Pre location: Impulse-Current-Method (ICE)

Transcription

1 1 ICE (Impulse current method three phased Ionisation delay time DIRECT MODE Output impedance of the generator 2 Surge generator as impulse source 3 High voltage test set as impulse source The current coupler Typical waveforms Evaluation of the measurement Improving the measurement accuracy Influence of a connection cable 6 2 Comparison mode I Straigt networks T eed Networks 8 The fault lies on the main conductor between generator and loop 9 Very long T eed networks Influence of a connection cable Influence of a connection cable (test lead) 13 3 Comparison mode II UnT eed Networks T eed Networks 15 The fault is located on a main conductor between generator and loop 15 The fault is located in a tee Influence of a connection cable 17 4 Differential comparison mode The differential coupler Straight networks T eed Networks 20 The fault is located on the main conductor between generator and loop 20 The fault is located in a tee Practical Application Influence of a connection cable 22 5 Loop On - Loop Off Method 23 6 Measuring system Centrix Phase selection Phase selection menu 24 Pg1

2 1. ICE (Impulse current method) three phased Four impulse current circuit methods have proven to be successful in prelocating high resistance and intermittent faults in power cables: Direct mode Comparison mode Differential-comparison mode Loop-on loop-off method 1.1. Ionisation delay time The ionisation delay time Dt is the period of time prior to the flashover at the fault during which the charge carriers are ionized. This can take up to some milliseconds. Many times a flashover only takes place after the voltage wave has passed the point of fault and has been reflected at the open cable end with the same polarity, which leads to a voltage doubling. When the voltage wave passes the fault again, usually a flashover occurs. In order to reduce the ionisation delay time Dt, it is recommended to use the maximum output energy of the generator without however exceeding the permissiblevoltage of the cable. Since the ionisation delay time can be some milliseconds, the test range to be set on the transient recorder always has to exceed the length of the faulty cable many times over DIRECT MODE A surge generator or a DC test set serve as an impulse source. Any tee between the start end of the cable and the fault will inevitably cause multiple reflections, which disturb the reflection process to be evaluated. Hence, the direct mode is only applied in cables without tees. Fig 1: Principle of operation Direct Mode The reflections in the faulty cable are recorded via a current coupler, which is connected to the faulty conductor and to a pulse reflection instrument with built-in transient recorder Output impedance of the surge generator In the direct mode, the current coupler is always positioned in direct vicinity of the generator (e.g. surge or shock wave generator - SWG). Each time a transient current wave travelling towards the start end of the cable passes the current coupler, it is almost simultaneously reflected with the Pg2

3 same polarity at the low resistance output impedance of the generator so that in this moment the outgoing and reflected waves superimpose (add up). In order to achieve a current doubling, the following must apply: Zgenerator << Zcable Zgenerator (output impedance of the generator) Zcable (characteristic impedance of the cable) Surge generator as impulse source As far as the transient phenomenon is concerned, the output impedance of the surge generator is zero. This enables a current doubling at the current coupler. High voltage test set as impulse source A current doubling is only possible by connecting a capacitor of some microfarad to the high voltage output. A surge generator, too, can serve this purpose, provided the output voltage is set to DC mode, while increasing the voltage. In this operation mode, the surge generator functions similar to a high voltage test set. The DC voltage potential at the fault increases until the breakdown voltage is reached. After a flashover appeared at the fault position, the whole energy stored in the cable and SWG discharges abruptly, thus generating a transient wave. Fig 2: High voltage test set with capacitor This test mode does not involve an ionisation delay time. In practice, the DC voltage discharge method delivers very clear, reproducible reflection diagrams. However, it will fail in the case of faults, which do not allow a charging of the cable due to a shunt resistance The current coupler The current coupler consists of an inductive pick up fixed to the high voltage output of the surge generator. The high-pass effect of the current coupler produces clear waveforms with steep rising edges. Since the current on the cable shield and in the conductor is the same and only the direction changes, it does not matter which conductor the current coupler is positioned over. Pg3

4 1.5. Typical waveforms Fig 4: Reflections at an open cable end Waveform No. 2: Fig 5: Reflection at a shorted cable end Waveform No. 3: Reflection in the event of a flashover at a discharge fault or an insulation resistance fault. Here, an ionisation delay time Δt may occur. Pg4

5 Fig 6: Reflection at a flashover fault Waveform No. 4: Similar to waveform no. 3 but with a longer ionisation delay time Dt. The fault only ignites after a reflection at the open cable end, whereby the transient voltage wave is reflected with twice the amplitude. Fig 7: Flashover fault: long ionisation delay time If the point A is difficult to determine, then the surge voltage has to be increased without exceeding the permissible voltage of the cable. The ionisation delay time Dt can thus be reduced. The resulting waveform corresponds to waveform no Evaluation of the measurement The evaluation of the waveforms is done between the negative reflections at the fault in consideration of the parasitic reflections. These parasitic reflections are caused by the inductance of the test leads of the generator. Pg5

6 X: Distance between fault and current coupler Equation 1: Calculation of the fault distance 1.7. Improving the measurement accuracy It is possible to eliminate the uncertainty of the factor v/2 and to obtain a more accurate test result by applying a surge wave to a healthy conductor (typical waveform no. 1). The total length L of the cable must be known. The parameter v/2 is changed until the cable length L is shown on the pulse reflection instrument. This v/2 value is then retained for all further measurements on this cable. Note: When using a high voltage connection cable: L = faulty cable + connection cable Influence of a connection cable When using a high voltage connection cable between the generator and the faulty cable, then the current coupler has to be placed as close to the generator as possible (point of reflection) in order to obtain a current doubling. Fig 8: Position of the current coupler with test lead In these measurements, do not forget to subtract the length of the test leads (transmission time t' or fictitious length l'). The length of the coaxial cable that connects the current coupler to the pulse reflection instrument has no effect on the measurement. 2. Comparison mode I A high voltage test set or a surge generator (with the surge button pressed) is used as an impulse source. This method is mainly used in T eed networks. Requirement: A healthy conductor must be available and it must be possible to charge the cable (no parallel resistance). Pg6

7 Fig 9: Schematic circuit diagram Comparison Mode I In this method, two traces recorded during a flashover at the fault are compared Straight (unbranched) networks First measurement: The generator charges the cable capacitance and an arc (flashover) develops at the fault position. This flashover generates a travelling wave, which propagates in both directions from the fault position. Fig 10: Measurement without loop at the cable end The reflection towards the cable end will travel at a velocity V and will be reflected between this point and the fault. This reflection will not influence the trace. The reflection towards the start end of the cable will also travel at a velocity V. At the contact points, a small part of the energy is reflected to the fault. The remaining part travels to the generator. When passing the current coupler, the transient current wave is recorded. The trace might look as follows: Fig 11: Measurement without loop at the cable end Pg7

8 Second Measurement A loop is connected at the cable end between the faulty conductor and a healthy conductor. Fig 12: Measurement with loop at the cable end An increase in voltage will lead to another flashover at the fault. The reflection towards the start end of the cable is also recorded. The start of the trace corresponds to that of the first measurement. In this case, however, the reflection towards the far end of the cable that did not have any influence in the first measurement will pass the loop and reach the start end of the cable via the healthy conductor. Only at the point where this reflection passes the current coupler, the trace will differ from the first measurement. The trace obtained could be as follows: Fig 13: Measurement with loop at the cable end 2.2. Evaluation of the measurement The distance covered by the transient current wave, which passes the loop at the cable end during the second measurement is calculated as follows: 2X + 2Y The distance covered by the transient wave, which is reflected at the fault in the first measurement is: 2X. The point of deviation of the two superimposed traces can be easily and accurately determined. The point corresponds to: 2X + 2Y - 2X = 2Y. Pg8

9 Fig 14: Comparison trace Hence, the distance of the fault from the far end of the cable is: Equation 2: Calculation of the fault distance 2.3. T eed Networks The fault is located on the main conductor between generator and loop. Fig 15: Fault on the main conductor First measurement without loop at "B": Despite the fact that the reflection in the direction of the start end of the cable has passed through the different tees, it reaches the coupler and is recorded in the transient recorder. Second measurement with loop at "B": The reflection in the direction of the far cable end, which did not have any influence on the first measurement passes the loop and returns to the start end of the cable via the healthy conductor. Pg9

10 At the point where this reflection passes the current coupler, the trace will deviate from the previous one. The tees do not have any effect on the principle. However, in each tee, only part of the signal will propagate in forward direction. Hence, the signals at the current coupler will be weakened and the deviations of the traces will be less distinct. As for a cable without tees, the measurement gives the distance of the loop to the fault. The fault is located in a tee Fig 16: Fault in in a tee The comparison of the traces recorded with and without loop at "B" is similar. The measurement gives the distance between the loop and the point of origin of the tee in which the fault lies. Now there is one uncertainty factor, since the fault can be located at any point on the tee "D-C". In order to eliminate this uncertainty, the loop has to be positioned at the end of the tee "C". Now the section "A-C" is the main current circuit and "D-B" a tee. Example: Given the following 3-phase 20 kv network with a discharge fault at the phase R at a flashover voltage of 14 kv. Fig 17: 20 kv Network The generator is connected at the start end "A" of the cable between the phases R (faulty conductor) and S (healthy conductor). The current coupler of the phase S (I2) is used. The Pg10

11 propagation velocity v/2 of 80 m/ms is an approximate value. First measurements: Trace "a" without loop. Trace "b" with loop at "E" between the phases R and S. Fig 18: Comparison with and without loop at far cable end The measurement shows the fault at a distance of 15 m from the tee "C" in the direction of "B". Considering the measuring error, especially in the propagation velocity v/2, the fault could also be located directly at "C" and possibly also in the tee "C-H". Second measurements: Trace "a" without loop. Trace "b" with loop at "H" between the phases R and S. Pg11

12 Fig 19: Comparison with and without loop at the end of the tee The measurement shows the fault at a distance of 172 m from "C" in the direction of "G". The uncertainty factor is now eliminated, since it is known that there is no other tee between these points. The main sources of error in these measurements are to be traced back to - an estimated propagation velocity v/2 - inaccurate network plans. Very long T eed networks In order to minimize the time spent for the location of faults in long cable networks, it is possible to pre-locate the fault roughly through a survey measurement so that the loop can be positioned as close to the start end of the cable as possible. Conducting a survey measurement First measurement: Surge wave on a healthy conductor. Fig 20: Measurement at a good conductor Second measurement: Surge wave on a faulty conductor Pg12

13 Fig 21: Measurement on a faulty conductor Evaluation of the measurement The superposition of the two traces shows a deviation. Fig 22: Comparison reflectogram Example: If tx + Δt = 32,5 μs, then it is known that the fault is located at a distance of less than 32.5 μs * 80 m/μs = 2600 m from the start end of the cable. For accurate measurements the ionisation delay time Δt has to be reduced and the propagation velocity on a cable with tees has to be reduced as follows: v/2» 80 - N (N: number of tees between the start end of the cable and the fault). In the subsequent accurate pre-location, the loop is positioned at the end of a tee whose origin is at least 2600 m from the start end of the cable. In this example, the loop is positioned at "B" or "C" Pg13

14 Fig 23: Positioning the loop after the survey measurement 2.4. Influence of a connection cable (test lead) If a high voltage test leads is used between the generator and the faulty cable, then the current coupler can be positioned either at the output of the generator, or at the connection point of the faulty cable. Fig 24: Positioning the current coupler In both cases, the propagation velocities in the high voltage test leads or in the coaxial cable do not influence the measurement. 3. Comparison mode II The impulse source is a high voltage test set or a surge generator (with the impulse button pressed). This method is mainly used in T eed networks. Prerequisite: At least one healthy conductor must be available and it must be possible to charge the cable (no parallel resistance). Two traces generated through rising cable voltage are compared to the resulting flashover. The method differs from the previous one only in the position of the current coupler and in the results obtained in T eed networks. Pg14

15 Fig 25: Basic circuit diagram Comparison Mode II The current coupler is connected directly at the output of the generator. Since the current in both conductors is the same, the coupler can be positioned over either of the two conductors. Fig 26: Positioning the current coupler 3.1. Straight Networks First measurement: The voltage in the cable is increased until a flashover is obtained. The trace could be as follows: Second measurement: A loop is connected at the far end of the cable between the faulty conductor and the healthy conductor. Fig 28: Measurement with loop at the far cable end. Pg15

16 An increase in voltage will cause another flashover at the fault. The trace could be as follows: Fig 29: Recorded reflectogram 2 A comparison between the two traces allows a determination of the distance of the loop at the far cable end to the fault. ty Fig 30: Comparison of the two-recorded reflectograms T eed Networks The fault is located on a main conductor between generator and loop. Fig 31: Fault in the main conductor Note: This method always indicates the distance between the loop and the fault. Pg16

17 The result of this measurement indicates a fault either at the point "X" or "X ". Hence, there is an uncertainty. It is to be noted that with the method described under 7.2.1, the fault would be prelocated directly at the point "X". The fault is located in a tee As in the preceding case, there is an uncertainty. Fig 32: Fault on a tee Note: When using the comparison mode I for fault location on tees, the fault distance indicated is the distance between the far cable end "B" and the point of origin of the tee "D". When using the comparison mode II, the fault distance indicated is the distance between the far cable end "B" and the fault position "X". A combination of the two comparison methods (I and II) enables a reduction of the number of measurements (shifting of the loop). Example: Given the following intentionally simple network: Fig 33: Network with fault on a tee A first measurement is carried out using the Comparison Mode I (coupler 1) with the loop at the end position "B". Y1 = BD (fault at the point of origin of the tee "DC"). A second measurement is conducted using the Comparison Mode II (coupler 2): Y2 = BX = BX' (fault at X or X'). Pg17

18 Conclusion: the fault is at X 3.3. Influence of a connection cable The generator can be connected to the start end of the cable by means of one or two high voltage test leads. To simplify matters, the current coupler is mostly positioned at the level of the generator. Fig 34: Connection via high voltage connection cables The propagation velocity in the high voltage connection cables has no influence on the measurement. 4. Differential comparison mode The impulse source is a surge generator. This method is mainly used in T eed networks. Prerequisite: a healthy conductor must be available. Fig 35: Basic circuit diagram of the differential comparison mode In this method, two traces that have been generated by a surge wave are compared. The surge generator is connected simultaneously to the faulty and healthy conductors The differential coupler The differential coupler consists of two identical couplers which are mounted over the conductors. Pg18

19 The software effects the difference formation. Fig 36: the differential coupler When a pulse or a reflection passes the two conductors simultaneously, then the difference of the two signals I1 and I2 will be Zero: A signal will not be coupled out. (Example: reflection at a tee). If however the reflection returns on one conductor only, then the same signal as with a single current coupler is obtained. (Example: Reflection at the flashover at the fault) Straight networks First measurement Fig 37: Connection for the first measurement If a surge wave passes at the level of the differential coupler, then it will trigger the recorder, notwithstanding the differential measurement. All identical reflections at the two conductors and at the tees will neutralize because of the differential formation. However, the reflection caused by the flashover at the fault and travels to the start end of the cable will generate a differential signal. The trace could be as follows: Pg19

20 Fig 38: Recorded reflectogram after the first measurement Second Measurement A loop is connected at the far cable end between the faulty and healthy conductors. Fig 39: Second measurement with loop A second surge wave generates another flashover at the fault and triggers the transient recorder again. The reflection travels in the direction of the start end of the cable and leads to a trace that is identical to the preceding one. The trace could be as follows: Fig 40: Recorded reflectogram after the second measurement Evaluation of the measurement The difference in the distance covered by the reflection, which passes the loop (2X+2Y) and the reflection without loop (2X) is twice as large as the distance of the loop to the fault (2Y). The corresponding time is "ty". Since the ionisation delay time Δt can vary between two measurements, the points M1 and M2 have to be superimposed for a comparison of the traces. Pg20

21 Fig 41: Superposition of the two traces Hence, the distance between the loop and the fault is: Equation 3: Calculation of the fault distance 4.3. T eed Networks The fault is located on the main conductor between generator and loop Fig 42: Network with T ees As in the comparison mode I, this measurement gives the distance between the loop and the fault, provided the fault lies on the main conductor between the generator and the loop. The fault is located in a tee Pg21

22 Fig 43: Fault indicated at the point of origin of a tee If the fault lies on a tee, then the measurement will give the distance of the loop to the base of this tee. In this case, there is an uncertainty factor, since the fault could lie anywhere on the tee "DC". In order to eliminate the uncertainty, the loop has to be mounted at the end of the tee "C". Now, "AC" is the main current circuit and "DB" a branch Practical Application Prerequisite: - Three linear couplers (I1, I2, I3) - The cable has two healthy conductors 1. Connect the surge generator and the three current couplers (I1, I2, I3) to the cable conductors (L1, L2, L3) 2. Connect a loop between the faulty and healthy conductor at the far cable end. Fig 44: Use of three couplers 3. Transmission of a surge wave and use of the differential coupler "I1-I2" (Corresponds to trace without loop). Fig 45: First measurement without loop 4. Transmission of a surge wave and use of the differential coupler "I1-I3" (Corresponds to a trace with loop). Pg22

23 Fig 46: Second measurement with loop 5. Comparison of the content of the memory after the points M1 and M2 have been shifted to superimpose. Fig 47: Both traces superimposed Conclusion: The advantage of this circuit is that if required, the traces with and without loop can be repeated several times in order to select the comparison that provides the best interpretation without having to connect and disconnect the loop. The safety factor cannot be denied (no additional contact with the cable required) Influence of a connection cable If a high voltage test leads is used between the generator and the start end of the cable, then the differential coupler can be positioned either at the level of the generator, or at the start end of the cable. Pg23

24 Fig 48: Position of differential couplers The propagation velocities in the high voltage test leads and in the coaxial cable do not have any influence on the measurement, since this is always carried out from the far end of the cable to the point of fault. 5. Loop On - Loop Off Method The impulse source is a high voltage test set or a surge generator. This method is used in unt eed networks. Prerequisite: the high voltage test leads has to be sufficiently long (l' >> 50 m). In this method, one compares to traces, first with loop and then without (loop on / loop off), whereby the loop is connected at the end of the high voltage test leads. Fig 49: Connection with high voltage test lead Pg24

25 Fig 50: Test result of the loop on / loop off method If the measurement is carried out with surge waves, then the points M have to be shifted to superimpose. If a voltage generator (e.g. surge generator with the impulse key pressed) is used for the measurement, then the left part of the waveform up to point "M" is not displayed. The measurement always gives the distance of the fault to the start end of the faulty cable. The fault distance is calculated as follows: Equation 4: Calculation of the fault 6. Measuring system Centrix 3 ICE 3PH 6.1. Phase selection For three-phase ICE measurement, the phase selection results from the following circuit diagram: In the phase selection menu a maximum of two phases can be activated for simultaneous measurement. If the operator selects three phases, the measurement cannot be started and an appropriate message is displayed. Pg25

26 6.2. Phase selection menu Connection of the test cable Simplified block diagram showing the currently activated phases (closed switch) and their polarity corresponding to the active measuring coil selection (in this case L1 and L2). Menu to select the phases applicable for measurement (in this case L1 and L2). Menu to select the measurement coil applicable for measurement (in this case P1). The phases measured by the respective coil and their polarity are shown in parentheses (corresponding to the active phase selection). For typical three-phased ICE measurements, two measurements have to be performed. For the first measurement, the test cable has to be connected as follows: Prior to the second measurement, the two selected phases (in this case L1 and L2) have to be looped at the far end of the cable as shown in the picture below: Pg26

27 Operation mode The operator can choose between two operation modes using the menu item: Operation mode Surge Charge Description The fault-flash over is forced by a capacitive discharge of an impulse current. The fault-flash over is forced by charging the cable with HV (with a surge capacitor connected in parallel). Pg27