JPH0245414B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for protecting a transformer using a relay, and particularly to a method for reliably detecting an internal fault in a transformer.

[Prior Art] Conventionally, differential relays have been widely used to detect internal faults in transformers connected to power systems.

A transformer is a means of passing energy, and installing a current transformer at the entrance and exit of the transformer, combining these detected currents to obtain a differential current, and detecting the presence or absence of an accident based on the presence or absence of this differential current is as follows: This is to see whether Kirhoff's first law holds true in electrical circuits, and is considered the most rational. In such differential protection of a transformer, normally there is no problem, except that the excitation current to establish the voltage (magnetic flux) of the transformer is almost zero and flows as a differential current as an error, but when the transformer is turned on, the Magnetic flux (voltage) of the device
Until this is established, a considerable amount of transient excitation current (excitation inrush current) flows and is indistinguishable from the fault current that would normally flow during an accident.

In order to avoid this problem, a so-called harmonic suppression method has conventionally been used that focuses on the waveform of the transient excitation current. That is, if the transient excitation current that flows when the transformer is turned on is analyzed by harmonics, there are many harmonic components, and this is used to prevent malfunctions caused by the transient excitation current that flows when the transformer is turned on.

However, in the conventional method using a differential relay as described above, a filter is required to process the detection signal, and a time delay is added to avoid the transient phenomenon, so the response to the fundamental wave is delayed and the relay operation time becomes longer. Furthermore, the reliability of detecting accidents inside the transformer is not sufficient.

To solve this problem, for example, Japanese Patent Application Laid-Open No. 55-18856
As disclosed in the above publication, a method has been developed to determine the internal induced voltage of a transformer, to speed up the operation of the relay, and to ensure the detection of faults inside the transformer. The technical contents described in JP-A-55-18856 are the primary side voltage, current, secondary side voltage of the transformer,
The current is taken in, the overall voltage is determined according to the number of windings on the primary and secondary sides given in advance, and by comparing these overall voltages on the primary and secondary sides, it is possible to determine whether the transformer is turned on or if there is an internal fault in the transformer. It is used for identification.

Then, the leakage impedance existing on both sides of the internal induced voltage of the transformer is determined, and the internal induced voltage is determined in terms of an electric circuit from this leakage impedance and the overall voltage. Furthermore, a method of determining an internal fault based on fluctuations in the difference between the internal induced voltage based on electrical constants and the internal induced voltage based on magnetic characteristics can also be considered. In this case, calculate the above voltage difference on both the primary side and the secondary side, ω ₁ = L ₁ (dI ₁ /dt) + R ₁ I ₁ − V ₁ + N ₁ (dΦm / dt) ε ₂ = L ₂ ( The variation of dI ₂ /dt) + R ² I ₂ − V ₂ +N ₂ (dΦm/dt) is independently monitored. However, in reality,
The leakage impedances L ₁ and L ₂ cannot be unambiguously separated as shown in the above equation from the results of characteristic tests with the transformer stopped, and even if they could be separated, they would contain errors that cannot be ignored. There is a possibility that there are. Therefore, it is difficult in practice to monitor the difference between the two voltages on both the primary side and the secondary side to determine an internal fault.

[Problems to be Solved by the Invention] As described above, the first method of conventional transformer protection methods has the problem of delayed operation.
Further, in the second method, since the internal induced voltage is obtained from an electric circuit, there is a problem that even if a fault occurs in the internal induced voltage, it cannot be distinguished from a fault in each leakage impedance. Furthermore, in the third method, the primary and secondary side operators ε ₁ are calculated based on the leakage impedance that may include errors.
and ε ₂ , it is not possible to reliably determine internal accidents, which poses a problem that it is not practical.

This invention was made to solve the above-mentioned problems, and aims to provide a transformer protection method that can operate quickly and reliably detect internal faults in the transformer.

[Means for solving the problem] The transformer protection method according to the present invention detects current values I ₁ and I ₂ and voltage values E ₁ and E ₂ on the primary side and secondary side of the transformer, respectively. , the normalized algebraic sum current I _D is determined from the current values I ₁ and I ₂ , and based on this algebraic sum current I _D and the excitation characteristics of the excitation impedance of the transformer, the inside of the transformer based on the magnetic characteristics is determined. induced voltage
EΦ is calculated, and on the other hand, the voltage values E ₁ and E ₂ on the primary and secondary sides of the transformer are compensated with the current values I ₁ and I ₂ , and the 2 obtained on the primary and secondary sides is
By averaging the two voltage values to find the transformer's internal induced voltage E _I based on electrical constants, and comparing the internal induced voltage EΦ and E _I found in this way,
This system is designed to detect internal accidents in transformers.

[Function] In this invention, the magnetic flux function, that is, the excitation characteristic, of the excitation impedance corresponding to the internal induced voltage of the transformer is determined in advance, and the internal induced voltage EΦ of the transformer is detected as a magnetic circuit, and the internal induced voltage EΦ of the transformer is detected as an electric circuit. The internal induced voltage E _I is determined by averaging the two voltage values obtained for the primary and secondary sides, and by comparing these EΦ and E _I , it is possible to reliably detect fluctuations in the voltage inside the transformer. Determine accidents.

[Example] Hereinafter, an example of the present invention will be described with reference to the drawings. FIG. 1 is a block circuit diagram for explaining the transformer protection method according to the present invention. In the figure, 1 is a transformer to be protected, and 2 and 3 are current transformers provided on the high voltage side and low voltage side of the transformer 1, respectively. The equivalent circuit of the transformer 1 is as shown in FIG. 2, and the leakage impedance 31 on the high voltage side and
an excitation impedance 32 connected to the excitation impedance 32 to flow an excitation current, a section 33 connected in parallel to this excitation impedance 32 to produce an ideal transformer action, and a leakage impedance 34 on the low voltage side connected to this ideal transformer section 33. It is made up of. and,
As is well known, the operation of the main body of the transformer 1 is such that when a voltage is applied to the primary terminal 1a on the high voltage side, current flows to the primary and secondary sides depending on the load on the low voltage side. At this time, a drop occurs due to the internal leakage impedance 31, and after subtracting that amount, the voltage is transformed by the excitation impedance 32, and after a drop occurs due to the secondary side leakage impedance 34, it is output to the secondary side terminal 1b. Voltage is starting to appear.

In FIG. 1, reference numerals 39 and 40 are voltage transformers provided on the high voltage side and low voltage side of the transformer 1, respectively. 41 to 47 are current transformers 2, 3 and transformer 39,
It is an auxiliary transformer that transforms 40 currents and voltages,
41 and 43 are for high voltage side current I ₁ , 42 and 44
is for the current I ₂ on the low voltage side, 45 and 46 are for the voltages E ₁ and E ₂ on the high and low voltage sides, respectively, and 47 is for the algebraic sum current _ID . Level converters 51 to 57 are provided corresponding to the auxiliary transformers 41 to 47, respectively, and adjust the output level of each of the auxiliary transformers 41 to 47.

Sample and hold amplifiers 61 to 67 sample and hold the values of I ₁ , I ₂ , I ₁ , I ₂ , E ₁ , E ₂ , and _ID via each level converter 51 to 57;
1 to 74 are each sample hold amplifier 61 to 64
Variable gain amplifiers 75 to 77 are buffer amplifiers that receive the output of the amplifier and amplify it with an appropriate gain.
Incidentally, the gain of each variable gain amplifier 71 to 74 is changed by a command output from a microprocessor 84, which will be described later.

81 switches the output of each amplifier 71 to 77.
This is a multiplexer that inputs to the AD converter 82. The AD converter 82 has the accuracy and speed necessary for the above operation. 83 is a register that temporarily stores the output of the AD converter 82. 84
is a microprocessor and AD converter 8
In addition to receiving input from 2 and performing conversion, calculation, and judgment processing, sample and hold amplifiers 61 to 67,
Variable gain amplifiers 71 to 74, multiplexer 8
1. A timing control signal is sent to the AD converter 82 and the like, and an internal fault in the transformer 1 is determined based on the above processing result, and the determination signal is output to the output logic 85.

Next, the operation of the embodiment of the present invention will be described with reference to the detailed flowchart of the operation of the microprocessor 84 shown in FIGS. 3a and 3b. The following flow is processed by a program start command issued periodically.

First, the high voltage side current I ₁ and low voltage side current I ₂ of the transformer 1 are read from the auxiliary transformers 41 and 42 (block 1
01), respectively multiplied by the CT ratio, CT, that is, current transformer 2
After obtaining the primary current, multiply by the transformer transformation ratio to obtain the normalized primary current. This is called correction of current I ₁ (block 102) and correction of current I ₂ (block 103). Next, proceeding to block 104, the algebraic sum current I _D is calculated from the normalized I ₁ and I ₂ by I _D =I ₁ +I ₂ . The value of this _ID is determined by the auxiliary transformer 47
It is more accurate than the value detected from , since the corrected I ₁ and I ₂ are used. Specifically, the normalized algebraic sum current I _D is I _D = I ₁ + (N ₂ / N ₁ ) I ₂ where N ₁ , N ₂ are expressed as the number of turns on the primary side and secondary side. Ru. Furthermore, in reality, since I ₁ and I ₂ are vectors of opposite polarity, the normalized algebraic sum current I _D becomes a known differential current, and usually takes a value close to 0.

This algebraic sum current I _D is calculated from the transformer 1 shown in Figure 2.
In terms of an equivalent circuit, it is approximately equal to the excitation current flowing through the internal excitation impedance 32. At this time, there is a small amount of leakage due to ground capacitance, but it can be ignored because it is small compared to the excitation current. Since the algebraic sum current I _D is equal to the excitation current, from the excitation characteristics given in advance using this _ID , the internal induced voltage EΦ, that is, the voltage V ₃₂ across the excitation impedance 32, can be calculated as EΦ=f(I _D ). Calculate (block 105). Here, f corresponds to the impedance value of the excitation impedance 32, and as is known, it varies nonlinearly as a function of the current value of the algebraic sum current _ID .

In this block 105, I D0 close to the algebraic sum current I _D and EΦ ₀ corresponding to this I _D0 are selected from a table given in advance, and EΦ=EΦ ₀ +f _' (I _D0 ) (I _D −I _D0 ) There is no problem in practical use even if the calculation is simplified by
This is because in the region where I _D is a large value, f′(I _D ) is relatively small, and even f′(I _D )△I _D , the error of EΦ with respect to the algebraic sum current I _D becomes small. It is from.

Next, the primary voltage E ₁ from the auxiliary transformers 45 and 46,
Read the secondary voltage E ₂ and also the auxiliary transformer 43, 4
₄ , the tap position of the tap changer is read, and the values of the numbers of turns N ₁ and N ₂ of the primary and secondary windings in the transformer ₁ are read from the given table (block 106).

Next, proceeding to block 107, the voltage on the high voltage side is
From E ₁ and current I ₁ and leakage impedance Z ₁ (corresponding to leakage impedance 31 in Figure 2), the internal induced voltage of transformer 1 based on the electrical constants on the primary side can be calculated.
Find Ee ₁ from Ee ₁ = E ₁ − Z ₁ I ₁ .

Next, in block 108, the voltage on the low voltage side is
From E ₂ and current I ₂ and leakage impedance Z ₂ (corresponding to leakage impedance 34 in Figure 2), the internal induced voltage of transformer 1 based on the electrical constants on the secondary side can be calculated.
Find Ee ₂ from Ee ₂ = E ₂ + Z ₂ I ₂ .

The internal induced voltages Ee ₁ and Ee ₂ thus obtained for the primary and secondary sides are corrected by cooling them using the transformer turns ratio (high voltage side: N ₁ , low voltage side: N ₂ ), and are calculated as Ee ₁ / Confirm N ₁ ≒Ee ₂ /N ₂ (block 109).

Note that standardization using the unit method (pu) is block 1.
07 and 108, in block 109 it is verified that Ee ₁ ≈Ee ₂ . It goes without saying that if there is a large difference between the two, it is assumed that there is a failure in the measurement system and a known alarm device (not shown) is activated.

Here, if the two induced voltages are equal, it can be assumed that there is no internal fault, so the most likely value for the internal induced voltage is found by averaging the two, that is, E _I = k・Ee ₁ + l・Ee ₂ (block 110). Here, Ee ₁
and Ee ₂ are values standardized by the unit method, and the values of the weighting variables k and l are, in common sense, k=l=1/2.

In this way, the internal induced voltage E _I obtained by applying current compensation to the terminal voltages E ₁ and E ₂ and the measured current I ₁
and the internal induced voltage EΦ based on the magnetic characteristics obtained from the normalized algebraic sum current I _D of I ₂ (block 111), they should of course be equal unless there is an accident. In addition, the internal induced voltage E _I based on the electrical constant does not decrease unless there is an accident and is constant, so if we select a value k ₀ that takes into account the variation and compare them (block 112), E _I will be constant. The value k is greater than or equal to ₀ . When either of these two conditions is met, the microprocessor 84
determines that there is no accident inside the transformer and outputs a suppression output of "1" (block 113).

As mentioned above, the current on both the input side and the output side,
From the measured voltage values I ₁ , I ₂ , E ₁ , E ₂ , find the internal induced voltage EΦ based on the currents I ₁ , I ₂ and the internal induced voltage E _I based on the voltages E ₁ , E ₂ , and calculate these. In addition to comparison, faults are determined by detecting fluctuations, which prevents unstable detection operations that occur in transient excitation states and realizes a transformer protection method with excellent responsiveness and reliability. I understand. In addition, in tests with the transformer in rest state, the electrical constants of the leakage impedances 31 and 34, which cannot actually be separated, are
By averaging and rounding the values obtained from the primary and secondary sides, errors are canceled out and the internal induced voltage E _I based on electrical constants is determined, making practical and reliable fault detection possible. It can be carried out.
If an accident occurs on the leakage impedance 31 and 33 sides, the internal induced voltage based on the electrical constants
If E _I fluctuates and a fault occurs on the excitation impedance 32 side, the internal induced voltage based on the magnetic characteristics
EΦ will fluctuate. Therefore, by comparing the two, internal accidents and their abnormal parts can be reliably detected, and the transformer can be promptly protected.

FIG. 3b is a flowchart following block 112 in FIG. 3a, in which the microprocessor 84 detects the second harmonic component from the detected current and voltage values to prevent malfunctions in abnormality detection at the time of power-on. This is to explain the case where it is done. This concept is a conventional method known as the so-called second high frequency suppression function, but since the harmonic analysis is performed by the arithmetic processing of the microprocessor 84 without using a filter, the response to the fundamental wave is reduced. This is quick and provides even greater effects in addition to the operation shown in FIG. 3a.

First, at the time of reading the above-mentioned currents I ₁ , I ₂ and E ₁ , E ₂ , the time of voltage 0, which is convenient for the following harmonic analysis, is read (block 114), and the values of I ₁ and I ₂ at this time are read. Let be (I ₁ ) ₀ and (I ₂ ) ₀ . Using this, (I _D ) ₀ = (I ₁ ) ₀ + (I ₂ ) ₀ ·N ₂ /N ₁ is calculated (block 115).

Below, (I _D ) ₀ , (I _D ) ₁ ..., (I _D ) ₁₁ are respectively ω
t=
Sequentially read and store every 2π/12 (block 11
5,116,117). These are y ₀ , y ₁ ,...,
If y ₁₁ , the DC component a ₀ , the fundamental wave components a ₁ and b ₁ , and the second high frequency components a ₂ and b ₂ are respectively a ₀ = (y ₀ +y ₁ +y ₂ +...+y ₁₁ )/12 a ₁ = [y ₀ − y ₆ + (y ₁ − _{y 5} − y ₇ + y ₁₁ )/2 +√3・(y ₂ − y ₄ − y ₈ − y ₁₀ )/2]/6 b ₁ = [y ₀ +y ₆ +y ₃ −y ₉ +(y ₂ +y ₄ −y ₀ −y ₁₀ )/2 +√3(y ₁ −y ₅ −y ₇ −y ₁₁ )/2]/6 a ₂ = [y ₀ +y ₆ −y ₃ −y ₉ + (y ₁ −y ₂ −y ₄ +y ₅ +y ₇ −y ₈ −y ₁₀ −y ₁₁ )/2]/6 b ₂ = [y ₀ +y ₃ +y ₆ +y ₉ +√ 3(y ₁ +y ₂ −y ₄ −y ₅ +y ₇ +y ₈ −y ₁₀ −y ₁₁ )/2]/6 (block 118).

Next, from among the DC components obtained in this way, the DC component _a0 is compared with a predetermined value k (block 119),
When the DC component _a0 is large, this is detected and the process proceeds to block 123, where a suppression output of "1" is generated.

Also, the absolute value of each harmonic component ai, bi is compared with a predetermined value k' (block 120), and if these are all small, the excitation current itself is small, so the process proceeds to block 123 and the suppression output is set to ``1''. ” may occur.

Furthermore, the ratio of mutually orthogonal components of each harmonic component ai and bi is calculated and compared with a predetermined value k' (block 12).
1), for example, if a ₂ < k ₁ b ₁ b ₂ < k ₁ a ₁ , the process proceeds to block 123 and the suppression output is "1".
Output. At the same time, the suppression output "1" is also output when a ₂ <k ₂ a ₁ b ₂ <k ₂ b ₁ is satisfied.

Furthermore, the effective value I〓f ₂ of the second harmonic component and the effective value I〓f ₁ of the fundamental wave component are respectively calculated as follows: I〓f ₂ =√ ₂ ² + ₂ ² I〓f ₁ =√ ₁ ² + ₁ ² and compare it with the predetermined value k'' (block 12
2) If I〓f ₂ ≧k ₁₂ I〓f ₁ , it is assumed that the second harmonic component is large, and the process proceeds to block 123, where a suppression output of "1" is generated.

If none of the comparison conditions in blocks 119 to 122 are satisfied, it is determined that an internal accident has occurred, and the process proceeds to block 124, where an operation output of "1" is generated.

In this way, malfunctions of the differential relay caused by transient phenomena when the transformer is healthy are completely prevented.

Further, according to the present invention, when an internal fault occurs in the transformer 1, the voltage drops and current flows;
Internal induced voltage assumed to occur based on I ₁ and I ₂
EΦ and the internal induced voltage E _I estimated based on the voltages E ₁ and E ₂ are not equal (EΦ>E _I ), and the equilibrium relationship (block 111) is broken. As a result, the process does not proceed to block 113 and the suppression output is ” does not occur. Therefore, even if it is difficult to find the second harmonic component in the differential current, the process proceeds to block 124 and produces the operating output "1". In the above embodiment, it goes without saying that the leakage impedances Z ₁ and Z ₂ used in blocks 107 and 108 in FIG. 3a are values corrected by on-line measured values. That is, the measured value E ₁ which can vary over time,
E ₂ , I ₁ , I ₂ , N ₁ and N ₂ , and the leakage impedance Z ₁₂ between the primary and secondary tested and measured before the transformer starts operation (online) (a value standardized by the unit method) ), for example, Z ₁ = −I ₂ · Z ₁₂ / (I ₁ − I ₂ ) + (E ₁ − E ₂ ) / (I ₁ − I ₂ ) Z ₂ = I ₁ · Z ₁₂ / It is calculated from I ₁ − I ₂ ) − (E ₁ − E ₂ )/(I ₁ − _{I 2} ). Therefore, Z ₁ and Z ₂ can be separated almost accurately. Here, Z ₁ +Z ₂ =Z ₁₂ .

Also, in this case, in blocks 107 and 108, voltages E ₁ and E ₂ and currents I ₁ and I ₂ are standardized to be expressed in unit system (Pu) using the number of turns N ₁ and N ₂ . Therefore, in block 109, Ee ₁ ≈Ee ₂ is determined. Note that the block 109 is simply a block for checking the measurement system and is not a feature of the present invention, so it may be deleted.

Furthermore, in block 110, the case where the internal induced voltages Ee ₁ and Ee ₂ are averaged by setting the values of k and l as k=l=1/2 was shown, but the values of k and l are as follows: k+l=1 It goes without saying that each can be set arbitrarily within the range that satisfies the following.

[Effects of the Invention] As described above, according to the present invention, the current values I ₁ and I ₂ and the voltage values E 1 and E ₂ on the primary and secondary sides of the transformer are detected, and the current value I ₁ is detected _. , I ₂ algebraic sum current I _D
The internal induced voltage EΦ _{of the} transformer _is determined based _on By determining the internal induced voltage E _I and comparing the two internal induced voltages EΦ and E _I , we are able to determine an internal fault in the transformer, thereby ensuring that no fault occurs due to a failure in the excitation impedance inside the transformer. This has the effect of providing a transformer protection method that can be detected.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram for explaining one embodiment of the present invention, FIG. 2 is an equivalent circuit diagram of a transformer,
FIGS. 3a and 3b are flowcharts for explaining the operation of an embodiment of the present invention. 1...Transformer, 2, 3...Current transformer, 39,
40...Voltage transformer, 84...Microprocessor, _I1 ...Primary side current value, _I2 ...Secondary side current value,
E ₁ ...Primary side voltage value, E ₂ ...Secondary side voltage value, I _D ...
...Algebraic sum current, EΦ,E _I ...Internal induced voltage, 10
5...Block for calculating EΦ, 110...Block for calculating E _I , 111... Block for comparing EΦ and E _I. In the figures, the same reference numerals indicate the same or corresponding parts.

Claims (1)

[Claims] 1. Calculate the normalized algebraic sum current I _D from the current values I ₁ and I ₂ on the primary and secondary sides of the transformer,
Based on this algebraic sum current _ID and the excitation characteristic of the excitation impedance of the transformer, the internal induced voltage EΦ of the transformer based on the magnetic characteristics is determined, The respective voltage values E ₁ and E ₂ are compensated by the current values I ₁ and I ₂ , and the values obtained from the primary and secondary sides are averaged to calculate the internal induction of the transformer based on electrical constants. A method for protecting a transformer, characterized in that an internal fault in the transformer is determined by determining a voltage E ₁ and comparing the internal induced voltage EΦ and E _I.