EP0864096A2

EP0864096A2 - Process for producing signals identifying faulty loops in a polyphase electrical power supply network

Info

Publication number: EP0864096A2
Application number: EP96946020A
Authority: EP
Inventors: Andreas Jurisch; Matthias Kereit
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 1995-11-27
Filing date: 1996-11-22
Publication date: 1998-09-16
Also published as: DE19545267A1; US6034592A; CN1203666A; CN1078710C; CA2238506A1; WO1997020219A2; WO1997020219A3; DE19545267C2

Abstract

The invention relates to a process for producing signals identifying faulty loops in a polyphase electrical power supply network using impedance starting, during which process the signals identifying the faulty loops are formed after checking for earth faults by comparing the amounts of the impedance values produced during impedance starting. To eliminate during said process with great reliability all the loops which are not actually faulty in spite of initial starting the invention proposes that actual faulty loops are determined while determining exclusively earth-faultless loops by comparing calculated virtual impedances (ULx-Ly/ILx; ULx-Ly/ILy) with impedances (ZLx-Ly) measured during impedance starting; by determining at least one loop with earth fault, faultless phase-to-ground loops are recognised and eliminated by comparing the amounts of the virtual impedance values (ZVLxE) obtained from the impedance values of the phase-to-ground loops, detected as faulty, with the smallest virtual impedance values (ZVLxE). To use the impedance values of the other loops determined as faulty but not eliminated, examination processes (23, 24, 25) designed in various ways according to the number of simultaneously established phase-to-ground loops are used.

Description

description

Method for obtaining faulty loops in signals characterizing a multiphase electrical power supply network.

The invention relates to a method for obtaining faulty loops in a signal characterizing multiphase electrical power supply network by means of impedance excitation, in which

- It is checked whether there is at least one loop with an earth fault among the loops detected as faulty and

- Then, by comparing the amounts of the impedance values obtained with the impedance excitation, the signals characterizing the faulty loops are formed.

A known method of this type is described on page 36 in the Siemens device manual “Digital Branch Protection 7SA511 V3.0, Order No. C53000-G1100-C98-1, 1995”.

In this known method, impedance excitation takes the form of a loop-related excitation method. After carrying out a first method step for earth fault detection, the conductor-earth loops are monitored for at least one earth fault detected and the conductor-conductor loops for no earth fault detected. A loop is considered to be excited if the corresponding impedance pointer determined lies within the excitation polygon that applies to the respective loop. If several loops are excited at the same time, an impedance comparison is carried out in which only those loops are classified as excited whose impedance is no more than 1.5 times the smallest loop impedance.

The invention is based on the object of a method for obtaining faulty loops in signals which characterize a multiphase electrical power supply network To be specified, in which all the loops which, despite the initial suggestion, are actually not faulty are eliminated with great certainty, so that, for example, associated distance protection only examines the loops which are actually affected by a fault.

To solve this problem, according to the invention in a method of the type specified at the outset

- If only loops that are free of earth faults are determined by comparing the virtual impedances calculated with regard to the detected conductor-conductor loops according to amount and phase with impedances determined in the case of impedance excitation, the loops that are actually faulty are determined, and

when at least one loop with earth faults is determined, a comparison of the amounts of virtual impedance values formed from the impedance values of the conductor-earth loops detected as defective with the smallest virtual impedance value detects and eliminates conductor-earth loops; - For the further processing of the impedance values of the other non-unlocked loops and detected as faulty loops, differently designed test methods are used with regard to the number of conductor-earth loops detected at the same time, of which the number of the loops determined in each case test procedures assigned by conductor-earth loops are carried out.

An essential advantage of the method according to the invention is that tests are carried out in succession according to various criteria to eliminate impedance scheme excitations, the subsequent test depending on the result of the previous test, so that it is only targeted the actually error-free loops are eliminated. A further advantage is seen in the fact that, with the method according to the invention, loops which have been excited in a relatively reliable, erroneous manner because they are error-free can be rejected. This is due to the The use of virtual impedances can be attributed, in the calculation of which only one of the currents used to determine the loop impedance is used, it being assumed that the other current has the same amplitude and is 180 ° in phase with the current used The use of virtual impedances avoids the disadvantages of using subsystem impedances according to the theory of symmetrical components, because the subsystem impedance of a faulty loop is larger than the subsystem impedance of a faultless one in certain error cases

Can be loop. For the definition of co-system impedances, reference is made to the book by R. Roeper "Short-circuit currents in three-phase networks", 6th edition, 1984, pages 48 to 53.

With regard to a very reliable and targeted elimination of excited but inherently error-free loops, the method according to the invention has proven to be advantageous if, in the case of a single detected conductor-earth loop, the test method is carried out in such a way that Comparison of the impedance values of all detected

Conductor-conductor loops with a predetermined multiple of the impedance value of the conductor-earth loop are eliminated those conductor-conductor loops whose impedance values are above the predetermined multiples of the impedance value of the conductor-earth loop, and then by one Compares the phase angle of the impedance values of the non-eliminated conductor-conductor loops and the conductor-earth loop, as well as comparing the impedance values of these loops with one another, a conductor-conductor loop that has not yet been eliminated or the conductor-earth loop is eliminated, and in the case of an eliminated conductor-earth loop, an additional test procedure is carried out with regard to this loop.

In order to obtain reliable work results, it has proven to be advantageous if a check is carried out before comparing the impedance values of all detected conductor-conductor loops becomes whether the magnitude of the respective impedance is greater than a predetermined threshold. This ensures that the phase angle is determined with sufficiently large impedances so that measurement errors are largely excluded.

In order to be able to recognize and eliminate error-free loops in the case of two conductor-earth loops which have been found to be defective in the excitation, likewise with high reliability, an advantageous embodiment of the method according to the invention is used in the case of two identified conductor-earth loops the phase difference between the impedances of the two conductor-earth loops is checked to determine whether it lies above a predetermined limit angle, and if the phase difference lies above this limit angle, the phase angle of the impedances of the conductor-earth Loops with the phase angle of a zero impedance eliminates the phase-earth loop with the greatest phase difference to the zero impedance; by comparing the impedance values of the detected conductor-conductor loops with a smallest impedance value of the conductor-earth loops, those conductor-conductor loops are eliminated whose impedance values are above a predetermined multiple of the smallest impedance value, and then by comparing the phase angles of the impedance values of the non-eliminated conductor-conductor loop and the conductor-earth loop and by comparing the impedance values of these loops with one another, a conductor-conductor loop or the conductor-earth loop that has not yet been eliminated eliminated, and in the case of an eliminated conductor-earth loop, the additional test procedure is carried out with regard to this loop.

With this embodiment of the method according to the invention, it has also proven to be advantageous with regard to the highest possible reliability if it is checked before the phase difference is checked whether the amount of the conductor Earth loops are each greater than a predetermined threshold value.

Both in the determination of actually faulty loops in the case of a single conductor-earth loop initially determined to be defective, and in the case of two conductor-earth loops initially identified as faulty, the impedance values are advantageously compared detected conductor-conductor loops using mitsystem impedances calculated according to the theory of symmetrical components.

To the course of carrying out the procedural invention ^¬ proceedings in error at first one or two alε ermit- which is erroneous in fact of two not yet eliminated wire loops telten phase-ground loops to be able to determine with certainty, according to a further advantageous embodiment, of the method according to the invention, in the case of a phase angle which results in each case when the phase angle is compared with a variable which lies above a ^{predetermined} further limit angle, the amounts of the impedances of the two respective loops are compared with one another; with a larger amount of the impedance of the conductor-conductor loop than that of the conductor-earth loop, the conductor-conductor loop is eliminated, while with a different size ratio, the conductor-earth loop is only eliminated if in the additional test methods, the value of the impedance of the conductor-conductor loop is smaller than the value of the co-system impedance of this loop calculated according to the theory of symmetrical components.

In another additional embodiment of the method according to the invention, in the case of three conductor-earth loops which are found to be defective, the virtual impedances calculated with regard to the conductor, which is detected as defective, according to magnitude and phase are compared with those in the case of impedance excitation, in the case of three Imp the actual faulty conductor-earth loops are determined.

To explain the invention, FIG. 1 shows an arrangement for carrying out the method according to the invention in the form of a block diagram, FIG. 2 shows a flowchart with which an overview of the course of an exemplary embodiment of the method according to the invention is given, FIG. 3 shows another flowchart, with which the sequence of the same exemplary embodiment of the method according to the invention is explained with a single excited conductor-earth loop, in FIG. 4 an additional flow chart with which the procedure is illustrated with two excited conductor-earth loops, and FIG. 5 shows a supplementary flowchart with which the process sequence is shown with three excited conductor-earth loops or with exclusively earth-free, excited conductor-conductor loops.

As the Figure 1 is shown in detail, are EMEM analog-to-digital converter 1 via converter Wil, Wι2 and Wι3 currents I to I _Lι _L3 in conductors of an unillustrated fed polyphase power supply network. In addition, an additional current transformer feeds the analog-digital converter 1 like a current variable proportional to the earth current in the power supply network. Further are emgangsεeitig at the analog-to-digital converter 1 of the voltages are obtained by U _L 3 / over the wall voltage ^¬ ler Wul to Wu3 not shown multiphase Energieversorgungsnetzeε proportional voltages U _u to the conductors; An additional voltage U _E is derived from the earth voltage on the multiphase power supply network via an additional voltage converter Wue. Digital values formed by means of the analog-digital converter 1 from the input variables mentioned are fed via a data bus 2 to a device 3 for impedance measurement. This device 3 is followed by a further data bus 4, with an excitation device 5, with which it is ascertained in more detail below whether the device 3 forms impedance values which indicate a faulty loop due to a comparison with an excitation polygon .

On the output side, the excitation device 5 is connected via a further data bus 6 to an arrangement 7 for Scheme excitation elimination, which is also supplied via an additional data bus 8 with measurement values which are assigned to the excited loops communicated via the data bus 6. In the

Arrangement 7 eliminates the loops which are actually error-free on the basis of a check carried out in this arrangement, so that only the loops that are actually faulty are communicated to a selection circuit 10 via a data bus 9. These actually faulty loops are assigned in the selection circuit 10 the impedance measured values supplied via a data bus 11, so that a distance protection measuring device connected to a further data bus 12 and not shown in FIG issues a tripping failure to an assigned, also not shown circuit breaker in the multiphase power supply network to be monitored.

FIG. 2 shows a flow diagram which gives an overview of the mode of operation of the arrangement 7 according to FIG. 1 m in an outline. If it has been ascertained in the excitation device 5 that an impedance has occurred in at least one loop of the multi-phase energy supply network to be monitored, the impedance polygon falling according to the amount and phase m em, then in a first Step 21 first checks whether an earth fault has occurred. If this is the case, then an amount comparison is carried out in a further checking step 22. In this comparison, only conductor-earth loops - detected by the excitation device 5 - are compared in terms of their impedance with the respective virtual impedance Z _V uε.

The respective virtual impedance ZV _LXE of the phase-to-earth loops detected by the excitation device 5 without taking into account the zero current using the phase current I _Lx in the phase-to-earth loop to be checked according to the following equation (1) vdugehgef uhgfe.

1 U _U. _E

ÜVLxE ⁼ ~ '_ (D In this equation, the quantity ZV _LXE denotes the virtual

Impedance of a loop between a conductor Lx, where x is 1 to 3, and earth; UU- _E denotes the voltage between the conductor of the respectively checked conductor-earth loop and earth, while Iu denotes the conductor current in the corresponding conductor Lx.

If the magnitudes of the impedances of the conductor-earth loops detected by the excitation device 5 exceed a threshold value, which is given by k times the magnitude (eg 1.5 times) of the smallest virtual impedance in each case, then the corresponding suggestion is reset. The corresponding conductor-earth loop is thus eliminated and is no longer considered to be excited.

This is followed by a further check of the loops which are still in the test procedure as being defective, depending on the number of loops originally recognized as being excited at the same time by the excitation device 5. The criteria for this are the signals which are output by the device 5 via the data bus 6. Based on one of the direction 5 generated excitation pattern that in the present case only a single conductor-earth loop has been excited, then a test path 23 is taken for the further check, while in a stimulation pattern with two excited conductor-earth loops a different test path 24 is followed; If the excitation pattern shows that three conductor-earth loops have been excited at the same time, an additional test path 25 is taken.

In the test path 23, a function block 26 checks that diY vcir, βU 'Aι »ftJ * <tι.n" tM «n» j T OI-t "*' ^{! τ,} t tttn whether a V-conductor-loop _Λ n is to be classified as actually defective or the detected conductor-earth loop. Correspondingly, a corresponding signal is sent from the arrangement 7 according to FIG. 1 to the selection circuit 10 via the data bus 9.

In the test path 24 there is a function block 27 with which it is determined whether in the case of two simultaneously excited conductor-earth loops, conductor-earth loops or conductor-conductor loops are to be treated as actually excited.

A function block 28 is present in the test path 25, in which a check for apparent impedance excitations is always carried out when either the earth fault detection in method step 21 has not resulted in an earth fault, or when — as just explained — three conductor-earth loops are indicated as being simultaneous has been detected by the excitation device 5. Even if the function blocks 26 and 27 do not provide any unambiguous statements, the function block 28 is additionally run through, so that after the process according to the invention according to FIG. 2 has been run through, in the end only signals characterizing such loops are sent to the selection circuit 10 according to FIG that actually denote faulty loops. The function block 26 according to FIG. 2 is shown in detail in FIG. 3, the assignment of FIG. 3 to FIG. 2 being identified by the circled numbers. The flow chart shown in FIG. 3 shows that in a step 30 it is first checked whether the amount of the impedance of the respective conductor-earth loop is greater than a predetermined limit value, in the present case 250 mΩ. This prevents a measurement error that is too large, which would result if the impedances were too small in the phase angle measurement of the loop impedances carried out subsequently.

After step or branching 30 with the impedance check just described, a further method step 31 then eliminates all of the conductor-conductor loops which have an impedance magnitude ZΓL-L greater by a certain factor K ₂ than the loading,. the smallest determined conductor-earth impedance Z _FL . _E is; v Xp can be 1.8, for example. A phase angle check is now carried out with the remaining conductor-conductor loops, provided that it was determined in a further method step 32 that the amounts of the remaining loop impedances are greater than 250 mΩ in the example. If this threshold value query fails in a conductor-conductor loop, then it is no longer possible to make a clear statement about the error in the function block shown in FIG. 3. It then branches to the function block 28 according to FIG. 2.

If the phase angles of the conductor-conductor impedance and the one detected conductor-earth impedance differ from one another by, for example, 30 °, as will be checked in the further method step 33, then it is assumed that one of the two excitations is one Mock excitation must be. Which of the two loops really has an error is determined in further method steps. If both phase angles are the same, a two-phase earth faults. In order to be quite sure in this regard, a further check is then carried out in a function block 34.

In the course of the further test, in order to determine which of the two loops is actually involved in the error, the amounts of the loop impedances in step 35 are compared with one another after step 33. If the magnitude of the conductor-conductor impedance is smaller than that of the conductor-earth impedance and the pointer of the conductor-conductor impedance lies in the 1st or 3rd quadrant of the pointer diagram, the conductor-conductor loop is not at Errors involved and can be eliminated. If these conditions are not met, the conductor-earth loop may only be eliminated if the measured conductor-conductor impedance can be classified as actually defective by comparison with the virtual impedances. In this case it is an overfunction of the earth fault detection. If the conductor-earth loop has been eliminated, after the function block according to FIG. 3 has been processed, the function block 28 according to FIG. 2 must also be run through. If the impedance of the conductor-conductor loop cannot be recognized as actually defective, this impedance must have an apparent impedance and the corresponding conductor-conductor loop is eliminated.

In step 36, following step 35, of checking the conductor-conductor impedance or verifying the conductor-conductor loop, it is assumed that the conductor currents of both conductors actually or supposedly involved in the fault are identical to the fault current smd. Under this assumption, two virtual loop impedances can be calculated for a conductor-conductor loop. Only one conductor current is used in each case for the calculation of a virtual impedance; the calculation is carried out according to the following equations (2) and (3) ι y Lx-Ly

2vι Lx-Ly

Those with Zyi _Lx-Ly the virtual impedance of the respective conductor-conductor loop when calculating with the current l _Lx through the first conductor involved in the fault and with _Zv _{2 Lx} - _L y the virtual impedance when calculating with the current I _Ly the second leader involved in the fault is designated; U _L χ-Ly smd in equations (2) and (3) denotes the voltages between the two conductors of the respective loop. The two virtual impedances calculated in this way do not differ significantly from one another and are these two to Mitsystem impedance according to the theory of symmetrical components of standardized loop impedances is not significantly greater than the impedance calculated by the device 3 according to FIG. 1, then it is certainly not an apparent impedance. If the amount of the co-system impedance is less than, for example, 0.3 times the amount of the virtual impedances, then the loop impedance cannot be verified. It follows that the corresponding conductor-conductor loop cannot be involved in the error; It is eliminated if the resistance and reactance have the same sign.

FIG. 4 shows the functional block 27 according to FIG. 2 m with its individual method steps. It can be seen that a first step 40 first checks whether the impedances of the two excited conductor-earth loops are large enough for a comparison of the phase angles smd. If this is not the case, then function block 28 according to FIG. 2 is run through.

Otherwise, it is then checked in a further method step 41 whether the error in the case of two conductor-earth loops created is a two-pole earth fault. This is decided by comparing the phase angles of the two phase-to-earth loops. If there is a difference of less than 30 °, then it could be a two-pole earth fault; a further check is then carried out in function block 28 according to FIG. 2. Otherwise, one of the two conductor-earth loops is not faulty, and one of the two loops must be eliminated. This is decided in consideration of the zero impedance in the following steps 42 and 42.

The zero impedance is calculated according to equation (4) below

Z ₀ = - ^ (4)

calculated in which Z_ ₀ denotes the zero impedance, U ₀ the voltage to earth and I _E the earth current. Only the phase angle of the zero impedance is important and is therefore only calculated. Subsequently, the differences between the phase angle of the two phase-to-earth impedances and the phase angle of the zero impedance are formed in step 42. Only the smallest phase angle φ _min between the phase angle of the zero impedance and the phase angles of the impedances of the phase-earth loop Z _{FL-E is taken} into account. If the smallest angle difference is greater than 45 °, for example, no clear statement can be made about this which of the two conductor-earth loops is not involved in the error. The test according to function block 28 according to FIG. 2 is then carried out immediately. Otherwise, the phase-to-earth loop is eliminated in method step 44, the impedance angle of which differs the most from the phase angle

Distinguishes zero impedance.

Further apparent excitation impedances are then eliminated in the manner described in detail in the explanation of FIG. 3, namely in a scope from method step 31. In block 28, shown in detail in FIG. 5, according to FIG. 2, a check is always carried out when device 5 does not report any conductor-earth loop as excited or when all three conductor-earth loops have been detected as faulty ; Furthermore, as has already been explained above in connection with FIGS. 3 and 4, the function block 28 according to FIG. 2 is always run through if it was not possible to make a clear statement in the function blocks 26 and 27 about the respectively error-prone loop. If errors have already been clearly identified in the function blocks 26 and 27 as two-pole errors, the function block 28 is ignored.

First, it is checked in a first test step 50 whether the excitation pattern generated by the excitation device 5 according to FIG. 1 corresponds to a phase-to-phase-earth fault; if this is the case, a two-pole earth fault is verified. Otherwise, in the test path according to FIG. 5, in a further test step 51, all loops that are still excited are verified with the aid of virtual impedances, namely in the following way:

For the elimination of conductor-earth loops, it is assumed that the conductor current of the faulty conductor and the zero-sequence current are identical to the fault current. Under this assumption, two virtual loop impedances can be calculated for each phase-to-earth loop; once the zero current I _E and the associated phase-to-earth voltage U _{Lx L become} a virtual impedance _Z _V ι LX E and em another time from the phase current _Iiχ ^and the associated phase-to-earth voltage U _Lx E a further virtual impedance Zy ₂ LX E calculated:

Z _{vl Lx} _ _E = - ^ = L ₍₅ ) _Z äV J, U -E. i _. ϊ _τ fcl (6)

If these virtual loop impedances are not significantly greater than the {Mitsystem-) impedance Zu, calculated by the excitation device 5. _E - labeled Z _R in FIG. 5 for reasons of space - then it is in this case with security not an apparent excitation.

In detail, this comparison is carried out in such a way that the magnitudes of the impedances are compared in connection with an angle comparison, as indicated in FIG. 1 and described in detail in the following equations (7) and (8):

<60 (7) and

1.5 | Z, υt-ε n arg UU-E - arg Z _U. _E <60 ⁽ (8)

In these equations (7) and (8), Z _{0 designates} the zero impedance and ZI the co-system impedance of the line; these impedances are not formed from measured variables, but describe them

Characteristics of the line; 1.5 is a freely specified factor _; vχ. Equations (7) and (8) take advantage of the fact that an impedance pointer formed with the loop current I _Lx differs significantly in magnitude and phase position from a pointer formed only with the phase current if there are faultless loops. In contrast, the pointers of faulty loops are almost congruent. However, the angle comparison is only carried out if the amounts of virtual impedance and co-system impedance do not differ too much from one another (factor 1.5). Without checking the angle, the because the loop was evaluated as not verifiable in this step 51.

With the help of this verification of the excitation, in the case of multiple earth shorts with different foot points in one direction from the installation location of a protective device operating according to the method of the invention, it is in principle not possible to clearly differentiate between fault and apparent impedances. For this reason, a dummy impedance elimination with an amount comparison is then carried out in a further test step 52.

To eliminate conductor-conductor loops, the excitation is verified taking into account the phase angle by taking the following relationships (9) and (10) into account:

-> 1. 2 fc? Lx-Ly n _arg zhLlϊL _ _arg z _u . _Ly <60 ° 1 (9) and arg Z Lx -Ly <60 '(10)

In these equations (9) and (10) Vix.iy / Iui denotes a first virtual impedance of the respective conductor-conductor loop and Hur.iy / Iι _Ϋ a further virtual impedance of the same conductor-conductor loop; Z _LX . _L >, is the je¬

Subject errors with different base points in one direction from the installation location of the protective device, in principle no clear distinction between error and apparent impedances is possible.

The first stage of the verification also consists here of a comparison of the amounts of the mitsystem impedance and the virtual eller impedance. If the amounts differ from virtual and mit-system impedance to εtly different from each other (here by factor 1.2), no angle test is carried out, the corresponding loop cannot be verified.

If the amounts of virtual and co-system impedance are not inadmissibly different for both conductor-earth and conductor-conductor loops, then the angle comparison between the Mitsyεtemimpedanz and the two virtual impedances. The respective angle comparison is only carried out if the impedance is large enough for a sufficiently precise angle determination.

When angle test all of the loops to be confirmed where the Dif ^¬ ference between the phase angle of the Mitsystemimpedanz and the phase angle of the virtual impedance smaller alε 60 ° lεt. If the amount comparison and the angle test with the first virtual impedance are positive, then the same tests are carried out again with the second virtual impedance. If the impedance of a loop cannot be verified, then this loop is only eliminated from the excitation pattern if the Mitsyεtemimpedanz is in the 2nd or 4th quadrant of the complex level.

If more than one loop is excited at this point and no double earth fault has been detected so far, the false excitation elimination is carried out for all loops with an amount comparison of the co-system impedances in step 52. As in step 22 according to FIG. 2, the factor k is 1.5. The additional comparison of the amounts is necessary since not all types of error can be clearly recognized when verifying the excitation.

Finally, a check is made as to whether all excited loops have been eliminated. If this is the case, then the last elimination is reversed by comparing the amounts.

Claims

claims

1. A method for obtaining faulty loops in a signal characterizing multi-phase electrical power supply network by means of impedance excitation, in which

- Then, by comparing the amounts of the impedance values obtained with the impedance excitation, the signals characterizing the faulty loops are formed, that is to say, d a ß

- When determining only earth fault-free loops by comparing virtual impedances calculated with regard to the detected conductor-conductor loops U_, _x -Ly / .lLy) according to amount and phase with impedances (Z _Lx . _Ly ) determined with the impedance excitation, those loops are determined as loops that are actually faulty, the virtual impedances of which are greater in magnitude than the determined impedance and for which the Difference in phase angle between these impedances is less than a predetermined angle and

- When determining at least one loop with earth faults, those conductor-earth loops detected are identified and eliminated as fault-free conductor-earth loops, the impedances of which are k times the magnitude of the smallest virtual impedance (ZV _LXE ) exceed the virtual impedances formed with respect to the detected conductor-earth loops.

2. The method of claim 1, d a d u r c h g e k e n n z e i c h n e t, d a ß

- In the case of a single detected conductor-earth loop, those conductor-conductor loops are eliminated by comparing the impedance values of all detected conductor-conductor loops with a predetermined multiple (K ₂ ) of the impedance value of the conductor-earth loop whose impedance values (Z _FL-IJ ) lie above the predetermined multiple of the impedance value (Z _FL - _E ) of the conductor-earth loop, and

- those of the non-eliminated conductor-conductor loops are then eliminated, - in which the difference between the phase angle of their impedance and the phase angle of the detected conductor-earth loop is smaller than a further predetermined angle (step 33),

- The amount of the impedance is smaller than the amount of the impedance of the phase-earth loop and

- whose impedance pointer lies in the 1st or 3rd quadrant of the vector diagram (step 35).

3. The method of claim 2, d a d u r c h g e k e n n z e i c h n e t, d a ß

- Before comparing the impedance values of all detected conductor-conductor loops, it is checked (step 30) whether the magnitude of the respective impedance is greater than a predetermined threshold value.

4. The method of claim 1, d a d u r c h g e k e n n z e i c h n e t, d a ß

- In the case of two detected conductor-earth loops, the phase difference between the impedances of the two conductor-earth loops is then checked (step 41) whether it lies above a predetermined further limit angle (30 °) and one above this limit angle (30 °) lying phase difference by comparing the phase angle of the impedances of the phase-earth loops with the phase angle of a zero impedance Z_ _0, the phase-earth loop with the greatest phase difference to zero impedance Z _{0 is} eliminated (steps 42 to 44) and

- By comparing the impedance values of the detected conductor-conductor loops with a predetermined multiple (K ₂ ) of the smallest impedance value of the conductor-earth loops, those conductor-conductor loops whose im- tolerance values (Z _FL - _L ) are above the predetermined multiple of the smallest impedance value Z _FL -E), and

5. The method of claim 4, d a d u r c h g e k e n n z e i c h n e t that

- Before the phase difference is checked, it is checked whether the amount of the conductor-earth loops is in each case greater than a predetermined threshold value (step 40).

6. The method of claim 1, d a d u r c h g e k e n n z e i c h n e t, d a ß

- In the case of three conductor-earth loops found to be defective, those conductor-earth and conductor-conductor loops detected whose impedance pointers differ significantly from their respective virtual impedance pointers are eliminated.

7. The method according to any one of the preceding claims, d a d u r c h g e k e n n z e i c h n e t, d a ß - used to compare the impedance values of the detected conductor-conductor loops according to the theory of symmetrical components Mitsystemimpedanzen.

8. The method according to any one of the preceding claims, d a d u r c h g e k e n n z e i c h n e t

with a phase angle of the non-eliminated conductor-conductor that results when comparing the phase angles Loop and the conductor-earth loop with a size lying above a predetermined additional limit angle, the amounts of the impedances of the two respective loops are compared with one another and, for a larger amount, the impedance of the conductor-conductor loop than that of the conductor-earth Loop the conductor-conductor loop is eliminated (step 35), whereas with a different size ratio the conductor-earth loop is only eliminated if the value of the impedance of the conductor-conductor loop is smaller than the value of the after The theory of the symmetrical components computed with the system impedance of this loop is.