JP3072958B2 - Partial discharge measurement method - Google Patents

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 for measuring a partial discharge of a power cable line, and more particularly to a partial discharge measuring method for determining measurement conditions.

[0002]

2. Description of the Related Art Voids in insulators of power cable lines,
As a test for detecting a partially defective portion such as a void, there is a partial discharge measurement method. Regarding this kind of technology, for example, JP-A-3-170076 and JP-A-3-170077
There is a description in the publication. The partial discharge measurement methods shown therein detect a noise pulse in a power cable line, simultaneously inject a calibration pulse into the power cable line, detect the pulse, and detect an S / N ratio (signal-to-noise ratio) from the frequency spectrum of both. ) Is determined to find a frequency having a high S / N ratio, and the partial discharge is measured at this frequency.

However, in these methods, since the measurement frequency is selected based on the S / N ratio measured at the calibration pulse injection point, when the measurement point is far from the expected partial discharge occurrence point, the partial discharge signal propagates. The selected frequency may not be the best frequency due to attenuation and the amount of attenuation depends on frequency. For example, if the propagation distance is long,
Even if the S / N ratio at the measurement point is large, if the frequency is high, the amount of attenuation increases, and the maximum S / N ratio at the measurement point may be obtained at another frequency. Discharge cannot be measured.

As a means for solving this problem, S / N (ω) = S
There is a method of calculating the S / N ratio using (ω) · exp [−L · α (ω)] / N (ω) to determine the measurement frequency.
Specifically, the partial discharge measurement is performed by the following means. (I) Inject a calibration pulse into the power cable line and measure the injected calibration pulse to obtain a first frequency spectrum. (Ii) Measure the noise in the state where the calibration pulse is not injected, and obtain the second frequency spectrum.

(Iii) The level at each frequency of the first frequency spectrum is S (ω), and the level at each frequency of the second frequency spectrum is N (ω), from the partial discharge measurement point to the partial discharge occurrence expected point. Is L, and the attenuation constant at each frequency is α (ω).
(Ω) = S (ω) · exp [−L · α (ω)] / N
(Ω) is calculated, and the frequency at which this value becomes the maximum value is obtained.
The frequency at that time is used as the partial discharge measurement frequency.

[0006]

However, the above-described method for determining the partial discharge measurement frequency is, for example, a three-line nine-phase (3
When performing partial discharge measurement on the power cable line of (phase alternating current x 3 lines = 9), since the measurement is performed for each phase, a maximum of 9
The type of measurement frequency has to be determined, the measurement takes a lot of time and is inefficient.

As another measuring system, a detecting section is provided at each of the insulated connecting section and the normal connecting section for each phase of the power cable line, and a detecting section for detecting a noise signal is also provided with the insulated connecting section. Some are usually provided at the connection part. In such a system, 9 × 2 + 9 × 2 = 36 detection units are required, which complicates the configuration and makes handling difficult.

It is an object of the present invention to provide a partial discharge measurement method capable of reducing the number of signals involved in measurement while securing the maximum detection sensitivity.

[0009]

In order to achieve the above object, the present invention provides a first frequency spectrum by injecting a calibration pulse into each phase of a power cable line,
In the state where the calibration pulse is not injected, the second
Of the first frequency spectrum, the level S (ω) at each frequency of the second frequency spectrum, the level N (ω) at each frequency of the second frequency spectrum, from the partial discharge measurement point to the partial discharge expected point. Based on the distance L and the attenuation constant α (ω) at each frequency,
The S / N ratio for each frequency is calculated as S (ω) ·
exp [−L · α (ω)] / N (ω), and the S / N ratios of the respective phases are compared with each other to determine the minimum S / N ratio from the respective phases for each frequency. , In which S
The frequency at which the / N ratio shows the maximum is selected, and that frequency is used as the partial discharge measurement frequency.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram showing a partial discharge measurement system to which a partial discharge measurement method according to the present invention is applied. FIG.
Shows only one line. In FIG. 1, the power cable lines 1a, 1b,
1c, each of which has an insulated connecting portion 2a,
They are connected via 2b and 2c. Insulated connection 2a,
A pair of metal foil electrodes 3a, 3b, 3c are adhered to both sides of the insulating portion on the surface of the sheath of 2b, 2c, and electrically connected to the respective cores of the power cable lines 1a, 1b, 1c. It is in a state of electrostatic coupling. The metal foil electrodes 3a, 3
b, 3c have a detection impedance of 4
a, 4b, 4c (functions as an AC load at the detection end)
Is connected.

In order to measure the S / N ratio of each phase, a calibration pulse generator 5 is provided.
b and 2c are connected via a metal foil electrode 3d which is sequentially adhered. Here, the calibration pulse oscillator 5 is reconnected for each phase, but three calibration pulse oscillators may be prepared and connected to each phase. Further, a spectrum analyzer 6 is sequentially connected to the detection impedances 4a, 4b, and 4c so that a detection waveform in a certain frequency band can be observed. A personal computer 7 as a processing device is connected to the spectrum analyzer 6.

Further, an antenna 8 is installed near the power cable line to be measured, and is used for detecting external noise. This detection noise is used for noise removal or noise suppression when obtaining the S / N ratio. Antenna 8
Is connected to a detection impedance 9, and the output signal is taken into the personal computer 7 via the spectrum analyzer 6.

Next, the measurement of the partial discharge with respect to the configuration shown in FIG. 1 will be described using the power cable line 1c as an example. First,
The calibration pulse generator 5 is set to a predetermined output level, and a calibration pulse is injected into the power cable line 1c through the metal foil electrode 3d. This calibration pulse is detected by the metal foil electrode 3c, and its electric signal is inputted to the spectrum analyzer 6 through the detection impedance 4c, and the frequency spectrum of the calibration pulse is obtained, and the level S (ω) of each frequency is obtained from the frequency spectrum. Can be

After that, the output of the calibration pulse from the calibration pulse oscillator 5 is stopped or removed, the background noise obtained from the antenna 8 is measured, and its frequency spectrum is obtained. Thus, the level N of each frequency
(Ω) is required. The levels S (ω) and N (ω) of each frequency thus obtained, the distance L from the predicted partial discharge occurrence point to the metal foil electrode 3c, and the attenuation constant α (ω)
The measurement frequency at the time of measuring the partial discharge is determined based on each of the above.

Assuming that the level of the partial discharge signal at each frequency at the point where the partial discharge is expected to occur is S (ω), the frequency spectrum intensity at the measurement point is given by the following equation: S (ω) · exp [−L · α (Ω)] (1) The ratio between the frequency spectrum intensity and the noise, that is, the S / N ratio is determined by the personal computer 7 using the following equation.

S (ω) · exp [−L · α (ω)] / N (ω) = SN (ω) (2) When finding a common measurement frequency for the power cable lines 1a, 1b, and 1c Are all phases 1a,
Minimum value S / N of S / N ratio {S / N (ω)} of 1b, 1c
_min (ω), then S / N _min (ω)
The personal computer 7 executes a process of searching for a frequency at which the / N ratio has the maximum value. This process is shown in FIG.
This will be described later with reference to FIG.

According to the above-described determination method, it is not necessary to set a measurement frequency for each phase, and partial discharge measurement for each phase can be performed at one measurement frequency. As a result, the number of signal lines can be reduced, and the amount of arithmetic processing can be reduced. The attenuation constant α (ω) is obtained by the following method (a) or (b). (A) A signal is input from one end of a power cable of a predetermined length, a signal output from the other end is measured, and an attenuation amount for each frequency is divided by a cable length to obtain an attenuation amount α (ω ). (B) A signal is input from one of two points (for example, adjacent insulated connection parts) having a predetermined interval in a power cable line, a signal output from the other is measured, and the amount of attenuation for each frequency is two points. The amount of attenuation α (ω) for each frequency is obtained by dividing by the interval (distance) between them.

FIG. 2 shows the S / N ratio of 0 to 10 MHz obtained for each of the three phases 1a, 1b, 1c.
(A) is the S / N ratio in the 1a phase, (b) is the 1b phase, and (c) is the S / N ratio in the 1c phase. FIG. 3 shows an S / N ratio created by comparing these three types of S / N ratio characteristics and selecting a value with the lowest S / N ratio for each frequency. The peak value of the S / N ratio created in this way has a value of 3.2 as shown in FIG.
MHz, which is the measurement frequency common to each phase in the partial discharge measurement. The determination of the optimum measurement frequency common to the three phases is performed using the spectrum analyzer 6 and the personal computer 7.

FIG. 4 is a comparison diagram showing that when the present invention is applied to multi-phase measurement, the number of signals to be handled can be reduced as compared with the prior art. Here, the comparison target is 1
In addition to the method of determining the measurement frequency for each phase,
For one phase and one phase of the detection block (slave station), one signal for the insulated connection, one signal for the normal connection, and one signal as external noise for each of the insulated connection and the normal connection. It is calculated as if it exists. As is clear from FIG. 4, although the number of signals does not change with respect to the detection signal, two signals are always applied to the external noise (noise) irrespective of the number of phases (if the signal at the ordinary connection is unnecessary). Only one signal is required), indicating that the number of signals to be handled can be reduced.

Next, a method for simply realizing the present invention will be described. When measuring the n-phase, the measurement frequency can be determined by the following procedure. (1) phase 1 for each phase n, S / N ratio is determined frequencies f _1, f ₂ ··· f _n that maximizes. (2) Frequency f in phase i (where 1 ≦ i ≦ n)
_The S / N ratio (SN _ij ) for _j (where 1 ≦ j ≦ n) is determined. (3) Next, the S / N ratio for the frequency f _j (S / N _j )
Take the minimum value of S / N _ij . (4) The maximum value of S / N _j is selected as the optimum value of the S / N ratio of the n-phase, and the frequency f _j corresponding to this frequency is set as the measurement frequency.

According to this method, it is not necessary to select a frequency having a high S / N ratio for each phase, and only one phase needs to be selected. Therefore, the time required for determining the measurement frequency can be reduced.

[0022]

As described above, according to the present invention, a first frequency spectrum is obtained by injecting a calibration pulse into each phase of a power cable line, and a second frequency related to noise is obtained without injecting the calibration pulse. A spectrum is obtained, a level S (ω) at each frequency of the first frequency spectrum, a level N (ω) at each frequency of the second frequency spectrum, and a distance L from a partial discharge measurement point to a partial discharge occurrence expected point. , And the attenuation constant α (ω) at each frequency, the S / N ratio for each frequency is calculated as S (ω) · exp [−L · α
(Ω)] / N (ω), and the S / N ratios of the respective phases are compared with each other to determine the minimum S / N ratio of the respective phases for each frequency. Is selected and the frequency is set as the partial discharge measurement frequency, so that it is possible to reduce the number of signals while obtaining the best sensitivity.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a partial discharge measurement system to which a partial discharge measurement method according to the present invention is applied.

FIG. 2: S / N of 0-10 MHz obtained for each phase
FIG. 4 is a characteristic diagram showing specific characteristics.

FIG. 3 is a characteristic diagram showing SN ratio characteristics for all three phases.

FIG. 4 is an explanatory diagram showing the effect of the present invention.

[Explanation of symbols]

 1a, 1b, 1c Power cable line 2a, 2b, 2c Insulated connection 3a, 3b, 3c Metal foil electrode 5 Calibration pulse oscillator 6 Spectrum analyzer 7 Personal computer 8 Antenna

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hiroshi Suzuki 5-1-1, Hidaka-cho, Hitachi City, Ibaraki Prefecture Inside the Power System Research Laboratories Hitachi, Ltd. 5-1-1, Nippon Electric Cable Co., Ltd. Power System Research Laboratory (72) Inventor Noriyuki Akiyama 5-1-1, Hidaka-cho, Hitachi City, Ibaraki Prefecture Nippon Electric Cable Co., Ltd. Power System Research Laboratory (56) References JP-A-8-43476 (JP, A) JP-A-7-209370 (JP, A) JP-A-7-43411 (JP, A) JP-A-6-66875 (JP, A) JP-A-3-279877 (JP JP, A) JP-A-3-175377 (JP, A) JP-B-6-8846 (JP, B2) JP-B-6-8851 (JP, B2) (58) Fields investigated (Int. Cl. ⁷ , (DB name) G01R 31/12

Claims (1)

(57) [Claims] 1. A first frequency spectrum is obtained by injecting a calibration pulse into each phase of a power cable line, and a second frequency spectrum relating to noise is obtained without injecting the calibration pulse. The level S (ω) at each frequency of the spectrum, the level N (ω) at each frequency of the second frequency spectrum, the distance L from the partial discharge measurement point to the predicted partial discharge occurrence point, and the attenuation constant α at each frequency ( ω), the S / N ratio for each frequency is determined for all phases by S (ω) · exp [−L · α (ω)] / N (ω), and the The S / N ratios are compared with each other to obtain a minimum S / N ratio from each of the phases for each frequency, and a frequency at which the S / N ratio shows a maximum is selected therefrom, and the frequency is measured by partial discharge. To make the frequency Characteristic partial discharge measurement method.