Method of measuring distance to point of single-phase short circuit onto ground in radial distribution circuits

FIELD: electrical power engineering.

SUBSTANCE: method concludes in registering voltage of damaged phase at buses of power supply in digital form and subsequent analyzing that voltage. Voltage to be measured is standardized from the moment of beginning of short-circuit in relation to breakdown voltage and sums of standard deviations

of measured standardized voltage u_m,i relatively to multiple calculated voltage curves u_c,i,k derived preliminary in digital form for damaged chain of lines of the circuits. Numerical modeling of process of single-phase short-circuit is carried out for different distances to supply point. Distance to point of short-circuit is found from minimal standard deviation corresponding to specific calculated curve without avoiding switching power supply of damaged line off.

EFFECT: improved precision.

3 dwg

The present invention relates to power engineering and can be used to determine distances to single-phase earth faults in distribution networks a radial structure.

There is a method of determining the distance of a single-phase earth fault in radial distribution networks /1/, which is based on measuring the time between the moment of sending in-line probing of the electrical pulse and the time of arrival at the start line of the pulse reflected from the circuit. Sending in a line pulse, measure the interval t_pthe time double the mileage of this impulse to the point of snapping. Distance to circuit found as l₃=vt_p/2, where v is the velocity of propagation of the pulse along the line.

The use of this method in automatic mode (when the measurements are performed on lines that are included under the operating voltage of the electrical network) is extremely difficult for a radial network, from the supply when the tire leaves a lot of lines, as in this case, there are multiple reflections from the intact lines, superimposed on the useful signal and distort it. The complexity of this method lies in the fact that the voltage measurement should be made during the arcing circuit - only in this case, the information is significant. The mountain is e same arc takes fractions of a millisecond, making it difficult to obtain the desired measurement data.

In addition, there is a method of determining the distance of a single-phase earth fault in radial distribution networks /2/taken as a prototype and used the network of lines, according to which to perform the measurement of time between the time (t_n) came to the beginning of the front line of a high voltage wave resulting from an electrical discharge in place of the circuit at a distance of l_Cfrom the beginning of the line, and the secondary moment of the parish (t_to) wave front after two reflections (at the beginning of the line and place the circuit). The time interval is defined as Δt=(t_to-t_n)=2l_C/v. From the last expression is determined by the distance to the circuit (considered similar to the first method) l_C=Δtv/2.

In this way the same way as in the above, there are reflected waves from the ends intact lines. With respect to cable networks distributed along the lines of the signal is greatly distorted because of the heterogeneity associated with the connection lines of different types and cross-sections and along the couplings, which makes the determination of the time interval Δt extremely difficult and sometimes impossible. This method is also difficult to apply in a radial network of lines, finding the, for example by transferring them under the operating voltage.

Analysis of the above prior art suggests that the task of the invention is to provide a simpler method of determining the distance of the circuit to ground when not disabled (under voltage) accession in radial distribution networks.

This is achieved by the known method of determining the distance of a single-phase earth fault in radial distribution networks based on the Desk in digital form of the faulted phase voltage on the power supply buses and subsequent analysis, the measured voltage (u_u,i, i is the number of measurement points) from the starting circuit with the first maximum of the derivative normalized with respect to the breakdown voltage (u*_u,ithe normalized voltage) and find the sum of the square errors (RMSE)(where n is the number of measurement points) measured normalized voltage u*_u,iin relation to the calculated curve of the voltage u_p,i,k(k is the number of the calculated curve), previously obtained in digital form, for invalid feeder of the network as a result of numerical simulation of single-phase circuit at different distances from the food item l_C,k; minimum standard deviation, the corresponding definition is calculated by (k-th) curve, determine the distance to the point of snapping.

Figure 1 shows one of the possible schemes of the network that implements the proposed method; figure 2 - examples of the estimated voltage curves for different distances l_Cto the point of earth fault in a radial distribution network, consisting of 10 cable lines of different length; figure 3 - standard deviation of the calculated curve of the voltage in relation to the curve of the measured voltage.

The device (1) contains a network diagram in which the main step-down substation (SDS) item power supply 1 through the transmission line connected 2-7 transformer substation 8-12 (TA) and distribution substation 13 (RP).

The method is as follows.

For the electrical network on the basis of all its parameters (topology, lengths and parameters of the lines) are preliminarily calculated transient single-phase earth fault (only the initial part of the order of hundreds of microseconds) for all chains and distances l_C,kfrom the start - point power 1 (AOP) to the endpoint of the branch-TP 8-12. For each such branch will receive in digital form m calculated waveforms (curves) transients (u_p,i,ki∈[1, n] is the number of design points, k∈[1, n] is the number of calculated waveforms) at various distances l_C,kfrom the GPP, the number to the verge (m) depending on the desired accuracy of determining the distance l _C,kmay vary within wide limits, for example from 20 to 30. The calculated waveforms obtained using computer simulation tools, for example using software systems EMTR /3/, EMTDC/PSCAD /4/, PSPICE /5/ etc. Modeling of processes in the network may also be physical or by using analog, analog-digital systems, however, the calculated data for further analysis is translated into a digital form.

Representing physical processes in the network in the form of waves of voltage and current, ground fault can be interpreted as a distribution from the point of damaging wave voltage (and current) on the shape close to rectangular, which due to its multiple reflections, refractions and overlays on the forced component of the voltage, generates a certain image of the transitional bus bar voltage GPP (1). The appearance of this voltage depends on the network topology, parameters and lengths of lines.

In the process of grounding in the real network on one of the outgoing lines 2-7 (figure 1) on the tires of the supply point register 1 voltage of the faulted phase (for example, by monitoring overvoltage /6, 7/) and allocate connection with the closure of any of the known methods, by implementing, for example, the principle of separate fixing polarities of the first half-wave high-frequency zero-sequence currents /8/ or comparing the I angles of phase shift between voltage and current zero-sequence on all outgoing lines /9/. To transmit information about the damaged feeder extending from RP (13), using telemetry.

Calculated curves of the transition process circuit previously obtained for damaged branches, and measurement data, if necessary, lead to a single step of sampling time (h), performing the interpolating functions. As the first measurement and calculation points take those in which there is a maximum decrease (in absolute value) voltage of pre-steady-state conditions, i.e. the point with the maximum of the derivative of the voltage (actually for discrete measurements with some fairly small discretization step derivative represented as finite differences of the first order). As a break in the General case may not occur at the maximum phase voltage (U_FM), for which the calculated stress curves, the measured voltage normalized with respect to the breakdown voltage (U_n)and calculate standard deviation(n is the number of measurement and calculation points). The minimum value of standard deviation determine the distance to fault damaged branches.

Figure 2 shows examples of the calculated waveform voltage in the primary is the moment of closing, obtained using the software /3/ for the radial network of ten cable lines long 0,5, 1, 2, 3, 4, 5, 6, 7, 9, and 10 km (the curves in figure 2, a-C correspond to the distances to the point of snapping 0,5, 1, 2, 3, 4, 5, 6, 7 km). The ten-cable consisted of two sections with variruete lengths, at the junction of which was modeled circuit. Figure 3 graphs shows the standard deviation obtained for different distances to the circuit and showing the reduction in RMSE as it approaches the “true” location circuit with measurement data (as “true” designated circuit with measurement data taken appropriate calculated values - figure 3 these distances are assumed to be 1, 3 and 7 km). Analysis charts allows you to easily set the distance to the point of snapping.

The sampling time (h) for reference points and measuring, as well as the computation time SKO Δt=nh depend on the lengths of the feeders in the network: from the minimum (τ_min) and maximum (total for all branches - τ_max) the constant of propagation of electromagnetic waves τ=l/v, I - length branches. Approximately you can acceptand the time interval to calculate the RMSE slightly greater than double the mileage of the electromagnetic the waves on the highest branches, ie Δt>2τ_max.

Thus, determination of the distance to the circuit without disconnecting the accession is simply by finding the minimum standard deviation of the measurement curve of the voltage of the faulted phase with respect to the set of the calculated curve in the initial stage of the transition process, obtained for different distances to the point of earth fault.

List of used sources

1. Measurement of the distances to faults on overhead and cable power lines and communication/ Velyaminovsky, Apostacy, Neastenam, Vccasino. -CNIEL, 1954, v.2.

2. Shalit G.M. Determination of faults in electrical networks, 1982.

3. The Electromagnetic Transients Program (EMTP). Rule Book 1, 2. DCG/EPRI, 1996.

4. A.M.Gole, O.B.Nayak, T.S.Sidhu and M.S.Sachdev. A Graphical Electromagnetic Simulation Laboratory for Power System Engineering Programs // IEEE Transactions on Power Systems, vol. 11, no. 2, May 1996, pp. 599-606.

5. Razevig E System circuit simulation and PCB design Design Center (PSpice), -M: IC Press, 1996.

6. Kutasov V.E., Larionov V., Ovsyannikov A.G. ON the results of monitoring strain in single-phase arcing fault to ground in the distribution cable networks // Electric station, No. 8, 2002.

7. Kadom C.P., Kutasov V.E., Lavrov Y.A., Ovsyannikov A.G., Sakhno, VA for detection and monitoring of cable networks medium CL is Sov voltage/ Electrical engineering, No. 11, 2000.

8. Laidudele, SUV, Whistance. Comprehensive protection against earth fault // Electric station, 1981, No. 7.

9. Reuters A.I., Batelco A.I Determinant damaged line single-phase earth faults in networks 6-35 kV with insulated neutral // Energetik, 2002, No. 4.

The method of determining the distance of a single-phase earth fault in radial distribution networks based on the Desk in digital form of the faulted phase voltage on the power supply buses and the subsequent analysis of this voltage, characterized in that in the above analysis of the measured voltagefrom the starting circuit with the first maximum of the derivative normalized relative to the breakdown voltage,(where- maximum phase voltage) and find the standard deviation(where n is the number of measurement and calculation points) measured normalized voltagein relation to the numerous theoretical curves stress,k=1,...,m, m≅20...30 previously obtained in digital form for invalid feeder of this network in the numerical modelirovaniya single-phase circuit at different distances from the point of supply to the point of snapping, minimum standard deviation corresponding to the calculated k-th curve, determine the distance to the point of snapping.