RU2372701C1 - Method for identification of feeder with single-phase earth fault and automatic load transfer in distribution networks - Google Patents

Abstract

FIELD: electricity.

SUBSTANCE: zero sequence currents without their own measurement and according selection of setpoints serve as monitored electric values, and only direction of supplying feeder current serves as basic parametre, which in its turn increases noise immunity. Continuous monitoring of zero sequence current condition in units, timely selective detection and disconnection of damaged point make it possible to avoid double earth faults. For practical realisation of suggested single-phase earth fault protection method, it is possible to use available devices and functional elements. Operations of measurement, multiplication with further integration, signal coding, information exchange with adjacent units, control of switches may be realised in a comparably simple manner with the help of microprocessor devices. It makes it possible at comparably low material expenses to provide for reliable, selective and fast-acting network protection against single-phased earth fault.

EFFECT: improved reliability of protected network operation.

3 dwg, 2 cl

Description

The present invention relates to electrical engineering and electric power industry and can be used in selective protection devices to determine a feeder with a single-phase earth fault in three-phase distribution branched networks of the middle voltage class with an isolated, compensated or grounded neutral through a resistor, followed by disconnection and, if necessary, trouble-free automatic input reserve.

A known method for determining the connection with a single-phase earth fault (OZZ) on the basis of measuring the zero sequence current (TNI) of each line (Krivenkov VV, Novella VN Relay protection and automation of power supply systems. 1981, Moscow, Energoizdat; Shabad MA Calculations of relay protection and automation of distribution networks. 3rd ed. L., 1985, Energoatomizdat). The selectivity of such protections is provided by detuning from the own capacitive current of the connection. The disadvantage of this method is that in some cases it is impossible to select the protection setpoint from the condition of the relay not working with an external OZZ and the response condition (sensitivity) on the protected connection. In practice, this leads to a "roughening" of the set point and, as a consequence, to false positives of working protections. In other cases, the selection of the setpoint is not possible, since the own capacitive current of the connection is equal to or exceeds the value of the sum of the currents of the intact connections. The tripping of the protection is random in nature, not related to the setting. The use of such protections is advisable to ensure the transverse selectivity of connections with low intrinsic capacitive currents (Shonin OB, Rogov PA Assessment of the selectivity of protection devices from single-phase earth faults in mine networks 6 (10) kV. Collection article "Notes of the Mining Institute ", St. Petersburg: SPGTI (TU), 2008). The current measurement is aggravated by the error of the zero sequence current transformer (TTNP), which depends on the amplitude and frequency of the current. For TTNP type TZLM and TZRL these errors make up 18-52% [Grechukhina I.V., Slyshalov V.K. Calculation of the electromagnetic field of the cable sheath in single-phase mode. IHEU, Bulletin of IHEU, Issue II, Ivanovo, 2001]. In this regard, the choice of setpoint based on averaged values leads to an incorrect assessment of sensitivity and selectivity.

There is a more sensitive and selective way to centrally determine the connection with the OZZ, based on comparing the consumer goods in transient or steady state and determining the damaged connection according to one of the signs, for example, according to the highest current (Fedoseev A.M. Relay protection of electrical systems. Textbook for universities. - M .: Energy, 1976). However, in the case of connection protection with a large intrinsic capacitive current, this method also does not provide selectivity. In addition, false alarms occur due to errors in current sensors. When a remote connection is closed in a branched network under conditions of operation, the protection of several load nodes is found and therefore it is not possible to ensure longitudinal selectivity.

The considered limitations are also valid for methods with inverse time-current characteristics.

There is a known method for determining the connection with OZZ, selected for the prototype, based on the principle of identifying the direction of power and allowing for transverse selectivity of protection (Shabad MA, Calculations of relay protection and automation of distribution networks. - St. Petersburg: PEIPK, 2003, p. 141). This method is implemented in a device for directional protection against single-phase earth faults in a network with isolated or compensated neutral (copyright certificate No. 1275622, H02H 3/16). However, angular errors of the order of 60 ° for TTNP type TZLM and TZRL lead to a shift in the zone of maximum sensitivity and subsequent non-selective operation. Ensuring longitudinal selectivity in an extensive network is impossible, since the conditions for operation in the remote connection and the feeders feeding it coincide.

The following methods are also known:

A method for determining a feeder with a single-phase arc fault to ground in radial distribution cable networks (RF 2254586, G01R 31/08, 12.24.2003), which consists in the fact that in the single-phase fault mode, the transient voltage is recorded on the buses of the main power point and the first half-waves of transient currents in the outgoing feeders and the polarities of the currents are compared, a feeder having a current polarity opposite to the others is considered damaged, and first, recognition of the closure mode based on the analysis of phase voltage on the buses of the main power point, after which the polarities of the first half-waves of the transient currents recorded in the conductors connecting the cable sheaths to the substation ground loop are compared. The disadvantage of this method is that the need to register a zero-sequence voltage requires the use of a voltage sensor, taking into account its errors, taking into account the recovery period of phase voltages, after the elimination of an earth fault, this leads to an increase in the dead zones of protection. Failure (blowing out) of even one fuse of electromagnetic voltage transformers leads to malfunctioning of the protection or its failure. When several electromagnetic measuring voltage transformers are used in the distribution network, fan disconnections of undamaged connections are possible due to ferroresonance phenomena. In the event of an earth fault through the transition resistance, failures of the protection operation may occur due to the low voltage of the zero sequence voltage. In addition, this method does not provide longitudinal selectivity.

A method of protecting a three-phase network from single-phase earth faults (RF 2309507, 05/10/2006), based on the use of a controlled electrical quantity - the pulsating power of the protected lines, allocated by measuring or calculating the instantaneous power of the three phases of the lines and at the same time its average value. To measure or calculate the instantaneous power of the three phases of the lines and its components, phase / line voltages of the network are used. In some cases, it is impossible to select the setpoint for the monitored parameter due to the "operational" instability of the network circuit, expressed in changes in the capacitive currents of the feeders and the total capacitive current of the network, in particular, as a result of disconnecting / connecting lines. The reliability of determining a damaged connection depends on the accuracy of measuring current and voltage under interference in the form of an unbalance of currents and voltages and errors of the corresponding sensors, the values of which can reach up to 60% depending on the signal level and signal frequency.

Other significant disadvantages of these methods include:

1. Inconsistency of the protection devices with automation devices leads to the development of emergency mode.

2. Sensitivity to changes in the network configuration leads to a mismatch of the settings in the original and new networks, arising, for example, due to the operation of automation devices.

3. The coincidence of the response conditions in the remote connection and the supply feeders do not allow selective determination of the damaged connection in the longitudinal section of branched networks, which in turn reduces the reliability of the emergency input of the reserve.

Significant disadvantages of other known methods include:

1. A number of methods are implemented only in networks with isolated neutral.

2. Using as a controlled quantity the steady-state value of the zero sequence current does not make it possible to identify feeders with short-term self-eliminating faults.

3. The need for short-term inclusion in the network of additional elements in accordance with the idea of the method eliminates the automatic operation of the protection and requires the constant presence of maintenance personnel, the use of special switching and measuring equipment, which leads to a relatively large time to determine the damaged connection.

4. The need for an artificial earth fault can cause emergency conditions, especially in branched networks.

5. In a number of ways, the protection operation can be random, since the higher harmonic components of the zero sequence current vary over a wide range due to a number of random factors, for example, the non-sinusoidality coefficient of the mains voltage, resonant processes, etc.

6. In a number of ways, reliable operation of the protections is not ensured in view of the small value of the higher harmonic components of the current of the zero sequence of connections.

7. When using "superimposed current", its flow is possible not only along the damaged line, but also along the undamaged lines with a relatively large capacity, especially when short circuits through a large transition resistance.

These shortcomings reduce the reliability, selectivity, sensitivity of the protection, increase the time to detect damage, increase the likelihood of false positives and emergency input reserve.

The technical result consists in increasing the reliability of the protected network.

The technical result is achieved by the fact that in a changing structure of a branched high-voltage network and changing parameters of a single-phase earth fault, the presence of unbalanced signals of current and voltage sensors of a zero sequence, a feeder with a single-phase earth fault is determined based on the control of electrical quantities in the protected network and their comparison, taking into account phase relations according to the invention: a damaged non-trunk connection in the load node is uniquely determined by comparing occurrences of transient and / or steady-state zero-sequence currents outgoing feeders with a reference signal - cable zero sequence current supplying unit load with subsequent generation of commands to disable the faulty feeder and transmission over the communication channel activation command for disabling parent nodes protections.

The technical result is also achieved by the fact that when damage is detected in the main cable and there is no blocking of operation from the downstream nodes, a team is formed to disconnect this cable from both sides and then connect the downstream node to the backup line.

This allows increasing lateral and longitudinal selectivity, increasing noise immunity, eliminating the need for direct measurement and comparison of currents and voltages, and eliminating zero-sequence voltage sensors.

The essence of the method is based on the conversion of phase relations between the zero sequence currents into logical variables, the exchange of information between load nodes (UN), the identification of a damaged feeder based on the analysis of logical variables. The essence of the method is that the directions of the zero sequence current (TNR) in the outgoing connections with respect to the direction of the current supplying the feeders are encoded as a logical unit when the compared currents are in phase or as a logical zero when they are out of phase or have a low level of compared currents. A feeder is considered to be damaged, the coding of the zero sequence current of which gives one, provided that the coding of the rest gives zero. To identify a short circuit on the trunk (inter-node) cable, an inter-node exchange of information is required, reflecting the distribution of currents in adjacent nodes. In case of damage to the inter-site cable in the "open ring" circuit, a command is sent for two-way disconnection of the feeder and input of the reserve.

The proposed method of protection against OZZ is explained using a fragment of a typical branched distribution network indicating the directions of zero sequence currents with OZZ on line 4 of the load node 4 shown in Fig. 1 and a functional block diagram of the proposed protection method presented in Fig. 2.

A fragment of a typical branched distribution network, shown in figure 1, contains the following blocks:

1, 2, 3.1, 3.2, 4 - load nodes (numbering is given from the power source);

1.0 - feeder feeding the load node 1;

1.1, 1.2, 1.3, 1.4, 1.5 - outgoing lines 1, 2, 3, 4, 5 of the load node 1;

1.6 - backup power supply lines of the load node 1.

Blocks 6 represent the circuit breakers of the respective feeders and intersectional circuit breakers. Blocks 7 represent the zero sequence current sensors of the corresponding feeders. Block 8 presents a damaged feeder with earth fault resistance R _З phase A.

On the example of block 1, it can be seen that each node has one supply feeder, several connections, an outgoing trunk cable and one emergency power line. On the example of block 2 it is shown that the supply line of each node is connected through a circuit breaker, all outgoing lines have circuit breakers and are equipped with zero sequence current sensors. It is assumed that the protection that implements the proposed method is installed in each node. Identification of a damaged connection is made in parallel in all nodes, which allows you to quickly identify a damaged line regardless of the number of nodes in the network. Protection ensures reliable operation in the node with two or more connected outgoing lines.

Functional block diagram of the proposed method of protection contains the following blocks:

1 - unit for measuring and converting currents of the zero sequence;

2 - block multiplication of signals with subsequent integration on a finite time interval;

3 - coding unit;

4 - zero-sequence current unbalance setting;

5 - unit for determining a damaged line node;

6 - block transmission and reception of information;

7 - communication channel;

8 - control unit switches;

9 - block time delay.

The method is as follows.

The inputs of the measurement and conversion unit 1 continuously receive signals from zero-sequence current sensors installed on the line supplying the load node (3I _0in ), and on the N outgoing lines of the load node (3I _0out1 , 3I _0out2 , ..., 3I _0outN ). In block 2, the out-of-phase / out-of-phase currents of the zero sequence of the outgoing lines are detected with respect to the current supplying the line node by multiplying the zero-sequence current signal of each outgoing line with the zero-sequence current signal of the supply line, followed by numerical integration over a finite time interval. The integration time is selected based on the condition of the detuning from interference, for example, caused by switching in the network. A positive signal at the output of block 2 indicates the in-phase currents, a negative signal indicates antiphase. In block 3, the signals are encoded by comparing each with the current unbalance setting (block 4). If the signal value is less than the setting, then it is taken to be zero, if it is more, then the signal is normalized to value 1. In block 5, the damaged node line is determined. A feeder is considered damaged if the encoding of the zero-sequence current signal gives one, provided that the encoding of the remaining current signals gives zero. When a damaged line is detected in the node, block 8 receives a signal with the number of the damaged line and blocks 6 and 7 send a blocking command to activate the protection of the higher node. Block 8 provides a signal for shutting down the detected damaged line in the absence of a blocking command from a subordinate node during a time delay (block 9), due to the time of detecting damage on the downstream node and the transmission time of the command. This method allows you to unambiguously identify and disable only a damaged connection in a branched distribution network. If damage is detected in the interstitial line of the "open ring" circuit, the main feeder is bi-directionally switched off and the reserve is automatically entered through blocks 6 and 7.

Only zero-sequence currents without their own measurement and the corresponding choice of settings serve as controlled electrical quantities, and only the direction of the current of the supply feeder is used as a basic parameter, which in turn increases the noise immunity. Continuous monitoring of the state of zero sequence currents in the nodes, timely selective detection and disconnection of the damaged connection eliminates the possibility of double earth faults.

The proposed method of protection against OZZ can be applied in branched networks of any configuration with different ratios between line capacities, allows you to quickly and selectively identify a damaged line and meets the requirements for protection of mine networks. Fast localization reduces the likelihood of electric shock, the likelihood of explosions, fires. Improving the selectivity and reliability of protection eliminates the development of emergency conditions, unjustified shutdowns of lines, interruptions in power supply. The method allows for trouble-free entries of the reserve. All this helps to increase production efficiency.

For the practical implementation of the proposed method of protection against OZZ can be used known devices and functional elements. To connect the protection to the network, you can use the existing zero-sequence current transformers on the lines. Zero-sequence voltage transformers commonly used in switchgears are not required to implement protection. Measurement, multiplication, followed by integration, signal coding, information exchange with adjacent nodes, control switches can be implemented relatively simply using microprocessor devices. This allows for relatively small material costs to provide reliable, selective and high-speed protection of the network from OZZ.

Based on the analysis of the structures and parameters of high-voltage branched cable networks of the mines of the Vorkutaugol association: Vorgashorskaya, Zapolyarnaya, Vorkutinskaya, distribution network models were developed in the Simulink environment of the MatLab v.7 system. In the developed model, the proposed method for determining the line with OZZ was implemented: a) using mathematical and logical blocks Simulink and b) using the S-function containing the program for detecting damage and controlling circuit breakers, written in Microsoft Visual C ++, according to the claimed method. The latter model works in a discrete time domain and thereby takes into account the specifics of microprocessor protection devices. The developed models have confirmed the claimed advantages of the algorithm for determining the SCR in stationary and transient circuit modes.

An experimental verification of the protection device that implements the proposed method was performed in the laboratory of the Department of Electrical Engineering and Electromechanics of St. Petersburg Polytechnic Institute (TU) named after G.V. Plekhanova. The purpose of the experiment was to check the selectivity of disconnecting a damaged line, the correct supply of blocking signals, and estimating the time to disconnect damaged lines. The protection device is implemented as a controller module, developed on the basis of the AT MEGA 16-16 processor.

The main parameters of the module controller:

- eight analog inputs, with a current measuring range from -200 mA to +200 mA;

- the device is sensitive to a current of 2 mA, taking into account the analog conversion of current-voltage to the corresponding voltage range 0-5 V and subsequent analog-to-digital conversion (20 kHz, 8 bits);

- eight discrete outputs, type "dry contact" at a voltage of 100 V and a switching current of 1 A;

- the execution time of one operation is 125 ns;

- RAM 16 kB;

- output of measurement data and intermediate calculations to a PC using the USB interface; data is displayed through a shell program developed for the controller module in Delphi v.7.0; indication of the status of switching devices is carried out using the information panel of the module controller;

- the ability to integrate into a single distribution network protection system based on the Ethernet interface;

The program that controls the operation of the controller module is written in the IAR Embedded Workbench language, downloaded and compiled using the PonyProg2000 program.

The signals of the current sensors supplied to the input of the controller module were reproduced by digital-to-analog converters (5 kHz, 8 bit) of the National Instruments NI USB 6218 data acquisition board according to the results of network simulation in a Simulink MATLAB environment. The same board was used to oscillograph test results. The output and measurement of signals were carried out in accordance with a program developed in the LabView v.8.5 environment.

Figure 3 shows the waveform obtained using the data acquisition board in one of the experiments, in which 1 is a signal-command to turn off the damaged line circuit breaker at 5 V, 2 is the zero sequence current of the damaged connection and 3 is undamaged.

Tests have confirmed that the protection device gives timely and correct commands to disconnect the damaged connection in the load node and timely and correct commands to block higher-level protections during short circuits in various feeders of the branched network.

Selective detection of a damaged connection was carried out 10-20 ms after the occurrence of damage, which corresponds to the disconnection of the damaged line already in the middle or at the end of the transient circuit. The protection device can operate both autonomously in the load node, and as part of an automatic control system in conjunction with the upper level.

The data and information confirm the possibility of industrial implementation of the invention.

Claims (2)

1. The method of relay protection against single-phase earth faults, which consists in the fact that based on the control of electrical quantities in the protected network and their comparison, taking into account phase relations, a feeder with a single-phase earth fault is determined, characterized in that in conditions of a changing structure of a branched high-voltage network and changes in the parameters of a single-phase earth fault, the presence of unbalanced signals of the current and voltage sensors of the zero sequence, a damaged non-trunk connection in the load node is one Significantly determined by comparing the directions of transient and / or steady-state currents of the zero sequence of the outgoing feeders with the reference signal - the zero-sequence current of the cable supplying the load node, with the subsequent development of a command to disconnect the damaged feeder and transmit the command through the communication channel to block the operation of the protection of the higher nodes. 2. The method of relay protection against single-phase earth faults according to claim 1, characterized in that when damage is detected in the main cable and there is no actuation lock from downstream nodes, it generates a command for two-way disconnection of this cable, followed by the connection of the downstream node to the backup line.

Images