1、 Abstract This paper mainly elaborated some new problems on the fault location for half-wavelength transmission lines,and studied its fault location method.The results of the research show that the accuracy of the traditional double-terminal fault location method is poor for half-wavelength transmis
2、sion lines.While adding the traveling wave detection units along the line and using the three-point method can effectively improve fault location accuracy.In addition,simulation results indicate that with the increase of the number of detection points,the accuracy of fault location is gradually impr
3、oved.But when the distance between two adjacent detection points is reduced to a certain extent,the location accuracy is no longer improved.It is reasonable to take 200-300km between adjacent detection points.For the capacitance tuning network,the arrangement of the measuring points also should ensu
4、res that the number of capacitors between the fault point and the three nearby measuring points is sufficiently small.Index Terms UHV half-wavelength transmission,fault location,multiple measuring points,wavelet transform.I.INTRODUCTION ROM the principle of traveling wave fault location,we can see t
5、hat the accuracy of fault location is related to the accuracy of the line length,wave velocity,and wave arrival time.The length of the half-wavelength transmission line is as long as 3000km.The range of the lines full-length change is greater;The distribution of parameters along the line is more une
6、ven,coupled with the influence of traveling wave dispersion,the change of wave velocity is more obvious;With the increase of travel distance,the attenuation effect of the line on the traveling wave is enhanced1.In particular,the attenuation of the traveling wave by shunt capacitors or series inducto
7、rs in the tuning network makes the detection of the wave head more difficult,and the accuracy of the obtained wave arrival time is lower.Compared with the conventional Guan Wang,is with the school of Electrical Engineering,Shandong University,PhD(e-mail:).Haoran Zhao,is with the school of Electrical
8、 Engineering,Shandong University,masters degree(corresponding author to provide phone:15863158581;e-mail:).transmission lines,there are some new problems in the half-wavelength transmission lines.The traditional method of fault location2-5 using the data obtained by the single terminal or double ter
9、minals of the line may not be applicable in the half-wavelength transmission lines.In the half-wavelength transmission line,if only the data obtained at double terminals of the line is used for fault location,the measurement span is too large.Adding traveling wave detection units along the line can
10、reduce the measurement span6-8.In tuning networks,increasing the detection point can also solve the problems brought about by capacitors or inductors.Theoretically,this method can improve the accuracy and reliability of fault location in half-wavelength transmission lines.In this paper,the applicabi
11、lity of this method in the half-wavelength transmission line is studied through theoretical analysis and simulation,and comparing with the traditional double-terminal fault location method.II.MAIN PROBLEMS OF FAULT LOCATION IN HALF-WAVELENGTH TRANSMISSION LINES A.Influence of Line Length Change on F
12、ault Location Accuracy The formula for traditional double-terminal fault location method is as follows:2)(2)(v t t Llv t t LlM NNN MM(1)From the formula,it can be seen that the error L of the line length will cause 2/L fault location error.Transmission lines are affected by sag,climate,and load curr
13、ents.The length error of a 3000-km half-wavelength transmission line may reach several tens of kilometers9.B.Wave Attenuation and Distortion During Propagation Due to the presence of line resistance,insulation conductance,and corona,the wave gradually attenuates and distorts during propagation.Which
14、 will make the wave head smooth,making it difficult to accurately and reliably identify the wave arrival time.The figure 1 shows the current traveling Research on Fault Location for UHV Half-wavelength Transmission Lines Guan Wang 1,2,Haoran Zhao 1,2,Qingshui Hao 3 1.School of Electrical Engineering
15、,Shandong University,Jinan 250061;2.Shandong Provincial Key Laboratory of UHV Transmission Technology and Equipment;3.Shandong Greenthink Power Technology Co.,Ltd.F 2018 China International Conference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201804230000019 Page1/5 1341wa
16、ve modulus waveform measured at a distance of xkm from the fault point when a fault occurs at a distance of 200km from the sending end of the line.As can be seen from the figure,as the travelling wave propagation distance increases,the wave head becomes more and more smooth,and it is difficult to ac
17、curately obtain the wave arrival time of the fault traveling wave at the end of the line.Fig.1.The current traveling wave modulus waveform measured at a distance of xkm from the fault point In addition,dispersion occurs as traveling waves propagate.If the travel distance of the fault wave is too lon
18、g,the high-frequency component with a faster speed attenuates seriously,and the fault location device may not be able to capture it,resulting in that the arrival time obtained is the arrive time of the component with lower frequency,the attenuation of which is relatively weak10.This leads to the fac
19、t thatthe actual wave velocity will gradually decrease as the travel distance of the fault wave increases.C.Attenuation of Traveling Wave by Shunt Capacitors and Series Inductors When the electrical length of the half-wavelength transmission line is shorter than half a wavelength,appropriate technol
20、ogy should be employed to tune the electrical length manually.In the tuning network,shunt capacitors and series inductors will cause the amplitude and waveform of the traveling wave on the line to change.If a fault occurs on the tuning line,when the fault traveling wave reaches the ends of the line
21、through capacitors or inductors,the attenuation of the wave headwill be severe,and it will be so difficult for the location device to accurately detect the arrival time of the wave head,which will greatly reduce the accuracy of the fault location.The following figure 2 shows the current traveling wa
22、ve modulus waveform measured at a distance of xkm from the fault point when a fault occurs at a distance of 200km from the sending end of the capacitance tuning line with natural length of 2400km.In this simulation,the fault wave reached the position 100km,400km,2200km away from the fault point and
23、past through 0,1 and 7 capacitors respectively.It can be seen from the figure that the more capacitors fault wave past,the more severe was the attenuation of it.When the fault wave reaches the end of the line through 7 capacitors,its arrival time is difficult to be clearly identified.In the T-type t
24、uning line(whose natural length is 2400km,and of which the tuning network is symmetrically installed at both ends),the effect of frequency on the line parameters is not taken into account.When the fault occurs at a distance of 200km from the sending end of the line,the waveform of the current travel
25、ing wave modulus measured at the head and end and middle point of the line is shown in Figure 3.When the fault wave reaches both ends of the line,the attenuation of the wave head is severe,and it is difficult to accurately obtain the wave arrival time.The condition in the-type tuning line is similar
26、 to that of the T-type tuning line.Fig.2.The current traveling wave modulus waveform measured at a distance of x km from the fault point Fig.3.The current traveling wave modulus waveform measured at the head and end and middle point of the line III.A TRAVELING WAVE FAULT LOCATION METHOD BASED ON MUL
27、TIPLE MEASURING POINTS Multi-point fault location method needs to add traveling wave detection unit along the line.When a fault occurs,Rogowski coils are used to transmit high-frequency signals in each traveling wave detection unit,and the analog signals are converted into digital signals and upload
28、ed to the master station.The master station obtains the wave arrival time of the fault traveling wave by wavelet transform analysis,and uses the wave arrival time obtained at each measuring point to judge the fault section,and then selects three measuring points closest to the fault point.The fault
29、location was calculated by the time at which the wave arrives at the three points and the length of lines between the three points.2018 China International Conference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201804230000019 Page2/5 1342A.Fault Section Identification In th
30、e existing multi-point fault location methods,some of which distinguished fault section by screening two measuring points which had detected the smallest wave arrival time.This method is applicable only if the measuring points are evenly distributed along the line.Some used N N discriminant matrix(N
31、 is the number of measuring points).In this method,the data redundancy is high and the amount of calculation is large.This paper provides a simpler method which also applies when the length of each section is different.As shown in Figure 4,the line is divided into n segments.The length of each segme
32、nt can be different.There are n+1 measuring points in total.When a fault occurs at time t0,the arrival time of the first wave head detected by each measuring point is tk(k=1,2,3.n+1).Calculate the difference between the arrival times of two adjacent measuring points in order tk=tk-tk+1(k=1,2,3.n).As
33、 can be seen from Figure 5,the slope of each section on both sides of the fault point should be approximately equal under normal conditions.This can be used to eliminate the time of mis-marking.Under normal circumstances,there will be three kinds of situations:(1)If tk0(k=1,2,3.n),the fault point lo
34、cated between measuring points Dn and Dn+1;(3)If tk-10,tk0(k=2,3.n),further judgement is needed.Calculate 1,11 kk ktlv and kk ktlv 1,2 separately,if 2 1v v,the fault point located between measuring points Dk-1 and Dk;if 2 1v v,the fault point is located between the detection points Dk and Dk+1.lk,k+
35、1M N1D2D1 kDkD1 kD2 kDnD1 nD Fig.4.Distribution of traveling wave detection units along the line 1D2D1 kDkD1 kDnD1 nDtl0tkt1 kt1t2t1 ktnt1 nt Fig.5.Wave arrival time at each measuring point when a fault occurs B.Three Point Fault Location Method Assuming that the fault section is Dk Dk+1,as shown in
36、 Figure 6.According to the traditional double-terminal traveling wave fault location principle,the following formula can be written by using the wave arrival times obtained at the detection points Dk and Dk+1:2)(1 1,1v t t ldk k k k(2)2)(1 1,2v t t ldk k k k(3)In order to eliminate the v in the equa
37、tion,introduce the data of a measuring point Dx closest to the fault section.For capacitance tuning lines,the data of measuring point Dx which has the least amount of capacitance between it and the fault point is preferentially introduced.If Dx is Dk+2,the following formula can be written by using t
38、he data obtained at Dk and Dk+2:2)(2 2,1v t t ldk k k k(4)Simultaneous formula(2)(4)can obtain:2)(2)(1,1 22,1 11 k kk kk k k klt tl t td(5)If Dx is Dk-1,the following formula can be written by using the data obtained at Dk-1 and Dk+1:2)(1 1 1,12v t t ldk k k k(6)Simultaneous formula(3)(6)can obtain:
39、2)(2)(1,1,1 12 k kk kk k k klt tl t td(7)k kl,1 1,k kl 2,1 k kl 1d 2d1 kDkD1 kD2 kDFig.6.Schematic diagram of partial line near the fault point IV.SIMULATION EXAMPLE The simulation structure of power system is shown in Figure 7.Frequency-dependent line model used in simulation.The wire type is 8 LGJ
40、500/35,using a triangular arrangement.The fault mode is set to phase A ground fault.The simulation sampling frequency is 1MHz.The db5 wavelet is used to analyze the waveform.mEnEM N Fig.7.Simulation Structure of Power System 2018 China International Conference on Electricity Distribution Tianjin,17-
41、19 Sep.2018CICED2018 Paper No.201804230000019 Page3/5 1343 A.Fault Location Simulation for Half-wavelength Transmission Lines The natural length of the transmission line is 3000km,and the measuring points are evenly distributed along the line with a distance of 300km.The fault points are located at
42、200km,760km,1400km,2350km,and 2860km away from the terminal M respectively.The failure occurs at time t=0.815s.Table I lists the simulation results of multi-point fault location and traditional double-terminal fault location when single-phase ground faults occur at different locations.In traditional
43、 double-terminal fault location,the velocity of the wave is taken as v=2.98108m/s.The results show that the traditional double-terminal fault location method has poor accuracy,while the multi-point fault location method has higher accuracy.TABLE I SIMULATION RESULTS OF MULTI-POINT FAULT LOCATION AND
44、 TRADITIONAL DOUBLE-TERMINAL FAULT LOCATION Location of fault point/km Multi-point Traditional double-terminal Location calculated/km Error/m Location calculated/km Error/m 200 200.4 400 201.912 1912 720 720.472 472 721.624 1624 1430 1429.762 238 1429.672 328 2350 2351.190 1190 2347.512 2488 2810 28
45、09.524 476 2808.816 1184 In the multi-point fault location method,the distance between the measuring points will affect the accuracy of the fault location.Table II lists the fault location simulation results under various measuring point spacing when a fault occurs at a distance of 720km from the te
46、rminal M.The results show that as the distance between measuring points decreases,the location accuracy increases until the measuring point spacing decreases to a certain extent.TABLE II THE FAULT LOCATION SIMULATION RESULTS UNDER VARIOUS MEASURING POINT DISTANCES location of fault point/km Spacing
47、of adjacent measuring point/km Locationcalculated/km Error/m 720 500 718.900 1100 300 720.472 472 200 720.238 238 100 720.238 238 B.Fault Location Simulation of Capacitance Tuning Transmission Lines This simulation mainly discusses the influence of shunt capacitance on fault location,so it adopted d
48、istributed line model.By converting the parameters of the above line model,the power frequency parameters of the line can be obtained as follows:TABLE III POWER FREQUENCY PARAMETERS OF 1000KV TRANSMISSION LINES Parameter Positive sequence Zero sequence R(/km)0.0081 0.1586 L(mH/km)0.847 2.154 C(F/km)
49、0.0138 0.00845 Assumed that the natural length of the transmission line is 2400km,and use the capacitance tuning method to compensate:Divide the line into 8 equal parts,place a tuning capacitor every 300km.According to the formula:NlCL lC)(002 22(8)we can obtain the compensation capacitance value C
50、2.0 0 9 F,The capacitance values at both ends of the line are C/21112.N is the number of segments.One measuring point is set every 300km.The specific structure is shown in the following figure:1D2D3D4D5D6D7D8D9DC C C C C C C2/C 2/C Fig.8.Simulation structure diagram of capacitance tuning lines Table
