JP4998706B2 - Transformer internal abnormality diagnosis method - Google Patents

Description

  The present invention relates to an internal abnormality diagnosis method for a transformer for detecting a partial discharge (internal abnormality) generated inside the transformer due to aged deterioration of the transformer or the like.

Conventionally, as a method for diagnosing the internal abnormality of a transformer, a method of analyzing a gas component contained in the insulating oil of the transformer (gas analysis in oil), an elastic wave (ultrasonic wave, etc.) generated by partial discharge inside the transformer is used. A method of detecting with an acoustic emission sensor (AE sensor, ultrasonic sensor), a method of detecting a discharge pulse signal flowing through the ground line when a partial discharge occurs with a current sensor, and the time difference between the signal from the AE sensor and the signal from the current sensor is the same For example, a method of determining a discharge when the cases are continuous (see, for example, Patent Document 1) has been put into practical use.
JP-A-4-194762

  However, the conventional method for diagnosing internal abnormality of a transformer has the following problems.

  First, the method of diagnosing internal abnormalities by gas analysis in oil has a problem that the diagnosis frequency cannot be increased because the diagnosis cost is high (for example, the diagnosis cost is 15,000 yen / unit), and the diagnosis cycle is prolonged. (For example, diagnosed every two years).

  In addition, although the method using the AE sensor is a simple method, there is a problem of noise, and it is difficult to determine whether the signal is caused by an internal abnormality or noise, so that the method has not been widely used.

  In addition, the current sensor method is not widely used because of the influence of noise such as picking up a pulse signal due to corona discharge outside the transformer.

  Further, as in Patent Document 1, a method of performing signal detection using both an AE sensor and a current sensor, and determining a partial discharge when signal detection in which the time difference between the signals from both sensors is the same continues multiple times. However, there are a problem that it is impossible to determine a false detection that picks up noise continuously by accident, and a problem that an initial discharge that does not discharge continuously cannot be detected.

  The present invention has been made to solve the above-described problems, and provides a transformer internal abnormality diagnosis method capable of performing abnormality diagnosis inside a transformer with higher accuracy and lower cost than conventional ones. It is aimed.

  The present invention has the following features to solve the above problems.

[1] A transformer internal abnormality diagnosis method for detecting a partial discharge which is an internal abnormality of a transformer, an acoustic emission sensor for detecting an elastic wave generated by the partial discharge, and a current sensor for detecting a current flowing through a ground line And a voltage detection terminal for detecting the voltage of the commercial power supply, the detection signal of the current sensor is synchronized with the commercial power supply voltage ( voltage applied to the transformer ) , and the detection signal of the current sensor and the acoustic emission A method for diagnosing an internal abnormality of a transformer, characterized in that a partial discharge is determined when a predetermined number or more of time differences from a sensor detection signal are the same within a certain time.

[2] The signal from the acoustic emission sensor and the signal from the current sensor are separated every 0.5 cycle from the time when the commercial power supply voltage is an instantaneous value of 0 V, and the detection signal By recording the generation time, when the detection signal of the acoustic emission sensor and the detection signal of the current sensor are in a range synchronized with the peak value of the commercial power supply voltage ( voltage applied to the transformer ) , partial discharge The method for diagnosing an internal abnormality of a transformer as described in [1], wherein the determination is made .

[3] The acoustic emission obtained by dividing the signal from the acoustic emission sensor and the signal from the current sensor every 0.5 cycle from the time when the commercial power supply voltage is an instantaneous value of 0 V, and is obtained every 0.5 cycle. the size P _AE detection signal by the sensor, after removal of the noise by using the magnitude P _CT of a detection signal from the current sensor, the occurrence time T _AE detection signal from the acoustic emission sensor, detected by the current sensor Using the signal generation time T _CT , the detection time T _CT of the detection signal by the current sensor is expressed as a distance from the origin O, the generation time T _AE of the detection signal by the acoustic emission sensor and the detection signal of the current sensor signal generation time difference Δt of the generation time _{T CT (= T} AE -T Draw a graph _T) from the longitudinal axis by the vector representation expressed in the rotation angle in the circumferential direction, according to [1] or [2], wherein the determining the partial discharge on the basis of the graph Transformer internal abnormality diagnosis method.

  According to the present invention, internal abnormality due to partial discharge in the transformer is cheaper than the conventional diagnostic method by gas analysis, which is expensive, and higher than the conventional partial discharge diagnostic method which has a problem in accuracy. Diagnosis can be made with accuracy. As a result, it is possible to accurately detect internal abnormality due to partial discharge in the transformer by daily inspection.

  Embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 shows an internal abnormality diagnosis system for a transformer used in an embodiment of the present invention. This diagnostic system includes an acoustic emission sensor (AE sensor, ultrasonic sensor) 2 that detects an elastic wave (ultrasonic wave) signal generated by partial discharge in the transformer 1 and an amplifier ( AE preamplifier) 3, current sensor 4 that detects a discharge pulse signal that flows through the ground line when partial discharge occurs, voltage detection terminal 7 that detects the voltage of commercial power supply 6, and calculation / image display based on the detection data The calculation / display device 5 is configured to determine partial discharge. Here, the commercial power supply voltage is detected because the voltage applied to the transformer 1 is in the same phase as the commercial power supply voltage, and thus the detection of the applied voltage is substituted by the detection of the commercial power supply voltage.

  FIG. 2 shows an example of signals sent from the AE sensor 2, the current sensor 4, and the voltage detection terminal 7 to the calculation / display device 5.

  In this embodiment, the detection signal of the current sensor 4 is synchronized with the commercial power supply voltage (voltage applied to the transformer 1), and the signal generation time difference between the detection signal of the current sensor 4 and the detection signal of the AE sensor 2 is detected. When Δt is the same, if there is a predetermined number or more (threshold number or more) within a certain time, it is determined that the partial discharge has occurred.

Specifically, an AE sensor obtained every 0.5 cycle is divided into 0.5 cycles (for example, 1/120 (S) in the case of 60 cycles) starting from the instantaneous value of commercial power supply voltage 0V. the size P _AE detection signal by 2, after removal of the noise by using the magnitude P _CT of the detection signal from the current sensor 4, and the occurrence time T _AE detection signal by the AE sensor 2, a detection signal from the current sensor 4 using the generation time _{T CT,} as shown in FIG. 3, draws a graph generation time _{T CT} with the aforementioned signal occurrence time difference Δt a ₍₌ T AE _{-T CT)} was the vector representation of the signal detected by the current sensor 4 When the area where the point density in the area where the graph is divided into meshes is equal to or higher than the threshold is within the voltage discharge determination range (the range synchronized with the peak value of the commercial power supply voltage), it is determined that the partial discharge is detected.

  The procedure for drawing the graph shown in FIG. 3 and determining partial discharge based on the graph will be described below.

First, in order to remove noise, a signal is extracted such that the product of the magnitude P _CT of the detection signal from the current sensor 4 and the magnitude P _AE of the detection signal from the AE sensor 2 is greater than or equal to a certain threshold value (or within a certain range). . This is because noise is roughly removed by distinguishing between the magnitude of the signal obtained by the internal discharge of the transformer 1 and the large noise and small noise generated in each substation.

Subsequently, a signal having a negative signal generation time difference Δt (= T _AE −T _CT ) is removed from the remaining signals. This is because the propagation speed of the elastic wave (ultrasonic wave) signal generated by the partial discharge in the transformer 1 is slower than the transmission speed of the discharge pulse signal flowing through the ground line by the partial discharge, and thus the current for detecting the discharge pulse signal. This is noise removal utilizing the fact that the sensor 4 detects the signal earlier than the AE sensor 2 that detects the elastic wave (ultrasonic wave) signal. That is, a signal with a negative signal generation time difference Δt is not a signal due to partial discharge.

Using the current signal generation time _TCT and the signal generation time difference Δt from the signal from which rough noise has been removed as described above, a vector display graph as shown in FIG. 3 is drawn.

In this vector display graph, as shown in FIG. 3A, the current signal generation time _TCT is represented by the distance from the origin O, and the signal generation time difference Δt is represented by the rotation angle from the vertical axis to the circumferential direction. Is.

As a result, the signal having the current signal generation time T _CT1 and the signal generation time difference Δt ₁ is displayed at the position of point A (T _CT1 , Δt ₁ ) in FIG. in T _CT2, signal signal occurrence time difference is Delta] t ₂ is FIGS. 3 (a) in the point B _{(T CT2,} Δt ₂₎ is displayed at the position of.

In the graph drawn as described above, when there is a partial discharge phenomenon in the transformer 1,
(A) The occurrence of partial discharge is synchronized with the peak value of the applied voltage.
(A) When the discharge points are the same, signals having the same signal generation time difference Δt are obtained.
Therefore, the drawn points are unevenly distributed as shown in FIG.

  Therefore, as shown in FIG. 3 (b), this uneven distribution is detected as the number of points in the area obtained by dividing the graph into a grid (mesh), and the area with a large number of points (high density) is applied voltage (commercial). If it is synchronized with the peak value of the power supply voltage), it is determined that the partial discharge has occurred.

  On the other hand, in the case of noise, since the signal generation time difference Δt is not constant, it is distributed over the entire graph, and uneven distribution as shown in FIG. 3B does not occur.

  Thus, in this embodiment, it is possible to overcome the point that the diagnostic method using the AE sensor and the current sensor is inherently easy to pick up noise.

  In addition, regarding the problem of the diagnostic method described in Patent Document 1, the problem that it is not possible to determine the false detection that picks up noise by chance or the initial discharge that does not discharge continuously cannot be detected. In the embodiment, the partial discharge is determined from the overall tendency of the detection signal, and further noise removal is possible by placing importance on the detection signal synchronized with the applied voltage peak value, and the partial discharge can be performed with higher accuracy. Discharge can be detected.

  Furthermore, by drawing the detection signal in a graph as shown in FIG. 3, the operator can visually confirm the determination result by the calculation / display device 5 finally.

  Therefore, in this embodiment, the internal abnormality due to the partial discharge in the transformer 1 is cheaper than the conventional diagnostic method by gas analysis, which is expensive, and the conventional partial discharge has a problem in accuracy. Diagnosis can be performed with higher accuracy than the diagnostic method. As a result, it is possible to accurately detect internal abnormality due to partial discharge in the transformer 1 through daily inspection.

  As an example of the present invention, an example in which an internal abnormality (partial discharge) of a transformer is diagnosed based on the above embodiment will be described.

The specifications of each diagnostic device used at that time are as follows:
AE sensor frequency band: 300 kHz to 2.2 MHz
AE preamplifier Frequency band: 100Hz to 20MHz
AE tester Maximum input voltage: 50mVp-p
Frequency characteristics: 100 kHz to 2 MHz (-3 dB)
Detection sensitivity: 50 μV to 5 mV
Current sensor Frequency characteristics: 20 kHz to 30 MHz
It is.

The diagnostic results are shown in FIGS. In each figure, (a) is a signal from each sensor, and (b) is a graph in which a current signal generation time _TCT and a signal generation time difference Δt are drawn in vector display.

  For FIG. 4, as shown in FIG. 4B, the drawn points are unevenly distributed, and the unevenly distributed area is synchronized with the peak value of the commercial power supply voltage. It was done.

  On the other hand, in FIG. 5, as shown in FIG. 5B, the drawn points are scattered on the graph, and it is determined that the discharge is not partial discharge.

  Thus, the internal abnormality diagnosis of the transformer could be performed accurately.

It is a figure which shows the internal abnormality diagnosis system in one Embodiment of this invention. It is a figure which shows an example of the signal from each sensor in one Embodiment of this invention. It is a figure which shows the vector display graph in one Embodiment of this invention. It is a figure which shows the diagnostic result in the Example of this invention (with internal discharge). It is a figure which shows the diagnostic result in the Example of this invention (no internal discharge).

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Transformer 2 AE sensor 3 AE preamplifier 4 Current sensor 5 Calculation / display apparatus 6 Commercial power supply 7 Voltage detection terminal

Claims (3)

A method for diagnosing an internal abnormality of a transformer that detects a partial discharge that is an internal abnormality of a transformer, an acoustic emission sensor that detects an elastic wave generated by the partial discharge, a current sensor that detects a current flowing through a ground wire, and a commercial A voltage detection terminal for detecting a voltage of the power supply , wherein the detection signal of the current sensor is synchronized with the commercial power supply voltage ( voltage applied to the transformer ) , and the detection signal of the current sensor and the detection of the acoustic emission sensor A method for diagnosing an internal abnormality of a transformer, characterized in that a partial discharge is determined when a predetermined number or more of time differences from a signal are the same within a certain time. The signal from the acoustic emission sensor and the signal from the current sensor are divided every 0.5 cycle from the time when the commercial power supply voltage is an instantaneous value of 0 V, and each detection signal and its generation time are divided every 0.5 cycle. When the detection signal of the acoustic emission sensor and the detection signal of the current sensor are in a range synchronized with the peak value of the commercial power supply voltage ( voltage applied to the transformer ) , it is determined that the partial discharge has occurred. The method for diagnosing an internal abnormality of a transformer according to claim 1. The signal from the acoustic emission sensor and the signal from the current sensor are separated every 0.5 cycle from the time when the commercial power supply voltage is an instantaneous value of 0V, and obtained by every 0.5 cycle. Detection by the acoustic emission sensor the size P _AE signals, said after removing noise by using the magnitude P _CT of the detection signal of the current sensor, the occurrence time T _AE detection signal from the acoustic emission sensor, the generation of a detection signal from the current sensor Using the time T _CT , the generation time T _CT of the detection signal by the current sensor is expressed by the distance from the origin O, the generation time T _AE of the detection signal by the acoustic emission sensor and the generation time T of the detection signal by the current sensor signal generation time difference Δt of the _{CT (=} T AE _{-T CT)} The internal abnormality of the transformer of Claim 1 or 2 which draws the graph by the vector display which represented the rotation angle to the circumferential direction from the vertical axis | shaft, and determines partial discharge based on the graph Diagnosis method.

Images