1 Introduction Transmission line fault traveling wave location can be divided into current traveling wave fault location and voltage traveling wave fault location according to the signal. Due to the poor transient response characteristics of capacitive voltage transformers (CVTs), they cannot meet the requirements of traveling wave measurement, so voltage traveling wave fault location is difficult to achieve. There is little research on voltage traveling wave fault location in China, and the main research is to use current traveling waves for fault location [1-3]. The key to transmission line fault traveling wave location is to find the arrival time of the traveling wave front. In China, software methods are usually used for this purpose. This method requires a high-speed acquisition system (if dual-end traveling wave location is used, GPS clock signals also need to be recorded) and complex analysis and calculation; when implemented in the field, it requires strong hardware and software support, which increases the cost of the traveling wave location device, and the detection accuracy is limited, making it difficult to detect faults with an initial phase angle of less than 10° [1,2]. A voltage traveling wave fault location system developed in Canada and operated in the BC Hydro transmission network has a field location error of less than 300m [4]. The system connects a reactor in series on the CVT ground line to extract the voltage traveling wave generated by the power system fault or disturbance. The traveling wave head is detected by the peak value and rise time criteria (the rise time is 0.7 to 8.3 ms, and the corresponding traveling wave frequency is 30 to 350 kHz). The fault location is directly determined by the arrival time of the traveling wave head, so there is no need for complex high-speed acquisition and information processing [4]. However, when this positioning system is installed, the primary system wiring needs to be changed, which may have a certain impact on the system operation. It is difficult to gain the approval of the power operation department in my country and it is difficult to promote its application. Therefore, it is necessary to develop a traveling wave sensor that has no direct electrical connection with the primary system and does not affect the system operation. This paper adopts the method of using the current traveling wave on the CVT ground line to reflect the line voltage traveling wave for fault location. A traveling wave sensor has been developed, which can realize voltage traveling wave fault location based on the entire power grid. [b]2 Extraction of traveling wave signal 2.1 Selection of measurement signal[/b] Due to the poor high-frequency transmission characteristics of CVT, the voltage traveling wave cannot be directly extracted from the secondary side of CVT. Analysis shows that the traveling wave signal of the current entering the ground on the CVT ground wire is i, which is the current flowing from the line or bus to the ground through the CVT; c is the CVT capacitance; and u is the line-to-ground voltage. Equation (1) can reflect the voltage traveling wave signal of the line or bus. The current on the CVT ground wire is the derivative of the line voltage. For a sine wave, the derivative is related to the frequency, and the higher the frequency, the larger the derivative value. Since the fault transient traveling wave front contains rich high-frequency components, the fault traveling wave front of the CVT ground current has a larger change than the fault traveling wave front of the line voltage, which is more conducive to fault detection. Taking a 500kV power grid shown in Figure 1 as an example, considering the actual structure and frequency response characteristics of the line, the EMTP simulation software is used for analysis. When the DE line is grounded, the current traveling wave, voltage traveling wave and CVT ground current traveling wave measured on the D side are compared, and the waveforms are shown in Figure 2. It can be seen that: (1) Since the D transformer bus has 3 outgoing lines (>2), and is affected by the substation's ground capacitance, the traveling wave front of the line current is significantly different from that of the line voltage, so it is easier to detect; (2) Due to the differential effect of the coupling capacitor, the signal of the sudden change of the traveling wave front of the ground current on the CVT is more obvious than that of the traveling wave front of the line voltage, so it is easier to detect; (3) Due to the influence of the wave trap, the sudden change of the ground current on the line CVT is greater than that on the bus CVT. The ground current on the line CVT is the easiest to find the traveling wave front. [b] 2.2 Selection of measurement points[/b] Generally, there are 2 grounding points for CVT, as shown in Figure 3. D1 is the grounding point of the primary side of the internal voltage transformer of the CVT, and D2 is the grounding point of the combined filter. Through parameter calculation, simulation analysis and high-voltage impulse test verification, most of the ground current of the CVT flows into the ground through D2, and the traveling wave sensor should be installed at point D2. To install a traveling wave sensor without shutting it down, a tap changer can be connected in parallel through the wire passing through the traveling wave sensor. [img=350,336]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/43-4.jpg[/img] [b]3 Development of Traveling Wave Sensor 3.1 Analysis of Traveling Wave Sensor Characteristics[/b] The traveling wave sensor is made by uniformly winding several layers of coils on a toroidal iron-cobalt-nickel alloy material with a uniform cross-section. As shown in Figure 3, the measuring conductor is located inside the large toroidal coil of the traveling wave sensor and has no direct potential connection with the secondary coil. To eliminate the influence of the magnetic flux linkage of the large coil, the coil is wound in an even number of layers, and the winding directions of adjacent layers are exactly opposite. The equivalent circuit of the sensor is shown in Figure 4. The primary input current i of the sensor generates an output voltage u1 on the load R. The self-inductance L of the coil and the inter-turn capacitance C0 of the coil form a filter circuit. The entire sensor functions as a bandpass filter with a lower cutoff frequency of 6 kHz and an upper cutoff frequency of 10 MHz. Within the passband frequency range, its transfer function remains constant. For a primary side current input of 1 A, the sensor output voltage is 1 V. Power system frequency signals and harmonic signals below the 100th order are filtered out by the traveling wave sensor, allowing direct fault location using its output signal, eliminating the need for a high-speed data acquisition system. This simplifies the fault location device and reduces its cost. [img=291,124]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/44-1.jpg[/img] The transmission path of the traveling wave sensor output signal is shown in Figure 3. After being limited by the surge arrester and TVS (Transient Voltage Suppressor) connected in parallel to the traveling wave sensor's auxiliary transformer, the voltage is divided by the resistor voltage divider circuit, outputting a traveling wave hardware start signal of -2.5 to 2.5V, which is sent to the traveling wave wavehead detection circuit for traveling wave wavehead identification. Its detection circuit is shown in Figure 5. [img=334,276]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/44-2.jpg[/img] [b]3.2 Detection of Traveling Wave Head[/b] The time interval between two consecutive recordings of the traveling wave starting by the traveling wave positioning device is only 0.1s, and the traveling wave positioning calculation is started by the relay protection trip signal (gatekeeping). The identification and analysis of the traveling wave and interference signal of the power system fault is very complicated and needs further research. Based on EMTP simulation analysis, high voltage impact experiment and literature [4], this paper uses the rate of change of the traveling wave, the rise or fall time and the amplitude of the traveling wave head to identify the traveling wave head. The set values ​​are: ① Rate of change is 0.01-0.1pu/ms (pu is the per unit value, and the rated value is taken); ② Rise or fall time is 1-10ms; ③ Amplitude is 0.1-0.5pu. Because the coupling capacitor amplifies the high-frequency components of the traveling wave, and the traveling wave sensor can filter out signals below 5kHz, the amplitude and rate of change of the traveling wave sensor output signal are relatively large. In most fault conditions, the sensor output signal is limited by a TVS, resulting in a rise or fall time <10ms and an amplitude >1V. Different fault traveling waves, after voltage suppression and voltage divider, produce signals with little difference, facilitating tuning. Corresponding to the traveling wave sensor output signal, the traveling wave front detection circuit shown in Figure 5 is typically tuned as follows: rate of change level 0.3V, hold time 10ms, and peak level 1V. This tuning has been verified in the field. The traveling wave front detection circuit can detect the polarity of the traveling wave front and can be further used for traveling wave direction protection. [img=352,229]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/44-3.jpg[/img] [b]3.3 Recording of Traveling Wave Head Arrival Time[/b] At the initial arrival time of the traveling wave head, the positive and negative trigger signals generated in Figure 5 are sent to Figure 6 to trigger the latch to store the time value of the GPS synchronization high-precision clock. The traveling wave detection signal generated in Figure 5 is also sent to Figure 6 to generate a read data signal, which reads the latch data into the CPU. Since the initial wave head of the fault traveling wave is used for positioning calculation, to improve the reliability of the recording, the arrival time of the three-phase traveling wave head is measured simultaneously. Among the three measured times, the earliest time is taken as the standard. Usually, the traveling wave of the fault phase has the largest change, and the measured wave head arrival time is the earliest. [b]4 Voltage Traveling Wave Fault Location Based on Traveling Wave Sensor 4.1 Fault Location Method[/b] A voltage traveling wave sensor is installed on the ground wire of the CVT bus of each substation in the power grid. When a fault occurs in the power grid, the arrival time of the traveling wave front is recorded. When the line fault trips, the dispatcher reads the arrival time of the traveling wave front recorded by each substation, and the fault location can be performed [5-8]. Fault location calculations can be performed by arbitrarily selecting one substation on each side of the fault point. For example, if substation 1 is selected on one side of the fault point in the power grid and substation 2 on the other side, the fault location formula is as follows: [img=288,36]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/44-5.jpg[/img] Where l1 is the distance from the fault point to substation 1; l is the shortest distance of the power line from substation 1 to substation 2 via the faulty line; is the propagation speed of the traveling wave on line 1-2; t1 is the arrival time of the traveling wave front detected by substation 1, and t2 is the arrival time of the traveling wave front detected by substation 2. Fault location can be performed by measuring the arrival times of the traveling wave fronts from any two substations at both ends of the faulty line. By fusing and processing the arrival time information of traveling waves measured by multiple substations in the entire power grid and performing fault-tolerant analysis, the influence of erroneous time recorded by some substations can be eliminated, and the robustness of fault location can be improved. In addition, the time difference recorded by the two ends of the non-faulty line can be used to measure the traveling wave velocity online. [b]4.2 Simulation Analysis of Sensor Output Waveform[/b] The traveling wave sensor model, TVS voltage limiting circuit model and voltage divider circuit model are added to the EMTP simulation model in Figure 1 to perform simulation analysis on various faults and measure the response of the traveling wave sensor under various faults and lightning strike conditions. The results show that the traveling wave sensor can generate an output signal under various fault conditions, including minor faults. The greater the voltage change at the fault point, the greater the sensor output signal. Due to space limitations, only the traveling wave waveforms measured by each substation when the fault occurs at a distance of 155.2km from the D side of the DE line are given here, as shown in Figure 7. It can be seen from the figure that: (1) Under general fault conditions, the traveling wave front signal output by the traveling wave sensor is truncated by the TVS. When the output signal reaches the full scale of 2.5V, the time for the amplitude to rise from 0V to 1V is less than 4ms, which meets the traveling wave initiation criterion; (2) When the voltage crosses zero instantaneously, the rising of the traveling wave front slows down and the amplitude is small, which does not meet the traveling wave initiation criterion. However, according to statistical data and theoretical analysis, the voltage crosses zero instantaneously faults are almost impossible to occur. This fault can only be located by other conventional methods [5,6]; (3) When the fault is near the voltage crosses zero (outside the range of -1.8° to 1.8°), the sudden change in the output signal of the sensor at both ends of the fault line can meet the fault initiation location requirements. However, it is difficult to initiate the traveling wave in power plants or substations more than 500km away from the fault point. Therefore, the location method based on the traveling wave sensor can locate the ground fault near the voltage crosses zero. [img=268,251]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/45-1.jpg[/img] [img=331,246]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/45-2.jpg[/img] [img=331,246]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/45-3.jpg[/img] [img=331,311]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/45-4.jpg[/img] 4.3 Fault Location Analysis Since the traveling wave sensor has a high sensitivity for detecting the traveling wave front, and the detected signal is the high-frequency component of the wave front, considering the dispersion characteristics of traveling wave transmission, the traveling wave velocity corresponding to a frequency signal near 1MHz is selected for fault location calculation. According to the line structure, the traveling wave velocity corresponding to a frequency signal near 1MHz is calculated to be approximately 296 × 10³ km/s. The simulation results for fault location at a distance of 155.2km from the D side of the DE line are shown in Table 1. The maximum location error is 0.73km, which is equivalent to the distance between two towers. This distance can be received on-site. [img=400,204]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/46-1.jpg[/img][align=left][img=400,177]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/zgdjgcxb/2002-6/46-2.jpg[/img] [b]5 Conclusion[/b] To reduce the cost of traveling wave positioning devices and improve positioning accuracy, this paper developed a specialized traveling wave sensor, realizing simple voltage traveling wave fault positioning without the need for high-speed data acquisition. The traveling wave positioning method has the following characteristics: (1) The traveling wave sensor is placed on the ground wire of the CVT and can measure the current traveling wave entering the ground through the CVT. It has no direct electrical connection with the primary equipment and is easy to promote and apply; (2) The traveling wave sensor has good frequency response characteristics and can effectively filter out frequency signals below 5kHz. Moreover, the differential of the coupling capacitor has an amplification effect on high frequency signals, which can effectively improve the sensitivity of measuring the traveling wave head and help to reduce the dead zone of traveling wave positioning near the voltage zero crossing; (3) Based on the voltage traveling wave fault positioning of the entire power grid, it has a certain fault tolerance capability, and the positioning result is accurate and robust; (4) The traveling wave sensor and the traveling wave detection circuit can detect the polarity of the traveling wave head and can be used for traveling wave direction protection. The traveling wave positioning device has passed the high voltage impact test and RTDS test, and the prototype is running well on site. References [1] Xu Bingyin. Techniques for fault location of transmission lines using transient traveling-waves [D]. Xi'an: Xi'an Jiaotong University, 1991. [2] Dong Xinzhou. Wavelet theory applied to study on fault location of transmission lines [D]. Xi'an: Xi'an Jaotong University, 1996. [3] Qin Jian. Study of wavelet transform applied in traveling-wave fault location of transmission line [D]. Beijing: Electric Power Research Institute, 1998. [4] Lee H, Mousa A M. GPS Traveling-wave fault locator systems: investigation into the anomalous measurements related to lightning strikes [J]. IEEE Trans on Power Delivery, 1996, 11(3): 1214-1223. [5] Zeng Xiangjun, Yin Xianggen, Chen Deshu, et al. Study on a new type of integrative fault location system for transmission line [J]. Automation of Electric Power Systems, 2000, 24(22):39-40. [6] Zeng Xiangjun. Research on advanced principles of power lines fault detection & fault location and their implementation with information fusion [D]. Wuhan: Huazhong University of Science and Technology, 2000. [7] Zeng Xiangjun, Yin Xianggen, Chen Deshu, et al. Research on GPS Traveling-wave fault location systems for transmission network [J]. Automation of Electric Power Systems, 1999, 23(10):8-10. [8] Zeng Xiangjun, Yin Xianggen, Chen Hao, et al. The development of GPS synchronous transient data acquisition systems. High Voltage Engineering. Engineering), 2000,26(2):56-58.[/align]