Identification of Lightning Overvoltage on Overhead Transmission Lines

With the development of power systems, the height and transmission capacity of overhead transmission lines are constantly increasing. Overhead transmission lines stretch for thousands of kilometers and are often attacked by various lightning overvoltages, causing power outages.

Therefore, the research on lightning overvoltage protection of overhead transmission lines is an important topic that must be studied in the construction of ultra-high voltage power grids and smart grids. From the perspective of the occurrence mechanism, the lightning overvoltage of transmission lines can be divided into induced lightning overvoltage, shielding lightning overvoltage (overhead lines equipped with lightning arresters), and counter-strike lightning overvoltage. Due to the different occurrence mechanisms and processes of the three types of lightning overvoltages, the protection measures taken are also different. Induced lightning overvoltage mainly harms overhead transmission lines of 35 kV and below. For overhead transmission lines of 110 kV and above, due to the high insulation level of the line and the shielding effect of the lightning arrester, the induced lightning overvoltage generally does not cause the insulator string to flash over; counter-strike and shielding lightning overvoltages are directly generated by the lightning current, and are also called direct lightning overvoltage. Counter-strike lightning overvoltage mainly relies on improving the insulation level of the line and reducing the grounding resistance of the tower to improve the lightning resistance level, while shielding overvoltage mainly relies on improving the line protection angle and other methods to reduce the probability of shielding. Most of the existing research on lightning overvoltage protection of transmission lines uses research methods such as empirical data and electrical geometry modeling, attempting to eliminate the threat of lightning overvoltage to overhead transmission lines during the design phase.

However, a large amount of operational experience shows that even if many perfect lightning protection measures are considered in the lightning protection design stage, there are still many discrepancies between the data, methods, and models used in the lightning protection design and the actual situation of the commissioned lines, resulting in errors between the actual line lightning accident tripping rate and the design value, and various lightning overvoltage accidents occur from time to time [2-4]. Therefore, it is very necessary to accurately identify the type of lightning overvoltage that occurs on the transmission lines that have been put into operation and provide reliable data for the lightning protection design of overhead transmission lines, which is very necessary to improve the efficiency of lightning protection design.

Reference [5] proposed a method for identifying lightning shielding failure and back-strike of DC transmission lines from the perspective of relay protection. This method is mainly aimed at DC transmission lines. Reference [6] proposed to measure lightning current waveform parameters by means of magnetic tape, magnetic steel rod, etc. to realize the identification of shielding failure and back-strike. However, since these measuring devices cannot make repeated measurements, the workload of obtaining data is large and judgments need to be made based on work experience, which may easily lead to misjudgment and missed judgment of overvoltage identification. References [7-14] conducted in-depth research on lightning signals and proposed to use the ratio of high-frequency and low-frequency energy of current, the ratio of zero-mode and line-mode components, the ratio of maximum slope of wave front and wave tail, waveform consistency coefficient, and modulus maximum value as judgment criteria to realize the identification of lightning interference and short-circuit signals. The starting point of the above-mentioned references is transient traveling wave protection. The main purpose is to identify the impact of lightning interference on transient traveling wave protection, rather than to identify the type of lightning overvoltage. At present, the research on off-site identification methods of lightning overvoltage types in AC systems is still a difficult point.

Since the sampling rate of existing substation recording and broadcasting devices is too low (usually only tens of kHz), it is difficult to completely and accurately record the lightning overvoltage waveform (usually the wave head and wavelength time are at the μs level), and the magnetic saturation characteristics of TV under overvoltage are not considered, the substation recording device cannot accurately obtain the lightning overvoltage signal. Reference [15] proposed an overvoltage pattern recognition method for distribution network, which uses the overvoltage signal obtained by the overvoltage monitoring device inside the substation. Due to the complex and diverse internal structure of the substation, when the lightning traveling wave is transmitted along the transmission line to the inside of the substation, it will undergo multiple complex refractions and reflections, causing the waveform to be distorted. Therefore, for the identification of the type of lightning overvoltage, even if the substation overvoltage monitoring device is used to obtain a real and reliable overvoltage signal inside the substation, it is difficult to use it to analyze and identify the type of lightning overvoltage occurring on the line.

Based on the above considerations, this paper studies the occurrence mechanism, process and wave head characteristics of induced, shielding and counter-strike lightning overvoltages in AC transmission lines from the perspective of transmission line current traveling waves, and proposes to use the time domain characteristic quantity of the transmission line current traveling wave head to realize the identification of lightning overvoltage types, providing statistical data support for transmission line lightning protection design, optimization of operation level, line maintenance, etc. Since the lightning current traveling wave of the transmission line is used as the analysis object, the influence of the refraction and reflection of the lightning current traveling wave inside the substation can be effectively avoided. EMTP simulation calculations show that the characteristics extracted in this paper can effectively identify the type of lightning overvoltage.

1. Mechanism of lightning overvoltage on transmission lines

1.1 Induced lightning overvoltage

When a thundercloud approaches the sky above a transmission line, an electric charge equal to the thundercloud's charge but with the opposite polarity will be induced on the overhead transmission line, which is called bound charge. When the thundercloud discharges to the ground, the bound charge is released because the charge in the cloud is quickly neutralized, inducing an overvoltage with the opposite polarity to the lightning current on the transmission line. The waveform and amplitude of the induced lightning overvoltage on the overhead line are related to multiple factors such as the conductor and lightning current parameters.

Since the distance between the three-phase conductors and the lightning strike point is basically equal, the induced lightning overvoltage on the three-phase overhead line has the same polarity, similar waveform and similar amplitude. Jankov gave the amplitude estimation formula of the induced lightning overvoltage on the overhead line based on the lightning current return stroke model and the coupled Agraw2al model:

In the formula, ku = k3 h; h is the height of the conductor from the ground; d is the distance from the lightning strike point to the conductor; the coefficients k0, k1, k2 and k3 are determined by the characteristics of the lightning current.

1.2 Counter-strike lightning overvoltage

When lightning current hits the top of a transmission tower, most of the lightning current flows into the earth along the tower. Due to the existence of the tower, lightning arrester wave impedance and grounding resistance, when the lightning current flows through the tower and enters the earth, it will produce a large voltage drop on the tower, causing the potential of the tower top and cross arm to rise sharply. When the potential difference between the two ends of the insulator string exceeds its impulse flashover voltage, the insulator string flashes, causing a grounding fault in the transmission line.

The lightning current acts on the tower and strikes back, including the tower potential rise and the insulator string breakdown, as shown in Figure 1. When the insulator string is not broken down, in addition to most of the lightning current entering the ground along the tower, a small part of the lightning current is diverted by the lightning arrester. According to the principle of space electromagnetic coupling, a current traveling wave will be coupled in the transmission line. At this time, there is no direct lightning current in the transmission line channel, only space electromagnetic coupling current. When the tower potential continues to rise, causing the insulator string to break down, the lightning current will be injected into the transmission line. At this time, there will be a large amount of lightning current in the transmission line, and the current of the transmission line will jump significantly compared with the space electromagnetic coupling component in the previous stage. When the strike back occurs, the Peterson equivalent circuit of the two processes of tower potential rise and insulator breakdown is shown in Figure 2. Assume that the tower wave impedance is Z1, the lightning arrester wave impedance is Z2, the grounding resistance is Rg, the transmission line wave impedance is Z, the lightning channel wave impedance is Z0, and the lightning current is i. The function of the insulator string is equivalent to the equivalent circuit switch S. When the insulator string is not broken down and S is not closed, there is no lightning current in the transmission line wave impedance Z. After the insulator string is broken down and S is closed, the lightning current is injected into the transmission line channel.

Figure 1 Schematic diagram of counterattack

Figure 2 Counterattack equivalent circuit

1.3 Shielding lightning overvoltage

When the lightning current directly hits the transmission line, the voltage of the transmission line to the ground rises sharply due to the injection of a large amount of lightning current. When the potential difference between the two ends of the insulator string is greater than the impulse flashover voltage of the insulator string, the insulator string flashes and the conductor discharges to the ground through the tower. When the shielding failure occurs, the lightning current first acts directly on the conductor. Therefore, the current traveling wave of the conductor during the shielding failure is all the lightning current component, and there is no electromagnetic coupling component similar to the counterattack. The schematic diagram of its occurrence process is shown in Figure 3.

Figure 3 Schematic diagram of bypass attack

When a shielding failure occurs, the Peterson equivalent circuit of the two processes of conductor potential rise and insulator string flashover is shown in Figure 4. The insulator string acts as the equivalent circuit switch S. When the insulator string is not broken down and S is not closed, the lightning current causes the transmission line potential to rise, and there is no lightning current in the tower and lightning conductor; after the insulator string breaks down and S is closed, the lightning current passes through the tower into the ground. Since the transmission line is directly affected by the lightning current, there is no electromagnetic coupling current component in the transmission line lightning current during the shielding failure process.

Figure 4 Shielding failure equivalent circuit

From the above analysis, it can be seen that when an induced lightning overvoltage occurs in a transmission line, the current traveling wave is an induced current, and the three phases are basically similar; when a counter-attack occurs, before the insulator breaks down, the current is an electromagnetic coupling current, and after the breakdown, the line current suddenly changes from an electromagnetic coupling current to a direct lightning current. When a shielding failure occurs, the line current is a direct lightning current component.

2 Lightning overvoltage electromagnetic transient simulation of transmission lines

2.1 Induced lightning overvoltage

This paper uses the EMTP electromagnetic transient simulation program to simulate and calculate the three types of lightning overvoltages of overhead transmission lines: induction, shielding and counterattack. Figure 5 is a simulation model of a 220 kV system. The transmission line is 30 km long. According to the tower size structure, the calculation formula proposed in the literature [17] is used to obtain the tower segmented wave impedance model, as shown in Figure 6. In the figure, ZA is the cross arm wave impedance, ZT is the support wave impedance, ZL is the bracket wave impedance, and Rg is the tower impulse grounding resistance. The insulator adopts the voltage-controlled switch model, and the line lightning arrester adopts the nonlinear model recommended by IEEE. The line span is 200 m. The lightning conductor is not eliminated during the calculation to consider the influence of the lightning conductor on the lightning current propagation process. The lightning current adopts the standard waveform of lightning protection design, with a wave head of 2.6 μs and a wavelength of 50 μs. In order to accurately simulate the transmission characteristics of the current traveling wave in the transmission line, the model uses multiple 200 m span lines and tower models in series.

Figure 5 220 kV transmission system model diagram

Figure 6 Tower structure and wave impedance model

2.2 Time domain waveform analysis

Using the above model, the three situations of transmission line suffering from induction, shielding and counter-strike lightning overvoltage were simulated respectively. The current signal of the transmission line was obtained by sampling 1 km after the lightning strike point. The current traveling waves of the three lightning overvoltages are shown in Figures 7 to 9.

Figure 7 Three-phase current waveform of induced lightning overvoltage

Figure 8 Reverse three-phase current waveform

Figure 9 Three-phase current waveform of shielding failure

It can be seen from Figures 7 to 9 that the three-phase current traveling wave of the induced lightning overvoltage is an induced current traveling wave. After attenuation, the three phases are still basically similar. Before the three-phase current traveling wave of the counter-attack overvoltage jumps sharply, there is an electromagnetic coupling current component, which has a small steepness and a long rise time; after the insulator string breaks down, due to the large amount of lightning current injected into the conductor, the current traveling wave amplitude jumps and the steepness increases. There is no electromagnetic coupling current in the winding overvoltage, and the current traveling wave jumps rapidly after the lightning strike occurs. Therefore, the similarity of the three-phase current traveling waves and the presence or absence of electromagnetic coupling current are important features for judging the type of lightning overvoltage.

3. Identification criteria for lightning overvoltage on transmission lines

From the above analysis, the difference between the current traveling waves of the three types of lightning overvoltages is mainly reflected in the similarity of the three-phase current traveling waves and the existence of electromagnetic coupling current traveling waves. The three-phase current traveling waves of the induced lightning overvoltage are more similar, while the three-phase current similarity of the direct lightning overvoltage is lower due to the direct injection of lightning current. Based on this feature, the identification criteria for induced lightning point overvoltage and direct lightning overvoltage are proposed:

Where S Thres is the threshold value of the induced lightning overvoltage amplitude criterion; Smin is the minimum similarity of the three-phase current traveling wave. For signals X (n) and Y (n), the similarity S is calculated as:

In actual calculation, in order to eliminate the interference caused by flashover and reduce the amount of calculation, the 4μs before the current wave peak is taken as the similarity calculation interval. When Smin is greater than the threshold value, it is determined to be an induced lightning overvoltage. Otherwise, it is determined to be a direct lightning overvoltage.

When a counter-strike lightning overvoltage occurs and the insulator string is not broken down, the current traveling wave only contains an electromagnetic coupling component with a low amplitude. After the insulator string is broken down, the lightning current is injected into the conductor, and the current traveling wave of the flashover phase is the lightning current with a large amplitude jump. Let Imax be the maximum amplitude of the current traveling wave of the struck phase. According to the simulation results, the current traveling wave amplitude of the electromagnetic coupling component is about 5%Imax. Therefore, before the amplitude of the current traveling wave during the counter-strike jumps sharply, there is a rising process with an amplitude of about 5%Imax. During this rising process, the current traveling wave has a low steepness and a long rising time. After the insulator is broken down, the lightning current is injected, the current traveling wave steepness increases, and the rising time is short. During the shielding failure, since the lightning current has been directly injected into the conductor before the insulator breaks down, the current traveling wave of the struck phase has no electromagnetic coupling current component, and the steepness is large and the rising time is short.

Assume t2% as the time point when the amplitude of the lightning current traveling wave reaches 2%Imax, t5%, t50% and so on. Since it is difficult to determine the exact starting point of the lightning current, t2% is used as the starting point for calculation. In order to avoid the influence of the reflected traveling wave of the lightning current, the time t1 taken for the current traveling wave to reach 50%Imax is used to characterize the rise time of the lightning current, and the time t2 taken for the current traveling wave to reach 5%Imax is used to characterize the rise time of the spatial electromagnetic coupling traveling wave, and their ratio is defined as ρ, and the calculation formula is:

For the counter-strike lightning overvoltage, due to the existence of electromagnetic coupling components, t2 is the rise time of the electromagnetic induction current traveling wave, and t1 is the rise time of the lightning current injected into the conductor. Since the electromagnetic coupling component has a low steepness, the rise time is long, while the lightning current has a relatively large steepness and a short rise time. Therefore, the value of parameter ρ will be too small when a counter-strike occurs. For shielding failure, due to the absence of electromagnetic coupling components, t2 and t1 are the time taken by the lightning current to reach 50%Imax and 5%Imax, respectively. Therefore, when a shielding failure occurs, parameter ρ will be > 1. Based on this feature, the identification criteria for shielding failure and counter-strike lightning overvoltage are:

Where ρThres is the threshold value for bypass attack and counterattack.

Summarizing the above analysis, the identification process of the three types of lightning overvoltages, namely, induction, shielding and counter-strike, is shown in Figure 10.

Figure 10 Lightning overvoltage identification flow chart

4 Simulation Verification

Based on the above simulation model, this paper simulates three types of lightning overvoltages, namely induction, shielding and counterattack. The lightning current adopts the standard waveform 2.6/50μs waveform of lightning protection design. The lightning current traveling wave is collected at four distances of 1, 1.5, 3 and 5 km after the lightning strike point, and the characteristic parameters such as Smin and ρ of the three types of overvoltages are calculated. The simulation results are shown in Table 1.

Table 1 Characteristic parameters of induced lightning, shielding strike and counterattack

According to Table 1, by setting S Thres to 0.8 and ρThres to 2.5, the criterion proposed in this paper can accurately judge the type of lightning overvoltage occurring in front of the line at the rear. The simulation results under different lightning current waveforms show that the extracted feature quantity is basically not affected by the dispersion of the lightning current waveform. The simulation results of this paper under different voltage level systems show that the method is applicable to 110, 220, and 500 kV voltage level systems. Due to the attenuation of the signal along the line, according to the simulation results, when the distance between the signal sampling point and the lightning strike point is > 8 km, the electromagnetic induction component current traveling wave basically decays to zero, and the criterion for the counter-strike and shielding lightning overvoltage fails. Since the attenuation degree of the three-phase conductors is basically the same, the criterion for the induced and direct lightning overvoltage is still valid. The method of obtaining the lightning current traveling wave signal can consider the method proposed in references [20] and [21].

5. Influence of other factors on the line

5.1 Impact of corona

Due to the effect of impulse corona, when an overvoltage occurs in an actual overhead transmission line, the line-to-ground capacitance increases, the current waveform will be distorted, and the steepness will decrease. The EMTP simulation software has not yet been able to simulate this process. Considering that the attenuation distortion degree of the three-phase conductors under the action of impulse corona is basically equal, the impulse corona will not affect the waveform similarity. As for the identification of shielding and counterattack, since the impulse corona mainly acts after the lightning current is injected into the conductor, the rise time of the current wave in the interval [5% Imax, 50% Imax] increases, while the rise time of the current wave in the interval [2% Imax, 5% Imax] is basically not affected by the impulse corona. Therefore, the characteristic quantity ρ of the actual line is generally larger than the simulation result. As long as the threshold is adjusted appropriately, the influence of the impulse corona can be avoided [22].

5.2 Impact of induced overvoltage during counterattack

When lightning strikes a pole tower and strikes back, before the insulator flashes over, the change in the electromagnetic field in space will generate an induced overvoltage on the transmission line, and the induced overvoltage will generate an induced current traveling wave on the transmission line. The current simulation software has not been able to simulate this process. The existence of the induced current traveling wave will slightly increase the amplitude of the current traveling wave before the insulator string flashes over. Since the steepness of the induced overvoltage is lower than that of the direct lightning, after considering the induced lightning overvoltage, the steepness of the current traveling wave before the insulator flashes over is still lower than that of the lightning current traveling wave, so the induced current traveling wave will not have a fundamental impact on the judgment criteria of the shielding and strike back proposed in this paper. Based on the simulation result that the amplitude of the electromagnetic coupling current traveling wave is about 5% Imax, this paper defines the calculation interval of the rise time ratio ρ as [2% Imax, 5% Imax] and [5% Imax, 100% Imax]. In practical applications, considering the influence of the induced current traveling wave, the calculation interval of ρ can be slightly adjusted according to the actual operating conditions to avoid the influence of the induced overvoltage of the lightning tower during the strike back.

5.3 Influence of Transmission Line Terminals

In actual transmission lines, both ends are often connected to lightning arresters, substations, power supplies, etc. The current wave will be refracted and reflected multiple times at the busbar. The identification feature used in this paper is taken from the wave head of the current wave of the transmission line. The characteristic parameter calculation interval proposed in this paper is 4μs before the peak of the current wave. As long as the current signal sampling point is > 600 m away from the busbar, it can be considered that the current wave head obtained does not contain the reflected current wave, and the extracted feature is authentic. As long as the appropriate measurement point is selected, the wiring at both ends of the transmission line will not affect the identification method proposed in this paper.

6 Conclusion

In order to avoid the influence of the internal wiring of the substation on the identification of the type of lightning overvoltage, this paper analyzes the waveform characteristics of the current traveling wave of the transmission line when the three types of lightning overvoltages occur, namely induction, shielding and counterattack, and verifies these characteristics using EMTP simulation. Based on the simulation results, the identification characteristic quantity of lightning overvoltage is proposed, and the influence of other factors on the characteristic quantity in the actual line is analyzed, and the following conclusions are drawn:

a) When an induced lightning overvoltage occurs, the three-phase current waveforms are basically similar because the spatial positions of the three-phase conductors are basically similar; when a counter-attack occurs, before the insulator breaks down, the current traveling wave is an electromagnetic coupling component with a small amplitude. The steepness is low and the rise time is long. After the insulator breaks down, the amplitude of the flashover phase current traveling wave rises sharply because the lightning current is directly injected into the conductor; when a shielding failure occurs, the lightning current is directly injected into the conductor, and the current traveling wave does not have an electromagnetic coupling component, and its amplitude rises sharply directly.

b) Based on the above characteristics, the minimum similarity Smin of the three-phase waveforms can be used to distinguish between induced and direct lightning overvoltages. For direct lightning overvoltages, the ratio of the rise time ρ of the lightning phase current traveling wave head in the two intervals of [2% Imax, 5% Imax] and [5% Imax, 50% Imax] can be used to identify the shielding and counter-strike lightning overvoltages.

c) For practical engineering applications, the reasonable setting of the calculation range of the current threshold and the rise time ratio ρ can avoid the impact of the corona and the induced overvoltage when the lightning strikes the tower; the reasonable setting of the current traveling wave signal sampling point can avoid the interference of the line terminal refraction and reflection waves on the identification parameters. The simulation shows that the characteristic quantity proposed in this paper is not affected by the lightning current waveform parameters and can effectively identify the lightning overvoltage such as induction, shielding, and counterattack.

References

[1] Zhang Weibei, He Jinliang, Gao Yuming. Overvoltage protection and insulation coordination[M]. Beijing: Tsinghua University Press, 2002.

[2] LIANG Yi-ming, GE Dong. Tower model in calculation of light-2 ning protection on multi-circuit transmission line with same tower [J]. High Voltage Engineering, 2006, 32 (1): 76-78.

[3] YANG Qing, ZHAO Jie, SIMA Wenxia, ​​et al. Lightning back flashover performance of t he yun2guang U HV DC t rans2mission lines [J]. High Voltage Engineering, 2008, 34 (7): 133021335.

[4] FU Jin, SIMA Wen2xia, LI Jian2biao, et al. Estimation on lightning shielding performance of t he t ransmission lines based on t he f ractal t heory [J]. High Voltage Engineering, 2009, 35(6): 127421278.

[5] SHU Hong2chun , ZHANG Guang2bin , SUN Shi2yun , et al . Identification of shielding failure and back st riking in U HVDC transmission lines [J ] . Proceedings of t he CSEE ,2009 , 29 (4) :13219.

[6] QI Chong. The differentiation of lightning shield failure and back flashover in high2voltage transmission lines[D] . Nanning: Guangxi University ,2007.

[7] LI Hai2feng , WANG Gang , ZHAO Jian2cang. Study on charac2 teristics and identification of t ransmission lines caused by indirect lightning st roke [ J ] . Proceedings of t he CSEE ,2004 , 24 (3) : 1142119.

[8] DUAN Jian2dong, REN Jin2feng, ZHANG Bao2hui, et al. Study of transient approach of discrimination lightning disturb2ance in ultra2high2speed protection [J]. Proceedings of the CSEE, 2006, 26 (23): 7213.

[9] WANG Gang, LI Haifeng, ZHAO Jiancang, et al. Identification of transient on transmission lines caused by direct lightningst rokes based on multi resolution decomposition[J]. Proceedingsof the CSEE, 2004, 24 (4): 1392144.

[10] SI Da2jun , SHU Hong2chun , CHEN Xue2yun , et al . Study on characteristics and identification of t ransient on t ransmission lines caused by lightning st rokes[J ] . Proceedings of t he CSEE ,2005 , 25 (7) : 64269.

[11] SHU Hong-chun, WANG Yong-zhi, CHENG Chun-he, et al. Analysis and fault identification of lightning electromagnetic transients in ±800 kV DC transmission lines[J]. Proceedings of the CSEE, 2008, 28 (19): 932-100.

Analysis of electrified romagnetic t ransient and fault detection on ±800 kV U HVDC t ransmission lines under light ning st roke[J]. Proceedings of the CSEE, 2008, 28 (19): 932100.

[12] GUO Ning2ming , QIN Jian. Locating method for short2circuitpoint for power t ransmission line under lightning fault st roke[J ] . Automation of Electric Power Systems , 2009 ,33 (10) :74277.GUO Ning2ming , QIN Jian. Locating method for short2circuitpoint for power t ransmission line under lightning fault st roke[J ] . Automation of Electric Power Systems , 2009 ,33 (10) :74277.

[13] DUAN Jian2dong, ZHANG Bao2hui, HAO Zhi2guo, et al. Identification of lighting and fault in EHV t ransmission linet ransient2based protection [J]. Automation of Electric Power Systems, 2004, 28 (18): 30235.DUAN Jian2dong, ZHANG Bao2hui, HAO Zhi2guo, et al. Identification of lighting and fault in EHV t ransmission linet ransient2based protection [J]. Automation of Electric Power Systems, 2004, 28 (18): 30235.

[14] GUO Ning2ming , QIN Jian , CHN Xiang2xun. Lightningst roke identification and its effect on t he t raveling wave fault location[J ] . Automation of Electric Power Systems , 2008 , 32(5) :76279.

[15] DU Lin, DAI Bin, SIMA Wen2xia, et al. Overvoltage identifi2cation in dist ribution networks based on support vector machine[J]. High Voltage Engineering, 2009, 35 (3): 5212526.

[16] Voislav J ankov. Estimation of the maximal voltage induced onan overhead line due to the near by lightning [J]. IEEE Transon Power Delivery, 1997, 12 (1): 3152324.

[17] Yamada T, Mochizuki A, Sawada J, et al. Experimental evaluation of a U HV tower model for lightning surge analysis [J]. IEEE Trans on Power Delivery, 1995, 10 (1): 3932402.

[ 18 ] IEEE Working Group 3. 4. 11 of Surge Protection Devices Com2mittee. Modeling of metal oxide surge arresters [ J ]. IEEETrans on Power Delivery, 1992, 7 (1): 3022309.

[19] Shi Wei, Guo Jie. Power System Overvoltage Calculation[M]. 2nd edition. Beijing: Higher Education Press, 2006.

[20] DU Lin, L IU Wei2ming, WANG You2yuan, et al. Data acqui2sition card wit h variable sampling speed for monitoring over2 voltage based on CPLD[J]. High Voltage Engineering, 2008, 34 (8): 158921593.DU Lin, L IU Wei2ming, WANG You2yuan, et al. Data acqui2sition card wit h variable sampling speed for monitoring over2 voltage based on CPLD[J]. High Voltage Engineering, 2008, 34 (8): 158921593.

[21] LI Xianzhi, DU Lin, CHEN Weigen, et al. A novel schemeof draw2out power supply utilized in transmission line state mo2nitoring[J]. Automation of Electric Power Systems, 2008, 32(1): 76280.

[22] Shu Hongchun, Si Dajun, Yu Jilai. Electromagnetic transient simulation of lightning-struck transmission lines[J]. Automation of Electric Power Systems, 2005, 29(17): 682-92.