation frequency is around 1.5–2.0 GHz. Furthermore, the amplitude of the extracting UHF signal from the bolt is larger when the PD source is located near to this bolt. 4 Experimental arrangements 4.1 Arrangement of UHF sensors Artificial defects are placed in the defect chamber with 0.4 MPa SF6. The arrangement of the UHF sensors is shown in Fig. 3a. An internal disk coupler (C1) is placed at the bottom of the defect chamber, and an external UHF sensor (C3) is placed on the spacer aperture. Fig 3Open in figure viewerPowerPoint Arrangement of UHF sensors (a) UHF sensor arrangement in defect chamber, (b) Extractor adapter of grading electrode's bolt, (c) Structure of C2 sensor (1, insulating spacer; 2, inner grading electrode; 3, extractor rod; 4, insulating sheath; 5, coaxial connector) The images of UHF sensors C1 and C3 are shown in Fig. 4. The commercial internal sensor C1 is a typical disk coupler and the commercial external sensor C3 is a patch antenna. The commercial UHF sensors are both manufactured by DMS Co., Ltd. The effective height calibration is usually used to represent the sensitivity of the UHF sensor, which is tested by using a giga-hertz transverse electromagnetic (GTEM) chamber with a pulse source [18]. The effective height curves of UHF C1 and C3 sensors are also illustrated in Fig. 4. It can be seen that the main frequency response ranges of the UHF C1 and C3 sensors are 400 MHz–1.4 GHz and 800 MHz–1.6 GHz, respectively. It is because the frequency of the EM wave leakage from the insulating spacer is much higher than the EM waves propagating inside the GIS. From Fig. 4, it also can be calculated that the average effective heights of the UHF C1 and C3 sensors are about 13.9 and 10.10 mm, respectively. Fig 4Open in figure viewerPowerPoint Images and effective height curves of UHF C1 and C3 sensors (a) C1 sensor, (b) C3 sensor An extractor adapter (C2) is designed to extract the UHF signal from the inner grading electrode, as shown in Figs. 3b and c. The grading electrode is electrically grounded to form a loop antenna to receive the UHF signal excited by PD. By screwing this adapter into one of the screw holes of different bolts, the PD signal can be extracted. An extractor rod is contacted with the inner grading electrode of the insulating spacer to extract the UHF signal. The insulating sheath is used to insulate the extractor rod from the flange of the insulating spacer. The coaxial connector is employed to transmit the UHF signal to a PD monitoring instrument and a digital oscilloscope. It is considered that free particles are usually dropped at the bottom of the GIS chamber and spacer, so the extractor adapter of the bolt is installed near the bottom of the spacer, which is blot A in Fig. 2a. It should be noticed that the effective height of UHF C2 sensor cannot be tested by the GTEM chamber due to the size and structure limitation. The DMS PDM (DMS Co., Ltd) is employed to display the phase-resolved PD (PRPD) and phase-resolved pulse sequence (PRPS) images. The frequency range of the DMS PDM is 100–2000 MHz with a level dynamic range of −75 to −5 dBm. The single port sampling rate: 15.4 kS/s with 256 measurement points per power cycle with 8–10 bit amplitude resolution. The UHF signal waveforms are recorded using a digital oscilloscope (Agilent DSO9404A). 4.2 Artificial defect models Three kinds of typical defects are set in the defect chamber of the GIS, which are the metal protrusion, free moving particles and floating potential defects, respectively. These defects are usually generated during the manufacturing process, assembly in the field and long-time operation. A copper needle with a 0.25 mm diameter and a 2.5 cm length is fixed on the high-voltage conductor, which is employed to simulate the defect of the metal protrusion on a high-voltage conductor. The bottom side of the copper needle is fixed by the aluminium tape smoothly. The needle is placed in the middle of the defect chamber of the GIS, as shown in Fig. 5a. Fig 5Open in figure viewerPowerPoint Artificial PD defects (a) Metal protrusion, (b) Free moving particles, (c) Floating potential The aluminium particles with an average 3 mm diameter are placed at the bottom of the defect chamber to simulate the free moving particles defect, as shown in Fig. 5b. The electrical floating element is usually generated by the incorrectly fixed or unscrewed elements of the GIS. In this study, the floating potential defect is built by using several copper pieces and insulating tapes. The insulating tapes are fixed on the high-voltage conductor, and the copper pieces are placed on the several layers of the insulating tapes to simulate the floating elements in the GIS. The floating potential defect is also placed in the middle of the chamber, as shown in Fig. 5c. 5 Experimental results and discussions 5.1 Metal protrusion With the increment of the applied voltage, the PRPD images detected by the inner grading electrode sensor (C2) under different PD development stages are shown in Fig. 6. It can be seen that C2 can detect the weak UHF signal of PD when the applied voltage amplitude is 37.5 kV, much less than the nominal operation voltage. It is because of the serious distortion of the electric field around the protrusion, the insulation strength of SF6 is much lower under a severe non-uniform electric field than that under a uniform electric field. Due to the effect of space charges on the head of protrusion, it is much easy to originate the negative corona discharge [19]. Thus, the PD pulses mainly appear around 270° in PRPD images. Fig 6Open in figure viewerPowerPoint PRPD images under metal protrusion on a high-voltage conductor (a) 37.5 kV, (b) 58.7 kV, (c) 85.3 kV, (d) 110 kV With the gradual increment of the applied voltage, a positive corona discharge is incepted and developed, and the amplitude and density of positive and negative PDs are both increased quickly. At an applied voltage of 58.7 kV, the amplitude of the UHF signal between the positive discharge and negative discharge is almost similar. The density of the positive discharge is much lower than that of the negative discharge. This is because the mechanism of the positive corona discharge is similar to the positive fluid discharge. The positive charges are easy to accumulate near the positive corona, which enhances the electric field ahead and tends to form a new electron avalanche. The electron avalanche promotes the propagation of the discharge channel. Therefore, the density of the positive corona discharge is relatively low while the amplitude is rather great. Similar to the Townsend discharge, the maintenance of the negative corona discharge is mostly relied on secondary electron emission around the cathode. The positive charges accumulating during the metal protrusion will enhance the local electric field around the cathode. Thus, a large number of electron avalanches are generated at the same time during the protrusion, which is easy to form diffuse plasma [19]. Finally, the 'rabbit ears' shape appears in the PRPD image and the discharge centre still around the peak of the applied voltage. This PRPD image indicates that the PD signal is very strong and transits to the dominant position stage of the positive corona. As described above, the positive corona promotes the propagation of the discharge channel. The large amounts of positive ions drift to the enclosure of the GIS, which shortens the gap between the protrusion and grounded enclosure of the GIS. At this stage, the insulation gap may breakdown. The UHF signal waveforms extracted from the C2 of the time domain and frequency domain under the applied voltage amplitude of 85.3 kV are shown in Fig. 7. Fig 7Open in figure viewerPowerPoint UHF signal waveform and its frequency spectrum with metal protrusion on a high-voltage conductor (a) Time domain, (b) Frequency domain The UHF signal at the time domain demonstrates that the damped oscillation is within 2 μs, and its amplitude is about 4 mV. The oscillation time is very long due to the low resonance frequency of this UHF sensor. From Fig. 7b, the peak frequency spectra of this UHF signal appear at 115 MHz, 0.4 GHz, 1.05 GHz and 1.4 GHz, respectively. The frequency spectrum of the UHF signal is mainly distributed below 1.5 GHz, which coincides with the main high-order waveguide mode simulation results in Section 2.2. In Fig. 7b, it can also be seen that the largest amplitude of the frequency spectrum appears at 115 MHz, which is close to the resonance frequency of 140 MHz in the S11 simulation, as shown in Fig. 2c. The experimental results validate the S11 simulation results in Section 3.2. 5.2 Free moving particles The PD development images of the free moving particles with the increment of the applied voltage are shown in Fig. 8. The PRPD images cannot be accumulated in a long time due to the instability of PD at lower applied voltage amplitude. It is because the particles may jump into the low-electric field strength area where it cannot jumped again under low applied voltage. Thus, the PRPD images are replaced by the distribution of PRPS images with 50 circles of the applied voltage. When the electric force is larger than the gravity force of particles, the particles will jump and leave the GIS tank. The PD occurs when the free moving particles drop onto the inner surface of the GIS tank, which is easy to occur at the zero-crossing point of the applied voltage. Therefore, the phase features of free moving particles concentrate around the zero-crossing point at lower applied voltage. Fig 8Open in figure viewerPowerPoint PRPD images with free moving particles (a) 44 kV, (b) 52.8 kV, (c) 77 kV When the applied voltage is increased to 52.8 kV, the electric force increases significantly compared with lower applied voltage. At this stage, almost all particles begin to dance in the GIS. Thus, the PRPD images show no obvious phase features. It indicates that the discharge transits to the 'dance' stage. The UHF signal amplitude is almost reaching the detection peak. With the applied voltage increasing to 77 kV, the UHF amplitude reaches the detection peak of the DMS instrument, and the PRPD image appears as the symmetrical phase feature between the positive and negative PD signal. The discharge activity is very strong in this stage. The PRPD images are similar as the floating potential discharge described in following Section 5.3, indicating that the moving particles transit from the 'jump' and 'dance' stage to the 'shuffling' stage [11]. The PDs are grouped around the zero-crossing point of applied voltage at this stage. The free moving particles may come into contact with the high-voltage conductor due to the electric force, and the discharge pulses occur at this moment. When the discharge transmits from the corona to leader discharge, the breakdown will occur in the GIS. The UHF signal waveform and its frequency spectrum at an applied voltage with an amplitude of 77 kV are shown in Fig. 9. Here, the amplitude of the UHF signal is about 34 mV, which is much larger than that of the metal protrusion on the high-voltage conductor in Fig. 7a due to the strong discharge intensity of free moving particles. Similar to Fig. 7a, the damped oscillation of the UHF signal under free moving particle defects is also very long since the resonance frequency of this UHF sensor. Fig 9Open in figure viewerPowerPoint UHF signal waveform and its frequency spectrum with free moving particles (a) Time domain, (b) Frequency domain The frequency spectrum in Fig. 9b is mainly centralised around 1–1.5 and 2.5 GHz. Due to the free moving particles located at the bottom of the GIS chamber, the PD tends to excite higher frequency bands of waveguide modes than that generated by the metal protrusion on the high-voltage conductor, which is in good agreement with the experimental results investigated by Meijer [11]. Furthermore, the main frequency spectra are also influenced by the S11 curve of C2 illustrated in Section 3.2. Similar to the metal protrusion defect, a peak amplitude of the frequency spectrum appears at about 115 MHz due to the resonance frequency of C2, which also indicates that the simulation results of S11 are in accordance with the experimental investigation in this study. 5.3 Floating potential The PRPD images of the floating potential defect at different stages with the increment of the applied voltage are shown in Fig. 10. At an applied voltage of 72.6 kV, the floating defect begins to discharge. However, the discharge at this stage is not stable. Since the floating potential defect is composed of a copper sheet and an insulating tape, the protrusion exists at the copper pieces. Therefore, the PD tends to occur at the protrusion of copper pieces excited by the induced voltage. Thus, similar to the mechanism of the corona discharge, the negative discharge is easy to be ignited. That is the reason why the density of the negative PDs is higher than that of the positive PDs under the lower amplitude of the applied voltage [19]. With the increment of the applied voltage to 88 kV, the discharge of the floating potential defect tends to be stable. However, the intensity and density of the discharge at the negative half-cycle of the applied voltage are still larger than that at the positive half-cycle. The reason is explained above, which is called as a negative corona stage here. Fig 10Open in figure viewerPowerPoint PRPD images with floating potential defect (a) 72.6 kV, (b) 88 kV, (c) 121 kV When the applied voltage amplitude increases to 121 kV, the amplitude and density of PD under the positive and negative half-cycle of the applied voltage are almost symmetrical, and the UHF signal is stable, which appears as the typical PRPD images of floating defects. Compared with the previous discharge stage, the high applied voltage at this stage causes a shift of the centre of the largest PD density zone from the peak point of applied voltage, which is caused by the space charge effect. The UHF signal waveform and its frequency spectrum at an applied voltage with an amplitude of 121 kV are shown in Fig. 11. It can be seen that similar to the protrusion defect, the amplitude of the UHF signal is very low about 6 mV. The damped oscillation is very long with about 1 μs due to the low resonance frequency of the C2 sensor. The main frequency spectrum is distributed at about 115 MHz, 500 MHz and 1.25 GHz. The main frequency spectrum distribution is <1.5 GHz, which is in accordance with the main waveguide simulation results in Section 2.2. The largest frequency spectrum of the floating potential defect appears at about 115 MHz, which is much larger than the other frequency band. This is caused by the resonance frequency of the C2 sensor as illustrated in the simulation. Fig 11Open in figure viewerPowerPoint UHF signal waveform and its frequency spectrum with floating potential defect (a) Time domain, (b) Frequency domain 5.4 Comparison of detection sensitivity with commercial sensors To investigate the detection sensitivity of the inner grading electrode UHF sensor more deeply, the detection characteristics of the novel sensor, commercial internal and external UHF sensors are contrasted under the floating potential defect. The arrangement of different UHF sensors is illustrated in Fig. 3a. C1 and C3 sensors are from DMS Co., Ltd. The effective height curves of the UHF C1 and C3 sensors are described in Section 4.1. C1 and C3 sensors are arranged at the bottom of the test defect chamber and the same insulating spacer with C2, respectively. At an applied voltage of 121 kV, the PRPD images of C1 and C3 are shown in Fig. 12. The UHF amplitudes of C1 reach the detection peak, and the weak UHF signal can be detected by the C3 sensor. The intensity and density of the UHF signal received by the C3 sensor are much larger than that of the C1 and C2 sensors. Similar to Fig. 10, the amplitudes and phases of positive and negative PD pulses are symmetrical, and the UHF signal is stable. The images of PRPD between C1 and C2 are similar. Both the PRPD images of C1 and C3 show the typical floating potential defect pattern. Furthermore, the detection sensitivity of C2 is much higher than that of C3; however, it is a little lower than that of C1. It is indicating that the C2 sensor can be employed to detect PD in field detection, which can achieve high PD detection sensitivity. Fig 12Open in figure viewerPowerPoint PRPD images with commercial sensors at 121 kV (a) C1, (b) C3 The UHF signal waveform and its frequency spectrum at 121 kV with a floating potential defect are shown in Fig. 13, respectively. Compared with Fig. 11, it can be seen that the UHF signal amplitude extracted from C2 is about twice that of C3, which is in accordance with the PRPD investigation. The main frequency spectrum of C1 is around 500 MHz, 1.25 GHz and 2.0 GHz, which is close to the resonance frequency spectra of the effective height curves of C1. Fig 13Open in figure viewerPowerPoint UHF signal waveforms and frequency spectra with commercial sensors (a) UHF signal waveforms with commercial sensor C1 at 121 kV, (b) UHF signal waveforms with commercial sensor C3 at 121 kV, (c) UHF frequency spectrum with commercial sensor C1 at 121 kV, (d) UHF frequency spectrum with commercial sensor C3 at 121 kV Since the C3 mounted at the casting aperture of the insulating spacer formed a slot aperture antenna, the resonance frequency of the slot aperture antenna can be calculated by [16] (1) where fa is the resonance frequency of the slot aperture antenna, c is the velocity of light, ɛ is the relative permittivity, and d is the length of the casting aperture. In our GIS, the length of the casting aperture of the insulating spacer is about 7 cm, and the relative permittivity of the insulating spacer is about 4.2. Thus, fa is about 1.04 GHz. From Fig. 13d, it can be seen that the main frequency spectrum of C3 is around 1, 1.75 and 2.5 GHz, respectively. The largest frequency spectrum of C3 appears at 1 GHz, which is very close to the resonance frequency of the slot aperture antenna fa. Therefore, it can be concluded that the detection characteristics of the UHF sensors are caused by the casting aperture of the insulating spacer, the frequency response characteristics of the UHF sensors and the propagation characteristics of the EM waves propagated through the insulating spacer. It should be mentioned that the detection sensitivity of C2 is much higher than that of C1 and C3 when PDs occur on the insulating spacer since the PD source is close to C2, indicating that C2 is more suitable to detect the defects on the insulating spacer. 6 Conclusion In this study, a novel sensor for GIS PD detection by using the inner grading electrode inside the insulting spacer is proposed, and the FDTD simulation and experiments are combined together to investigate the frequency response and detection characteristics of this PD sensor. The cut-off frequencies of the main high-order waveguide modes are distributed around 250 MHz–1 GHz. The resonance frequencies of the proposed PD sensor are extremely abundant below 3 GHz, and the major resonance frequencies are around 140 MHz, 0.45–1.2 GHz and 2.2–3 GHz. The S11 of bolt C is about 2.5 dBa lower than bolts A and B by about 1.5–2.0 GHz. The PD signal and its features with different defects can be extracted by employing the proposed PD sensor. The PDs with a metal protrusion on a high-voltage conductor can be divided into negative corona in the dominant position stage, the coexistence of positive and negative corona stages, positive corona in the dominant position stage and 'rabbit ears' shape stage. The PDs of free moving particles can be divided into 'jump', 'dance', and 'shuffling' stages. The PDs of the floating potential defect transits from the protrusion corona discharge stage to the typical floating discharge stage. A resonance frequency of the proposed sensor appears at around 115 MHz in the experimental results, which is close to the simulation value of the resonance frequency at 140 MHz. The PD detection sensitivity of the proposed sensor is much higher than that of the external sensor; however, it is a little lower than that of the internal sensor. 7 Acknowledgments This work was supported in part by the National Natural Science Foundation of China (51607140), and by the China Postdoctoral Science Foundation funded project (2015M580848). 8 References 1Hikita, M., Ohtsuka, S., Teshima, T. et. al.,: 'Examination of electromagnetic mode propagation characteristic in straight and L-section GIS model using FD-TD analysis', IEEE Trans. Dielectr. Electr. Insul., 2007, 14, (6), pp. 3095– 3102 2Kaneko, S., Okabe, S., Muto, H. et. al.,: 'Electromagnetic wave radiated from an insulating spacer in gas insulated switchgear with partial discharge detection', IEEE Trans. Dielectr. Electr. Insul., 2009, 16, (1), pp. 60– 68 3Onomoto, M., Kunitake, Y., Ohtsuka, S. et. al.,: 'Motion and size estimation of a free moving metallic particle in GIS based on propagation properties of acoustic waves', Electr. Eng. Jpn., 2005, 150, (1), pp. 26– 33 4Lundgaard, L.E.: 'Partial discharge-part XIV: acoustic partial discharge detection-practical application', IEEE Electr. Insul. Mag., 1992, 8, (5), pp. 34– 43 5Kawada, M.: 'Fundamental study on locating partial discharge source using VHF-UHF radio interferometer system', Electr. Eng. Jpn., 2003, 144, (1), pp. 32– 41 6Judd, M.D., Farish, O., Hampton, B.F.: 'The excitation of UHF signal by partial discharge in GIS', IEEE Trans. Dielectr. Electr. Insul., 1996, 3, (2), pp. 213– 228 7Judd, M.D., Farish, O., Coventry, P.F.: ' UHF couplers for GIS - sensitivity and specification'. Proc. 10th Int. Symposium on High Voltage Engineering, Montreal, Canada, 1997 8Okabe, S., Yuasa, S., Kaneko, S. et. al.,: 'Simulation of propagation characteristics of higher order mode electromagnetic waves in GIS', IEEE Trans. Dielectr. Electr. Insul., 2006, 13, (4), pp. 855– 861 9Hoshino, T., Kato, K., Hayakawa, N. et. al.,: 'Frequency characteristics of electromagnetic wave radiated from GIS apertures', IEEE Trans. Power Deliv., 2001, 16, (4), pp. 552– 557 10Kaneko, S., Okabe, S., Muto, H. et. al.,: 'Detecting characteristics of various type antennas on partial discharge electromagnetic wave radiating through insulating spacer in gas insulated switchgear', IEEE Trans. Dielectr. Electr. Insul., 2009, 16, (5), pp. 1462– 1472 11Meijer, S., Smit, J.J.: 'UHF defect evaluation in gas insulated equipment', IEEE Trans. Dielectr. Electr. Insul., 2005, 12, (2), pp. 285– 296 12Hikita, M., Ohtsuka, S., Teshima, T. et. al.,: 'Electromagnetic (EM) wave characteristics in GIS and measuring the EM wave leakage at the spacer aperture for partial discharge diagnosis', IEEE Trans. Dielectr. Electr. Insul., 2007, 14, (2), pp. 453– 460 13Neumann, C., Krampe, B., Feger, R. et. al.,: ' PD measurements on GIS of different designs by non-conventional UHF sensors'. Proc. Session 2000 of CIGRE, Paris, France, 2000, pp. 1– 9 14Reid, A.J., Judd, M.D., Stewart, B.G. et. al.,: 'Partial discharge current pulses in SF6 and the effect of superposition of their radiometric measurement', J. Phys. D. Appl. Phys., 2012, 39, (19), pp. 4167– 4177 15Taflove, A., Susan, C: ' Computational electrodynamics: the finite-Difference time-domain method' ( Artech House Publishers, 2005), pp. 629– 630 16Balanis, C.A.: ' Antenna theory, analysis and design' ( Wiley Publication, 2005), pp. 231– 275 17Reid, A.J., Stewart, M., Judd, M.D.: ' FDTD modeling of UHF partial discharge sensor response'. Proc. Int. Conf. On Sustainable Power Generation and Supply, Nanjing, China, 2009, pp. 1– 4 18Judd, M.D., Farish, O.: 'A pulsed GTEM system for UHF sensor calibration', IEEE Trans. Instrum. Meas., 1998, 47, (4), pp. 875– 880 19Raizer, Y.P.: ' Gas discharge physics' ( Springer-Verlag, Berlin, Heidelberg, 1991) Citing Literature Volume12, Issue1January 2018Pages 90-97 FiguresReferencesRelatedInformation
Referência(s)