The invention concerns a method of fault location in transmission lines, a especially in HVDC (High Voltage Direct Current) mixed lines, where part of the line is overhead line and part is a cable line. The fault location is based on estimation arriving times of travelling waves induced by fault that are propagating along the line from faulted point to the measurement point, located at one end of the line.

Background of the invention

In transmission lines, different methods for fault detection and fault location may be used in order to locate accurately faulted place. For instance the most accurate solution in transmissions lines are utilizes the traveling wave principle - synchronized in time domain measurements from both ends of line. The computer device in a form of a fault locator monitors the traveling waves that are induced during appearing of faults. The waves are propagated from faulted place, with the velocity close to the speed of light in overhead lines, in both directions in transmission line. In the end of the transmission line the waves are detected and its times of arrival at the measuring points are determined.

From patent application WO2011127967A1 there is known a method and apparatus for determining the time of arrival of fault wave at a measurement point of power transmission system. The method is useful for localization a faulted place in transmission lines which is a homogeneous line. For long transmission mixed lines like overhead line and submarine cable or underground cable it is desirable to determine where such a fault occurs directly after the fault. The distance to the faulted place is proportional to the difference between times of arrival the first fault waves at measurement points in two stations and to the speed of the fault wave in transmission line. The distance to the fault can be determined with the accuracy of the synchronization device used for detection time of fault waves arriving. Usually the synchronization error is close to 200ns. This invention is in general a double-ended method. A single-ended transmission line fault locator is known from patent US 4 766 549. A system for locating faults in a transmission line is presented. The system for locating faults in a high voltage direct current (HVDC) transmission lines uses measurements taken from only one end of the transmission line. The system comprises means for detecting transients, ie. traveling waves, generated by a fault located in the transmission line and means for sorting by polarity detected transients for transients produced by a fault and transients reflected from the one end of the transmission line. Further the system comprises means for determination the arrival times of transients at one end of the transmission line and means for calculation the difference between the arrival times of such transients in order to determine the location of the fault. Such system is operating in a homogeneous type of transmission line - pure overhead line or pure cable line, but suffer some mal functions when the transmission line is consisted from two or more different segments, overhead and cable segments, where each segment has different surge impedances and different wave propagation velocities, what causes additional wave reflections at junction point of such two segments. The present invention offers an improvement over the above discussed single-ended techniques.

The present invention is utilized a creation of a traveling waves schemes what is known for example from a literature " Transmission Lines and Wave Propagation", by Philip C.Magnusson, Gerald C.Alexander, Vijai K.Tripathi, Andreas Weisshaar; 4th Edition, 2001, CRC Press LLC [chapter 3] for calculation the traveling wave pulses in a theoretical faulted point and theoretical arrival times of such pulses to the measuring points with a use a Bewley lattice diagrams.

Summary of invention

The essence of the invention consists on:

• creation a reference database, comprising theoretical time schemes for at least first, second and third theoretical traveling wave pulses Ρ , P2', P3' generated in theoretical faulted points for known theoretical distances XI established by user for known line parameters and calculation of times of arrival Tl', Τ2', T3' of such pulses in the measuring point (SI),

• detection the presence of a faulted pulses in unknown real point of the line for an unknown distance through measuring traveling wave pulses PI, P2, P3 in one end of the faulted line and calculation of arrival times of first Tl, second T2 and third T3 of traveling pulses PI, P2, P3 in a computer device for unknown distance for faulted line, where Tl is equal 0 when a magnitude of the first pulse PI has a value bigger than a threshold Th given by the user, and creation a real time scheme for pulses PI, P2, P3 generated in faulted point,

• comparing a real time scheme of pulses PI, P2, P3 generated in faulted point with theoretical time scheme of pulses PI', P2', P3' generated in known theoretical faulted points and matching by polarity checking of real time scheme for unknown point with all theoretical time schemes,

• indication the known distance for theoretical time scheme, which is the best matching time scheme with the real time scheme for unknown distance,

• indication the unknown distance as a value equal to the known distance what determines the location of faulted point.

Preferably theoretical time scheme for first, second and third theoretical traveling wave pulses PI', P2', P3' generated in theoretical faulted points are based on Bewley lattice diagrams for such pulses and theoretical arrival times Τ , Τ2', T3' calculated for theoretical known distances given by user, are taken for creation theoretical time schemes.

Preferably the process of matching real time scheme for unknown faulted point with theoretical time scheme is based on calculating a time difference ΔΤ2ι _< between the real arrival times T2 and theoretical arrival time T2' for second traveling wave pulses P2 and P2', . and a time differenceAT3k between the real arrival times T3 and theoretical arrival time T3' for third traveling wave pulses P3 and P3'. Preferably a sum of absolute values of time differences ΔΤ2ι _< and ΔΤ3ι _< is calculated for selection of a minimum value of a sum∑(k).

Preferably the unknown distance X of faulted point is determined by selection of a minimum value of an absolute sum of time differences.

Preferably after the determination the unknown distance of faulted point an alarm is triggering for faulted line and accurate value of fault location is known.

The advantage of the invention is an improvement over the prior art and allows for increasing an accuracy of a fault location in transmission lines, especially lines comprise many different segments having different surge impedance. The invention enables the fault location without using communication channel or GPS (Global Positioning System) like device between two terminals, because the only one end measurement is applied.

Embodiment of the invention

The inventive method is presented in exemplary embodiment on the drawing where: fig. 1 - depicts the transmission system in which the presented invention is operated, fig. 2 - depicts a set of steps for the realization of the inventive method, fig. 3 - depicts a Bewley lattice diagram for times of arrival of theoretical traveling waves propagated in a faulted point, placed in an overhead part of the transmission line, fig. 4 - depicts a Bewley lattice diagram for times of arrival of theoretical traveling waves propagated in a faulted point, place in a cable part of the transmission line, fig.5 - depicts a theoretical time scheme for pulses generated for fault at a known place XI and times of arrival such pulses to the measuring device, fig.6 - depicts a real time scheme for pulses generated in real faulted place X and times of arrival such pulses to the measuring device, fig.7 - depicts an example of comparison of times schemes taking from fig.5 and fig. 6. As shown in. fig 1 the system comprises a HVDC transmission line 1 having an overhead line section OHL and a cable line section CBL. The joint point between the two types of lines is indicated as J. The total length of the transmission line 1 is indicated as TL. The line 1 has a terminal A having a measurement point SI and a terminal B having a measurement point S2. Between the points SI and S2 the place of a fault through fault resistance RF is shown as point S3 which can be placed anywhere in the line sections OHL or CBL. In the terminal A, a fault locator 2 is placed having an electrical connection with the measurement point SI. The fault locator is a computer device equipped with standard means eg. processor, RAM, ROM, power supply etc., means for measuring and processing variable parameters of the line 1, not presented on the drawing, and means for detecting the fault in the form of a detection module 3. The detection module 3 is connected with an operational module 4 for generating and storing a special time scheme for data receive from module 3, concerning the time of arrival traveling waves from faulty point S3. The fault locator 2 is also equipped with a data-based creation module 5 in which a reference database is automatically created on the base of known constant and variable data of the transmission line 1 and on a base that the fault is placed in a known theoretical distance, which data is gathered in a special table, what will be explain later. Database creation module 5 is adapted for receiving the permanent data from user in terminal A and variable data from the point S3 by connection 6. Both modules: the operational module 4 and the data-based creation module 5 have their outputs to the input of a fault location module 7 for determination the fault location by comparison real data stored in module 4 with the theoretical data calculated and stored in module 5. The output of the module 7 is connected with a device 8 for presented the result of fault location in transmission line or a device for triggering an alarm if it needed, having a form of the know devices used in such situation (display, phone, warning light or sound, etc.). The fault locator 2 is placed in the terminal A but it could be situated in terminal B and in such situation the measuring point is S2 instead of the point SI. The method concerns only one single end fault location for the HVDC line. The method according to the invention comprises the following steps. Step 1A

In step 1A a special reference database is created in a module 5 of the fault locator 2 as a time scheme for a fault location in a known theoretical distance XI. Assuming the creation of theoretical traveling wave pulses PI', P2', P3' in a theoretical faulted point S3', theoretical arrival times Τ , 12', T3' of these pulses to the measuring point SI is calculated for a known fault distance XI, taking into account a polarity ( + or -) of wave pulses (fig.5). The known distance of fault XI is given by a user as equal intervals for the HVDC line having a total length TL. The transmission line 1 is divided for many equal intervals k= l...n-l, having an initial point SI and many variable theoretical fault points S3' in the distance XI, presented in meter: for example 100m, 200m, 300m, 5000m, ... 59900m, etc. A computer program, which is implemented in the processor of the fault locator 2 creates a table 1 for k(l), k(2) ... k(n-l) , where k is an ordinal number of known interval and indicates theoretical faulted point in line in the known distance XI, n is a natural number. For creation the table 1, having a known distance XI, using a Bewley lattice diagrams for fault in overhead line section and fault in a cable line section (fig.3 and 4) a theoretical time schemes ( left side of fig.3 and 4) are created . All theoretical k time schemes for each known distance XI are gathered in an exemplary table 1.

Table.l

K XI [m] PI' P2' P3' Tl' [\is] T2' [μ ₅ ] T3' [μ ₅ ]

1 100 - - - 0,33 0,66 1,33

2 200 - - - 0,66 1,33 2,66

3 300 - - - 1 2 4

297 29700 - - + 99 2 4

303 30300 - + - 102 4 8 ...

500 50000 - + - 233,33 133,33 266,66

n-1 59900 - + + 299,33 1,33 2,66

In order to better understanding the creation of the table 1 , the example of calculation of k time scheme for Table 1 is presented below on k=499 for the fault in overhead line section OHL and k=800 for the fault in cable line section CBL. The joint point J is the end of overhead line section OHL or the start point for the cable line section CBL. This example is based on measured current signal at one end of HVDC line. The parameters of the line 1 are known in terminal A or B.

For overhead line section OHL the following parameters are known:

Surge impedance ZL=400 Ω

Propagation velocity of traveling waves VL= 300000 [km/s]

Total length of overhead line section OHL = 30km

For cable line section CBL the following parameters are known:

Surge impedance for cable line: ZC=50 Ω

Propagation velocity of traveling waves: VC=150000 [km/s]

Total length of cable line CBL = 30km

The traveling wave pulses Ρ , P2', P3' in a theoretical faulted point S3' and theoretical arrival times Τ , Τ2', T3' of such pulses to the measuring point SI are calculated, using Bewley lattice diagrams, what is explained on fig.3 and 4.

According to drawing presented in fig.3 a backward traveling wave TWN of current signal has a negative polarity. In opposite to a forward traveling wave TWM has a positive polarity. After fault ignition in the line OHL in a point S3 that is close to junction J, the both waves TWN and TWM propagate along line with the speed VL in different direction. The wave TWN which reaches the terminal A at the measuring point SI, has the arrival time Tl' that is calculated according to equation T1'=X1/VL and the pulse PI' has the negative polarity. Then, the wave TWN is reflected back as TWN' and it is propagated to the fault point S3. Mutually with the propagation of TWN the forward traveling wave TWM reaches the junction point J and has the time TJ, that is calculated according to equation TJ=(OHL- X1)/VL It has a positive polarity. Then a first part of wave TWM is transmitted through the point J and becomes TWM", still has a positive polarity and is traveled to S2 after time TC equal TC=CBL/VC. A second part of a wave TWM is reflected from the point J and becomes TWM' and is traveling to fault place S3 after time equal to TJ and changes the polarity from positive to negative because of reflection coefficient is equal to (ZC-ZL)/(ZL+ZC) and is negative. When wave TWM' reaches S3 then part of wave TWM' is transmitted through S3' and becomes TW2 with a negative polarity and is traveled to SI after time T2' equals T2'=2*TJ. Second part of wave TWM' is reflected from point S3' and becomes wave TWM'" and reaches point J again after time TJ. Again wave TWM'" is reflected from point J and becomes TWM"" and is traveling to fault place S3 after time equal to TJ. When wave TWM"" reaches S3 then part of wave TWM"" is transmitted through fault point S3 and becomes wave TW3 which is traveled to SI at time T3' equals T3'=2*TJ. This situation is for example described in table 1 for k= 297 that is related to known fault place Xl=29700m (Table 1). The value of T2' is equal 2μ≤ because second wave P2' is traveling after first wave PI' with time delay equal 2* Ι=2*1μ5. The value of T3' is 4μ≤ because third wave P3' is traveling after first wave PI' with time delay equal 4*^=4*1μ≤.

According to drawing presented in fig. 4 a backward traveling wave TWN of current signal has a negative polarity. In opposite to a forward traveling wave TWM has a positive polarity. After fault ignition in S3' the both waves TWN and TWM propagate along line with the speed VC in different direction. The wave TWN reaches the J point after the time TJ1 that is calculated according to equation TJl=(Xl-OHL)/VC, and has the negative polarity. Then the first part of wave TWN is reflected back as wave TWN', which has negative polarity because of reflection coefficient is equal to (ZL-ZC)/(ZL+ZC) and it is propagated to the fault point S3 which reaches again after time TJ1. The second part of wave TWN is transmitted through point J, becomes wave TWl and is traveling in line OHL by time TL1=0HL/VL to terminal A from point J. TW1 reaches the point SI as the pulse PI' with the same polarity as wave TWN at arrival time Τ that is calculated according to equation T1'=TJ1+TL1. Mutually with the propagation of wave TWN the forward traveling wave TWM is traveled to the terminal B and reaches point S2 with the arrival time TCI equal to TC1=(0HL+CBL-X1)/VC, and has a positive polarity. Then wave TWM is reflected from point S2 and becomes wave TWM' has positive polarity and is traveled to S3, which reaches after time TCI. Next part of wave TWM' is transmitted through S3 and becomes TWM" and is traveling to point J after time equal to TJ1 from point S3. When wave TWM" reaches point J then part of wave TWM" is transmitted through point J and becomes wave TW2 with positive polarity P2' which is traveling by time TJ1 from point J to SI and reaches SI at time T2' equals T2'=2*TC1 after time Tl' .

When wave TW1 is traveling to point SI from point J, the second part of wave TWN' is reflected from point J and becomes wave TWN' and reaches point S3 again after time TJ1. Then part of wave TWN' is reflected from point S3 and becomes TWN" which has negative polarity and is traveling to point J by time TJ1. When wave TWN" reaches point J then part of wave TWN'" is transmitted through point J and becomes wave TW3, which has still negative polarity and is traveling from point J to terminal A by time TL1 and reaches the point SI at time T3' equals T3'=2*TJ1 after time ΤΓ. This situation is for example described in table 1 for k= 500 that is related to known fault place Xl=50000m (Table 1) ) The value of T2' is equal 133,33μ$ because second wave P2' reaches point SI after first wave with time delay equal 2*ΤΟ=2*66,66μ5. The value of T3' is 266,66μ$ because third wave P3' is traveling after first wave Ρ with time delay equal 2*Τ.1=2*133,33μ5.

Step IB

In step IB first a fault is detected in a known way using HVDC current signals of the transmission line 1 in the point SI. Next a polarity of traveling wave pulses PI, P2, P3 and the time of arrival Tl, T2, T3 of the such pulses to the point SI (fig.6) are determined. The pulses PI, P2, P3 are related to a fault in unknown points in the line 1, for example in the point S3. An initial arrival time Tl of the pulse PI is registered when the value of the measured magnitude of pulse PI is bigger than the threshold Th, which is depended of the total length of the line and is given by the user. The Tl is a starting point of next pulses 92, P3. Then next pulses P2, P3, are registered and the second arrival time T2, and the third arrival time T3 for these pulses are calculated and stored in a detection module 3. On the base of data presented above, a time scheme with 12, T3 for unknown distance X is created (fig.6). From time scheme is clear that PI and P3 have a negative polarity and P2 has a positive polarity.

Step 2

In this step, time scheme created in step IB for unknown faulted points in distance X is compared with the all theoretical time schemes, gathered in the table 1. The process of matching a real time scheme with the theoretical time scheme is started. In first step the polarity of second and third pulse P2 and P3 is checking, because the first pulse PI has always negative polarity. If the polarity of P2 and P3 of time scheme for unknown X is matched with the polarity of P2' and P3' for time scheme from database created in step 1A (fig.5 and fig.6), then arrival times Τ2', T3' for this case are taken from the table 1 for calculation together with the real arrival times T2, T3 of pulses P2 and P3 from the time scheme for unknown faulted place created in step IB. In the table one it is occurs for k equal 500, what means a fault located in CBL section.

Next time differences AT2 _k =T2-T2' and ΔΤ3ι _< =Τ3-Τ3' (fig.7) are calculated according to the formulas:

AT2 _k =T ₂ -T ₂ - (1) and

AT3 _k =T ₃ -T ₃ ' (2)

And next a sum∑(k) of absolute values of time differences ΔΤ2 and ΔΤ3 is calculated for second and third traveling wave pulses, according to the formula:

∑(k) = IAT2 _k l +ΙΔΤ3κΙ (3) After that the result of the sum∑(k) is stored in the module 7, and process of comparison starts again for next k from table 1 and is calculating until k is equal to n-1. For the all time schemes from database matched with time scheme for unknown XI the sum of ∑(k) is calculated and stored. Then the minimum value of the stored sum ofI(k) is selected in a known way. Then the only one related number k, having a minimum value of ∑(k) is indicated.

Step 3

In step 4 a known fault location distance XI is determined with one time schemes created in step 1A and selected in step 2 for indicated k. The k indicated in Step2 is connected with related value of XI that is stored in table 1.

Step 4

The unknown fault location distance X is determined from the best matching time scheme for the minimum value of the sum∑(k) and the unknown distance X is equal to value XI, which value is indicated an unknown fault location and triggering the alarm.