JP2008067535A - Method for detecting abnormality in gas insulated power apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect internal abnormality with high sensitivity and more surely. <P>SOLUTION: A material 4 for absorbing decomposition gas generated in a grounded tank 2, filled with a decomposed gas 1, is contained in an enclosed container 5 independently of the grounded tank 2 and the enclosed container 5 and the grounded tank 2 are interconnected so that the decomposed gas is introduced into the sealed container 5 and is then adsorbed by an adsorbent 4 for removal. When abnormality detection is performed, the gas that has not yet touched the adsorbent 4 is sampled, from a sampling port 13 provided in the interconnection passage 9 from the grounded tank 2, to the adsorbent 4, and abnormality in the grounded tank 2 is detected based on the detection of SO<SB>2</SB>F<SB>2</SB>. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

The present invention relates to a gas insulated power device using, for example, SF ₆ gas or a mixed gas containing SF ₆ gas as a main insulating medium or arc extinguishing medium, for example, a gas insulated switchgear (GIS), a gas circuit breaker (GCB) ), A cubicle type gas switchgear (C-GIS), a gas insulation transformer, a pipeline gas-insulated power transmission line (GIL) and the like.

  Since gas-insulated power equipment uses an insulating gas above atmospheric pressure as an insulating medium, a high-voltage central conductor (main circuit) that becomes an electric circuit is used together with a solid support insulator (spacer) and a metal ground tank (equipment jacket) It is housed inside and has a sealed structure. Therefore, it has various advantages such as being unaffected by the external environment, being able to reduce the size of the device, and being safe in terms of maintenance, and is extremely frequently used in Japan. On the other hand, there is a problem that it is difficult to monitor the internal state from the outside of the device, and even if an abnormality occurs in the conduction of the main circuit or the device insulation, it is difficult to detect the abnormality. Therefore, there is a strong demand for the development of technology for detecting the internal state of equipment, particularly the insulation performance from the outside.

  For example, when a corona discharge occurs in a gas insulated switchgear, an electromagnetic wave is radiated. Therefore, there is an insulation abnormality detection device that detects the corona discharge by receiving this electromagnetic wave (Japanese Patent Laid-Open No. 01-235865). In this insulation abnormality detection device, the antenna for detecting the corona discharge is disposed at a position where the reception sensitivity is high with respect to the electromagnetic wave generated by the corona discharge, and the noise detection antenna is disposed at a position where the reception sensitivity is sufficiently low with respect to the electromagnetic wave generated by the corona discharge. , And taking the difference between the signal received by the corona discharge detection antenna and the signal received by the noise detection antenna, only the corona discharge signal is extracted to detect an abnormality.

  There is also a partial discharge detection device that detects sound caused by an abnormality (Japanese Patent Laid-Open No. 5-45402). In this partial discharge detection device, an AE (acoustic emission) sensor is attached to a sealed container that houses electrical equipment, and the output of the AE sensor is input to a band-pass filter and a preamplifier. The AE sensor has a characteristic of resonating at a frequency at which the intensity of the frequency spectrum of the AE wave generated by the partial discharge is maximized, and converts sound waves generated when the partial discharge is generated into an electric signal. Of these electrical signals, only the electrical signal in the frequency region where the internal noise of the preamplifier is minimized with the resonance frequency of the AE sensor as the center is passed through the bandpass filter, and the external discharge is removed to detect the partial discharge. ing.

Furthermore, there is a discharge detection device that chemically detects various derived gases (hereinafter referred to as decomposition gas) decomposed and generated from SF ₆ gas by partial discharge or arc discharge due to abnormality in energization or insulation (Japanese Patent Laid-Open No. Sho 50). -129938). In this discharge detection device, a detection element made of a glass epoxy laminated substrate whose resistance value is reduced by reacting with decomposition gas is disposed inside a sealed container of a gas-insulated electric apparatus filled with SF ₆ , and the resistance of the detection element Monitor the value. When discharge occurs, SF ₆ is decomposed to generate active gas, and the detection element reacts with the SF ₆ decomposition product gas. Therefore, the discharge can be detected by measuring the decrease in the resistance value of the detection element.

  However, a gas insulating device is installed in the method of detecting an abnormality by receiving an electromagnetic wave accompanying the occurrence of the abnormality with the antenna and the method of detecting an abnormality by detecting an AE wave accompanying the occurrence of the abnormality with an AE sensor. Under the environment of substations, the performance is not fully demonstrated due to the presence of background noise. This is because even if the reception sensitivity / sensor sensitivity is increased in order to increase the detection sensitivity of the abnormality, the background noise is also detected, and the discrimination between the true information indicating the abnormality and the background noise (so-called S / N ratio) is increased. This is extremely difficult. Here, in the insulation abnormality detection device that detects the abnormality by receiving the electromagnetic wave accompanying the occurrence of the abnormality with the antenna, the background noise is canceled by providing the noise detection antenna. Since the background noise of the place where it is provided is not measured, it is considered that the influence of the background noise cannot be completely eliminated.

In this regard, a technique for detecting a decomposed gas chemically is no problem of such background noise, moreover, even an extremely small abnormalities such as partial discharge, because decomposition gas of SF ₆ gas that is normally accumulated There is an advantage that the concentration gradually increases and the detection becomes easy. However, since many of the decomposition gases of SF ₆ gas corrode metals, and many of them have harmful effects on the equipment, usually an adsorbent for adsorbing and removing the decomposition gas is sealed inside the equipment, The concentration is set to a level that does not adversely affect the equipment.

Japanese Patent Laid-Open No. 01-235865 JP-A-5-45402 JP 50-129938 A

As described above, in the gas-insulated power device, since the adsorbent that adsorbs the decomposition gas is enclosed in the device, even if it is attempted to detect an energization abnormality or an insulation abnormality based on the decomposition gas of SF ₆ gas, the detection element The cracked gas is adsorbed by the adsorbent before the cracked gas is detected by sensors such as the above, so that the cracked gas cannot be detected well, and its practical use is difficult. That is, since the cracked gas in the gas is detected based on the gas after coming into contact with the adsorbent, the detection sensitivity of the cracked gas is inferior, and it is difficult to detect abnormal occurrences well.

  In addition, when detecting cracked gas, there is a demand to select an appropriate monitoring gas and to make the detection more reliable.

  An object of this invention is to provide the abnormality detection method of the gas insulated power apparatus which can detect an internal abnormality more reliably.

In order to achieve this object, the abnormality detection method for gas-insulated power equipment according to claim 1 is characterized in that the adsorbent that adsorbs the decomposition gas generated in the ground tank in which the insulating gas is sealed is sealed separately from the ground tank. While being accommodated in a container, the inside of the sealed container communicates with the inside of the grounded tank, and the decomposition gas is guided into the sealed container to be adsorbed and removed by the adsorbent. The gas before coming into contact with the adsorbent is collected from the collection port provided in the communication path until and the occurrence of abnormality in the ground tank is detected based on the detection of SO ₂ F ₂ .

When an abnormality such as energization abnormality or insulation abnormality occurs in the ground tank, decomposition gas is generated from the insulating gas, and the concentration of the decomposition gas increases. The ground tank and the sealed container communicate with each other, and the decomposition gas generated in the ground tank is adsorbed and removed by the adsorbent in the sealed container. For this reason, the density | concentration of the decomposition gas in a ground tank reduces. Of the generated cracked gas, SO ₂ F ₂ has a long life and exists stably. For this reason, SO ₂ F ₂ is suitable for use as a monitoring gas for abnormality detection. Further, since the gas before coming into contact with the adsorbent is collected from the collection port, a gas containing more SO ₂ F ₂ can be used for analysis. Further, the inside of the ground tank is not opened, and the gas can be collected while maintaining the airtightness in the ground tank.

The abnormality detection method for a gas-insulated power device according to claim 2 is characterized in that the adsorbent that adsorbs the decomposition gas generated in the ground tank in which the insulating gas is sealed is housed in a sealed container separate from the ground tank. When the closed container is detachably connected to the grounded tank, the closed container communicates with the grounded tank, and the decomposition gas is guided into the sealed container to be adsorbed and removed by the adsorbent. The closed container is separated from the ground tank with the ground tank side communication path closed, and the adsorbent is analyzed to detect the occurrence of abnormality in the ground tank based on the detection of SO ₂ F ₂ .

When an abnormality such as energization abnormality or insulation abnormality occurs in the ground tank, decomposition gas is generated from the insulating gas, and the concentration of the decomposition gas increases. The ground tank and the sealed container communicate with each other, and the decomposition gas generated in the ground tank is adsorbed and removed by the adsorbent in the sealed container. For this reason, the concentration of the cracked gas in the ground tank decreases, and the cracked gas accumulates in the adsorbent. Of the generated cracked gas, SO ₂ F ₂ has a long life and exists stably. For this reason, SO ₂ F ₂ is suitable for use as a monitoring gas for abnormality detection. The closed container is separated from the ground tank, the adsorbent is taken out and analyzed, and the occurrence of abnormality in the ground tank is detected based on the detection of SO ₂ F ₂ . Since the sealed container is cut off with the ground tank side communication path closed, airtightness in the ground tank can be maintained. That is, the adsorbent can be taken out and analyzed while maintaining the airtightness in the ground tank.

Furthermore, the abnormality detection method for the gas-insulated power device according to claim 3 is based on the detection of at least one of SF ₄ , SOF ₄ and SOF ₂ in addition to SO ₂ F _2. Is to detect the occurrence of.

The lifetime of SO ₂ F ₂ is long, whereas the lifetime of SF ₄ , SOF ₄ and SOF ₂ is relatively short. Therefore, I notice that if at least _{SF 4,} SOF 4, any one of the gas of the SOF ₂ is detected with the _SO 2 _{F 2} is abnormal is recently, _SO 2 _{F 2} is detected However, when none of SF ₄ , SOF ₄ , and SOF ₂ is detected, it can be seen that the occurrence of abnormality is not recent and that a certain amount of time has passed since the occurrence of abnormality.

The abnormality detection method for a gas-insulated power device according to claim 1, wherein the adsorbent that adsorbs the decomposition gas generated in the grounded tank in which the insulating gas is sealed is housed in a sealed container separate from the grounded tank and sealed. While communicating between the inside of the container and the grounded tank, the cracked gas is introduced into the sealed container and adsorbed and removed by the adsorbent.On the other hand, when detecting an abnormality, the communication path between the grounded tank and the adsorbent Since the gas before coming into contact with the adsorbent is sampled from the provided sampling port and the occurrence of abnormality in the ground tank is detected based on the detection of SO ₂ F ₂ , The occurrence of abnormality can be reliably detected. In addition, since it is possible to perform detection by collecting a gas containing a larger amount of SO ₂ F ₂ before coming into contact with the adsorbent, it is possible to detect the occurrence of abnormality in the ground tank with higher sensitivity. Furthermore, since the gas can be collected while maintaining the airtightness in the ground tank, an abnormality occurring in the ground tank can be detected without stopping the operation of the gas insulated power device.

Further, in the abnormality detection method for a gas insulated power device according to claim 2, the adsorbent that adsorbs the decomposition gas generated in the ground tank in which the insulation gas is enclosed is housed in a sealed container separate from the ground tank. When the closed container is detachably connected to the grounded tank, the closed container communicates with the grounded tank, and the decomposition gas is guided into the sealed container to be adsorbed and removed by the adsorbent. With the ground tank side communication path closed, the sealed container is separated from the ground tank and the adsorbent is analyzed to detect the occurrence of abnormality in the ground tank based on the detection of SO ₂ F ₂ . The occurrence of abnormality can be detected more reliably based on the gas having a long life. Further, even if the concentration of SO ₂ F ₂ is small, SO ₂ F ₂ is accumulated in the adsorbent, so that SO ₂ F ₂ can be detected, and the occurrence of abnormalities in the ground tank is further increased. Sensitivity can be detected. Furthermore, since the adsorbent can be recovered while maintaining the airtightness in the ground tank, an abnormality occurring in the ground tank can be detected without stopping the operation of the gas-insulated power device.

Furthermore, in the abnormality detection method for gas-insulated power equipment according to claim 3, an abnormality in the grounded tank is detected based on detection of at least one of SF ₄ , SOF ₄ and SOF ₂ in addition to SO ₂ F _2. Therefore, it is possible to estimate the time of occurrence of abnormality based on the difference in gas life.

  Hereinafter, the configuration of the present invention will be described in detail based on the best mode shown in the drawings.

  FIG. 1 shows a gas-insulated power apparatus that implements an example of an abnormality detection method for gas-insulated power equipment according to the present invention. This gas-insulated power device accommodates a conductor 3 in a grounded tank 2 in which an insulating gas 1 is sealed in an electrically insulated state, and adsorbs a decomposition gas of the insulating gas 1 generated in the grounded tank 2. Is removed by adsorption. The adsorbent 4 is accommodated in a sealed container 5 separate from the ground tank 2 and the sealed container 5 is detachably connected to the ground tank 2 so that the inside of the sealed container 5 and the ground tank 2 communicate with each other. A first opening / closing valve is provided to close the ground tank side communication path 28 when the container 5 is separated.

  The ground tank 2 is provided with an air supply / exhaust pipe (gas pipe) 6 used for evacuating the inside and sealing the insulating gas 1 and an opening / closing valve 7 for opening and closing the air supply / exhaust pipe 6. In the present embodiment, the ground tank side communication path 28 is the supply / exhaust pipe 6 of the ground tank 2, and the first on-off valve is the on-off valve 7 provided on the supply / exhaust pipe 6. That is, the existing air supply / exhaust pipe 6 and on-off valve 7 normally provided in the ground tank 2 are used. For this reason, attachment of the airtight container 5 is easy. Further, the sealed container 5 can be attached as it is without changing the design of the existing grounded tank 2, and in particular, it can be retrofitted to gas-insulated power equipment that has already been installed and operated. Furthermore, the gas-insulated power device can be returned to its original state by removing the airtight container 5 attached later. In addition, when applying to the gas-insulated electric power apparatus already installed and operated, the adsorbent provided in the ground tank 2 is removed.

  The hermetic container 5 is provided with a hermetic container side communication path 10 and a second on-off valve 11 that closes the hermetic container side communication path 10 when the hermetic container 5 is separated. The closed container side communication path 10 is connected to the ground tank side communication path 28. The sealed container side communication path 10 and the ground tank side communication path 28 are detachably connected by abutting each other's flanges 10a and 6a and fixing them with bolts.

  A sampling port 13 is provided in the communication path 9 from the ground tank 2 to the adsorbent 4 to collect gas before coming into contact with the adsorbent 4. In the present embodiment, a sampling port 13 is provided in the sealed container 5. However, the position where the sampling port 13 is provided is not limited to the sealed container 5, and may be a position where the gas before contacting the adsorbent 4 can be sampled. By providing the sampling port 13 in the sealed container 5, the sampling port 13 can be easily installed and handled as an integrated unit, which is easy to handle. The sampling port 13 is provided with an opening / closing valve 14, and the sampling port 13 is closed at times other than when gas is sampled to ensure airtightness in the ground tank 2 and the sealed container 5.

The conductor (main circuit) 3 is a high-voltage center conductor, for example, and has a cylindrical shape, for example. The conductor 3 is arranged at the center position of a ground tank (equipment jacket) 2 having a cylindrical shape, for example, and is supported by a support insulator (spacer) 8. Insulating gas 1 is, for example, SF ₆ gas, mixed gas containing SF ₆ gas. However, the gas is not limited to these gases, and for example, N ₂ gas, CO ₂ gas, C ₃ F ₈ gas, c-C ₄ F ₈ gas, CF ₃ I gas, CF ₄ gas, and a mixed gas thereof may be used. .

  During operation of the gas insulated power device, the first and second on-off valves 7 and 11 are opened to allow the inside of the ground tank 2 and the inside of the sealed container 5 to communicate with each other. Moreover, the on-off valve 14 is closed. When an abnormality such as an energization abnormality or an insulation abnormality occurs in the ground tank 2, a decomposition gas is generated from the insulating gas 1, and the concentration of the decomposition gas increases. The ground tank 2 and the sealed container 5 are in communication with each other, and the decomposition gas naturally diffuses to reach the sealed container 5 and is adsorbed and removed by the adsorbent 4. For this reason, the concentration of the cracked gas in the ground tank 2 can be reduced. Many of the cracked gases are corrosive gases that corrode metals. Since the decomposition gas is adsorbed and removed by the adsorbent 4, the concentration of the decomposition gas does not become so high that the ground tank 2, the conductor 3 and the like are corroded, and these corrosions can be prevented.

  When the sealed container 5 is separated from the ground tank 2, the air tightness in the ground tank 2 can be maintained by closing the ground tank side communication path 28 using the first on-off valve 7. For this reason, the sealed container 5 can be separated without stopping the operation of the gas-insulated power device, and the adsorbent 4 can be taken out and replaced, repaired, or regenerated. Moreover, the airtightness in the airtight container 5 can be maintained by closing the airtight container side communication path 10 by the second on-off valve 11, and an abnormality is caused based on the gas adsorbed on the adsorbent 4 as described later. When the detection is performed, the adsorbent 4 in the separated sealed container 5 can be analyzed without contacting the outside air.

Abnormalities in the ground tank 2 can be detected as follows. That is, according to the abnormality detection method for gas-insulated power equipment of the present invention, the adsorbent 4 that adsorbs the decomposition gas generated in the ground tank 2 in which the insulating gas 1 is sealed is placed in a sealed container 5 separate from the ground tank 2. While accommodating, the inside of the sealed container 5 and the inside of the ground tank 2 are communicated, and the decomposition gas is guided into the sealed container 5 to be adsorbed and removed by the adsorbent 4. The gas before coming into contact with the adsorbent 4 is sampled from the sampling port 13 provided in the communication path 9 up to the material 4, and the occurrence of abnormality in the ground tank 2 is detected based on the detection of SO ₂ F ₂ To do.

When an abnormality such as an energization abnormality or an insulation abnormality occurs in the ground tank 2, SO ₂ F ₂ , SF ₄ , SOF ₄ , SOF ₂ or the like is generated as a decomposition gas from the insulating gas 1. The abnormality detection method of the present invention detects an abnormality occurrence in the ground tank 2 that cannot be seen from the outside, based on the detection of SO ₂ F ₂ .

The inventors of the present invention have come to know that the cracked gas generated in the ground tank 2 has a long and stable life of SO ₂ F ₂ among the cracked gases. In the experiment, for example, as shown in FIG. 16 (c), when cracked gas is generated by continuous charging for 8 hours, 279 hours (11.6 days) have passed after the application of the voltage for generating the partial discharge is stopped. , SO ₂ F ₂ was present stably. In particular, SO ₂ F ₂ having an FTIR wave number of 1503 cm ⁻¹ was stably present at a high concentration.

The present invention has been made on the basis of such knowledge. Since SO ₂ F ₂ exists stably over a long period of time, the ground tank 2 that uses SO ₂ F ₂ as a monitoring gas and cannot be seen from the outside. This is to detect the occurrence of abnormality in the interior. By using SO ₂ F ₂ having a long life as the monitoring gas, the detection becomes more reliable. Further, since the new SO ₂ F ₂ is generated each time a partial discharge occurs, the concentration increases in life long SO ₂ F ₂ are accumulated. This also makes the detection of SO ₂ F ₂ more reliable. As described above, SO ₂ F ₂ is suitable for use as a monitoring gas for abnormality detection.

In this abnormality detection method, the gas before contacting the adsorbent 4 can be collected by opening the on-off valve 14 and collecting gas from the collection port 13. For this reason, it is possible to detect SO ₂ F ₂ which is a kind of cracked gas by using a gas containing a larger amount of cracked gas, and to detect the occurrence of abnormality in the ground tank 2 with higher sensitivity. Can do. In addition, since gas can be collected while the ground tank 2 is sealed, abnormality detection can be performed without stopping the operation of the gas-insulated power device. The collection of gas for detecting SO ₂ F ₂ is performed, for example, regularly or irregularly.

Further, the first and second on-off valves 7 and 11 may be opened at all times during the operation of the gas-insulated power device, and the adsorption and removal of the decomposition gas by the adsorbent 4 may be continued. The first and second on-off valves 7 and 11 are closed before a predetermined time, so that the adsorption gas 4 cannot be adsorbed and removed by the adsorbent 4, and the concentration of the decomposition gas is increased. good. In this case, the detection of SO ₂ F ₂ that is a monitoring gas can be performed with higher sensitivity.

Also, an abnormality in the ground tank 2 can be detected as follows. That is, according to the abnormality detection method for gas-insulated power equipment of the present invention, the adsorbent 4 that adsorbs the decomposition gas generated in the ground tank 2 in which the insulating gas 1 is sealed is placed in a sealed container 5 separate from the ground tank 2. While accommodating, the closed container 5 is connected to the ground tank 2 so as to be separable, the inside of the closed container 5 and the inside of the ground tank 2 are communicated, and the decomposition gas is guided into the sealed container 5 and is adsorbed and removed by the adsorbent 4. , the abnormality detection in the case of performing the a grounded tank side communication channel 28 in a closed state in the grounded tank 2 based on the detection of SO ₂ F ₂ performs an analysis of adsorbent 4 to disconnect the closed container 5 from the grounded tank 2 It detects the occurrence of an abnormality in

By decomposing the cracked gas by the adsorbent 4, the cracked gas is accumulated in the adsorbent 4. The first and second on-off valves 7 and 11 are closed, the sealed container 5 is disconnected from the ground tank 2, and the adsorbent 4 is taken out and analyzed in a state where the outside air is shut off. The SO ₂ F ₂ is detected by analyzing the adsorbent 4, thereby detecting the occurrence of abnormality in the ground tank 2. Since the decomposition gas is accumulated in the adsorbent 4, the detection of the decomposition gas is easy even if the concentration of the decomposition gas in the ground tank 2 is low, and the occurrence of abnormality in the ground tank 2 is made more sensitive. Can be detected.

Here, in order to analyze the gas, it is necessary to release the decomposition gas adsorbed on the adsorbent 4. For example, the gas adsorbed on the adsorbent 4 is released by heating the adsorbent 4. Done. The gas released from the adsorbent 4 is analyzed by, for example, a Fourier transform infrared spectrophotometer (FTIR), a gas chromatograph, or the like, and in addition to SO ₂ F _{2 and} , in some cases, SO ₂ F ₂ , SF ₄ and SOF ₄ And SOF ₂ are checked for the presence of at least one gas. However, the gas analyzing device is not limited to these, and other devices may be used.

  In the present embodiment, the above two methods, that is, the method of collecting and analyzing the gas before contacting the adsorbent 4 and the method of analyzing the gas adsorbed on the adsorbent 4 are used in combination. However, only one of these two methods may be used.

  Further, the sampling port 13 may be omitted when a method for analyzing the gas before contacting the adsorbent 4 is not used.

  In addition, when the sealed container 5 is separated from the ground tank 2, the second opening / closing valve 11 may be omitted when the adsorbent 4 may touch the outside air.

In addition to SO ₂ F ₂ as an abnormality detection monitoring gas, it is also possible to detect the occurrence of an abnormality in the ground tank 2 based on detection of at least one of SF ₄ , SOF ₄ and SOF _2. it can.

The present inventors have come to know that the life of SF ₄ , SOF ₄ , and SOF ₂ among the cracked gases is short in a new study on the cracked gas generated in the ground tank 2. In the experiment, for example, as shown in FIGS. 16A and 16B, it is observed that the concentration of SF ₄ , SOF ₄ , and SOF ₂ immediately decreases when the application of the voltage for generating the partial discharge is stopped. Was confirmed to be relatively short. Therefore, notice that at least _{SF 4,} SOF 4, occurrence of abnormality when any one of the gas of the SOF ₂ is detected with the _SO 2 _{F 2} is recently, _SO 2 _{F 2} is detected However, when none of SF ₄ , SOF ₄ , and SOF ₂ is detected, the occurrence of abnormality is not recent, and it can be seen that a certain amount of time has passed since the occurrence of abnormality.

  The above-described embodiment is an example of a preferred embodiment of the present invention, but is not limited thereto, and various modifications can be made without departing from the scope of the present invention.

  For example, you may apply to the gas insulated power apparatus shown in FIG. In addition, the same code | symbol is attached | subjected to the member same as the member of the above-mentioned gas insulated power apparatus, and those description is abbreviate | omitted.

  In this gas insulated power device, the communication path 9 that connects the inside of the ground tank 2 and the inside of the sealed container 5 is used as a circulation path having the forward path 15 and the return path 16, and the gas in the ground tank 2 is sealed from the forward path 15 to the sealed container. A circulation device 22 that circulates from the return path 16 to the ground tank 2 after being guided to 5 and brought into contact with the adsorbent 4 is provided. In the present embodiment, the forward path 15 and the return path 16 are constituted by an inner pipe 17 and an outer pipe 18 arranged in a double tubular shape.

  The double tubular communication passage 9 is formed in the ground tank 2 using the existing air supply / exhaust pipe 6. That is, by inserting the inner pipe 27 into the air supply / exhaust pipe 6, a double pipe is provided as the ground tank side communication path 28, and the sealed container side communication path 10 is constituted by the inner pipe 19 and the outer pipe 29. It is constituted by a double pipe, and the supply / exhaust pipe 6 and the outer pipe 29, and the inner pipe 27 and the inner pipe 19 are detachably connected. That is, the inner pipe 17 is constituted by the inner pipe 27 and the inner pipe 19, and the outer pipe 18 is constituted by the air supply / exhaust pipe 6 and the outer pipe 29.

  The double tube outer tube 29 opens into the space 20 in the sealed container 5, and the inner tube 19 opens into the adsorbent housing chamber 21 in the sealed container 5. Further, a circulation device 22 for sending the gas in the space 20 into the adsorbent accommodating chamber 21 is installed at the entrance of the adsorbent accommodating chamber 21. The circulation device 22 is an electric fan driven by a motor (not shown), for example. The sampling port 13 is provided in the middle of the double tube outer tube 29.

  When the circulation device 22 is started, the space 20 in the sealed container 5 has a negative pressure, and the adsorbent storage chamber 21 has a positive pressure, so that a forced gas flow is formed. The gas in the ground tank 2 is sucked into a space between the air supply / exhaust pipe 6 and the inner pipe 27 (hereinafter referred to as an air supply / exhaust pipe outer passage), and a space between the outer pipe 29 and the inner pipe 19 of the double pipe (hereinafter referred to as the internal pipe 19). (Referred to as a double pipe outer passage) and sucked into the space 20 in the sealed container 5. Then, after being sent into the adsorbent accommodating chamber 21 by the circulation device 22 and contacting the adsorbent 4, the space in the inner pipe 19 of the double pipe (hereinafter referred to as the double pipe inner passage) → inside the air supply / exhaust pipe 6 The inner space of the inner pipe 27 (hereinafter referred to as an air supply / exhaust pipe inner passage) is forcedly circulated into the ground tank 2. The cracked gas generated in the ground tank 2 reaches this adsorbent 4 on this flow and is adsorbed and removed.

  In this way, the communication path is a circulation path, and the circulation device 22 is provided to forcibly circulate the gas, so that the decomposition gas is adsorbed more quickly than when waiting for the decomposition gas to reach the adsorbent 4 by natural diffusion. In addition to being able to be removed, it is possible to prevent the decomposition gas from remaining in the ground tank 2.

  Moreover, since the gas before contacting the adsorbent 4 flows through the double pipe outer passage, the gas in the state before contacting the adsorbent 4 can be collected from the sampling port 13.

  In the above description, the supply / exhaust pipe outer passage and the double pipe outer passage are defined as the forward path 15, and the double pipe inner path and the supply / exhaust pipe inner passage are defined as the return path 16. The double pipe inner passage and the supply / exhaust pipe inner passage may be the forward path 15, and the supply / exhaust pipe outer passage and the double pipe outer passage may be the return path 16.

  In the above description, the circulation device 22 is provided at the inlet of the adsorbent accommodating chamber 21, but the circulation device 22 is moved to another position as long as it is a position where a forced circulation flow of gas can be generated. It may be provided.

  Here, the double pipe structure of the communication path 9 will be described. As the double pipe structure of the communication passage 9, for example, a double pipe is used when the gas insulated power device is operated, but when the sealed container 5 is separated from the ground tank 2, a single pipe structure is adopted. Can be considered. An example is shown in FIG. The inner tube 17 of the communication passage 9 is formed of a flexible tube, for example. A tube winding device 23 is provided in the sealed container 5. By winding the inner tube 17 with the tube winding device 23, the inner tube 17 can be pulled out from the outer tube 18. Further, the wound inner tube 17 can be fed into the outer tube 18 by the tube winding device 23. By communicating the inner pipe 17, the communication path 9 has a double pipe structure. That is, the inner tube 27 and the double tube inner tube 19 are formed of a single tube.

  The first on-off valve 7 and the second on-off valve 11 are, for example, full-bore type ball valves that can ensure the cross section and linearity of the air supply / exhaust pipe 6 or the outer pipe 29 in the open state as valves. By opening the first and second on-off valves 7 and 11, the inner pipe 17 can be fed into the outer pipe 18 constituted by the outer pipe 29 and the air supply / exhaust pipe 6. In addition, the first and second on-off valves 7 and 11 can be closed by winding up the inner pipe 17. However, valves other than the full bore type ball valve may be used as the first and second on-off valves 7 and 11.

  During operation of the gas-insulated power device, the first and second on-off valves 7 and 11 are opened, and the inner pipe 17 is sent into the outer pipe 18 so that the communication path 9 includes the forward path 15 and the return path 16. It can be a circuit. When the sealed container 5 is separated from the ground tank 2, after the inner tube 17 is wound up by the tube winding device 23, the first and second on-off valves 7 and 11 are closed and the outer supply pipe 6 is removed from the supply / exhaust tube 6. The tube 29 may be removed. When the separated sealed container 5 is attached to the ground tank 2, after connecting the outer pipe 29 to the supply / exhaust pipe 6 in the atmosphere of the insulating gas 1, the first and second on-off valves 7 and 11 are opened, and the tube The inner tube 17 may be sent out into the outer tube 18 by the winding device 23.

  The electric tube winding device 23 is remotely operated from the outside of the sealed container 5. However, a manual tube winding device 23 may be used in place of the electric tube winding device 23. In this case, the tube winding device 23 is driven by operating a handle provided with a gas seal. Let

  Further, instead of using a flexible tube as the inner tube 17, the inner tube 17 having a bellows structure may be used, and the inner tube 17 may be inserted into or pulled out from the outer tube 18 by expanding and contracting. As a means for expanding and contracting the bellows-structured inner tube 17, for example, it is conceivable to insert a rod into the inner tube 17 to connect the tips thereof, and to extend and contract the bellows-structured inner tube 17 by operating the rod. Alternatively, the inner tube 17 can be expanded and contracted as a whole by forming the inner tube 17 with a telescopic cylinder structure, that is, a plurality of pipes having slightly different diameters, and sequentially storing and pulling each pipe into and from the adjacent pipe. The structure may be inserted into the outer tube 18 or pulled out by expanding and contracting. As a means for expanding and contracting the inner tube 17 having this structure, for example, it is conceivable to insert a rod into the inner tube 17 to connect the tips thereof, and to expand and contract the inner tube 17 by operating the rod.

  Further, instead of the extendable inner tube 17, for example, a non-expandable inner tube 17 such as a metal pipe may be used. That is, the long inner tube 17 may be pulled out in the axial direction as it is. In this case, the structure can be simplified. However, since a space in which the inner tube 17 can be pulled out in the axial direction in the direction of arrow A in FIG. 3 is required, it is effective when such a space can be secured.

  Moreover, as a double pipe structure of the communication path 9, in addition to the structure in which the closed container 5 is separated from the ground tank 2 described above, a single pipe cannot be used and a double pipe is always used. It is also possible to adopt the structure that is used. In this case, as the first and second on-off valves 7 and 11, for example, as shown in FIG. 4, the through holes in the balls 7 a and 11 a of the full bore type ball valve have the same diameter as the double pipe of the communication passage 9. As a double structure to be connected, it can be considered that both the forward path 15 and the return path 16 can be opened and closed simultaneously.

  In the above description, the communication path 9 is configured using the existing air supply / exhaust pipe 6 in the ground tank 2, but the communication path 9 may be provided separately from the air supply / exhaust pipe 6. An example of this is shown in FIG. In this example, a hole is formed in an existing handhole flange (opening flange) lid 24 in the ground tank 2, and a double pipe 30 serving as a ground tank side communication path 28 is fixed to the sealed tank side communication path 10. The double pipe 31 is connected. In this case, since the diameter of each double pipe 30 and 31 can be freely selected to some extent, the degree of freedom in designing the communication path 9 is improved, and the formation of the communication path 9 is easy. However, the fixing position of the double pipe 30 is not limited to the hand hole flange cover 24.

  Further, for example, the present invention may be applied to the gas insulated power device shown in FIG. In addition, the same code | symbol is attached | subjected to the member same as the member of the above-mentioned gas insulated power apparatus, and those description is abbreviate | omitted.

  This gas-insulated power device is constituted by pipes 25, 26, 32, and 33 in which an outward path 15 and a return path 16 are separately provided. For example, the pipes 25 and 26 are fixed by providing holes in the hand hole flange lid 24. The pipes 32 and 33 are fixed to the sealed container 5. The pipe 32 is connected to the pipe 25 and the pipe 33 is connected to the pipe 26 so as to be separable.

  The handhole flange cover 24 has a larger diameter than the air supply / exhaust pipe 6 and can be installed with a distance between the forward path 15 and the return path 16 as compared with the case where the communication path 9 has a double pipe structure. For this reason, it is possible to circulate the gas more efficiently in the ground tank 2 and to further prevent the decomposition gas from remaining in the ground tank 2. Further, the pipes 25 and 26 are easily fixed to the handhole flange lid 24, and the manufacture of the gas insulated power device is simple. The pipes 25 and 26 may be fixed to portions other than the hand hole flange lid 24. Moreover, you may comprise either one of each piping 25 and 26 using the air supply / exhaust pipe 6. FIG. Further, when two or more handhole flange lids 24 are provided in the same gas section of the gas-insulated power device, a pipe 25 serving as the forward path 15 and a pipe 26 serving as the return path 16 are provided in different handhole flange lids 24. May be fixed. In this case, since the forward path 15 and the return path 16 can be further separated from each other, the gas can be circulated more efficiently in the ground tank 2, and the residual cracked gas in the ground tank 2 can be further prevented. it can.

An experiment was conducted to examine the change over time of the cracked gas.
(1) First, the generation mechanism of the cracked gas will be described.

By heating and recombination, as shown in Chemical Formula 1, SF ₆ gas dissociates into sulfur atoms and fluorine atoms to form many groups, ions, and neutral molecules.

Here, when the energy ΔE is not input, most of the sulfur atoms and fluorine atoms return to the original SF ₆ by recombination, but a part is combined with oxygen, moisture, and the constituent materials of the apparatus to become a decomposition gas. . The main gas generated in Chemical Formula 1 is SF ₄ with X = 4. The cracked gas accompanying corona discharge or sparkover tends to form sulfur fluoride containing oxygen when SF ₄ reacts with oxygen atoms or moisture. The generation mechanism of these cracked gases is shown in FIG. The broken line is a reaction that occurs only during the period in which the SF ₆ gas is being decomposed. In particular, when oxygen atoms or the like are present in the apparatus during recombination, SOF ₄ , SOF ₂ , SO ₂ F _2, etc., which are fluorides containing oxygen, are generated as shown in chemical formulas 2 to 5. Here, among these cracked gases, cracked gas species suitable for the present invention were experimentally examined from the viewpoints of detection sensitivity, change with time, and the like.

(2) Next, an experimental method will be described.
The SF ₆ cracked gas generation experiment by partial discharge was performed using a small-scale basic experiment tank that is easy to handle and a large-scale actual tank that can simulate an actual machine. Since the basic experimental tank has a small volume of 15.5 liters, it can generate a high concentration cracked gas relatively easily. Therefore, it is suitable for identifying the type of cracked gas, analyzing its properties, quantitatively evaluating the relationship between the partial discharge charge amount and cracked gas concentration, and evaluating the adsorbent unit described above. The actual tank has a large capacity of 900 liters and can achieve the same level of pressure as the actual machine. In any of the experiments, in order to avoid contamination of impurities as much as possible, evacuation for 1 hour or more was performed, and after replacement with SF ₆ gas several times, SF ₆ gas of a specified pressure was filled. The experiment was performed at room temperature (25 ° C.). The following describes each experimental system and the equipment used.

(2.1) Basic Experiment Tank Experiment System First, the experimental arrangement will be described. FIG. 8 shows an electrode layout in the basic experiment tank. The tank consists of an insulating tube made of epoxy resin and a metal flange. A high voltage is applied to the upper flange. The inner diameter is 240 mm and the height is 360 mm. The gas volume in the state where an electrode, a partial discharge detection circuit, etc. are installed inside is 15.5 liters (actual measurement value). The SF ₆ gas pressure was 0.1 MPa (abs.). The experiment was conducted in a shield room having a shielding performance of 40 dB against electromagnetic waves.

  The electrode was a flat plate electrode having a flat portion diameter of 130 mm on the high voltage side, a needle electrode having a diameter of 0.5 mm and a length of 35 mm on the ground side, and a gap length of 15 mm. FIG. 9 shows the electric field of the same electrode system. This needle electrode is insulated from the lower ground-side flat plate electrode by an insulating spacer, and the current accompanying the partial discharge generated by the needle flows to the ground through a high-frequency shunt (50Ω). The potential difference of the high frequency shunt resistance at this time is measured with an oscilloscope (Tektronix TDS7254B, analog band 2.5 GHz, 20 GS / s). The input impedance of the oscilloscope was 50Ω, sampling was 800 ps / pt, record length 25 × 106 words, and measurement time was 20 ms (one cycle of commercial frequency 50 Hz). At this time, the charge amount Q of each partial discharge pulse is expressed by Equation 1.

Where Q: charge amount of partial discharge pulse, V: observation voltage of oscilloscope, Rs: shunt resistance (50Ω).

Further, the total charge amount Qtotal in one cycle at a commercial frequency of 50 Hz is expressed by Equation 2.
However, t1-t0 = 20 ms.

  The partial discharge varies greatly depending on the state of the needle tip. In this test, in order to generate a stable partial discharge, a needle electrode which was aged by applying a voltage of about 80% of the minimum breakdown voltage (V50-3σ) of the gap for about 8 hours was used.

  The test voltage was alternating current (commercial frequency 50 Hz), and was applied from a corona-free type test transformer (100 kV, 10 kVA) manufactured by Tokyo Transformer.

Next, analyzer arrangement will be described.
FIG. 10 shows the gas flow path and analyzer arrangement in the basic experiment tank. In FIG. P is a vacuum pump; P is a dry pump; G is a vacuum gauge. It has a closed loop configuration with a gas circulation system, and an adsorbent unit, a gas circulation dry pump, an analytical gas sampling cylinder, and two types of analyzers are arranged in the loop, and a general-purpose port for a gas detector tube is provided. Details of the analyzer will be described later, but the FTIR (Fourier Transform Infrared Spectrophotometer) can identify and quantify decomposed gas species, and the DILO gas analyzer is installed mainly for the purpose of measuring moisture. did. This time, in order to quantify the target gas (SO ₂ F ₂ ) and to calibrate the absolute value of FTIR, the target gas is separately collected in an analytical gas sampling cylinder in FIG. 10 and analyzed using a gas chromatograph. did. A bypass is provided in the gas input / output port of each device, and the device can be disconnected from the system (entered into the system) as necessary. The gas flow rate for forced circulation was set to about 3 liters / min.

(2.2) Full-scale tank experiment system First, the experimental arrangement will be described. The electrode arrangement of the actual scale tank is shown in FIG. The actual scale tank has an inner diameter of 1000 mm and a height of 1000 mm. The gas volume is about 900 liters. The SF ₆ gas pressure was 0.5 MPa (abs.). All experiments were performed in a shielded room.

  The electrode was a flat plate electrode having a flat portion diameter of 360 mm on the high voltage side, a needle electrode having a diameter of 0.5 mm and a length of 28 mm on the ground side, and a gap length of 15 mm. Although the needle length is slightly different from that of the basic experimental tank, the gap length is the same, and as a result, the potential distribution between the gaps is almost the same as that of the basic experimental tank (Fig. 9). As in the case of the basic experimental tank, this needle electrode is insulated from the lower ground-side plate electrode by an insulating spacer, and the current associated with the partial discharge generated by the needle flows to the ground through the high-frequency shunt. The potential difference of the high frequency shunt resistance at this time is measured with an oscilloscope (Tektronix TDS7254B, analog band 2.5 GHz, 20 GS / s).

  The test voltage was alternating current (commercial frequency 50 Hz), and was applied from a corona-free test transformer (200 kV, 25 kVA) manufactured by Haefely.

Next, analyzer arrangement will be described.
The analyzer arrangement (gas piping) in the actual tank is almost the same as that of the basic experiment tank shown in FIG. The gas is circulated from a gas port installed facing the side of the tank (see FIG. 11).

(2.3) Gas analyzer and analysis method (a) FTIR (Fourier transform infrared spectrophotometer)
When infrared rays are irradiated to gas molecules, the infrared rays corresponding to vibration energy between atoms forming the molecules are absorbed. By examining the degree of absorption, the structure of the compound can be estimated and quantified (infrared spectroscopy). At present, as a device for performing infrared spectroscopy, a wave number monitor using a laser beam, an interferometer having a moving mirror, and a Fourier transform infrared spectrophotometer (FTIR) having a computer processing unit are mainly used. SF ₆ and the cracked gas whose components are close to it are global warming gases, have a feature of absorbing infrared rays called an atmospheric window, and can be detected by FTIR. The outline of the equipment used is as follows.

・ Name: MATTSON Infinity Gold
・ Internal configuration: Michelson interferometer, corner cube mirror, laser quadrature system ・ Infrared light source: High-intensity air-cooled ceramic mid-infrared light source ・ Beam splitter: Ge / KBr
-Spectral range: 7000-700 cm ^-1 (resolution: 0.5-32 cm ^-1 )
・ Gas cell: Ultra low capacity ・ Long pass type gas cell (pass length: 2.0 m)

(B) Gas analyzer manufactured by DILO The generation of decomposition products, water content, and SF ₆ gas purity can be measured by the ion mobility spectrum method. In this experiment, the water content was mainly measured. The outline of the equipment used is as follows.

Decomposition product measurement function Measurement range: 0 to 5000 ppm Vol.
Accuracy: 2% (100 / 5000ppm Vol.)
・ Moisture measurement function Measurement range: dew point -50 ~ + 10 ℃
Accuracy: ± 2% Dew point + 10-40 ° C range
± 4% Dew point -40 ° C or less · Volume percentage% (purity) measurement function Measurement range: 0 to 100 Vol. % SF ₆
Accuracy: ± 1% SF ₆ -N ₂ -mixing standard

(C) Gas analysis by gas chromatograph Gas analysis by gas chromatograph was conducted at Toray Research Center, Inc. Details of the analysis method are disclosed in Japanese Patent No. 3318473, so an outline of this method will be described here. Excellent points of the present approach is that it is possible to separate the SF ₆ decomposition gases present in trace amounts in the SF ₆ gas (typically difficult to separate in order to configure the original SF ₆ gas are similar) As mentioned above, it is possible to quantify a gas for which standard gas is difficult to obtain, and it has excellent resolution. That is, a standard gas can be generated from the analysis target gas generated in the test without separately requiring a standard gas, and the gas can be quantitatively analyzed. A series of analysis procedures are as follows. The cracked gas (in this case, SO ₂ F ₂ gas) in SF ₆ gas is separated by a dedicated capillary column and technique and quantified by an atomic emission detector (GC / AED). This gas is diluted and used as a standard gas, and a calibration curve of an electron capture gas chromatograph (GC / ECD) which is a high resolution analyzer is obtained. Thereafter, the decomposition gas generated in the test can be quantitatively analyzed by GC / ECD for which calibration is performed. Resolution is 0.5ppb Vol. The error is within 0.5% (ppb is 1/1000 of ppm).

(D) Gas detector tube The gas detector tube shows the concentration of the target gas when the gas reacts with the indicator and changes color. It is used for periodic inspections of equipment because gas analysis can be performed easily even at the power site. This time, a Kitagawa type gas detector manufactured by Komyo Chemical Co., Ltd. and a dedicated detector tube were used. Since the gas detection tube reacts not only with the target gas but also with various other gases (interfering gases), when the generation of various cracked gases is expected as in this case, the indicated value indicates the absolute amount of the target gas. It is not a thing. However, since there is no reaction with clean SF ₆ gas, it is possible to determine the presence or absence of decomposition gas (the presence or absence of partial discharge). This time, the gas detector tube shown in Table 1 was used.

(3) Next, experimental results and discussion will be described.
The purpose of the basic experimental tank was to verify the accumulation effect of the cracked gas, identify the cracked gas species, obtain the characteristics of the cracked gas over time, quantitative evaluation of partial discharge and cracked gas, and verify the effect of the external adsorbent unit The experiment was conducted. The SF ₆ gas pressure was 0.1 MPa (abs.), And the voltage application time was continuous 8 hours. The actual scale tank was assumed to be an actual machine, a relatively small partial discharge was generated, cracked gas was accumulated for a long time, and cracked gas detection was verified. The SF ₆ gas pressure was 0.5 MPa, and the voltage application time was continuous 72 hours. The details of the partial discharge in each experiment will be described later. The outline is shown in Table 2, and the partial discharge current waveform observed as an example is shown in FIG. The table also shows the moisture content of SF ₆ gas after the experiment. In the experiment of the basic experimental tank, the water content is relatively high, which is thought to be because the moisture adsorbed on the wall of the experimental tank (made of epoxy) was released into the gas. In this experiment, as will be described later, sufficient cracked gas is generated for analysis even when the water content is 58 ppm or less (full-scale experimental tank). Further, alternating current (commercial frequency 50 Hz) was used as the applied voltage, and partial discharge was sampled 10 times per hour. The values in Table 2 indicate the maximum value during the experiment for the maximum charge amount, the average value for the average charge amount per pulse, and the average value of the acquired data (for one cycle of commercial frequency (50 Hz)) for the others. .

(3.1) Identification of decomposition gas species by FTIR and verification of accumulation effect When FTIR is used, the analysis result is expressed as a spectrum of absorbance with respect to wave number (wave number [cm ^-1 ]). Therefore, it is necessary to identify the gas species from this spectrum, but since the generated cracked gas is known to some extent this time, we compared it with the measured data from the library data and literature data of these gases.

FIG. 13 shows the spectrum of SF ₆ obtained by measuring the cracked gas from the library (lib) and the literature (Ref) as a solid line, and actually measuring (mes) as a reference. The horizontal axis is wave number and the vertical axis is absorbance. The solid line parallel to the vertical axis in FIG. 13 indicates the wave number used for identifying the gas type, and the gas type corresponding to the upper part of FIG. 13 (the reason for determining the wave number will be described later).

The absorbance depends on the FTIR body, the gas cell used, and the gas type (wave number). In the case of the same gas type (wave number), the absorbance is relative to the concentration (that is, in the case of the same wave number (same gas type), The higher the absorbance, the higher the concentration, but when the wave numbers are different (different gases), the higher the absorbance, the higher the concentration is not necessarily). Therefore, when quantifying gas, a method such as calibration using a standard gas is required. In this experiment, it was difficult to obtain standard gas, so the decomposition gas (concentration of 3 types) generated in the experiment was quantitatively analyzed in a gas chromatograph, and the absorbance was calibrated. The gas species that was calibrated was SO ₂ F ₂ that was excellent in stability over time. Originally, when a certain kind of gas is converted into another gas by decomposition or recombination, and the ratio of the gas changes, the absorbance of the original gas decreases in FTIR as described above. However, in this experiment, the decomposition gas generated by the partial discharge is extremely small, and the SF ₆ gas occupies most of the decomposition gas after decomposition, so the spectrum corresponding to the SF ₆ gas does not change. Therefore, in the FTIR measurement, the background is usually measured, then the sample is measured, and the gas is evaluated from the difference. This time, SF ₆ gas (new gas from the cylinder) is used for the background. , Only the spectrum of cracked gas is output. FIG. 14 shows the measured emissivity when 0.1 MPa SF ₆ is used as the background. The emissivity is called a “window” in FTIR measurement, and the portion where the emissivity is zero has no sensitivity (the “window” is closed). From FIG. 14, it can be seen that the wave number portion corresponding to the large absorbance peak of SF ₆ indicated by the wavy line in FIG. 13 has no sensitivity because the concentration of SF ₆ gas is high. Therefore, for the measurement of cracked gas, the cracked gas spectrum was selected from the open part of the window. Table 3 shows the types of cracked gas and the corresponding wave numbers, and FIGS. 13 and 14 show solid lines. From the generation process of cracked gas (see Fig. 7), HF is expected to be generated, but HF is sensitive to FTIR wave number 3500-4000, and this area is sensitive to moisture (H ₂ O). Therefore, measurement is difficult.

FIG. 15 shows an example of a spectrum of cracked gas measured before, 4 hours, 6 hours, and 8 hours before voltage application. At any wave number, a peak almost consistent with the library and literature data is recognized, and the gas species can be identified well. It can also be seen that the concentration of the cracked gas increases with the passage of time (the cracked gas is accumulated). In the measurement in FTIR, there are cases where the measurement becomes highly sensitive (indicating greater absorbance) by increasing the pressure of the gas to be analyzed. However, the pressure of the base gas (SF _{6 in} this case) also increases, and as a result, the “window” also narrows. However, in the current target gas, SOF ₂ and SO ₂ F _{2 at} a gas pressure of 0.5 MPa are used. An improvement in sensitivity was observed. In contrast, the wave number of SOF ₄ has lost sensitivity. For reference, the results when measured at an analysis gas pressure of 0.5 MPa (abs.) Are shown in (4) below.

(3.2) Characteristics of cracked gas over time In the previous section, it was found that cracked gas accumulates (concentration increases). As a method of analyzing gas directly, it can be easily and immediately analyzed using a gas detector tube etc. in the field, but when performing detailed analysis, the gas is taken home and analyzed later. In this case, it is necessary to analyze gas species with a long life (stable existence). In either method, when directly analyzing the gas, it is necessary to temporarily stop the external adsorbent unit and accumulate the decomposition gas. However, it is preferable that the external adsorbent unit be stopped as short as possible. Therefore, in order to easily detect an abnormality, it is considered that a gas that is efficiently generated by partial discharge and has a long life is suitable for the analysis target gas (a gas that enables highly sensitive abnormality detection). The same applies to the case of analyzing cracked gas from the adsorbent. However, if a short-lived gas is detected, it indicates that a partial discharge has occurred in the immediate vicinity. Thus, by comprehensively considering the detected cracked gas species and the detected amount, it is possible to estimate the internal state of the device, such as the scale, generation period, and location of partial discharge. In addition, by performing trend management together, more detailed status management becomes possible.

  Here, in order to evaluate the change over time of the cracked gas, the cracked gas is generated by continuous power application for 8 hours, and the change over time of each cracked gas up to 279 hours (11.6 days) after the voltage application is stopped is obtained by FTIR. did. In order to facilitate measurement, a relatively high concentration of cracked gas is generated, and the partial discharge amount is set to an extent that, on average, 2000 or more partial discharge pulses are generated in one AC cycle. (See Table 2 above for details.)

FIG. 16 shows the change over time of each cracked gas. From the figure, it can be seen that the various cracked gases are steadily accumulated (increased) as the voltage is applied. In addition, SF ₄ , SOF _4, and SOF ₂ (FIGS. 16A and 16B) are decreased immediately after the voltage application is stopped, and it can be seen that the lifetime is relatively short. On the other hand, SO ₂ F ₂ continues to increase even after the voltage application is stopped, equilibrates from the point of about 30 hours, and exists stably until the point of 279 hours after the end of the measurement. As shown in FIG. 7, this is considered to be because SO ₂ F ₂ is not directly generated by partial discharge, but is generated by converting SF ₄ , SOF ₄ and SOF ₂ . Particularly, _SO 2 _{F 2} wavenumber 1503cm ^-1, of the currently selected gas species and wavenumber, can be said to be promising in most absorbance is high (more sensitive) FTIR Analysis, as described above _SF 4, The presence / absence of SOF ₄ and SOF ₂ can also be an important factor in device status diagnosis. Details of the partial discharge generated in this test are shown in (5) below.

(3.3) Quantitative Evaluation of Partial Discharge and Decomposed Gas In the general method for detecting partial discharge using electrical signals, parameters of partial discharge are estimated from the detected signals. In this case, the detected electrical signal corresponds to the charge amount (C / pulse) of an individual (single pulse) partial discharge. On the other hand, since the temporal accumulation of cracked gas is an important factor in this method, the corresponding partial discharge parameter is considered to be the total charge amount during the accumulation period. Here, in order to clarify the relationship between the cracked gas and the partial discharge amount, the characteristics of the cracked gas were acquired by setting the three levels of partial discharge amounts shown in Table 4. The generation time (voltage application time) of the cracked gas is 8 hours.

FIG. 17 shows the number of pulses per cycle (20 ms) of commercial frequency (50 Hz), total charge amount, average charge amount, and maximum charge amount over time, and Table 4 quantifies parameters related to partial discharge amount. The SO ₂ F ₂ concentration and various gas concentrations immediately after voltage application obtained by the gas detector tube are shown. In this experiment, partial discharge was sampled 10 times per hour. The values in Fig. 17 and Table 4 are the maximum charge amount, the average charge amount is the maximum value and average value per pulse, and the partial discharge. The number of pulses and the total charge amount are shown as average values per cycle of commercial frequency (50 Hz). The generation rate of SO ₂ F ₂ obtained in this experiment was about 1300 to 4800 μmol / C.

  From FIG. 17 and Table 4, when comparing the partial discharge amounts up to “levels 1 to 3”, the average charge amount is almost equal, and even in the case of “level 3” indicating the highest numerical value even in the maximum charge amount, There is only about twice as much as "1". However, the number of pulses is about 20 times “level 1”, and as a result, the total charge amount is about 25 times. (5) shows the polarity of partial discharge in each experiment, a histogram of partial discharge charge amount per pulse, and the like.

As for the analysis using the gas detector tube, in all cases, after the voltage application is stopped, the sensitivity decreases with time, the HF detector tube becomes undetectable about 40 hours after the stop, and the SO ₂ detector tube is stopped after the stop. It became 1/3 or less of the initial value in about time.

FIG. 18 shows the relationship between the total charge amount and the FTIR absorbance. In the figure, the absorbance on the vertical axis is the maximum value in the measurement. That is, (a) SF ₄ , SOF ₄ , and (b) SOF ₂ are values immediately after the voltage application is stopped, and (c) SO ₂ F ₂ has a concentration that gradually increases after the voltage application is stopped ( FIG. 16 (c)), and the value was 70 hours after the voltage application was stopped when the concentration was stable. All the cracked gases have a significant correlation between the partial charge total charge amount and the absorbance (decomposed gas concentration), and it can be seen that the larger the total charge amount, the higher the cracked gas concentration.

For SO ₂ F ₂ , quantitative analysis by gas chromatography was performed 70 hours after the voltage application was stopped when the concentration was stable. FIG. 19 shows the total charge amount and the SO ₂ F ₂ concentration obtained by quantitative analysis. The relationship between the two can be expressed by the following empirical formula (Formula 3).

_{However, C 'SO2F2: SO 2 F} 2 concentration [ppm], Q _total: AC1 total charge amount in the cycle [pC], voltage application time: 8 hours, the gas volume: 15.5 liters.

Since the concentration of SO ₂ F ₂ is known by quantitative analysis, calibration is performed for the most sensitive wave number (1503 [cm ⁻¹ ]: see FIG. 18C) among the absorbances of SO ₂ F ₂ of FTIR. FIG. 20 shows a measurement curve obtained and a quantitative transition of concentration over time. From the figure, the amount of SO ₂ F ₂ generated increases rapidly after several hours from immediately after the voltage application, and is approximately twice as high as when the voltage application is stopped in about 60 to 70 hours after the voltage application is stopped. It can be seen that the value converges. The amount of SO ₂ F ₂ generated immediately after voltage application does not increase because a certain amount of time is required until SF ₄ , SOF ₄ and SOF ₂ are converted to SO ₂ F ₂ (see FIG. 7). It seems that there is a delay. This can be inferred from the fact that the amount of SO ₂ F ₂ gradually increases even after the voltage application is stopped. _Accordingly, in the amount of SF 4, SOF ₄ and SOF ₂ is constant, if the conversion ratio of the _SO 2 _{F 2} is constant, after the delay time described above, _SO 2 _{F 2} is believed to increase the constant It is done. Here, if it is based on the above-mentioned assumption, the general formula which considered application time and gas volume about Formula 3 is given by Formula 4.

_However, C SO2F2: _{SO 2 F} 2 concentration [ppm], Q _total: total amount of charge in the AC1 cycles [pC], t: application time [time], Vg: gas volume at 0.1MPa terms [l] is there.

  As described above, Equation 4 is based on several assumptions, and although it is necessary to improve the accuracy of each parameter, it is a basic equation for estimating the concentration of cracked gas.

  A summary of the transition of FTIR absorbance from the start of voltage application to about 80 hours for each cracked gas is shown in (6) below.

(3.4) Verification of the external adsorbent unit In this section, the external adsorbent unit (the adsorbent 4 is housed in a sealed container 5 separate from the ground tank 2 and unitized; In order to verify the purification of the cracked gas by the closed container 5 and the grounded tank 2 in communication), the high-concentrated cracked gas having a known cracked gas concentration (level 3 gas, SO ₂ F in the previous section) _Two experiments were carried out: the case where the adsorbent unit was applied to ₂ concentrations (560 ppm) and the case where the partial discharge and the operation of the adsorbent unit were performed simultaneously. In the previous test, SO ₂ F ₂ was quantitatively analyzed for the gas after adsorption, and the effect of the adsorption unit was quantitatively evaluated. As the adsorbent, a synthetic zeolite adsorbent “Zeoram” manufactured by Tosoh Corporation, type: F-9, spherical product, size 4-8 # was used. In each case, the adsorbent was 50 g (about 100 cc), and the amount of gas sent to the adsorbent unit was about 3 liters per minute.

When the adsorbent unit is applied to the level 3 gas (SO ₂ F ₂ concentration 560 ppm) in the previous section (operating for about 1 hour) in FIG. 21, and when partial discharge and the adsorbent unit are operated simultaneously in FIG. The analysis result by FTIR is shown. Table 5 summarizes the results.

FIG. 21 shows that the decomposition gas is almost adsorbed by operating the adsorbent unit. Moreover, as shown in Table 5, the indicated value of the SO ₂ gas detector tube before and after the operation is changed from 100 ppm to 0 (no change) ppm, and the quantitative analysis result of SO ₂ F ₂ is from 560 ppm to 0.011 ppm to about 1/50000. It was decreasing. The water content in SF ₆ was also less than the lower limit of quantification (<56 ppm). In addition, the partial discharge amount when the partial discharge and the operation of the adsorbent unit are performed simultaneously (FIG. 22) is as shown in Table 2. From the parameters related to partial discharge, level 3 in the previous section (SO ₂ F ₂ concentration 560 ppm) It is estimated that the same level of cracked gas is generated. However, it can be seen from FIG. 22 that the FTIR is so clean that there is no sensitivity. As a result of the inspection by the gas detection tube at the end of the experiment, there was no sensitivity for SO _{2 and} it was less than the lower limit of quantification (0.17 ppm) for HF (the reagent near the gas inlet was only slightly discolored). Yes, the color is different from the color specified by HF). From the above, it has been found that the gas in the equipment can be kept sufficiently clean by installing an adsorbent unit outside and forcibly circulating the gas. (5) shows the details of the partial discharge generated in this test.

(3.5) Detection of cracked gas in an actual scale experimental system In this experiment, an actual machine was assumed, and 0.5 MPa (abs.) Of SF ₆ gas was sealed in a large actual scale tank (gas capacity 900 liters). Went. Therefore, the amount of gas used is about 290 times that of the basic experimental tank. The generated partial discharge was about 1/2000, about 27 pC in the case of the basic experimental tank, based on the total charge amount, and the voltage application time was continuous 72 hours. FIG. 23 shows temporal transitions of the number of pulses per commercial frequency (50 Hz) per cycle (20 ms), total charge amount, average charge amount, and maximum charge amount for the partial discharge generated in the experiment. Table 6 summarizes the partial discharge and the gas analysis results. Both measurements were performed 10 times per hour, but the maximum charge amount and average charge amount were the maximum and average values per pulse, and the number of partial discharge pulses and total charge amount were per commercial frequency (50 Hz) per cycle. It was shown as an average value. According to Table 6, the impression that the maximum charge amount is relatively large is received, but this occurs very rarely during the 72-hour experiment, and the average charge amount is extremely small at 4.6 pC. FIG. 24 shows the histogram of the charge amounts of all partial discharge pulses measured in this test. The vertical axis of the figure is standardized by the largest number (3.8 pC or less), and it can be seen that there is almost no partial discharge exceeding 20 pC.

Also in this test, SO ₂ F ₂ was quantitatively analyzed for the gas for about 70 hours after the gas detector tube and the voltage application were stopped. The decomposition gas could not be detected by FTIR. In the gas detector tube, immediately after the voltage application was stopped, in HF, the indicator was slightly discolored, indicating the presence of decomposition gas, although it was less than the lower limit of determination (0.17 ppm). However, when the inspection with the gas detector tube was performed again 24 hours later, there was no change in any of the detector tubes. As a result of quantitative analysis of _SO 2 _{F 2} by gas chromatography, 0.029 [ppm] was confirmed. From the above, it was found that the partial discharge can be sufficiently detected by accumulation for about 72 hours under the present experimental conditions (gas volume, partial discharge level).

In addition, the concentration of SO ₂ F ₂ under these experimental conditions is calculated as follows using Equation 5.

  Although there is a slight difference between the calculation result and the actual measurement value, Formula 4 is proposed based on the limited experimental data at the present time, considering that the order handled here is ppb (1/1000 of ppm) By doing so, it is considered possible to obtain a more accurate empirical formula by repeating experiments in the future.

(3.6) Detection of cracked gas from adsorbent unit In the above section (3.4), in order to verify the effect of the adsorbent unit, high concentration cracked gas was adsorbed on the adsorbent (Table 5). In Experiment 1), the decomposition gas was detected from this adsorbent. Since the adsorbent used this time (synthetic zeolite adsorbent) is a physical adsorption type, the "heated expulsion method" was used to expel the gas adsorbed on the adsorbent. This is a method of heating the adsorbent to 300 ° C. to drive out the decomposition gas, and quantitative analysis was performed on SO ₂ F ₂ in the discharged gas. The amount of the adsorbent used for expulsion was 20 [g], and the capacity of the heating container used for expulsion was 0.028 liter. As a result, the component concentration in the gas phase portion was 1.5 ppm. Therefore, the volume of SO ₂ F ₂ in the heating container (V _SO2F2 ) is expressed by Equation 6.

[Equation 6]
_VSO2F2 = 1.5 × 10 ⁻⁶ × 0.028 = 4.2 × 10 ⁻⁸ [liter]

When this is converted into mass (M _SO2F2 ), Equation 7 is obtained.
[Equation 7]
M _SO2F2 = 4.2 × 10−8 / 24.46 × 102.07
= 0.18 × 10 −6 [g]
= 0.18 [μg]
However, 24.46: volume of 1 mol of ideal gas at 25 ° C. [liter], 102.07: molecular weight of SO ₂ F ₂ .

Since the amount of adsorbent used for expulsion is 20 g, the separation amount of SO ₂ F ₂ from the adsorbent per unit mass is about 0.01 [μg / g]. Further, since the total amount of the adsorbent was 50 g, 0.5 μg can be separated from the entire adsorbent. On the other hand, since SO ₂ F ₂ before passing through the adsorbent is 560 ppm and the total gas amount is 15.5 liters, the mass of SO ₂ F _{2 is} similarly determined to be 36.2 [mg]. Therefore, it is considered that SO ₂ F ₂ separated from the adsorbent is only 0.0014 [%] of the whole, and most of the SO ₂ F ₂ remains adsorbed on the adsorbent. This may be due to the lack of heating temperature in the heat removal method. However, SO ₂ F ₂ expelled from the adsorbent has a concentration of 1.5 ppm (1500 ppb) even at present, and has sufficient sensitivity when considering the resolution of gas chromatograph (0.5 ppb). It can be said that.

(4) Next, the spectrum when the gas pressure is 0.5 MPa in the FTIR analysis will be described.
FIG. 25 shows a spectrum when the gas pressure is 0.5 MPa in the FTIR analysis. From FIG. 5A, it can be seen that SOF ₄ has no sensitivity. However, the sensitivity of SOF ₂ and SO ₂ F ₂ was improved compared to the case where the gas pressure was 0.1 MPa.

(5) Next, the details regarding the partial discharge generated in the basic experiment tank will be described.
Fig. 26 shows the details of the partial discharge generated in the test conducted using the basic experimental tank (Section (3.2): Evaluation of degradation gas aging characteristics, Section (3.4): Verification of external adsorbent unit). (Gas pressure 0.1 MPa, continuous application time 8 hours). This figure shows the temporal transition of the number of pulses, total charge amount, average charge amount, and maximum charge amount in one cycle (20 ms) at a commercial frequency of 50 Hz. In all cases, measurements were performed 10 times per hour, and the maximum charge amount was the maximum value during measurement, and the others were average values. In the figure, the largest partial discharge amount (level 3 in section (3.3)) in which the amount of SO ₂ F ₂ produced can be quantified is also shown for reference. As can be seen from the figure, the partial discharge amount equivalent to “Level 3” was generated for “degradation gas aging characteristics evaluation” and “verification of adsorbent unit”, and the cracked gas was also generated to the same extent. It is thought that there is.

FIG. 27 shows a histogram representing the polarity of partial discharge and the distribution of charge amount per pulse in each experiment. The histogram is a cumulative value of all measured values, and is a standardized value with the highest number of levels 3 in which cracked gas (SO ₂ F ₂ ) can be quantified (ie, the maximum at level 3). The value is 1). In either case, it can be seen that negative partial discharge is dominant (note that the needle electrode is placed on the ground side).

(6) Next, the change with time of the cracked gas (analysis result by FTIR) will be described.
28 to 30 show temporal transitions of the FTIR absorbance for each cracked gas when the partial discharge level is changed.

It is a schematic block diagram which shows 1st Embodiment of the gas insulated power apparatus which implements the abnormality detection method of this invention. It is a schematic block diagram which shows 2nd Embodiment of the gas insulated power apparatus which implements the abnormality detection method of this invention. It is a schematic block diagram which shows the example of the structure which expands / contracts the inner pipe | tube of a communicating path. The 1st and 2nd on-off valve is shown, (A) is sectional drawing of an open circuit state, (B) is sectional drawing of a closed state. It is a schematic block diagram which shows the modification of the gas insulated power apparatus of FIG. It is a schematic block diagram which shows 3rd Embodiment of the gas insulated power apparatus which implements the abnormality detection method of this invention. It is a figure which shows the generation | occurrence | production mechanism of cracked gas. It is a figure which shows electrode arrangement | positioning. It is a figure which shows the equipotential surface of a needle point, and electric potential distribution (r = 0 is rotational symmetry). It is a figure which shows arrangement | positioning of gas piping and each apparatus. It is a figure which shows electrode arrangement | positioning. It is a figure which shows an example of the observed partial discharge current. Is a diagram illustrating the FTIR spectra of various SF ₆ decomposition gas (library and literature value). _{SF 6 (0.1MPa (abs.)} ) Is a diagram showing an emissivity in the case of the background. An example of the spectrum of cracked gas is shown, (a) is a figure about wave numbers 700-950, (b) is a figure about wave numbers 1200-1400, (c) is a figure about wave numbers 1400-1600. Shows the time course of the decomposition gas, which is a diagram for (a) drawing of SF ₄ and SOF ₄ is, (b) the figure for the SOF _2, (c) the _SO 2 _{F 2.} FIG. 7 shows the temporal transition of the partial discharge amount, (a) is a diagram regarding the number of partial discharge pulses, (b) is a diagram regarding the total charge amount, (c) is a diagram regarding the average charge amount, and (d) is the maximum charge. It is a figure about quantity. The relationship between the total charge amount and the FTIR absorbance is shown, (a) is a diagram for SF ₄ and SOF ₄ (absorbance immediately after voltage application is stopped), and (b) is a graph for SOF ₂ (absorbance immediately after voltage application is stopped). a diagram of (c) is _SO 2 _{F 2} (absorbance after the voltage application is stopped for about 70 hours). It is a diagram showing the relationship between total charge and _SO 2 _{F 2} concentration. FIG. 5 shows the temporal transition of the amount of SO ₂ F ₂ produced, where (a) is a diagram for level 1, (b) is a diagram for level 2, and (c) is a diagram for level 3. FIG. Results of the validation of the adsorbent unit to high concentrations cracked gas, is a diagram for (a) drawing of SF ₄ and SOF ₄ is, (b) the figure for the SOF _2, (c) the _SO 2 _{F 2} . The partial discharge occurrence and the result of the synchronous operation of the adsorbent unit are shown, (a) is a diagram for SF ₄ and SOF ₄ , (b) is a diagram for SOF ₂ , and (c) is a diagram for SO ₂ F ₂ . It is. FIG. 5 shows the time course of partial discharge (basic experiment using an actual scale tank), where (a) is a diagram regarding the number of partial discharge pulses, (b) is a diagram regarding the total charge amount, and (c) is a diagram regarding the average charge amount. (D) is a diagram regarding the maximum charge amount. It is a histogram of a partial discharge charge amount. FTIR analysis shows a spectrum when the gas pressure is 0.5 MPa, (a) is a diagram for SF ₄ and SOF ₄ , (b) is a diagram for SOF ₂ , and (c) is a diagram for SO ₂ F ₂ . FIG. FIG. 7 shows the temporal transition of the partial discharge amount, (a) is a diagram regarding the number of partial discharges, (b) is a diagram regarding the total charge amount, (c) is a diagram regarding the average charge amount, and (d) is the maximum charge amount. It is a figure about. The histogram of the charge amount of each partial discharge is shown, (a) is the histogram for level 1, (b) is the histogram for level 2, (c) is the histogram for level 3, and (d) is the decomposition gas aging characteristic. A histogram for evaluation, (e) is a histogram for verification of the adsorbent unit. The analysis result by FTIR (partial discharge amount: level 1, SO ₂ F ₂ concentration: 5.7 ppm) is shown, (a) is a diagram for SF ₄ and SOF ₄ , (b) is a diagram for SOF ₂ , and (c ) Is a diagram for SO ₂ F ₂ . The analysis results by FTIR (partial discharge amount: level 2, SO ₂ F ₂ concentration: 33 ppm) are shown, (a) is a diagram for SF ₄ and SOF ₄ , (b) is a diagram for SOF ₂ , and (c) is it is a diagram for SO ₂ F _2. The analysis results by FTIR (partial discharge amount: level 3, SO ₂ F ₂ concentration: 560 ppm) are shown, (a) is a diagram for SF ₄ and SOF ₄ , (b) is a diagram for SOF ₂ , and (c) is it is a diagram for SO ₂ F _2.

Explanation of symbols

1 Insulating gas 2 Ground tank 4 Adsorbent 5 Sealed container 9 Communication path 13 Sampling port

Claims (3)

The adsorbent that adsorbs the decomposition gas generated in the grounding tank in which the insulating gas is sealed is housed in a sealed container different from the grounded tank, and the sealed container and the grounded tank communicate with each other, When the cracked gas is introduced into the sealed container and adsorbed and removed by the adsorbent, when an abnormality is detected, the adsorption is performed from a sampling port provided in a communication path from the ground tank to the adsorbent. A method for detecting an abnormality in a gas-insulated power device, comprising: collecting a gas before contacting a material and detecting the occurrence of an abnormality in the grounded tank based on detection of SO ₂ F ₂ . The adsorbent that adsorbs the decomposition gas generated in the grounding tank in which the insulating gas is sealed is housed in a sealed container different from the grounded tank, and the sealed container is detachably connected to the grounded tank. In the closed container and the ground tank, the cracked gas is introduced into the sealed container and adsorbed and removed by the adsorbent. On the other hand, when abnormality detection is performed, the ground tank side communication path is closed. Then, the closed container is separated from the grounded tank, the adsorbent is analyzed, and an abnormality in the grounded tank is detected based on the detection of SO ₂ F _2. Detection method. In addition to the SO ₂ F _2, SF ₄ and SOF ₄ and based on the detection of at least one gas of SOF ₂ and detecting the occurrence of abnormality in said ground tank according to claim 1 or 2 The method for detecting an abnormality of the gas-insulated power equipment described.