Method and device for determining a short-circuit direction

Abstract

Method (500) for determining the direction of an earth fault of a conductor (110a) in a compensated power supply network (100), comprising: - Determine the polarity of the gradient of the voltage of the earth-faulted conductor (110a) before the earth fault occurs, and - determining the polarity of the DC component of the zero-sequence current in a section of the earth-faulted conductor (110a) or of the total current of the conductors 110a, 110b and 110c after the earth fault has occurred, and - Determining the direction of the earth fault with respect to the direction of the undisturbed current flow based on the determined polarity of the zero current or the total current and the polarity of the slope of the voltage of the earth-faulted conductor (110a).

Description

[0001] The invention relates to a method for determining the direction of a single-phase earth fault in compensated power supply networks and to a correspondingly configured device for determining the direction of the earth fault.

[0002] Power grids are referred to as resonant-earthed or compensated grids if their star point, for example, the star point of the supplying transformer, is connected to earth via a compensation reactor, also known as a Petersen coil or earth-fault suppression coil. Typically, but not exclusively, such a compensation reactor is used in three-phase power grids at voltage levels above 1 kV and up to 110 kV. Such a compensation reactor has no influence on the grid during normal operation because the currents of the three phases add up to zero in symmetrical operation, which is typically desired and approximately achieved, and the star point of a three-phase power grid is thus at earth potential.Only in the event of a fault, in this case a single-pole earth fault, does a current flow through the compensation reactor due to the potential difference between the star point of the network and the earth potential.

[0003] A significant proportion of all faults in three-phase power supply networks are due to earth faults, for example due to cable damage or other faulty or damaged insulation of a conductor. Single-phase earth faults are the most common cause of faults. In the event of a conductor earth fault, for example due to an insulation fault in a cable network, a current circuit is created in such a compensated network from the earth-faulted conductor via the earth fault to earth and via the compensation reactor. Due to the high capacitance of a cable network in particular, a very high capacitive earth fault current would flow through the fault location. In a network without a compensation reactor, this earth fault current could be so large that the touch voltage at the fault location would be correspondingly high, endangering human safety and requiring the network to be shut down immediately.In a compensated network, however, the compensation reactor causes an additional inductive current to flow across the fault location, which ideally completely compensates for the capacitive current at the fault location. The magnitude of the inductive current should therefore be as equal as possible to the magnitude of the flowing capacitive earth fault current, so that the active current at the earth fault location is as small as possible. The inductance of the compensation reactor is dimensioned so that its impedance is equal to the network capacitance. In this way, the compensation reactor creates an inductive current to compensate for the large capacitive earth fault current.

[0004] In the event of such a ground fault, the location of the fault must be determined as quickly as possible so that the fault can be rectified as quickly as possible. This is typically done by determining the direction of the fault from a measuring point—i.e., in the direction of the load flow or opposite to the load flow, as viewed from the measuring point—so that the location of the ground fault can be narrowed down. The load flow is equal to the current flow in an undisturbed state.

[0005] Known from the state of the art are, for example, the so-called wiper method or the wattmetric cos(φ) method or the OU method from DE 102 25 058 A1. These methods each require the zero-sequence current and the zero-sequence voltage. The wiper method is a transient method and evaluates the first transients of the zero-sequence current and zero-sequence voltage, so these values ​​must be permanently recorded so that they are available in the event of a fault. This is correspondingly complex for overhead line indicators. For the stationary cos(φ) method, the zero-sequence current and zero-sequence voltage must be available with very high accuracy at the time of the earth fault; this is hardly possible in practice.The still well-known so-called pulse location requires a device for feeding a pulse into the star point of the network and, moreover, can only be used for low-resistance earth faults, is unreliable in the case of under-compensation of the network and requires that the earth fault persists stationary for approximately 30 seconds.

[0006] EP 3 046 197 B1 describes a method and a device for detecting the earth fault direction in a compensated or isolated three-phase electrical network, in which pairs of values ​​of a zero-sequence voltage and a zero-sequence current are measured, the active or reactive energy is calculated, a voltage and a current flag are combined by a Boolean operation, whereby the presence of an earth fault is determined depending on the result, and a decision is made at least on the basis of the sign of the active or reactive energy whether the earth fault direction is to be reported as "forward" or "reverse".

[0007] AT 507 454 B1 describes a method for detecting the direction of an earth fault from the charging oscillation of the single-ignition, the re-ignition and the intermittent earth fault for fault resistances from 0 ohms up to the range of a few kOhms in radial, ring-shaped or meshed AC and three-phase systems with an isolated star point or with a low- or high-resistance earthed star point by continuously sampling and storing the zero-sequence voltage and the zero-sequence current and determining an earth fault and its type.

[0008] DE 197 32 103 C2 describes a method for indicating the direction of earth fault currents in medium-voltage three-phase systems. The earth fault current is detected using a transformer with a ring-shaped iron yoke. A core-balance transformer generates a true-angle 50 Hz signal of the earth fault current, and the earth fault direction is determined by comparing the phase angles of the earth fault current and voltage.

[0009] Therefore, the object is to propose a method and a corresponding device with which the direction to the location of the earth fault can be determined from the measurement location. This object is achieved by the method and the device according to the independent claims. The method and the device are described in more detail below with reference to the figures.

[0010] It shows Fig. 1 a simplified representation of a phase of a compensated network with an earth fault in one phase; Fig. 2 an RL model for modelling fault occurrence; Fig. 3a a voltage curve for a single-phase earth fault t0; Fig. 3b shows a zero-sequence current curve for a single-phase earth fault; Fig. 4 a simplified representation of a compensated ring network; Fig. 5 a flowchart showing the steps for detecting the direction of a conductor earth fault.

[0011] Fig. Figure 1 shows a simplified representation of a known phase network 100, here three-phase with the corresponding conductors 110a, 110b, and 110c. The network is fed on one side by a source 120, which is represented in the figure by the voltage sources 120a-120c. The network supplies a load 160. The inductances of the conductors 110a-110c are schematically represented by the line inductances 130a-130c, and the capacitances of the conductors 110a-110c with respect to ground are represented by the capacitances C E 140a-140c. The phase network 100 is connected at its star point N via a compensation choke 150, which is also called Petersen coil L P or quenching coil, connected to earth.

[0012] The single-pole fault of conductor 110a with earth, i.e. in the case of an earth fault of conductor 110a with earth, can essentially be described as an ohmic resistance 170 R Fthrough which the earth-faulted conductor 110a is connected to earth.

[0013] In contrast to a conventional network, the network schematically illustrated here comprises at least one device 190, which is configured and provided to determine the direction of a ground fault of a conductor 110a-110c. In the schematically illustrated phase network 100, at least one such device 190 is arranged to provide the measured values ​​described below.

[0014] In the following, it is assumed that under normal circumstances, i.e., without a ground fault of a conductor, the current in the phase network 100 flows from the sources 120a-120c to the load 160 and thus also in the conductor 110a in the direction of the arrows shown. In the event of a ground fault, here of the conductor 110a, a new circuit is created in which the current flows from the voltage source 120 via the conductor 110a, the inductance 130a of the line, the resistance 170 R Fof the earth fault, the earth and the compensation reactor 150 flows.

[0015] Fig. Figure 2 schematically shows a model of a circuit 200 at the moment of a ground fault between conductor 110a and ground. As briefly described above, in the event of a ground fault, a current i(t) flows from source 220 through inductance 230 and the resistance R240 of the ground fault to ground. The inductance 230 is the sum of the line inductance of the ground-faulted conductor along the length from the source to the location of the ground fault and the compensation reactor of the network. The ground fault is represented in the RL model by the closing of switch S at time t=t0.

[0016] As is well known, in the case of an earth fault, a large current i(t) initially flows with a transient compensation process, which is characterized by a homogeneous i h (t)and a particulate fraction i p (t) can be described as: i(t)=ih(t)+ip(t)

[0017] One loop of the stitch thus leads to L⋅di(t)dt+R⋅i(t)=u⌢⋅cos(ωt+φu)

[0018] Where û denotes the peak voltage, ω the frequency of the voltage and φ u the phase angle of the voltage.

[0019] Assuming that the transient compensation process can be neglected at a later time, only the particulate part of the steady-state current waveform remains, so that from [2] for t → oo results in: L⋅dip(t)dt+R⋅ip(t)=u⌢⋅cos(ωt+φu)

[0020] The homogeneous solution is obtained by subtracting [2] from [3]: L⋅dih(t)dt+R⋅ih(t)=0

[0021] Since the particular solution in [3] describes the steady state, it can be assumed that the switch S in Fig. 2 was closed an infinitely long time ago. The particular solution can then be described with the following cosine oscillation: ip(t)=l⌢⋅cos(ωt+φi) where for the current i l⌢=u⌢R2+(ωL)2 and for the phase angle of the current φ i φi=φu−tan−1(ωLR) applies.

[0022] The general approach is chosen to describe the homogeneous solution: ih(t)=k⋅e−RL⋅t

[0023] This approach satisfies the condition that the homogeneous solution must have decayed to infinity. The unknown k is determined via the initial condition, namely that before the switch S is turned on, i.e. before the earth fault occurs, in which no current flows through the earth fault, and at the time t=t0 of the earth fault, in which Fig. 2 the current i(t) is zero, so that i(t0)=ih(t0)+ip(t0)=!0

[0024] Inserting [8] into this equation gives the unknown k as k=−l⌢⋅cos(ωt0+φi)⋅eRL⋅t0

[0025] The homogeneous solution for the current i(t) is obtained by inserting k in [8] to ih(t)=l⌢⋅cos(ωt0+φi)⋅e−RL⋅(t−t0)

[0026] Equation

[11] thus describes the decaying DC component of the current i(t), which from t=t0 in the circuit of the Fig. 2 and thus flows through the earth-faulted conductor and across the earth fault. The magnitude of the DC component depends on the time of fault occurrence and is obtained by inserting [7] into

[11] to ih(t)=−l⌢⋅cos(ωt0+φu−tan-1(ωLR))⋅e−RL⋅(t−t0)

[0027] Since the compensation choke typically has a very large inductance value, represented by L in equation

[12] above, the tan -1- Term in

[12] can be simplified to 90°. Thus, when defining the phase reference, the time of the earth fault determines whether a DC component is present and, if so, how high this is. If the earth fault occurs exactly at the maximum or minimum of the voltage, ie the switch S in Fig. 2 is closed exactly at the maximum or minimum of the cosine-shaped voltage waveform, the cosine term in

[12] is zero. It follows directly that there is no DC component if the earth fault occurs exactly at the time of the voltage maximum or minimum (negative voltage maximum), see equation [8]. In all other cases, the earth fault causes a decaying DC component of the zero-sequence current of phase 100 or the sum current of the phase currents of conductors 100a, 100b, and 100c.

[0028] This means that in the event of an earth fault in a conductor of a compensated three-phase network, a current flows between the location of the earth fault and the compensation reactor 150. The DC component of the current via the earth fault can be measured in a three-phase network by detecting the zero-sequence current. If this is not possible, e.g. with overhead line indicators, the sum current of the individual conductor currents can be determined. The sum current of a three-phase network is always three times the zero-sequence current and therefore has the same characteristic curve. In practice, the so-called Holmgreen circuit is often used in secondary network stations to evaluate the zero-sequence current or the sum current. The voltage gradient in the conductor affected by the earth fault at the time shortly before the earth fault occurs determines the polarity of the DC component of the current, i.e. the direction of current flow.

[0029] The polarity of the DC component of the current across the earth fault, and thus the polarity of the DC component in the total current of a three-phase network, corresponds to the polarity of the gradient of the phase voltage affected by the earth fault shortly before the earth fault occurs, if the location of the earth fault is in the direction of the load 160 and thus in the direction of the undisturbed current flow, seen from the measuring point 190. Otherwise, if the location of the earth fault is in the direction of the source, seen from the measuring point 190, the DC component in a phase network is zero.

[0030] For a ring network, see also Fig. 4, the same applies to a ground fault as to a phase network, only if the location of the ground fault is in the direction of the load, i.e., in the direction of undisturbed current flow. If the location of the ground fault in a ring network is in the direction of the source, as seen from the measuring point, the polarity of the DC component of the zero-sequence current or the total current is opposite to the polarity of the voltage gradient shortly before the ground fault occurs and is not zero.

[0031] Fig. Figure 3a shows a voltage waveform in a three-phase compensated phase network in which a single-phase ground fault occurs at time t=t0. It is assumed that at the time of the ground fault, i.e., at time t=t0, the voltages 310a-310c are symmetrical and sinusoidal. The ground fault of the phase with voltage 310a occurs at time t=t0, i.e., while voltage 310a is rising on its positive slope. Voltage 310a collapses at the time of the ground fault and, after a brief transient settling, is at zero, while the voltages 310b and 310c of the two phases not affected by the ground fault remain sinusoidal with increased amplitude.

[0032] Fig. Figure 3b shows the corresponding curve of the DC component 320 of the transient curve of the earth fault current 330 across the fault location, which flows through the earth to the compensation reactor of the compensated network. It is assumed that the device measuring the current flow through the conductor is connected to the Fig. 1. The DC component of the current is greater than zero here, meaning the current flows in the same direction as the current flow during normal operation, i.e., without an earth fault. From this, it can be concluded that the earth fault is in the forward direction, in the direction of the load, as seen from the measuring location. If the location of the earth fault were between the source and the measuring location, the DC component 320a of the current would be zero in a phase network and would have a different polarity in a ring network, i.e., it would be less than zero.

[0033] Fig. Figure 4 shows a schematic representation of a compensated network 400 with the outgoing lines AE and an annular section 410 formed over the outgoing lines C and D, wherein the connection between the conductors of outgoing line C and the conductors of outgoing line D is schematically illustrated with a closed switch S 420. The voltages in the conductors of outgoing line C are thus equal to the voltages in the connected conductors of outgoing line D.

[0034] Just like the one in Fig. 1, this network has voltage sources 120, line inductances 130, and line capacitances 140, and the network supplies at least one load 160. Furthermore, the network is connected to its star point N via a compensation choke L P 150 with grounded.

[0035] The network here has (at least) the devices 190A-190E for determining the direction of an earth fault at the locations shown, with the devices 190A designating those on the conductors of outgoing circuit A, the devices 190B designating those on the conductors of outgoing circuit B and so on.

[0036] In the case of an earth fault, which is schematically represented in the figure by the resistance R F 420 on a conductor in outgoing circuit D, a circuit is created on the one hand via the earth-faulted conductor in outgoing circuit D and via the earth and the compensation reactor 150. Furthermore, the earth fault creates a circuit via the switch 420 of a conductor in outgoing circuit C, so that the conductor in outgoing circuit C connected to the earth-faulted conductor via the switch 420 is also earth-faulted, albeit not directly, but only via the closed switch 420.

[0037] Assuming that at the time of occurrence of the earth fault in a conductor of the outgoing circuit D the voltage waveform is as in Fig. 3a, ie at the time of the earth fault, the voltage of the earth-faulted conductor increases and is greater than zero, the earth fault direction indicators 190 can determine and display the direction to the earth fault based on the above-described.

[0038] Arrows 440A-440E indicate the directions for the earth-faulted conductor determined by devices 190, i.e., the earth-fault direction indicators, and based on the respective measuring location of the earth-fault indicator 190. Arrow 440D1 points in the direction of undisturbed current flow, i.e., forward, so that the location of the earth fault, based on the earth-fault direction indicator 190D-1, is in the direction of the load. The two earth-fault direction indicators 190D-2 and 190D-3, see corresponding earth-fault direction arrows 440D2 and 440D3, point opposite to the load flow direction (from the source toward the load during normal operation), i.e., in the reverse direction, since the location of the earth fault 430 is backward with respect to the undisturbed load flow direction and with respect to the respective measuring location of the earth-fault direction indicators.

[0039] The ground fault direction indicators 190C-1 and 190C-2, which are located in the network's outgoing feeder C, which is not itself affected by the ground fault, indicate the direction of the load flow in outgoing feeder C, i.e., the forward direction, whereas the ground fault direction indicator 190C-3 indicates the reverse direction for the location of the ground fault. Accordingly, the method, or rather, the ground fault direction indicators, can also be used to indicate the direction of a ground fault in ring networks.

[0040] Fig. Figure 5 shows the steps of the method for determining the direction of a ground fault in a three-phase network. It is assumed that each of the three phases is routed in a conductor. Furthermore, the method described below assumes that only one of the three phases, i.e., one of the three live conductors, has the ground fault.

[0041] The method 500 begins in the first step 510 by determining the presence of a ground fault. For this purpose, a known method can be used, which in one embodiment determines the current through the compensation reactor. In the case of a sudden increase in the current through the compensation reactor, a ground fault in a conductor of the network typically exists.

[0042] In step 520 of method 500, the conductor that has the ground fault is then determined. For this purpose, the voltage of the conductors can be monitored, for example. As described above with reference to Fig. 3a, the magnitude of the voltage of a conductor subject to an earth fault drops steeply in the event of an earth fault in the conductor. Accordingly, a voltage drop in a conductor 110a can be used to conclude that there is an earth fault. Alternatively, the voltage curve of the other two conductors 110b and 110c can be used to conclude that there is an earth fault in the conductor 110a. As also described above with reference to Fig. As described in Figure 3a, the voltage of the two conductors 110b, 110c not affected by the ground fault increases abruptly at the time of the ground fault. This allows the ground fault of a conductor to be determined even if the voltage curve of the conductor affected by the ground fault occurs at the time of the zero crossing.

[0043] In a further step, see step 530, it is determined whether the voltage of the earth-faulted conductor 110a rose or fell shortly before the occurrence of the earth fault, mathematically expressed as du / dt > 0 or du / dt < 0, i.e. it is determined whether there was a voltage rise or a voltage drop in the earth-faulted conductor immediately before the occurrence of the earth fault. In other words, the polarity of the voltage gradient is determined. This step can be carried out using means known per se. In one embodiment, the voltage curve of a conductor can be saved by sampling and storing the determined values ​​for an evaluation time shortly before the occurrence of an earth fault. The saved values ​​of the voltage curve only need to be saved for a very short period of time.For a mains frequency of f=50Hz, it is therefore sufficient if the sampled voltage values ​​are stored for a time interval of 1 / (2f) = 1 / 100 second, since the sampled values ​​of this time interval represent at least one half-wave of the voltage curve.

[0044] If the slope of the voltage is zero, see 540, i.e. du / dt = 0 is determined, i.e. the voltage of the conductor subject to the earth fault has passed through a positive or negative voltage maximum shortly before the earth fault occurs, no statement can be made and it is not possible to identify the direction of the earth fault, see 550.

[0045] Otherwise, if it could be determined whether the voltage of the conductor now subject to the earth fault rose or fell shortly before the earth fault occurred, i.e., the polarity of the voltage gradient is positive or negative, the direction of the earth fault can be determined with reference to the measurement location of the zero-sequence current. To do this, as described above, the DC component of the zero-sequence current in the conductor section subject to the earth fault is determined at a time shortly after the earth fault occurred, see step 560. Because only the polarity of the DC component of the current through the conductor is determined after the earth fault occurred, it is sufficient if the current flow measurement is performed after the earth fault occurred, so that the corresponding measuring device 190 only needs to measure the current through the conductor subject to the earth fault after the earth fault occurred.Accordingly, a corresponding measuring device 190 must only determine the current flow after an earth fault has occurred and only then must it be activated accordingly.

[0046] The direction of the earth fault, seen from the measuring point of the current after the earth fault, can then be determined on the basis of the determined polarity of the gradient of the voltage immediately before the earth fault and the polarity of the DC component of the current flow after the earth fault, see 570. If the polarity of the gradient of the phase voltage is the same as the polarity of the DC component of the zero sequence current, then the earth fault lies in the direction of the previous, i.e. undisturbed, current flow 580. Otherwise, if the polarity of the gradient of the phase voltage is not the same as the polarity of the DC component of the current flow in the earth-faulted conductor, then the direction of the location of the earth fault, seen from the measuring point of the current through the conductor, lies opposite to the direction of the undisturbed current flow, see 590.

[0047] In this way, with reference to the measurement location at which the polarity of the DC component of the zero-sequence current through the earth-faulted conductor section is determined shortly after the earth fault occurs, it is possible to determine the direction of the earth fault location relative to the undisturbed current flow direction. The order of the steps of the method described above can be interchanged as appropriate or performed simultaneously. Thus, the steps of determining the polarity of the voltage gradient and the step of determining the polarity of the current flow through the earth-faulted conductor can be performed simultaneously or in any order.

[0048] The method described above is also suitable for determining the direction to the location of a ground fault in a single-phase compensated network. The step of determining the earth-faulted conductor is therefore obsolete in such a network. List of reference symbols 100 strand network 110a, 110b, 110c Conductors of the branch network 120 voltage source 120a, 120b, 120c voltage sources 130a, 130b, 130c line inductances 140a, 140b, 140c line capacities 150 compensation choke, Petersen coil 160 load 170 Earth fault with resistance R F 180 star point 190 Device for determining the direction of an earth fault, earth fault direction indicator 200 circuit 210 Switch S 220 source, voltage source 230 Inductance L 240 ohmic resistance R 310a-310c conductor voltages 320 DC component 330 Exchange share 410 ring-shaped network section 420 switches 430 Earth fault or earth fault resistance 440A-440E Earth fault direction indicator arrows 510-580 procedural steps

Claims

[1] Method (500) for determining the direction of an earth fault of a conductor (110a) in a compensated power supply network (100), comprising: - Determine the polarity of the gradient of the voltage of the earth-faulted conductor (110a) before the earth fault occurs, and - determining the polarity of the DC component of the zero-sequence current in a section of the earth-faulted conductor (110a) or of the total current of the conductors 110a, 110b and 110c after the earth fault has occurred, and - Determining the direction of the earth fault with respect to the direction of the undisturbed current flow based on the determined polarity of the zero current or the total current and the polarity of the slope of the voltage of the earth-faulted conductor (110a). [2] The method of claim 1, further comprising determining (520) the ground faulted conductor (110a). [3] Method according to one of the preceding claims, wherein the voltage of the earth-faulted conductor (110a) is determined and recorded during a predefined time interval before and during the occurrence of the earth fault. [4] Method according to one of the preceding claims, wherein the DC component of the zero current is determined as the sum of the phase currents. [5] Method according to claim 4, wherein the network (100) is an overhead line network. [6] Device for determining the direction of an earth fault of a conductor (110a) of a compensated power supply network (100), wherein the device comprises a device for determining the voltage of the conductor immediately before the occurrence of an earth fault of the conductor (110a) and a device (190) for determining the polarity of the slope of the DC component of the current in the earth-faulted conductor after the occurrence of the earth fault, and which is further adapted to carry out the method according to any one of claims 1-4. [7] Electrical power supply network (100) which is earthed at its star point via a compensation choke (160) and comprises at least one device according to claim 6.

Images