Method for determining the position of a fault precursor in an operational high-voltage cable

Description

FIELD OF THE INVENTION

[0001] The present invention relates to the general field of locating fault precursors in a high voltage cable in operation. STATE OF THE ART

[0002] High voltage cables, typically High Voltage A ("HTA") cables, are part of the electrical distribution system and transmit a three-phase High Voltage signal. The high voltage signal is periodic (typically with a period of 20 milliseconds for a frequency of 50 Hz) and of high amplitude (typically 20 kVolts).

[0003] In high-voltage cables, different types of failures in electrical conduction can appear over time.

[0004] A first type of possible failure is the appearance of a partial discharge site, which is a localized defect in the cable insulation. This site produces a very low amplitude electrical discharge that only impacts a small portion of the insulation. Partial discharge occurs when the high-voltage signal exceeds a so-called onset voltage value, and turns off when the voltage falls below this onset voltage, which is then called the extinction voltage. Partial discharges can occur several times per period.

[0005] A second possible type of failure is the occurrence of a fault precursor, also called a "pre-fault" or "self-extinguishing fault." This second type of failure is more serious than the occurrence of a partial discharge site: when the high-voltage signal reaches a certain threshold, generally at the maximum network voltage, the fault precursor can generate a total discharge. There is then a sudden drop in the voltage in the cable to zero and a current of significant amplitude flows in the fault. In the case of the self-extinguishing fault, the disturbance ends before the next half-cycle, and the high-voltage signal again follows the expected curve. This total discharge does not necessarily occur at each half-cycle of the high-voltage signal.

[0006] Finally, a third type of failure is the occurrence of a clear fault. This third type of failure is more serious than the previous two. When the high-voltage signal reaches a certain threshold, the clear fault produces a total discharge. There is then a sudden drop in the voltage in the cable to zero. The disturbance lasts for several successive half-periods without interruption. In this case, the disturbance leads to the opening of the protective circuit breaker on the high-voltage link to put an end to the disturbance.

[0007] If no intervention is carried out on the cable, a partial discharge site can, after a certain time, become a precursor to a fault and then finally become a clear fault.

[0008] Therefore, it is important to monitor high voltage cables in order to assess their health and intervene before a clear fault occurs.

[0009] There are methods for detecting and locating faults in the electrical conduction of the high-voltage cable. Some detection methods (called "on-line" methods) are implemented while the high-voltage cable is "in operation". A high-voltage cable is "in operation" when it is connected to the rest of the distribution network and is powered by the high-voltage signal from the distribution network.

[0010] Other detection methods (known as "off-line" methods) require the high-voltage cable to be "off," meaning it must be isolated from the rest of the distribution network. A team of technicians must then be deployed on site to disconnect the cable and conduct the tests using a dedicated power supply.

[0011] To monitor high-voltage cables and assess their health, online high-voltage cable fault detection and location processes have the advantage of not requiring a team of technicians, of not being time-limited to the duration of the team's intervention, and of continuing electrical distribution using all the high-voltage cables in the electrical distribution system.

[0012] Among the "online" methods for detecting and locating a partial discharge site in a high-voltage cable, methods using transponders are known. A transponder is a device capable of sending a signal after a predefined activation time.

[0013] For example, a method for detecting and locating a partial discharge site may follow the following chronology. When a partial discharge occurs in a high-voltage cable that connects two High Voltage A / Low Voltage ("HTA / LV") substations, two electrical pulses are generated from the discharge site and propagate in opposite directions each to a respective HTA / LV substation. At a first HTA / LV substation, a first pulse is detected. This activates a transponder that sends a warning signal to the second HTA / LV substation. The transponder activation time is of the order of 100 µs. At the second HTA / LV substation, a direct pulse generated by the partial discharge is detected successively, followed by the warning signal. A calculation is then carried out from the measurements of the two detection times of the direct pulse and the warning signal, and the transponder activation time, to determine the location of the partial discharge.

[0014] However, such a detection and localization method only concerns partial discharge sites, and is not suitable for the detection and localization of fault precursors.

[0015] In fact, the electrical disturbance due to a fault precursor is too significant to allow detection of the indicator signal.

[0016] Prior art includes US 5,682,100 A, US 5,416,418 A And JP S61-225666 A each disclosing a method for determining a position of a fault in a high-voltage cable connecting a first substation and a second substation.

[0017] Methods for detecting and locating fault precursors are also known. These methods generally require time synchronization of clocks in HTA / LV substations, and use GPS or GSM systems for this purpose. However, these methods have a number of disadvantages.

[0018] On the one hand, for detection methods using GPS systems, the difficulties of capturing a GPS signal in urban environments make their application more difficult in cities and limit their interest in using them on electricity distribution networks. To obtain precision in the location of the fault precursor, very high stability oscillators are required, which are therefore expensive to purchase.

[0019] On the other hand, detection methods using GSM systems are dependent on the proper functioning of the GSM link. The processing of information is linked to a subscription with the provider, resulting in additional subscription costs linked to the number of links to be monitored, and the fact that the provider has access to the network operator's data.

[0020] There is therefore a need for an "on-line" method for detecting and locating fault precursors in a high-voltage cable, which does not require the use of complex synchronization systems. STATEMENT OF THE INVENTION

[0021] A general aim of the invention is to propose a method for detecting and locating fault precursors in a high-voltage cable in operation, which does not require the use of complex synchronization systems.

[0022] The aim is achieved within the framework of the present invention by means of a method for determining a position of a fault precursor in a damaged high voltage cable connecting a first station and a second station, one of the first station and the second station being a source station suitable for supplying the high voltage cable with a high voltage signal and the other of the first station and the second station being a secondary station suitable for receiving the high voltage signal, comprising the following steps: a step S1 detection by the first station of a first signature of an electrical current anomaly in the high voltage signal, a step S2 detection by the second station of a second signature of the electrical current anomaly in the high voltage signal, a step S3 of emission by the second station of a witness signal to the first station in the high voltage cable, the witness signal being emitted by the second station with a predetermined delay from a time of detection of the second signature, the predetermined delay taking a value between 15% and 50% of the period of the high voltage signal, a step S4 of detection by the first station of the witness signal received by the first station, a step S5 of determination of the position of the fault precursor in the high voltage cable, as a function of a time of detection of the first signature by the first station and a time of detection of the witness signal by the first station.

[0023] Thanks to the transmission by the second station of the witness signal to the first station, the proposed method does not require the use of a synchronization system.

[0024] Furthermore, thanks to the particular value of the delay with which the indicator signal is emitted, it can be detected without being drowned in an electrical disturbance due to the fault precursor. Indeed, the delay is on the one hand sufficiently large so that the indicator signal is sent after the end of the electrical disturbance, and on the other hand sufficiently small so that the indicator signal is sent before the following half-period when a new electrical disturbance could occur.

[0025] Such a method is advantageously supplemented by the following different characteristics or steps taken alone or in combination: the high-voltage cable comprises three phases, the step of detection by the first station of a first signature comprising the measurement of three electrical signals supplying the three phases, the calculation of the sum of the measurements of the three electrical signals and the comparison of the sum with a threshold; the first station is connected to a plurality of second stations via a plurality of high-voltage cables, each cable being supplied by the same high-voltage signal, the step of detection of the first signature comprising the following sub-steps: a step S1A of detection by the first station of a plurality of first signatures of the electrical current anomaly in the high-voltage signals, a step S1B of identification, among the first signatures, of the first main signature representative of a current irregularity of maximum amplitude, of maximum initial variation and opposite to the other first current irregularity signatures,a step S1C of identifying, among the plurality of high-voltage cables, the damaged high-voltage cable supplied by the high-voltage signal comprising the identified first main signature, and in which the position of the fault precursor is determined in the damaged high-voltage cable by implementing steps S1 to S5 with the first main signature as the first signature; the first station is connected to a plurality of second stations via a plurality of high-voltage cables, each cable being supplied by the same high-voltage signal, and the second stations are configured to each emit a witness signal with a predetermined characteristic which differs from predetermined characteristics of the witness signals emitted by the other second stations, the method comprising, after step S4 of detection by the first station of the witness signal received by the first station,the additional steps: a step P1 of detecting the predetermined characteristic of the indicator signal, a step P2 of determining a possible position of the fault precursor in the high voltage cable, as a function of a time of detection of the first signature by the first station and a time of detection of the indicator signal by the first station., step S5 of determining the position of the fault precursor in the high voltage cable taking into account for each indicator signal received the possible position of the fault precursor in the high voltage cable; the predetermined characteristic is chosen from a particular delay value, a time distribution between several pulses forming the indicator signal, or a combination of these characteristics; The invention also relates to a first device for determining the position of a fault precursor in a damaged high voltage cable connecting two stations,comprising a first system for acquiring and processing a high-voltage signal passing through the high-voltage cable, the first acquisition and processing system being adapted to detect a signature of a current anomaly and a witness signal, and a first time counting system controlled by the first acquisition and processing system.

[0026] Advantageously, but optionally, the first acquisition and processing system of the first device can be adapted to detect a plurality of signatures of a current anomaly passing through a plurality of high-voltage cables all connected to one of the two stations, each cable being powered by a respective high-voltage signal, and adapted to detect among this plurality of signatures a current irregularity signature of maximum amplitude, of maximum initial variation and opposite to the other current irregularity signatures.

[0027] The invention finally relates to a second device for determining the position of a fault precursor in a damaged high-voltage cable connecting two stations, comprising an electromagnetic pulse detector, an apparatus for generating and injecting a warning signal into the high-voltage cable, the second determination device being adapted so that the detection of an electromagnetic pulse by the electromagnetic pulse detector triggers the injection of a warning signal by the apparatus after a precise delay.

[0028] Advantageously, but optionally, the apparatus of the second device may be configured to emit a control signal with a predetermined characteristic among a particular delay value, preferably between 15% and 50% of the period of the high-voltage signal, a time distribution between several pulses forming the control signal, or a combination of these characteristics. DESCRIPTION OF FIGURES

[0029] Other characteristics and advantages of the invention will emerge from the following description, which is purely illustrative and non-limiting, and must be read in conjunction with the appended drawings in which: [ Fig. 1 ] there Figure 1 schematically represents a general structure of an HTA departure from a source station supplying a high voltage cable. Fig. 2 ] there Figure 2 schematically represents a witness signal. Fig. 3 ] there Figure 3 schematically represents a method for detecting and locating a fault precursor in accordance with an embodiment of the invention. Fig. 4 ] there Figure 4 schematically represents a timeline of the steps and events occurring during a method of detecting and locating a fault precursor in accordance with an embodiment of the invention. Fig. 5 ] there Figure 5schematically represents a general structure of an HTA departure from a source station supplying a plurality of high voltage cables. DETAILED DESCRIPTION OF THE INVENTION General structure of an HTA departure from a source station supplying an underground connection of a distribution network

[0030] There Figure 1 represents a general structure 1 of several HTA departures Dep1, Dep2, DepN from a source station 2 supplying a plurality of underground connections of a distribution network.

[0031] Part 1 of the electrical distribution system comprises a source substation 2 in which a set J of three bars is supplied with a three-phase high voltage signal. High voltage cable

[0032] From the Dep1 departure, a high voltage cable 7 runs which connects a plurality of HTA / BT stations 3, 5, 9.

[0033] The high voltage cable 7 is formed of three high voltage phases 71, 72, 73 and is supplied by a three-phase high voltage signal.

[0034] High voltage cable 7 may be partly or entirely underground.

[0035] The single-pole high-voltage phases 71, 72, 73 of the distribution networks each have a coaxial structure consisting of three parts: a conductive core, a dielectric insulator which surrounds the conductive core and a metal screen which surrounds the insulator.

[0036] The dielectric insulator is either synthetic, such as polyethylene, or based on impregnated paper for older technologies.

[0037] The coaxial structure of the cables is similar to a transmission line and transient phenomena affecting the connection, such as failures in electrical conduction (partial discharges, fault precursors), generate electromagnetic waves which propagate in the cables.

[0038] The propagation speed of transient phenomena depends on the nature of the dielectric. For commonly used cables, the propagation speed of waves varies between 150 and 180 m / µs.

[0039] Upstream of the implementation of the detection method, the propagation speed Vp of the transient phenomena in the high voltage cable 7 is determined.

[0040] The high voltage cable 7 connects on the one hand the source station 2 to a first end of the high voltage cable 7 and on the other hand a final HTA / LV station 9 to a second end of the high voltage cable 7.

[0041] Before implementing the detection process, the length L of the high-voltage cable 7 between these two ends is identified. This length L is typically between 2 and 10 kilometers.

[0042] In the event of a failure in the electrical conduction deemed sufficiently serious, it is necessary to carry out an on-site intervention to replace the high-voltage cable 7 or a part of the high-voltage cable 7 (for example a junction). The replaced cable or the replaced part of the cable comprises the three high-voltage phases 71, 72, 73. It is therefore not necessary to determine on which phase 71, 72 and 73 the failure in the electrical conduction is located. Source post

[0043] A source substation or HTB / HTA substation allows the electrical voltage to be lowered for transmission to users. Thanks to an HTB / HTA transformer, the electrical voltage can be lowered in the source substation, for example, from 225,000 volts (High Voltage Type B) to 20,000 volts (High Voltage Type A).

[0044] A type A voltage output of an HTB / MV transformer supplies a set J of three bars. A plurality of HVA feeders Dep1, Dep2... DepN are connected to set J of bars. Each HVA feeder supplies a three-phase high-voltage cable. The HVA feeder Dep1 supplies the high-voltage cable 7. Typically, the source substation 2 can supply up to 30 three-phase high-voltage cables. The source substation typically comprises two HTB / MV transformers. Each transformer supplies a set of HVA feeders representing approximately half of the HVA feeders of the source substation.

[0045] The source station 2 comprises a first device D1 for determining the position of a fault precursor.

[0046] The first device D1 comprises an acquisition and processing system 23.

[0047] The acquisition and processing system 23 comprises one or more electromagnetic pulse detectors, such as a current torus 21, a high-frequency current sensor 22 or a high-frequency torus.

[0048] The acquisition and processing system 23 includes sufficient detectors to measure transient phenomena: in intensity on each of the high voltage cables which are supplied by source station 2, and in voltage for each HTB / HTA transformer of the source station connected to a set of HTA departures.

[0049] The detector(s) can be placed in the common earthing of the cable screens of the outgoing high voltage cable(s).

[0050] The acquisition and processing system 23 is adapted to recognize transient phenomena such as, for example, current anomalies due to a fault precursor or a series of electrical pulses.

[0051] The determination device D1 is suitable, when detecting a transient phenomenon, for measuring the value of the high voltage signal in the phase of the high voltage cable where the transient phenomenon is detected.

[0052] The first device D1 comprises a time counting system 24 controlled by the acquisition and processing system 23. The time of detection of the transient phenomena by the acquisition and processing system 23 can thus be determined.

[0053] The acquisition and processing system 23 also comprises a memory in order to record data, such as for example the detection times or other characteristics of the detected transient phenomena. HTA stations / Intermediate BTs

[0054] The high voltage cable 7 passes through a plurality of HTA / LV stations 3, 5 (High voltage type A / Low voltage for example 230 Volts) which are distributed in series on the high voltage cable 7.

[0055] Each HTA / LV substation includes an HTA / LV transformer 35, 55 connected to the underground link 7 via a switch. The transformer 35, 55 includes a primary 36, 56 powered by the three-phase of the high-voltage cable, and a secondary 37, 57 which distributes the energy to the subscribers. HTA station / BT at the end of the high voltage cable

[0056] The final HTA / BT substation 9 includes an HTA / BT transformer 95 which itself includes a primary 96 and a secondary 97 similar to that described for the other HTA / BT substations 3 and 5.

[0057] The final HTA / BT station 9 includes a second device D2 for determining the position of a fault precursor.

[0058] The second device D2 comprises at least one electromagnetic pulse detector 92, such as a high-frequency current sensor.

[0059] The second device D2 includes enough detectors to measure transient phenomena arriving through the high voltage cable, in voltage or in intensity.

[0060] The detector(s) may be placed in the common ground of the high voltage cable screens 7.

[0061] The second device D2 also comprises an apparatus 98 for generating and injecting a control signal into the high voltage cable 7.

[0062] The signal can be injected by means of the detector 92 or a torus 92b. The second device D2 is adapted so that the detection of an electromagnetic pulse by the electromagnetic pulse detector 92 triggers the injection of a control signal by the device 98 after a precise delay. The precision on the delay is typically of the order of ten nanoseconds.

[0063] Device 98 may be a repeater system or a transponder system.

[0064] The indicator signal generated by the device 98 may be a series of pulses.

[0065] It is possible to configure the 98 device before its installation in the HTA / LV substation so that the indicator signal includes two or three pulses or even more.

[0066] As shown in the Figure 2 , an ST witness signal can be composed of three square pulses I 1 , I 2 and I 3 .

[0067] The square pulses I 1 , I 2 and I 3 can all have the same amplitude AI .

[0068] The amplitude of the pulses injected by the device 98 into the high voltage cable 7 is typically worth a few amperes

[0069] The apparatus 98 is adapted so that the rising edge of a generated pulse is spread temporally over a duration as short as a few nanoseconds. The durations between the rising edge of one square pulse and the next square pulse are not necessarily equal.

[0070] More generally, it is possible to configure the device 98 to vary the durations t 1 (respectively t 2 ) elapsed between the rising edge of the first square pulse I 1 and the rising edge of the following square pulses I 2 (respectively I 3 ). These durations are chosen to be at least equal to several tens of microseconds, for example 50 microseconds. Method for determining and locating the fault precursor

[0071] The high voltage cable 7 supplies three-phase power to all the various HTA / BT stations 3, 5, 9 in order to distribute the electrical energy to the subscribers.

[0072] In the example illustrated on the Figure 1 , a fault precursor 10 is present on one of the high-voltage phases of the high-voltage cable 7.

[0073] At a time ti where the voltage in the high voltage phase 71 exceeds in absolute value a certain threshold, a total discharge occurs at the level of the fault precursor.

[0074] This creates an electrical disturbance or electrical current anomaly in the high voltage signal passing through the high voltage cable 7.

[0075] The electrical disturbance generates two electrical pulses which propagate in opposite directions in the high voltage cable 7: a first electrical pulse Sa which propagates from the fault precursor 10 towards the source substation 2 at the first end of the high voltage cable 7, and a second electrical pulse Sb which propagates from the fault precursor 10 towards the final HTA / LV substation 9 at the second end of the high voltage cable 7.

[0076] The acquisition and processing system 23 is configured to implement a method for determining a position of the fault precursor in the damaged high-voltage cable by following the steps illustrated in the Figure 3 .

[0077] There Figure 4is a timeline of events that occur during the process.

[0078] During a first step S1, the acquisition and processing system 23 of the first device D1 detects the first pulse Sa. The acquisition and processing system 23 in fact comprises sufficient detectors to measure the transient phenomena.

[0079] Simultaneously with the propagation of the first electrical pulse Sa, the second pulse Sb propagates towards the final HTA / BT substation 9, which triggers step S2 presented below. It should be noted that depending on the position of the fault precursor, the first pulse Sa may arrive at the source substation 2 before or after the second pulse arrives at the final HTA / BT substation 9.

[0080] The order of these steps S1 and S2 is therefore not fixed from one configuration to another in the implementation of the method.

[0081] Thanks to the device D1 placed at the source station 2, the detection of the first electrical pulse Sa makes it possible to find a first signature of the electrical current anomaly in the high voltage signal, namely the total discharge due to the fault precursor.

[0082] The time counting system 24 makes it possible to associate a detection instant ta with the first electrical pulse Sa. The measurement of the first electrical pulse Sa and its detection instant ta can be recorded by the acquisition and processing system 23.

[0083] The first electrical pulse Sa measured in the first device D1, by means of a High Frequency toroid, typically has a time span less than a quarter of the period of the high voltage signal.

[0084] In the case of a pre-fault signal, the time extent of the first electrical pulse Sa is typically 3 milliseconds.

[0085] In the case where source station 2 supplies several high-voltage cables, a first signature of the current anomaly in the associated high-voltage signal is obtained for each high-voltage cable. The electrical conduction failure in one high-voltage cable in fact influences the other high-voltage cables.

[0086] It is then necessary to compare the different first signatures of the anomaly to determine in which high voltage cable the fault precursor is located.

[0087] During a step S1A, the acquisition and processing system 23 detects the different first signatures measured for each of the high-voltage cables. These different first signatures can be recorded by the acquisition and processing system 23.

[0088] During a step S1B, a search among these first signatures is carried out by the acquisition and processing system 23. During this step S1B, the first signature representing the current irregularity of greatest amplitude, of greatest initial variation, and of which the variation is opposite, that is to say of opposite sign, to the current variations represented in the other first signatures is identified.

[0089] During a step S1C, the acquisition and processing system 23 determines in which high-voltage cable the fault precursor is located; this is the high-voltage cable for which the first signature identified during step S1B was obtained.

[0090] The first signature obtained in fact for the high voltage cable containing the fault precursor in fact presents: the highest amplitude of variation in intensity - more precisely this highest amplitude is close to the sum of the amplitudes of the first signatures of the other cables connected to source station 2 -; a current variation of polarity opposite to that of the other first signatures; an initial slope in the current variation greater than that of the other first signatures.

[0091] During step S1C, the acquisition and processing system 23 can also record in the memory the instant ta when the first electrical pulse Sa was detected at the source station 2 and the measurement of this pulse Sa.

[0092] It should be noted that the detection of the first electrical pulse Sa can be completed by a step of confirmation of the nature of the failure in the conduction.

[0093] During step S1, the acquisition and processing system 23 can measure the three electrical signals supplying the three high-voltage phases 71, 72, 73.

[0094] The determination device D1 is in fact suitable, when detecting a transient phenomenon, for measuring the value of the high voltage signal in the phase of the high voltage cable where the transient phenomenon is detected.

[0095] The determination device calculates the sum of the three voltages measured at source station 2 in the phases. If the sum of the three voltages is greater than an identified threshold, then it is a total discharge type failure.

[0096] Comparing the sum to the threshold makes it possible to ensure that the detected disturbance is an abnormal event that we wish to study, such as a total discharge due to a fault precursor, and not a trivial event such as a change in load on the network.

[0097] During a step S2, the electromagnetic pulse detector 92 of the second device D2 detects the second pulse Sb.

[0098] Thanks to the D2 device placed at the final HTA / BT station 9, the detection of the second electrical pulse Sb makes it possible to find a second signature of the electrical current anomaly in the high voltage signal, namely the total discharge due to the fault precursor.

[0099] The detection of the second electrical pulse Sb is carried out at a detection time tb.

[0100] During a step S3, the device 98 emits a warning signal St in the high-voltage cable 7.

[0101] The indicator signal St is emitted after a certain delay or time delay Δt elapsed from the instant of detection tb of the second pulse Sb by the transponder system 98.

[0102] The delay Δt is chosen to be equal to a value between 15% and 50% of the period of the high-voltage signal. As mentioned above, this corresponds in the case of a high-voltage signal with a frequency of 50 Hz to a delay Δt between 3 milliseconds and 10 milliseconds.

[0103] The delay Δt is predetermined upstream of the implementation of the detection method and parameterized in the transponder. The value of the delay Δt is recorded in the memory of the acquisition and processing system 23.

[0104] The control signal St consists of a series of square pulses, for example three pulses, as described previously in relation to the Figure 2 .

[0105] The control signal St propagates and reaches the source substation 2. Given the significant length L of the high-voltage cable 7, the control signal St is strongly attenuated during propagation. For an amplitude of several tens of volts of the control signal sent from the final HTA / LV substation 9, a control signal with an amplitude of a few tens or hundreds of millivolts is received at the source substation 2 depending on the distance separating the substation 9 and the source substation. During a step S4, the control signal St is detected within the first device D1 of the source substation 2.

[0106] The measurement of the witness signal St and its detection time tf can be recorded in the memory by the acquisition and processing system 23.

[0107] During a step S5, the acquisition and processing system 23 determines the position of the fault precursor 10 in the high-voltage cable 7.

[0108] To establish this position, the acquisition and processing system 23 takes into account the following information: the detection time ta of the first pulse Sa, the detection time tf of the control signal, the delay Δt recorded in the memory of the acquisition and processing system 23, the propagation speed Vp of the transient phenomena in the high voltage cable 7 predetermined and recorded in the memory of the acquisition and processing system 23, and the length L of the high voltage cable 7 predetermined and recorded in the memory of the acquisition and processing system 23.

[0109] From this different information, the distance D separating the source station 2 and the fault precursor 10 can be calculated according to the following relation: D = 0.5 × L - V × tf - L / V − Δ

[0110] Or again D = L - 0.5 × V × tf - ta - Δt . Technical effects and advantages

[0111] The method for determining the position of the fault precursor 10 in the high voltage cable 7 as just presented involves a controlled delay Δt.

[0112] This controlled delay in sending the indicator signal St by the second device is equal to or close to the duration of the first electrical pulse Sa and the second electrical pulse Sb. This duration corresponds to the typical time of electrical disturbance of the high voltage cable 7 due to the total discharge produced at the fault precursor 10.

[0113] This disturbance can be significant, with variations in measured intensity reaching one kiloampere, particularly at the very start of an electrical disturbance.

[0114] Measuring a witness signal St emitted by the second device D2 which would arrive at the first device D1 before a few milliseconds, typically 2 or 3 milliseconds would be difficult because the witness signal would be drowned out by the disturbance.

[0115] By imposing a delay of at least 3 milliseconds, the witness signal St leaves the second device D2 and arrives at the first device D1 after most of the high-frequency component of the electrical disturbance has been damped. This allows a much better signal-to-noise ratio in the detection of the witness signal by the first device D1.

[0116] It should be noted that imposing a delay greater than 50% of the period of the high voltage signal, for example 10 milliseconds for a frequency of 50 Hz, implies that the indicator signal is sent at the following half-period. However, at the following half-period the voltage of the high voltage signal is again sufficiently different from zero to trigger a total discharge of the fault precursor 10, or to trigger another failure in the electrical conduction of the high voltage cable. The risk that the indicator signal is drowned in another electrical disturbance is significant, so that a delay greater than 50% of the period of the high voltage signal is not of major interest.

[0117] It thus becomes possible to propose a method for detecting and locating fault precursors in a high-voltage cable in operation.

[0118] Furthermore, the proposed method does not use time synchronization of clocks located at the ends of the high-voltage cable 7. Adaptation of the process when the source station supplies several high-voltage cables or supplies high-voltage cables in bypass

[0119] In the case where the source station 2 supplies a plurality of cables or high-voltage branch cables, it is possible to adapt the method for determining the position of the fault precursor.

[0120] There Figure 5 illustrates this situation in the case of two high-voltage cables in derivation. The elements of the Figure 5 which have the same numbering as in the Figure 1 are identical to the elements of the Figure 1 as presented above in the description.

[0121] The portion of the electrical network illustrated in the Figure 5has a high voltage cable 6 at the source station 2. Within a branch ZD zone, the high voltage cable is separated into two high voltage cables 8 and 12 which are therefore branched.

[0122] Each high voltage cable 6, 8 and 12 is powered by a three-phase high voltage signal.

[0123] High voltage cable 8 supplies an HTA / BT station 4 and a second final HTA / BT station 11 at the end of high voltage cable 8.

[0124] These two HTA / BT stations 4 and 11 comprise a transformer 45, 115 itself comprising a primary 46, 116 and a secondary 47, 117 similar to those described for the other HTA / BT stations 3, 5 and 9. The high voltage cable 8 can supply a plurality of HTA / BT stations distributed in series between the branch zone ZD and the second final HTA / BT station 11.

[0125] The second final HTA / BT station 11 at the end of the high voltage cable 8 comprises a third device D3 for determining the position of a fault precursor, similar to the second device D2 presented previously.

[0126] The third D3 device includes in particular: at least one electromagnetic pulse detector 112, a device 118 for generating and injecting a control signal into the high voltage cable 8.

[0127] The first apparatus 98 and the second apparatus 118 are configured so that the indicator signals they emit have one or more different characteristics.

[0128] This feature can be: a particular delay value: the first delay Δt of the first device 98 is chosen to be different from the second delay Δt2 of the second device 118, a number of pulses forming the indicator signal: the first device 98 can produce an indicator signal comprising three pulses and the second device can produce an indicator signal comprising four pulses, a time distribution between the pulses forming the indicator signal: the durations elapsed between the rising edge of the first square pulse and the rising edge of the following square pulses forming the indicator signal can differ between the indicator signal emitted by the first device and the second device, and a possible combination between these characteristics.

[0129] These characteristics are chosen to be recognized by the acquisition and processing system 23 when detecting the witness signal, even after the witness signal has traveled the entire length of a high voltage cable.

[0130] It is possible to record upstream of the implementation of the detection method in the memory of the acquisition and processing system 23 of the source station 2 which witness signal characteristics are chosen for which device.

[0131] Generally speaking, in the case where the source substation 2 supplies a plurality of cables or high voltage branch cables, there are a plurality of final HTA / LV substations at the end of the lines. In each of these final substations, it is possible to place a device for determining the position of a fault precursor, the apparatus of which is configured to emit a warning signal whose characteristics are different from all the other warning signals emitted by the other apparatuses. As in the case specified above, it is possible to record upstream of the implementation of the detection method in the memory of the acquisition and processing system 23 of the source substation 2 which warning signal characteristics are chosen for which apparatus.

[0132] If a fault precursor 10 is present on one of the high-voltage phases of the high-voltage cable 8, it is possible at a certain time during operation of the electrical distribution network that a total discharge occurs at the level of the fault precursor 10.

[0133] There Figure 5 illustrates an example where the fault precursor 10 is located on the high voltage cable 8 between the bypass zone ZD and the second final HTA / BT station 11.

[0134] A first electrical pulse Sa propagates towards the source station 2. When this first pulse Sa reaches the bypass zone ZD, a first part of the electrical pulse Sa propagates towards the source station 2. The detection of this first part of the electrical pulse at the source station 2 follows the steps described previously.

[0135] A second part of the first electrical pulse Sa propagates towards the first final HTA / LV station 9. The device D2 detects the second part of the electrical pulse Sa and a first indicator signal is emitted towards the source station 2 by the first device 98.

[0136] A second electrical pulse Sb propagates towards and reaches the second final HTA / LV station 11. The device D3 detects the second pulse and a second indicator signal is emitted towards the source station 2 by the second device 118.

[0137] The first witness signal and the second witness signal reach source station 2, where they are detected.

[0138] During a step P1, the predetermined characteristic, or predetermined characteristics, of each witness signal received is detected.

[0139] The witness signal has particular characteristics, in terms of shape and delay Δt, which are in fact recorded in the memory of the acquisition and processing system 23. These characteristics can therefore be recognized by the acquisition and processing system 23, so that it is possible to guarantee the origin of a witness signal.

[0140] During a step P2, a possible position of the fault precursor in the high-voltage cable is determined for each witness signal received. As previously, the determination uses the time of detection of the first signature by the first station and the time of detection of the witness signal by the first station. The information necessary for determining the possible position and which depends on the high-voltage cable on which the determination device that emitted the witness signal received is located is selected from the memory of the acquisition and processing system 23. the delay associated with the received witness signal, the length of the high voltage cable or high voltage cables traveled by the received witness signal.

[0141] The distance separating the source station 2 and the fault precursor is then determined using the method and mathematical formula described above.

[0142] In the example of the Figure 5 , the results of the calculations give: for the first indicator signal emitted in the first final HTA / BT substation 9, the fault precursor is located at a first possible position which is the position of the bypass zone ZD. for the second indicator signal emitted in the second final HTA / BT substation 11, the fault precursor is located at a second possible position which is the position of the fault precursor 10.

[0143] In step (S5) the position of the fault precursor in the high voltage cable is determined taking into account for each witness signal received the possible position of the fault precursor in the high voltage cable. For this, the following principles are used: If the fault precursor is located between the source substation 2 and the branch zone, the two possible positions are identical and are the position of the fault precursor.

[0144] If the fault precursor is located between the deviation zone and a final HTA / LV substation, the two possible positions are not identical.

[0145] The possible position given by the final “direct” HTA / LV station on the branch of which the fault precursor is located is the position of the fault precursor.

[0146] The possible position given by the final “indirect” HTA / BT station on the branch of which the fault precursor is not located is the position of the bypass zone.

[0147] In the example of the Figure 5 , we retain the second possible position as the position of the fault precursor.

[0148] Depending on the situation, it is always possible to determine the position of the fault precursor 10.

Claims

1. A method of determining a position of a fault precursor (10) in a deteriorated high-voltage cable connecting a first station and a second station, one of the first station and the second station being a source station (2) adapted to supply the high-voltage cable with a high-voltage signal and the other of the first station and the second station being a secondary station (9) adapted to receive the high-voltage signal, comprising the following steps (S1) detection by the first station of a first signature of an electric current anomaly in the high-voltage signal, (S2) detection by the second station of a second signature of the electric current anomaly in the high-voltage signal, (S3) emission by the second station of a witness signal to the first station in the high-voltage cable, the witness signal being emitted by the second station with a predetermined delay from an instant of detection of the second signature, (S4) detection by the first station of the control signal received by the first station, (S5) determination of the position of the fault precursor in the high-voltage cable, as a function of a time of detection of the first signature by the first station and a time of detection of the warning signal by the first station, the method being characterized in that the predetermined delay takes on a value of between 15% and 50% of the period of the high-voltage signal.

2. A method according to claim 1, in which the high-voltage cable comprises three phases, the step (S1) of detection by the first station of a first signature comprising measuring three electrical signals feeding the three phases, calculating the sum of the measurements of the three electrical signals and comparing the sum with a threshold.

3. Method according to one of claims 1 or 2, in which the first station is connected to a plurality of second stations via a plurality of high-voltage cables, each cable being fed by the same high-voltage signal, the step (S1) of detecting the first signature comprising the following sub-steps : (S1A) detection by the first station of a plurality of first signatures of the electric current anomaly in the high-voltage signals, (S1B) identification, among the first signatures, of the first main signature representative of a current irregularity of maximum amplitude, maximum initial variation and opposite to the other first current irregularity signatures, (S1C) identification, among the plurality of high-voltage cables, of the deteriorated high-voltage cable fed by the high-voltage signal comprising the identified first principal signature, and in which the position of the fault precursor is determined in the damaged high-voltage cable by carrying out steps (S1) to (S5) with the first main signature as the first signature.

4. A method according to any one of claims 1 to 3, in which the first station is connected to a plurality of second stations via a plurality of high-voltage cables, each cable being supplied with a same high-voltage signal, and the second stations are each configured to emit a test signal with a predetermined characteristic which differs from predetermined characteristics of the test signals emitted by the other second stations, the method comprising, after the step (S4) of detection by the first station of the test signal received by the first station, the additional steps of (P1) detecting the predetermined characteristic of the control signal, (P2) determining a possible position of the fault precursor in the high-voltage cable, as a function of a time of detection of the first signature by the first station and a time of detection of the witness signal by the first station, the step (S5) of determining the position of the fault precursor in the high-voltage cable, taking into account the possible position of the fault precursor in the high-voltage cable for each control signal received.

5. Method according to claim 4, in which the predetermined characteristic is chosen from a particular delay value, a time distribution between several pulses forming the control signal, or a combination of these characteristics.

6. A second device (D2) for determining the position of a fault precursor (10) in a damaged high-voltage cable (7) connecting two substations, the second device being configured to implement the method according to one of claims 1 to 5 and comprising a detector (92) of a second signature of the electric current anomaly in the high-voltage signal, an apparatus (98) for generating and injecting a witness signal into the high-voltage cable, the second determination device (D2) being configured so that detection of the second signature by the electromagnetic pulse detector (92) triggers injection of a test signal by the apparatus (98) with a predetermined delay from an instant of detection of the second signature, the second determination device being characterized in that the predetermined delay takes a value of between 15% and 50% of the period of the high-voltage signal.

7. Second device according to claim 6 in which the device (98) is configured to emit a control signal with a predetermined characteristic from among a particular delay value, a time distribution between several pulses forming the control signal, or a combination of these characteristics.

8. System for determining a position of a fault precursor (10) in a deteriorated high-voltage cable (7) connecting a first station and a second station, comprising : a second determination device (D2) according to one of claims 6 or 7, and a first device (D1) for determining the position of a fault precursor (10) in the high-voltage cable (7), the first device being configured to implement the method according to one of claims 1 to 5 and comprising a first acquisition and processing system (23) for the high-voltage signal passing through the high-voltage cable (7), the first acquisition and processing system (23) being configured to detect the first signature of a current anomaly and a test signal n, and a first time-counting system (24) controlled by the first acquisition and processing system (23) configured to determine the position of a fault precursor in the high-voltage cable as a function of an instant of detection of the first signature and the witness signal.

9. A system according to claim 8 in which the first acquisition and processing system (23) is configured to detect a plurality of signatures of a current anomaly passing through a plurality of high-voltage cables all connected to one of the two substations, each cable being supplied with a respective high-voltage signal, and configured to detect among this plurality of signatures a current irregularity signature of maximum amplitude, maximum initial variation and opposite to the other current irregularity signatures.