Method, protective device and protective system for detecting a fault on a line of an electrical power supply system

Abstract

A method detects a fault on a line of an electrical power supply system, in which current values are measured at the line ends of the line and the current values are used to check whether there is a fault on the line. In order to perform monitoring of the line such that faults can be detected comparatively quickly and sensitively, it is proposed that voltage values are also measured at the line ends. The respective measured current and voltage values are used to ascertain respective comparison current values indicating the current flowing at a comparison location on the line and the respective comparison current values are used to check for the presence of a fault on the line, and a fault signal is generated if the check has resulted in a fault present on the line being detected.

Claims

1. A method for detecting a fault on a line of an electrical power supply system, the method comprises the steps of: measuring first current values at a first line end of the line; measuring further current values at least at a further line end of the line; using the first and the further current values to check whether there is a fault on the line; measuring first voltage values at the first line end of the line; measuring further voltage values at the at least one further line end of the line; using the first current and voltage values measured at the first line end to ascertain first comparison current values indicating a current flowing at a comparison location on the line; using the further current and voltage values measured at the at least one further line end to ascertain further comparison current values indicating a current flowing at the comparison location on the line; using the first and the further comparison current values to check for a presence of a fault on the line; and generating a fault signal if the check has resulted in a fault present on the line being detected.

2. The method according to claim 1, which further comprises determining the first and the further comparison current values by using a mathematical model of the line for a fault-free case.

3. The method according to claim 2, which further comprises estimating the first and the further comparison current values by using the first and the further current and voltage values and also the mathematical model of the line, wherein an estimation is performed by considering a propagation of travelling waves along the line.

4. A method for detecting a fault on a line of an electrical power supply system, the method comprises the steps of: measuring first current values at a first line end of the line; measuring further current values at least at a further line end of the line; using the first and the further current values to check whether there is a fault on the line; measuring first voltage values at the first line end of the line; measuring further voltage values at the at least one further line end of the line; using the first current and voltage values measured at the first line end to ascertain first comparison current values indicating a current flowing at a comparison location on the line; using the further current and voltage values measured at the at least one further line end to ascertain further comparison current values indicating a current flowing at the comparison location on the line; using the first and the further comparison current values to check for a presence of a fault on the line; and generating a fault signal if the check has resulted in a fault present on the line being detected; wherein the first and the further comparison current and voltage values are ascertained by performing an estimation that uses a mathematical model of the line for a fault-free case and that considers a propagation of travelling waves along the line; and wherein the estimation of the first and the further comparison current values includes subjecting the first and the further current and voltage values to filtering, wherein a filter is used to reproduce a transfer function dependent at least on a propagation constant of a travelling wave along the line and/or a transfer function dependent on a characteristic impedance of the line.

5. The method according to claim 1, which further comprises: using the first and the further comparison current values to form a fault magnitude; and detecting the fault on the line if the fault magnitude exceeds a prescribed threshold value.

6. The method according to claim 1, which further comprises subjecting the first and the further current and voltage values to filtering using a high-pass filter.

7. The method according to claim 1, which further comprises: performing ascertainment of the first and the further comparison current values at each of the first and the further line ends; and transmitting the first and the further comparison current values ascertained at the first and the further line ends between the first and the further line ends by means of an Internet protocol-based data communication.

8. The method according to claim 7, wherein the line has more than two line ends and results in a transmission of the first and the further comparison current values between the line ends being effected by means of multicast messages.

9. The method according to claim 1, which further comprises assigning the first and the further comparison current values timestamps.

10. The method according to claim 9, which further comprises synchronizing timers used for timestamping at the first and the further line ends by means of a time pulse of a central timer or a time synchronization protocol.

11. The method according to claim 1, which further comprises performing the check to determine whether there is a fault on the line at each of the first and the further line ends.

12. A method for detecting a fault on a line of an electrical power supply system, the method comprises the steps of: measuring first current values at a first line end of the line; measuring further current values at least at a further line end of the line; using the first and the further current values to check whether there is a fault on the line; measuring first voltage values at the first line end of the line; measuring further voltage values at the at least one further line end of the line; using the first current and voltage values measured at the first line end to ascertain first comparison current values indicating a current flowing at a comparison location on the line; using the further current and voltage values measured at the at least one further line end to ascertain further comparison current values indicating a current flowing at the comparison location on the line; using the first and the further comparison current values to check for a presence of a fault on the line; performing the check to determine whether there is a fault on the line at each of the first and the further line ends; generating a fault signal if the check has resulted in a fault present on the line being detected; wherein the first and the further comparison current values are received at each of the first and the further line ends that have been ascertained at the other line ends; and wherein received comparison current values and locally ascertained comparison current values are time aligned with one another by means of a renewed sampling and a subsequent interpolation.

13. The method according to claim 10, wherein the time synchronization protocol is a precision time protocol.

14. A protective device for detecting a fault on a line of an electrical power supply system, the protective device comprising: a measuring device set up for recording first current values measured at a first line end of the line; a communication device set up for receiving further current values measured at least at a further line end of the line; an evaluation device set up to use the first and the further current values to check whether there is a fault on the line; said measuring device set up to record first voltage values measured at the first line end of the line; said communication device set up to receive further voltage values measured at the at least one further line end of the line; said evaluation device set up to use the first current and voltage values measured at the first line end to ascertain first comparison current values indicating a current flowing at a comparison location on the line; said evaluation device set up to use the further current and voltage values measured at the at least one further line end to ascertain further comparison current values indicating a current flowing at the comparison location on the line; and said evaluation device set up to use the first and the further comparison current values to check for a presence of a fault on the line and to generate a fault signal if the check has resulted in a fault present on the line being detected.

15. A protective system for detecting a fault on a line of an electrical power supply system, wherein the protective system comprising: a communication connection; and at least two protective devices according to claim 14 connected to one another via said communication connection.

16. The protective system according to claim 15, wherein: the protective system is set up to detect the fault on the line having more than two ends; and one of said protective devices is provided at each end of the line, wherein each of said protective devices are connected to one another via said communication connection.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

(1) FIG. 1 is a schematic view of a first exemplary embodiment of a line of a power supply system;

(2) FIG. 2 is a schematic view of a second exemplary embodiment of the line of the power supply system;

(3) FIG. 3 is a circuit showing a depiction of electrical parameters of a line section Δx to explain a propagation of a current signal on a basis of a travelling wave principle;

(4) FIG. 4 is a schematic block diagram of an exemplary embodiment of an electrical protective device;

(5) FIG. 5 is a graph showing profiles of current and voltage during a fault on a line;

(6) FIG. 6 is a logic diagram to explain a determination of a comparison current value;

(7) FIG. 7 is a schematic block diagram of an exemplary embodiment of an electrical protective device in a more detailed depiction;

(8) FIG. 8 are graphs showing exemplary current profiles without filtering;

(9) FIG. 9 are graphs showing exemplary current profiles with filtering; and

(10) FIG. 10 is a logic diagram to explain the determination of a transfer function of a digital filter.

DETAILED DESCRIPTION OF THE INVENTION

(11) Referring now to the figures of the drawings in detail and first, particularly to FIG. 1 thereof, there is shown a schematic view of a protective system 10 for detecting a fault on a line of an electrical power supply system, which is not depicted in more detail in FIG. 1. In this regard, FIG. 1 shows an electrical line 11 of the power supply system in a simplified depiction. The line 11 may be a single-phase or polyphase line. The line 11 is bounded at its line ends 11a and 11b by circuit breakers 12a, 12b and can be disconnected from the rest of the power supply system thereby, in particular in the event of a fault. The line ends 11a, 11b are also provided with measurement points at which current and voltage values are recorded using current transformers 13a, 13b and voltage transformers 14a, 14b, which are depicted merely by way of example in FIG. 1. The current transformers 13a, 13b and the voltage transformers 14a, 14b may be what are known as conventional or unconventional transformers. On the secondary side, the transformers deliver current values i and voltage values u, which may be analog or digitized values.

(12) The current transformers 13a, 13b and the voltage transformers 14a, 14b have protective devices 15a, 15b connected to them at the respective line ends 11a, 11b for the purpose of detecting a fault on the line 11. The protective devices 15a, 15b use a measuring device to record the current and voltage measured values, and perform digitization and/or preprocessing if need be. In this case, the respective measured values are also assigned a timestamp precisely indicating the time at which they are recorded. For this purpose, the protective devices 15a, 15b have internal timers whose timings are synchronized to one another by means of popular methods, e.g. GPS time pulses, IRIG-B, Precision Time Protocol (PTP) according to IEEE 1588. On the basis of the time-stamping, the measured values picked up at the two line ends 11a, 11b are comparable with one another.

(13) The protective devices 15a, 15b may be electrical differential protective devices, for example, which monitor the operating state of the line 11 on the basis of the recorded current and voltage values and, in the event of a fault, transmit a fault signal T to their respective circuit breaker 12a, 12b in order to prompt it to open its switching contacts.

(14) The protective devices 15a, 15b comprise one evaluation device each, which is respectively set up to monitor the line 11 for a fault (e.g. short, ground fault). Moreover, the protective devices 15a, 15b have one communication device each, via which they are connected to one another by means of a communication connection 16. The communication connection 16 may be any suitable wired or wireless communication connection in the form of a dedicated connection or a communication network. In the example of FIG. 1, the communication connection 16 is configured as a communication network to which the protective devices are connected via network components 17, for example routers or switches. The communication connection 16 can be used by the protective devices 15a, 15b to exchange data messages, the content of which is used for detecting a fault on the line 11. Moreover, they can use the communication connection 16 to exchange messages that are used for synchronizing the timings of the timers of the protective devices.

(15) To monitor the line 11 for faults, the respective protective devices 15a, 15b use the current and voltage values measured at their respective line end 11a, 11b to determine respective comparison current values indicating the current flowing at a comparison location 18 chosen on the line. The comparison current values determined by the different line ends 11a, 11b can finally be used to infer any fault. If a fault is detected, an evaluation device 42 delivers a fault signal F that can be used for actuating the circuit breaker 12a associated with the protective device 15a. Moreover, the fault signal can also be transmitted via the communication device to the other protective devices 15b, 15c in order to trip the respective circuit breakers 12b, 12c there. Instead of or in addition to circuit breaker tripping, the fault signal can alternatively be used as an excitation signal for further protective applications of the protective device 15a and/or can be output to a network operator as an alarm signal. The specific approach for detecting a fault that is present is explained later on.

(16) FIG. 2 shows another exemplary embodiment of a protective system 20. In this case, mutually corresponding or comparable components are denoted by the same reference signs. The protective system 20 shown in FIG. 2 differs from the protective system 10 of FIG. 1 in particular in that the line 11 now has three ends 11a, 11b, 11c instead of two. At the third line end 11c too, a circuit breaker 12c is provided that can be used to isolate the line 11 from the rest of the power supply system. For this purpose, the end 11c is also provided with a protective device 15c, which evaluates current and voltage values measured at the line end 11c and, in the event of a fault, generates a fault signal T in order to prompt the circuit breaker 12c to open its contacts. Furthermore, the protective device 15c is also connected to the communication connection 16 via a network component 17. The protective devices 15a, 15b, 15c are connected to one another via the communication connection 16, which according to FIG. 2 is configured as a communication network.

(17) The protective devices 15a, 15b, 15c are used, using the current and voltage values measured at their respective line end 11a, 11b, 11c, to determine respective comparison current values indicating the current flowing at a common comparison location 18. The comparison location 18 chosen in the example of FIG. 2 is the node at which the three line ends meet. On the basis of the comparison current values from all line ends, it is possible to infer the presence of a fault.

(18) The protective devices 15a, 15b and 15c of FIGS. 1 and 2 can apply what is known at the travelling wave principle for detecting faults. This exploits the circumstance that a current signal propagates on the line 11 at approximately the speed of light. Therefore, if the current and voltage at one point on the line are known, it is possible to infer the current at another point on the line. An approach for sensitive fault detection is described below. First of all, a brief explanation of the principle of travelling waves needs to be provided for this purpose. For this, a propagation model for travelling waves along the line 11 is used.

(19) The algorithm in question is obtained by using “long-line theory”. This is the model-based mapping of an electrical line in the form of what are known as “distributed parameters”. This is depicted by way of example in FIG. 3.

(20) FIG. 3 reveals that the network parameters such as inductance per unit length L.sub.0, capacitance per unit length C.sub.0, resistance per unit length R.sub.0 and conductivity per unit length G.sub.0 are distributed along the line. Based on this line model, using Kirchhoff's laws for the section Δx of the line, the following equations can be obtained for the voltage u and the current i:

(21) $\begin{array}{cc}a)u\left(x+\Delta x,t\right)=R_0\Delta x.Math.i\left(x+\Delta x,t\right)+L_0\Delta x\frac{\partial i\left(x+\Delta x,t\right)}{\partial t}+u\left(x,t\right)& \left(1\right)\\ b)i\left(x+\Delta x,t\right)=G_0\Delta x.Math.u\left(x,t\right)+C_0\Delta x\frac{\partial u\left(x,t\right)}{\partial t}+i\left(x,t\right).& \left(2\right)\end{array}$

(22) Mathematical transformations allow equations (1) and (2) to be converted into the following form:

(23) $\begin{array}{cc}a)\frac{\partial u\left(x,t\right)}{\partial x}=R_0.Math.i\left(x,t\right)+L_0\frac{\partial i\left(x,t\right)}{\partial t}& \left(3\right)\\ b)\frac{\partial i\left(x,t\right)}{\partial x}=G_0.Math.u\left(x,t\right)+C_0\frac{\partial u\left(x,t\right)}{\partial t}.& \left(4\right)\end{array}$

(24) These equations (3) and (4) are partial differential equations for a homogeneous line and are usually referred to as “telegraph equations”. They can be generalized to apply to any numbers of conductors.

(25) By considering equations (3) and (4) in the Laplace domain, assuming x as a parameter (distance of the comparison location from the measurement point), it is possible for many effects occurring the line to be interpreted much more easily:

(26) $\begin{array}{cc}\frac{\partial u\left(x,s\right)}{\partial x}=R_0.Math.i\left(x,s\right)+s{L}_0.Math.i\left(x,s\right)& \left(5\right)\\ \frac{\partial i\left(x,s\right)}{\partial x}=G_0.Math.u\left(x,s\right)+s{C}_0.Math.u\left(x,s\right).& \left(6\right)\end{array}$

(27) The derivation of equations (5) and (6) according to parameter x produces the following:

(28) $\begin{array}{cc}\frac{{\partial }^{2}u\left(x,s\right)}{\partial x^2}=Z\left(s\right)Y\left(s\right).Math.u\left(x,s\right)& \left(7\right)\\ \frac{{\partial }^{2}i\left(x,s\right)}{\partial x^2}=Y\left(s\right)Z\left(s\right).Math.i\left(x,s\right).& \left(8\right)\end{array}$

(29) Equations (7) and (8) can be solved for a voltage and current separately using differential equation theory:
U(x)=e.sup.−γ(s)x.Math.A.sub.1+e.sup.γ(s)x.Math.A.sub.2  (9)
Z.sub.c(s).Math.I(x)=e.sup.−γ(s)x.Math.A.sub.1−e.sup.γ(s)x.Math.A.sub.2  (10).

(30) When solving equations (9) and (10), it is possible to calculate the unknown parameters A.sub.1 and A.sub.2 from the initial conditions:
A.sub.1=½(U.sub.1(s)+Z.sub.c(s).Math.I.sub.1(s))  (11)
A.sub.2=½(U.sub.1(s)−Z.sub.c(s).Math.I.sub.1(s))  (12)
where U.sub.1 and I.sub.1 are the initial conditions when x=0.

(31) Furthermore, equations (9) and (10) contain what is known as a characteristic impedance Z.sub.c and the propagation constant γ, these being able to be calculated from the line parameters:
γ(s).sup.2=Z(s)Y(s)  (13)
Z.sub.c(s)=γ(s).sup.−1.Math.Z(s)  (14).

(32) In this case, Z represents the series impedance and Y represents the shunt admittance of a section of the line. The values are indicated in relation to length in each case.

(33) The following forms are therefore obtained for equations (9) and (10):
U(x,s)=½e.sup.−γ(s)x.Math.(U.sub.1(s)+Z.sub.c(s).Math.I.sub.1(s))+½e.sup.γ(s)x.Math.(U.sub.1(s)−Z.sub.c(s).Math.I.sub.1(s))   (15)
Z.sub.c(s).Math.I(x,s)=½e.sup.−γ(s)x.Math.(U.sub.1(s)+Z.sub.c(s).Math.I.sub.1(s))−½e.sup.γ(s)x.Math.(U.sub.1(s)−Z.sub.c(s).Math.I.sub.1(s))   (16).

(34) Equations (15) and (16) represent a voltage-related and current-related propagation model for travelling waves along the line 11 that can be used for determining comparison currents at a comparison location using the travelling wave method.

(35) FIG. 4 will be taken as a basis for explaining the approach for fault detection using a possible example. In this regard, FIG. 4 shows a schematic block diagram of an exemplary embodiment of a protective device, for example the protective device 15a (cf. FIGS. 1 and 2), for detecting a fault on a line. The other protective devices 15b, 15c are of corresponding design.

(36) The protective device 15a has a measuring device 41, the evaluation device 42 and a communication device 43. The evaluation device 42 may be e.g. a microprocessor that accesses an applicable piece of device software that is in a memory of the protective device 15a. Alternatively, it may also be an arithmetic block with hardware-determined programming, e.g. an ASIC or FPGA.

(37) First of all, the measuring device 41 is used to measure the local currents and voltages present at the applicable line end 11a and to generate corresponding current and voltage values. These measured values are converted into samples of the current and voltage signals. An example of the recorded current and voltage measured values can be seen in FIG. 5; the occurrence of a fault can be seen at approximately t=0.02 s.

(38) In the case of three-phase power supply systems, it is also optionally possible for the respective signals to be handled by means of a transformation (e.g. Clarke transformation, in the case of which the signals are expressed as “alpha”, “beta” and “zero” components), e.g. in order to decouple the phase-related components. The resultant signals are likewise shown in FIG. 5. The transformation can take place in the measuring device 41 or in the downstream evaluation device 42.

(39) If the subsequent evaluation is intended to take place in the frequency domain, a conversion of the transformed current and voltage values into the frequency domain can take place. This is preferably affected by means of a Fast Fourier Transformation (FFT) or Discrete Fourier Transformation (DFT).

(40) A determination of comparison current values indicating the current on the line at a previously chosen comparison location (distance x from the measurement point) on the line also takes place by means of the evaluation device 42 using the current and voltage values—if need be transformed and converted to the frequency domain. The locally determined comparison current is firstly used for the fault detection in the evaluation device and is secondly delivered to the communication connection 16 (cf. FIGS. 1 and 2) via the communication device 43.

(41) The communication connection 16 is used by all protective devices 15a, 15b, 15c (cf. FIGS. 1 and 2) to exchange the respective locally determined comparison current values among one another. Given knowledge of the comparison current values from all line ends, the fault detection in the evaluation devices of the individual protective devices 15a, 15b, 15c can be performed.

(42) An approach that can be used to determine the comparison current values on the basis of the locally measured current and voltage values is explained below.

(43) Rearrangement of equations (15) and (16), where the reference location chosen is the beginning of the line and the comparison location chosen is x.sub.0, produces the following set of equations (17):

(44) $\begin{array}{cc}a)\left(\begin{array}{c}{U}_{x0}\left(s\right)\\ I_{x0}\left(s\right)\end{array}\right)=\left(\begin{array}{cc}\cosh \left(\gamma \left(s\right).Math.x_0\right)& Z_c\left(s\right).Math.\sinh (\gamma \left(s\right).Math.x_0\\ \frac{1}{Z_c\left(s\right)}.Math.\sinh (\gamma \left(s\right).Math.x_0& \cosh \left(\gamma \left(s\right).Math.x_0\right)\end{array}\right).Math.\left(\begin{array}{c}{U}_1\left(s\right)\\ I_1\left(s\right)\end{array}\right).& \left(17\right)\end{array}$

(45) This set of equations (17) describes, in the frequency domain, how the signals U.sub.x0 and I.sub.x0 at the comparison location x.sub.0 can be calculated from the signals U.sub.1 and I.sub.1 from the first line end.

(46) In order to be able to ascertain the signals from the sequences of samples, the equations can be transformed into the z domain. This is shown by way of example using the prediction equation for ascertaining the estimated comparison current I.sub.x0(s) at the comparison location from the current and voltage signals I.sub.1(s) and U.sub.1(s) from the first line end:
i.sub.x0(z)=Y.sub.c(z).Math.G.sub.1(z).Math.u.sub.1(z)+G.sub.2(z).Math.i.sub.1(z)  (18).

(47) The transfer functions used in equation (18) are defined as follows in this instance:

(48) $\begin{array}{cc}{G}_1\left(z\right)=F^{-1}\{\sinh \left(\gamma \left(s\right).Math.x_0\right\}& \left(19\right)\\ G_2\left(z\right)=F^{-1}\{\sinh \left(\gamma \left(s\right).Math.x_0\right\}& \left(20\right)\\ Y_c\left(z\right)=F^{-1}\left\{\frac{1}{Z_c\left(s\right)}\right\}.& \left(21\right)\end{array}$

(49) Advantageously, the transfer functions G.sub.1(z) and G.sub.2(z) taking into consideration the propagation constant γ can be mapped by an FIR filter and the transfer function Y.sub.c(z) taking into consideration the characteristic impedance Z.sub.c can be mapped by an IIR filter. Therefore, the estimated comparison current at the comparison location of the line can be determined from the current and voltage signal from the first line end using a logic arrangement as depicted in FIG. 6. The blocks Y.sub.c(z), G.sub.1(z) and G.sub.2(z) in this case are a realization of equations 18-21 and form the comparison current value i.sub.x0 at the common comparison location x.sub.0 as seen from the first line end. The estimated comparison current i.sub.x0 at the comparison location x.sub.0 is formed from the current and voltage values i.sub.1 and u.sub.1 measured at the first line end by applying transfer functions G.sub.1(z) and G.sub.2(z) and also Y.sub.c(z) (e.g. by filtering) and by means of addition in an adder 61. The logic arrangement shown in FIG. 6 may in this case be implemented in the form of a piece of hardware or a piece of processor-executed software in the protective device 15a.

(50) The transfer function Y.sub.c(z) can be calculated from the frequency response of the characteristic impedance Z.sub.c by means of a least squares estimator, for example. If it is possible to dispense with modelling the frequency dependency of the parameters R′, G′, L′ and C′ and the conductance G′ can be assumed to be very small (which is always the case for frequencies below 50 kHz and overhead lines), then the convolution product of Y.sub.c with the local voltage of the line can also be formed using an arrangement shown by way of example in FIG. 10. The transfer functions G.sub.1 and G.sub.2 can be formed by inverse FFT from the terms of the hyperbolic functions using the argument γl.

(51) Therefore, according to equation 18, a travelling wave model of the line is used in order to calculate the comparison current i.sub.x0(k) at a common comparison location on the line from the transient current and voltage signals from the local line end. The variable k in this case represents the k-th sample. On the basis of the comparison currents from all line ends that are determined for the common comparison location, it is then possible to infer the presence of a fault.

(52) The common comparison location in the case of a two-ended line (cf. FIG. 1) may be the middle of the line or the other end of the line in each case, for example. In the case of a three-ended line (cf. FIG. 2), this common comparison location is preferably the point on the line at which the T line branches off. Accordingly, for lines having more than three ends, the comparison location can be stipulated at one of the nodes.

(53) When data windows of approximately 1-2 ms duration are used for measurement, a fault magnitude A for the line can be formed with sufficient accuracy and can be used to assess the presence of a fault. The fault magnitude can be formed according to the following formula, for example:

(54) $\begin{array}{cc}A=\sqrt{\frac{\underset{j=0}{\overset{M}{.Math.}}{\left(\underset{T=1}{\overset{N}{.Math.}}\left(i_{x0}^T\left(k-j\right)\right)\right)}^2}{P.Math.\underset{j=0}{\overset{M}{.Math.}}\underset{T=1}{\overset{N}{.Math.}}{\left(i_{x0}^T\left(k-j\right)\right)}^2+P_0}}.& \left(19\right)\end{array}$

(55) The meanings of the terms in formula 19 are as follows:

(56) a) A: fault magnitude;

(57) b) i.sub.x0.sup.T: comparison current as seen from the line end T;

(58) c) x.sub.0: comparison location on the line;

(59) d) M: number of samples in the data window;

(60) e) N: number of line ends to be taken into consideration;

(61) f) j: count variable for the samples in the data window;

(62) g) T: count variable for the line end to be examined;

(63) h) k: present sample;

(64) i) P: selectable weighting factor;

(65) j) P.sub.0: selectable stabilization value.

(66) The proposed fault magnitude uses the comparison currents that have been determined for the common comparison location x0 by means of measurements at the respective line ends. Its denominator has a “hold component” formed by squaring and summing the individual comparison currents. The result obtained as the hold component is the sum of the rms values of the comparison currents predicted by the ends, weighted with a factor P and stabilized with a factor P0. The numerator of the proposed fault magnitude has a “trip component” formed by adding the individual comparison currents with the correct arithmetic sign and subsequently squaring them. Adding the comparison currents with the correct arithmetic sign in the numerator of the fault magnitude A should yield a value close to zero for a fault-free line. When there is a fault on the line, the prediction model used is no longer concordant with the real line and the prediction error is no longer zero. The formula for calculating the trip signal A provides a signal proportional to the prediction error. If A reaches a value above a prescribed threshold value, a fault on the line to be protected is inferred. The threshold value can preferably assume the value 1.

(67) In the fault magnitude A, ix0T(k−j) represents the instantaneous value of the comparison current at the common comparison location, as calculated from the line end T, with a delay of J samples based on the present sample k. The value M represents the length of the data window. The value N represents the number of line ends taken into consideration.

(68) FIG. 7 depicts a block diagram of a protective device 15a in a more detailed view. The design of the protective devices 15b, 15c may be correspondent.

(69) As already described for FIG. 4, the protective device 15a contains a measuring device 41, an evaluation device 42 and a communication device 43. The measuring device 41 is supplied with locally recorded measurement signals for currents and voltages, said signals being converted into current and voltage values. The local input signals u.sub.1(k) and i.sub.1(k) are preferably filtered by means of a high-pass filter 71 in this case. FIGS. 8 and 9 depict the action of a 1.sup.st order high-pass filter on the current signals from two line ends and on the difference current by way of example. The high-pass filter is assumed to have the rated frequency as corner frequency. A different time constant of the primary transformers in the signal path results in different DC components in the secondary signals of the primary transformers during the current and voltage measurement at the individual line ends. In the example of FIG. 8, the behavior of a linearized transformer is depicted by means of the curve 81 and the behavior of a toroidal core transformer is depicted in the curve 82. The difference signal Δi for the instantaneous values of the two transformers would result in more than a 30% difference current based on the amplitude of the current signal. As can be seen in FIG. 9, even with the simple 1.sup.st order high-pass filtering, the difference current can already be lowered to approximately 2% for the same transformers. Thus, the influence of different transformers in the instantaneous values can be eliminated almost completely by means of high-pass filtering.

(70) The filtered current and voltage values are transferred to the evaluation device 42, where they are processed in the blocks 72a and 72b by means of digital filtering as described in FIG. 6. Adding the filter results determines the comparison current i.sub.x0.

(71) The locally ascertained comparison current value i.sub.x0 is transferred to a transmit buffer store 73a. From there, it is firstly transferred to the communication device 43 in order to be transmitted to the other protective devices 15b, 15c. Secondly, it is supplied to an interpolation device 74.

(72) Moreover, the comparison current value is provided with a timestamp of the local relative time of the protective device 15a and mapped onto the time axis of the Master Clock of the common absolute time kept in sync between the protective devices 15a, 15b, 15c e.g. by means of the PTP protocol. The timestamp of the samples in the local relative time is determined by using a hardware latch 75, which, at the time of sampling of the local samples, stores the present value of the device relative time provided by an internal timer 76. Since the PTP protocol likewise uses the device relative time of the internal timer 76 as the reference to the global time in the device combination (what is known as the “Master Clock”), the local relative time of the protective device 15a can be mapped onto the global time of the master clock in the device combination. This mapping is performed by means of the mapping block 77. The value pairs ix0(k) and the corresponding PTP time are accumulated in the data block that is to be transmitted next. In this case, it is e.g. possible for the comparison current values at a common comparison location and the associated timestamps to be collected for a period of 0.5 ms. An applicable message is sent using the IP network stack 79 via a physical Ethernet port 78 of the protective device 15a, e.g. as a UDP (User Datagram Protocol) data packet. Preferably, a multicast address is used for this purpose in order to send the data packet to all other protective devices 15b, 15c in the differential protection system at the same time. In return, the data packets of the other protective devices 15b, 15c in the differential protection system are received via the Ethernet port 78. All received data packets are buffered in a received buffer store 73b as time/current value pairs for further use.

(73) Using the Ethernet port 78, the protective device 15a is also used to perform the PTP protocol in order to obtain an absolute time that is synchronous between all protective devices 15a, 15b, 15c in the differential protection system. The fundamental way in which the PTP protocol works is explained e.g. on the Internet under the link http://www.ni.com/newsletter/50130/en/. A prerequisite for the use of the PTP protocol is support for the hardware timestamping of the protocol by all network components 17 (cf. FIGS. 1 and 2), involved in the communication between the protective devices 15a, 15b, 15c. This prerequisite is met in modern MPLS (Multiprotocol Label Switching) networks. Many network operators equip their networks with this technology. Thus, a high level of availability of the necessary communication networks between substations can be assumed.

(74) The respectively timestamped comparison current values are provided to the interpolator 74 of the evaluation device 42 from the transmit buffer store 73a and the receive buffer store 73b. Since the times of sampling in the devices may be asynchronous in relation to one another, the comparison current values need to be recalculated for a common reference time. This can be done e.g. using the method for resampling samples described in international patent disclosure WO 2015/110150 A1.

(75) The comparison current values whose timings have been brought into line with one another can subsequently be used to perform a check, in a fault detection block 79, on whether there is a fault on the line. This can be accomplished by using equation 19 to calculate the fault magnitude A, for example. If the fault magnitude exceeds a threshold value 70, which may have e.g. the value 1, then the evaluation device 42 delivers the fault signal F.

(76) The described evaluation can be performed in each protective device 15a, 15b, 15c of the differential protection system separately. The locally generated fault signals are then used to trip the respective local circuit breaker 12a, 12b, 12c.

(77) In the described exemplary embodiment of a method for detecting a fault on a line of an electrical power supply system, the hitherto standard communication is replaced with a UDP/IP communication contemporary for wide area networks. At the same time, a line model on a travelling wave basis is used in order to combine the speed advantage of a wideband IP network in respect of the low communication latencies arising in the network with the speed advantage of an instantaneous value current comparison that becomes possible by means of a travelling wave line model and simultaneous high-pass filtering, and therefore to allow much faster differential protection that avoids the known disadvantages of previous differential protection solutions.

(78) Although the invention has been illustrated and described in more detail by means of preferred exemplary embodiments above, the invention is not limited by the disclosed examples, and other variations can be derived therefrom by a person skilled in the art without departing from the scope of protection of the patent claims that follow.