JPH07209370A - Partial discharge measuring method - Google Patents

Abstract

PURPOSE:To select an optimal correction frequency regardless of series resonance caused by a lead wire for injecting pulse by determining a peak frequency caused by resonance of a correction pulse injecting circuit based on a frequency spectrum being detected when the correction pulse is infected. CONSTITUTION:Correction pulses are injected from a correction pulse injecting circuit 8 connected between metal shield layers 5 separated by an insulator 7 of an insulating cable joint box 1 or between the metal shield layer 5 and the earth. A frequency for measuring a partial discharge pulse is then selected from a frequency spectrum being detected. A peak frequency caused by resonance of the correction pulse injecting circuit 8 is then determined from a frequency spectrum being detected when the correction pulse is injected. Subsequently, a frequency having a high S/N ratio is selected from frequencies lower than 80% of the peak resonance frequency.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a partial discharge measuring method used for diagnosing deterioration of a power cable, etc. Method for measuring partial discharge.

[0002]

2. Description of the Related Art Partial discharge measurement used for diagnosing deterioration of a power cable (CV cable) is shown in FIG.
It is measured by the measuring circuit shown in.

In this measuring circuit, a terminal connecting portion 3 is provided at the other end of a cable 2 whose one end is connected via a cable insulating connecting box (hereinafter referred to as an insulating connecting box) 1, and the terminal connecting portion 3 is charged with a high voltage. Attach terminal 4. Further, the detection impedance 6 is connected in parallel between the metal shield layers 5 of the insulation connection box 1 which are separated by the insulator 7 and are connected to the metal sheath of the cable 2.

Then, an AC high voltage is applied to the high voltage charging terminal 4, and a high frequency voltage generated at both ends of the detection impedance 6 by the partial discharge pulse generated in the insulator of the cable 2 is measured by the measuring instrument M.

At this time, the measurement frequency of the high frequency voltage is
Noise frequency measurable several to several tens MH _Z sensitive are chosen less.

However, it is considered that the on-site noise spectrum differs from place to place. For example, as shown in Japanese Patent Application Laid-Open No. 3-170076, the noise frequency spectrum at the site and the frequency at the time of injection of the calibration pulse are measured. Measure spectra and calculate S / N from those spectra
A method of selecting and measuring a frequency with a high ratio has been considered.

A method of injecting a calibration pulse at this time is disclosed in Japanese Patent Laid-Open No. 1674843 or No. 2-167.
As shown in Japanese Patent No. 484, there is a method of injecting a calibration pulse between the metal shield layers 5 of the insulation junction box 1.

That is, a calibration pulse injection circuit 8 is provided in parallel with the detection impedance 6 between the metal shield layers 5 separated by the insulator 7 of the insulation junction box 1, and the frequency spectrum of the calibration pulse injected from the outside is provided. Assuming that the spectrum of the internal partial discharge actually generated in the cable 2 and the insulation junction box 1 to be measured matches, the frequency spectrum of the noise generated in the cable 2 is first obtained without applying the voltage, and the noise is calculated. Of the calibration pulse when the calibration pulse is injected from the calibration pulse injection circuit 8 (noise + frequency spectrum of the calibration pulse), and selects a frequency having a high S / N ratio. The partial discharge measurement is performed at the frequency.

[0009]

However, the above-mentioned measurement method using the frequency spectrum has a problem that the spectrum cannot be measured accurately because the calibration pulse injection circuit is connected.

This is because a pulse injection lead wire of a certain length (about several meters or less) is required in actual calibration, and series resonance occurs due to the inductance of the lead wire and the output capacitor of the calibration pulse injection circuit. To do. When the series resonance is present, the impedance is lowered at the resonance frequency and the calibration pulse injection charge amount is increased. Then, the increase in the charge amount appears as a peak of the calibration pulse frequency spectrum as shown in FIG.

On the other hand, when an actual partial discharge occurs,
Since there is no pulse injection lead wire, there is no such resonance frequency f _k (peak of frequency spectrum) and the spectrum becomes flat up to a high frequency.

If the method disclosed in Japanese Patent Laid-Open No. 3-170076 is applied as it is to this case, the signal of the resonance frequency f _k becomes large at the time of injection of the calibration pulse. Of S
Since / N becomes the highest, the resonance frequency is selected and measured.

By the way, as will be described later, in consideration of the actual length of the pulse injection lead wire, a voltage increase of several to several tens of times easily occurs at the series resonance point.

That is, when the resonance frequency f _k is selected and the partial discharge is calibrated, a detection sensitivity several times to several tens times better than the actual detection sensitivity is apparently obtained, so that the actual magnitude of the partial discharge is obtained. Would be underestimated by a few to one tenth.

Therefore, an object of the present invention is to provide a partial discharge measuring method capable of selecting an optimum calibration frequency even in the presence of series resonance due to a pulse injection lead wire.

[0016]

In order to solve the above problems, according to the present invention, a calibration pulse connected between metal shield layers or between a metal shield layer and a ground separated by an insulator of a conventional cable insulation junction box. In the partial discharge measurement method in which a calibration pulse is injected from the injection circuit and the measurement frequency of the partial discharge pulse is selected from the frequency spectrum detected at that time, the resonance of the injection circuit is detected from the frequency spectrum detected when the calibration pulse is injected. Find the frequency peak,
A frequency having a high S / N ratio is selected from the frequencies of 80% or less of the obtained peak frequency of the resonance frequency.

Further, when the calibration pulse is injected, a calibration error due to the resonance frequency of the injection circuit is obtained from the frequency spectrum detected, and this error is corrected by adjusting the injection amount of the calibration pulse to correct the error near the resonance frequency. A method of selecting a frequency as a measurement frequency is performed, or an external impedance is connected to the calibration pulse injection system to change the resonance frequency characteristic of the calibration pulse injection circuit obtained from the frequency spectrum detected when the calibration pulse is injected. After that, a method of selecting a frequency having a high S / N ratio as a measurement frequency from the frequencies below the resonance frequency was performed.

[0018]

In the partial discharge measuring method configured as above,
Series resonance caused by the inductance component of the injection lead wire of the calibration pulse injection circuit connected between the metal shield layer of the cable insulation junction box or between the metal shield layer and the ground is several to ten times more than the measured frequency spectrum at the time of the calibration pulse injection. Therefore, during calibration, the optimum measured frequency can be selected by removing the increased spectral frequency from the measurement frequency, correcting the increase, or changing the resonance characteristic.

As a method thereof, for example, in a method of selecting a frequency having a high S / N ratio from the frequencies of 80% or less of the resonance frequency of the series resonance circuit, accurate calibration cannot be performed because the voltage increases at the resonance frequency. Select a frequency lower than the resonance frequency. At this time, in order to obtain a practically effective calibration value, S /
Select a frequency with a high N ratio.

In the method of adjusting the injection amount of the calibration pulse from the calibration pulse injection circuit, correcting the calibration error due to the resonance frequency of the calibration pulse injection circuit, and selecting the frequency near the resonance frequency as the measurement frequency, the resonance By reducing the injection amount of the calibration pulse according to the increase ratio of the spectrum voltage of the frequency and suppressing the generation of the peak of the resonance frequency, the measurement frequency can be selected even in the vicinity of the resonance frequency.

Further, in the one in which the external impedance is attached to the calibration pulse injection circuit to change the resonance frequency characteristic of the calibration pulse injection circuit, by attaching the external impedance to the injection circuit, the series resonance of the injection lead wire is suppressed. The Q is adjusted so that the resonance frequency does not have a large peak so that the measurement frequency can be selected.

[0022]

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

In the embodiment, in order to deepen understanding, first, an equivalent circuit at the time of calibration using a calibration pulse is shown, and a mechanism of series resonance is analyzed. And
A realistic parameter is assumed from the analysis result, and an example will be described based on the parameter.

FIG. 4 shows a first example of Japanese Unexamined Patent Publication No. 2-167483.
A calibration circuit for injecting a calibration pulse between the metal shielding layers 5 of the insulating junction box 1 shown in the figure is shown (Japanese Patent Laid-Open No. 2-167).
The method disclosed in Japanese Patent No. 484 uses a calibration electrode and uses a metal shielding layer 5
A calibration pulse is injected in the meantime.
(Substantially the same as that of the publication No. 7483).

The calibration circuit comprises a calibration pulse injection circuit 8 between the metal shielding layers 5 separated by the insulator 7 of the insulation junction box 1.
This is the same as the insulation junction box 1 of the partial discharge measurement circuit shown in FIG. 11 in which the detection impedance and the detection impedance are connected in parallel.

That is, the calibration pulse injection circuit 8 has an output capacitor 9 having a capacitance C, and its pulse injection lead wire (hereinafter referred to as lead wire) 10 has an inductance component L ₁ .

The metal shield layer 5 and the cable conductor 2'are connected via a surge impedance r, and the cable conductor 2'connected to the surge impedance r has an inductance component L ₂ .

By the way, at the measurement frequency that is being considered, the attenuation in the cable 2 is large, and therefore, if there is a cable length of the actual line scale, the reflection at the cable end can be ignored.
The impedance between the cable conductor 2'and the metal shielding layer 5 is considered to be surge impedance r (resistance component).

Therefore, the pulse injection system of FIG. 4 can be represented by an equivalent circuit of L ₁ , L ₂ , C and r as shown in FIG. In addition, the circuit can be expressed as a series circuit of L, C, and R as shown in FIG. 6 (however,
L = 2L ₁ + L ₂ , R = r + r).

Here, the voltage e (t) generated at R in FIG.
By the Laplace transform.

[0031]

[Equation 1]

Equation (1) is a parameter

[0033]

[Equation 2]

The waveform varies depending on whether the sign is positive or negative.

When σ <0, α + β = R / L (α> β) αβ = 1 / LC

[0036]

[Equation 3]

And is a single non-oscillating pulse.

When σ = 0

[0039]

[Equation 4]

And again a single non-oscillating pulse.

When σ> 0

[0042]

[Equation 5]

Then, the damping vibration waveform shown in FIG. 7 is obtained.

First, it will be shown that resonance exists even if the value of the most realistic parameter in which resonance does not occur is considered.

Output capacitor 9 of calibration pulse injection circuit 8
Is at most C = 10 pF (too small to be affected by stray capacitance).

The length of the lead wire 10 is required to be at least about 0.3 m in consideration of the actual injection. In the lead wire 10 having a thickness that can be normally handled, the inductance L ₁ is 1
Since it is about μH / m, the inductance L ₁ of the lead wire 10 in FIG. 5 is 0.3 μH. At this time, assuming that the inductance L ₂ of the cable conductor 2 ′ is 0.4 μH / m, the combined inductance of the equivalent circuit is L = 1 μH.

Further, since the surge impedance r of the cable 2 is at most about 50Ω, it is considered as 50Ω per one, and even if R = 50 × 2 = 100Ω, σ = (1 / LC-R ² / 4L ² ). = 9.75 × 10 ¹⁶ >
0, and it can be seen that in a normal calibration, the injected simulated pulse signal has the resonance of the injection system.

Next, the voltage increase ratio at the resonance frequency is obtained.

FIG. 8 shows the frequency spectrum of the pulse appearing at R in FIG. 6 when there is no L (corresponding to the actual partial discharge).

Here, the spectrum of the step voltage (1
It is considered that (proportional to / f) was measured through a CR high-pass filter. At this time, the detected signal has a cutoff frequency f _c (1 in the above parameters) determined by CR.
59MH _Z) until, as shown in FIG. 8a, with a flat spectrum.

FIG. 9 shows the frequency spectrum of the pulse appearing at R in FIG. 6 when L is present (corresponding to actual calibration). At that time, the detected signal has a peak at the LC resonance frequency f _k , as shown in FIG. 9b, and coincides with FIG. 8 at a frequency sufficiently smaller than f _k .

Therefore, assuming that the spectrum size in this region is 1, the spectrum size at f _k is the same as the Q value.

Here, the resonance frequencies f _k and Q are obtained as follows.

[0054]

[Equation 6]

For example, when f _k and Q values are calculated with the above parameters,

[0056]

[Equation 7]

Therefore, when measured at f _k, there is an error about 3 times that of the actual partial discharge. Also, it can be seen that as L becomes larger, the resonance frequency f _k becomes smaller and Q becomes larger.

Hereinafter, a realistic value of LCR is assumed.

Considering that the lead wire length is actually about 1 m, and considering that the surge impedance r of the actual CV cable is often 20 to 30 Ω, the practical parameters of LCR are as follows. It becomes a value.

L = 4 μH C = 10 pF R = 50Ω At this time, the resonance frequencies f _k and Q are as follows.

Resonance frequency f _k = 25 MHz Voltage rise ratio Q = 12.6 (Example 1) FIG. 1 shows the frequency peak due to the resonance of the calibration pulse injection circuit 8 from the frequency spectrum detected when the calibration pulse is injected. And the calculated resonance frequency f
S / N from 80% or less of the peak frequency of _k
The frequency spectrum figure which concerns on the partial discharge measuring method which selects the frequency with a high ratio as the measurement frequency of a partial discharge pulse is shown.

This spectrum diagram is obtained by the realistic parameters obtained by the above assumption.

At this time, the calibration pulse injection method may be performed directly between the metal shield layers 5 of the insulating connection box 1 or may be performed by capacitive coupling via the calibration metal electrode.

The spectrum diagram is plotted with the Y axis as Q and the X axis as frequency f.

Therefore, the magnitude of the spectrum at the resonance frequency f _k is indicated by the value of Q. Also, the Q
It is possible to know the calibration error with respect to the spectrum of the step voltage from the value of.

Therefore, for example, it can be seen from FIG. 1 that measurement can be performed at a frequency of 0.7 × f _k or less if a calibration error of about 1.4 times is allowed. Further, if the allowable error is doubled, it is possible to measure at a frequency of 0.8 × f _k or less.

Now, it is considered that the calibration error is practically useful if it is within a double error, and the resonance frequency f _k is 8
Calibration can be performed by measuring at a frequency of 0% or less.

In the actual measurement, if a high S / N frequency is selected in this frequency range, an accurate calibration value can be obtained while avoiding the voltage increase of the resonance frequency f _k .

(Embodiment 2) Next, a calibration error due to the resonance frequency f _k of the calibration pulse injection circuit 8 is obtained from the frequency spectrum detected when the calibration pulse is injected, and this is adjusted by adjusting the injection amount of the calibration pulse. A partial discharge measuring method for correcting an error and selecting a frequency in the vicinity of the resonance frequency f _k as the measurement frequency will be described.

Even in the vicinity of the resonance frequency f _{k in} FIG. 1, calibration can be performed by correcting the calibration error due to the calibration pulse injection system. Three correction methods will be described below, but other correction methods may be used.

Specifically, the injection charge amount is reduced by reducing the injection output of the calibration pulse in proportion to the reciprocal of the voltage increase ratio Q. The method of obtaining the Q value in that case is
, Shown in.

If the value of each parameter of LCR is obtained in advance and is substituted into the equation (3), the voltage increase ratio Q at the resonance frequency f _k can be obtained. Then, if the amount of the injection pulse is determined so as to be proportional to the reciprocal of the voltage increase ratio Q, the peak of the resonance frequency due to the injection pulse is suppressed, so that calibration without error can be performed.

For example, as assumed above, Q = 12.6.
If the injection frequency of the simulated pulse at that time is 100 PC, the correct calibration can be obtained by injecting the simulated pulse of 100 / 12.6 = 7.9 PC. It is possible.

A calibration pulse is injected and the resonance frequency f _k is obtained from the frequency spectrum. At this time, if the value of CR is obtained in advance, L can be obtained from equation (2), and if that value is substituted into equation (3), then Q can be obtained. After that, if the injection amount of the injection pulse is determined in the same manner as, correct calibration is possible.

By injecting a calibration pulse, the voltage increase ratio Q is calculated from the peak value V _{k at the} resonance frequency f _k according to the frequency spectrum and the value V _{X at a} frequency sufficiently lower than f _k (about f _k / 2) as follows. Ask.

Q = V _k / V _{X In} this way, if the voltage rise ratio Q is obtained and the value is used to determine the injection amount of the injection pulse, correct calibration is possible.

(Embodiment 3) FIG. 2 shows a case where an external impedance RX is connected to the calibration pulse injection circuit 8 to change the resonance frequency characteristic of the calibration pulse injection circuit 8 obtained from the frequency spectrum detected when the calibration pulse is injected. After that, a partial discharge measuring method is shown in which a frequency having a high S / N ratio is selected as a measuring frequency from the frequencies below the resonance frequency f _k . Further, FIG. 3 shows an equivalent circuit thereof.

Here, in FIG. 2, the calibration pulse is directly injected between the metal shield layers 5 of the insulating junction box 1, but
It may be injected by capacitive coupling through the calibration metal electrode.

In this method, the external impedance RX is connected without changing the parameters of the LC, and as shown in FIG. 3, by increasing R as R = 2r + RX, the peak of the resonance frequency does not occur. The frequency characteristic of series resonance is changed as described above. In the embodiment, the resistance component is used as the external impedance, but an inductance or capacitance component may be connected instead.

The value of the external impedance RX can be set as follows. For example, if the practically effective allowable error is doubled, Q may be selected to be 0.5 to 2.

That is, if Q = 0.5 is solved from the equation (3), R
= 1265Ω, and if Q = 2 is solved, R = 316Ω.

Here, the external impedance RX is, R
= 50Ω, that is, the surge impedance r of the cable 2 is assumed to be r = 25Ω, so if 50Ω of the surge impedance is subtracted from R to select 266 to 1215Ω,
A practically valid calibration is possible at all frequencies below f _k .

In the actual measurement, S within this frequency range
Select a frequency with a high / N.

As described above, by carrying out these methods, it is possible to select a frequency having a high S / N ratio as the measurement frequency of the partial discharge pulse regardless of the inductance component of the injection lead wire 10.

[0085]

The present invention makes it possible to perform calibration in partial discharge measurement in a frequency range of several MHz to several tens of MHz by taking into consideration the inductance of a calibration pulse injection lead wire, which has not been neglected in the past, by the above method. You can do it right.

Therefore, the measurement frequency in the calibration of the partial discharge measurement can be correctly selected.

[Brief description of drawings]

FIG. 1 is a spectrum diagram showing the operation of Example 1.

FIG. 2 is an explanatory view of the operation of the first embodiment.

FIG. 3 is an equivalent circuit diagram of the third embodiment.

FIG. 4 is an operation explanatory view for explaining a generation mechanism of series resonance.

FIG. 5 is an equivalent circuit diagram of a calibration circuit.

FIG. 6 is an equivalent circuit diagram of a calibration circuit.

FIG. 7 is an operation diagram illustrating resonance characteristics of a calibration circuit.

FIG. 8 is a frequency spectrum diagram when a calibration pulse is injected without an injection lead wire.

FIG. 9 is a frequency spectrum diagram at the time of injection of a calibration pulse with an injection lead wire.

FIG. 10 is a frequency spectrum diagram measured during calibration pulse injection.

FIG. 11 is a schematic diagram showing a partial discharge measurement circuit.

[Explanation of symbols]

 1 Cable Insulation Junction Box 2 Cable 2'Cable Conductor 5 Metal Shielding Layer 6 Detection Impedance 7 Insulator 8 Calibration Pulse Injection Circuit 9 Output Capacitor 10 Pulse Injection Lead Wire r Surge Impedance

Claims (3)

[Claims] 1. A calibration pulse injection circuit, which is connected between metal shield layers separated by an insulator of a cable insulation junction box or between the metal shield layer and the ground, injects a calibration pulse, and a part from a frequency spectrum detected at that time is injected. In the partial discharge measuring method for determining the measurement frequency of the discharge pulse, the frequency peak due to the resonance of the injection circuit is obtained from the frequency spectrum detected when the calibration pulse is injected, and 80% or less of the peak frequency of the obtained resonance frequency. A partial discharge measuring method in which a frequency having a high S / N ratio is selected as the measurement frequency of the partial discharge pulse from among the frequencies. 2. A calibration pulse is injected from a calibration pulse injection circuit connected between the metal shield layers separated by the insulator of the cable insulation junction box or between the metal shield layer and the ground, and a part of the frequency spectrum detected at that time is injected. In the partial discharge measurement method that determines the measurement frequency of the discharge pulse, find the calibration error due to the resonance frequency of the injection circuit from the frequency spectrum detected when the calibration pulse is injected, and adjust this injection error by adjusting the injection amount of the calibration pulse. The partial discharge measuring method in which the frequency near the resonance frequency is corrected and is selected as the measurement frequency. 3. Injecting a calibration pulse from a calibration pulse injection circuit connected between the metal shield layers separated by the insulator of the cable insulation junction box or between the metal shield layer and the ground, and a part from the frequency spectrum detected at that time is injected. In the partial discharge measurement method for determining the measurement frequency of the discharge pulse, connecting an external impedance to the calibration pulse injection system,
After changing the resonance frequency characteristic of the calibration pulse injection circuit obtained from the frequency spectrum detected when the calibration pulse is injected, S / is selected from frequencies below the resonance frequency.
A partial discharge measurement method in which a frequency with a high N ratio is selected as a measurement frequency.