Search

To lower the lightning impulse withstand voltage of gas insulated switchgear (GIS) while maintaining the high reliability of its insulation performance, it is important to define in an organized way the insulation characteristics for non-standard lightning impulse voltage waveforms that represent actual surge waveforms in the field and compare them with the characteristics for the standard lightning impulse waveform quantitatively. In the previous work, the dielectric breakdown voltage - time characteristics were measured under several different conditions on the quasi-uniform SF6 gas gap and partly the cone-shaped insulating spacers that represent an insulation element of GIS for six kinds of non-standard lightning impulse waveforms associated with lightning surges and disconnector switching surges. This paper, describes detailed examination in non-uniform electric field, and the pattern in which a disconnector switching surges are superposed on residual DC elements generated through the operation of disconnectors (bias voltages). This is intended to simulate more severe conditions in an actual system. Consequently, it was found that the presence of a bias voltage in the nonuniform electric field resulted in lower breakdown voltage values for single-frequency oscillation waveforms both in the positive and negative polarities. However, the breakdown voltage values under positive-polarity waveforms posing more severe insulation conditions were higher than the ones under the standard lightning impulse waveforms. Despite the presence of a bias voltage, it was possible to evaluate these breakdown voltage values in the positive polarity using the concept of a duration examined in the past.

SF6 gas, an insulation medium used for gas insulated switchgear (GIS), is known for its high global warming potential, hence the search for an effective alternative from an environmental perspective. As a potential alternative, the authors are focusing on N2 gas. To use this N2 gas for actual GIS, the insulation characteristics for actual overvoltage waveforms generated in the field (non-standard lightning impulse waveforms) must be obtained. To do so, the earlier study experimentally obtained and evaluated the insulation characteristics for single-frequency oscillation waveforms as a representative among nonstandard lightning impulse waveforms under the quasi-uniform electric field, which was considered to be a basis for the insulation design of GIS. The present study obtained the insulation characteristics for double-frequency oscillation waveforms to investigate surge waveforms generated in substations in more detail. Consequently, it emerged that negative polarity waveforms were more severe in terms of insulation, as was the case for single-frequency oscillation waveforms. The insulation performance was consistently higher than that for the standard lightning impulse waveforms, which were 1.08 to 1.09 times. In other words, it was confirmed that the insulation specification could be rationalized by about 10%, even for complex oscillation waveforms, such as those of double-frequency oscillation. Furthermore, a study was conducted on the influence of the gas pressure and gap length for single-frequency oscillation waveforms. Consequently, it emerged that the results under the basic experimental conditions were likely to be applicable to actual GIS under wider-ranging conditions.

SF6 gas, an insulation medium used for gas insulated switchgear (GIS), has a high global warming potential, hence the search for an effective alternative is required from an environmental perspective. The authors focus particularly on CO2 gas, which has a low global warming potential, as one of these potential alternatives. In order to apply this CO2 gas to actual equipment, the insulation characteristics for overvoltage waveforms generated in an actual field (called non-standard lightning impulse waveforms) must be obtained. Accordingly, in previous studies, the insulation characteristics for single-frequency oscillation waveforms were experimentally obtained in the quasi-uniform electric field, which is the basis when performing the insulation design of GIS, and its evaluation method was examined. In this paper, to further elaborate the evaluation method established thus, the insulation characteristics in relation to complex oscillatory waveforms generated in actual substations, e.g. double-frequency oscillation waveforms and single-frequency oscillation waveforms on which a dc component (bias voltage) that remains due to the superimposed operation of a switching device, were obtained. Consequently, the results obtained indicated that negative polarity waveforms were more severe in terms of electrical insulation under the double frequency oscillation waveforms as well as under the single-frequency oscillation waveforms. Therefore, when considering a dielectric strength, it is reasonable to conduct experiments using negative-polarity waveforms for the complex oscillatory waveforms, such as the double frequency oscillation waveforms. Furthermore, a study was performed on a method of evaluating insulation characteristics for non-standard lightning impulse waveforms of the CO2 gas gap via a comprehensive approach using the results including the previous ones. As a result, as in the case of studies conducted previously on SF6 gas, the insulation characteristics, including complex oscillatory waveforms generated in actual power systems, could be expressed by a single characteristic line by using the parameter of duration.

The soundness of a power transformer under an operating voltage is evaluated in partial discharge (PD) test of long-duration induced ac voltage test. The acceptance criterion for this PD test is 500 pC according to IEC standard; however, few basic data backing this criterion are available. To establish a clear criterion for this PD test, the authors initially conducted a study on the harmful PD level of materials themselves constituting the insulation of an oil-immersed transformer and a gas-filled transformer. This PD level was evaluated based on the rate of decline in the lightning impulse breakdown voltage using the insulating materials themselves and the element models simulating the insulating structure which had been exposed to a PD. Consequently, it emerged that insulations of both types of transformers were degraded if exposed to a PD of 7,000 pC to 10,000 pC. With the safety factor for this PD value taken into account considering the long-term operation and the structural difference of an actual transformer, 5,000 pC was deemed as the harmful PD level at the PD occurrence position. Subsequently, using a winding model of a transformer, PD signal propagation characteristics inside the winding were investigated through actual measurement and analysis. As a result, it was found that the PD having occurred inside the winding is measured as the signal significantly damped depending on the position of occurrence. The transmission rate was 2.2% for an oil-immersed transformer and 2.8% for a gas-filled transformer in the respective lowest cases. What should be controlled in the PD test of an actual transformer is the value of the harmful PD level at the PD occurrence position multiplied by the transmission rate at the PD detection position. Therefore, the conclusion was reached that the acceptance criterion in the test should be set to 5,000 pC × 2.2% = 110 pC or less for an oil-immersed transformer and 5,000 pC × 2.8% = 140 pC or less for a gas-filled transformer, respectively.

The technical soundness of an oil-immersed power transformer under an operating voltage is evaluated in partial discharge (PD) test of long-duration ac withstand voltage test. To establish clear control criteria for this PD test, it is necessary to study PD propagation characteristics inside the transformer windings. In this paper, an analysis model was constructed using constants in the oil for a 500 kV-class transformer to analytically obtain PD signal propagation characteristics. For this analysis model, a circuit model for high frequencies up to several MHz was used in order to obtain the characteristics of highfrequency PD propagating inside the winding in detail. Using this model, PD propagation characteristics were investigated in terms of the PD occurrence position, measurement position, and front time as parameters. As a result, it emerged that the propagation pattern was determined by the winding constants as viewed from the PD injection position and the terminal conditions. When the charge amount was the same, the steeper the PD signal front time, the higher the peak voltage value was likely to be because the waveform became more oscillatory. Conversely, assuming an actual test using the ERA device (PD detector developed by Electrical Research Association), it was found that the front time had little impact on the charge amount measured because the oscillatory high-frequency components were removed. As with the previous studies, it emerged that the PD signal transmission rate was lowest when a PD occurred between sections and the signal was detected as one significantly damped to a level of several percent depending on the conditions.

SF6 gas has been widely used in electrical power equipment such as gas insulated switchgear (GIS) due to its superior insulation and interruption characteristics. However, in 1997, SF6 gas has been designated a greenhouse gas subject to emission restrictions at COP3 (The 3rd session of the Conference Of the Parties to the United Nations Framework Convention on Climate Change). Therefore, it is important to study the possibility of using environmentally friendly gases as alternative insulation gases for practical gas-insulated apparatuses, such as GIS. In the selection of an alternative gas, it is thought that high-pressure natural gases, such as Air, N2, and CO2 are promising environmentally friendly candidates. The authors are focusing on N2 gas as one of its potential alternatives. In order to use this N2 gas for actual GIS, the insulation characteristics for actual overvoltage waveforms generated in the field (nonstandard lightning impulse waveforms; non-standard-LIWs) must be obtained. The present study, with single-frequency oscillation waveforms as a representative basis, experimentally obtained the insulation characteristics by changing the polarity, frequency, and damping rate of the applied voltage in the non-uniform electric field represented by the presence of metallic particle in the equipment. As a result, the breakdown voltage was lower for the positive polarity under the non-uniform electric field. In this polarity, the breakdown voltage was 1.06 to 1.74 higher than that by standard lightning impulse waveforms even if the frequency and damping rate were changed. These results support the rationalization of insulation for actual surge waveforms under the quasi-uniform electric field.

To establish clear control criteria for a partial discharge (PD) test of long duration ac withstand voltage test of a gas-filled power transformer, the authors have conducted a study on the harmful PD level of materials themselves using the insulating material- and structural-models of an actual gas-filled power transformer. In the PD test on an actual transformer, since the PD having occurred inside the transformer is measured from the outside, the measured signal is damped due to various influential factors, such as the position of occurrence, measurement position, and internal structure of the equipment. In particular, in the event of the PD between sections or turns, which might actually occur, significant damping of the signal is anticipated in certain cases due to the influence of the position of occurrence and the internal structure of the equipment, which is a propagation path. Therefore, this damping must be taken into consideration to establish control criteria for the PD. In this paper, to understand the characteristics of this damping, such as its magnitude, two types of winding, interleaved disk and continuous disk, were produced as transformer models to investigate the propagation characteristics of PD signals with the position of occurrence, measurement position, and wavefront duration of the PD as parameters. Consequently, it emerged that, where the PD occurs between sections or turns, the transmission rate at the measurement position was generally low because the PD signal circulates within the PD occurrence parts and does not come outside. Assuming measurement at the bushing test terminal, the lowest transmission rate was 2.8% and 2.0% between sections and turns, respectively. A study was also conducted on the occurrence of the PD against ground, despite the low anticipated potential of actual occurrence, and the lowest transmission rate was 4.0%. Accordingly, the PD having occurred inside the winding of the transformer can be detected with a magnitude of 2% or more; however, from a reverse perspective, the signal is detected as one significantly damped to a level of about 1/50.

SF6 gas, an insulation medium used for gas insulated switchgear (GIS), has a high global warming potential, hence the search for an effective alternative is required from an environmental perspective. As one of its potential alternatives, the authors are focusing on N2 gas, which has relatively good insulation characteristics among natural gases and a low environmental impact. However, to use this N2 gas for actual GIS, the insulation characteristics for actual overvoltage waveforms generated in the field (non-standard lightning impulse waveforms; non-standard-LIWs) must be obtained. The present study, with single-frequency oscillation waveforms as a representative basis among non-standard LIWs, experimentally obtained the insulation characteristics of an N2 gas gap by changing the frequency and the damping factor. Consequently, the breakdown voltage (BDV) was lower for the negative polarity under a quasi-uniform electric field, hence the conclusion that it was reasonable to conduct experiments using negative polarity waveforms in order to discuss the dielectric strength. In addition, even if the frequency and damping factor changed, the BDV remained relatively constant, and that is consistently higher than that for the standard lightning impulse waveforms (standard-LIWs) at a level of 1.14 to 1.24 times. As a result, it was also confirmed for GIS using N2 gas that the insulation specification could be rationalized by about 20% by converting lightning impulse waveforms from non-standard to equivalent standard-LIWs of 1.2/50 μs. Furthermore, insulation characteristics were obtained under the pattern in which a disconnector switching surges are superposed on residual dc elements generated through the operation of disconnectors (bias voltages). Consequently, the bias voltage had only a minor influence on the insulation characteristics and it emerged that insulation characteristics were likely to be evaluated using a single polarity waveform.

The soundness of a gas-filled power transformer under an operating voltage is evaluated in partial discharge (PD) test of long-duration ac withstand voltage test. To establish clear control criteria for this PD test, the authors initially conducted a study on the harmful PD level of materials themselves used in a gas-filled transformer. In the actual PD test, since the PD having occurred inside the transformer is measured from the outside, the measured signal is damped due to various influential factors. Since this damping must be taken into consideration to establish control criteria, a preceding study investigated the propagation characteristics of PD signals through actual measurement using transformer winding models. As a result, it was found that the PD having occurred inside the winding is detected as the signal significantly damped depending on the position of occurrence. This paper studies whether PD signal propagation characteristics can be analytically evaluated using the equivalent circuit of the transformer winding. This analytical evaluation, if feasible, can be applied to various winding structures and will make it possible to evaluate the propagation characteristics in further detail as well as the validity of the actual measurement results. The study was conducted on the interleaved and continuous disk windings to make a comparison between the analysis and measurement results of the voltage waveform and the propagation pattern at various locations inside the winding with the position of occurrence and measurement position of the PD as parameters. Consequently, it emerged that the analysis using the equivalent circuit could adequately simulate the propagation pattern and waveform shape of the PD signal. A detailed study using this equivalent circuit also clarified the fact that the PD signal propagation waveform was determined by the winding constant and the terminal conditions. Furthermore, the analytical and measurement results of the transmission rate were almost identical, whereby it was analytically confirmed that the transmission rate significantly decreased depending on the position of occurrence of the PD. It also emerged that, since the frequency characteristics of the PD measurement device may influence the transmission rate, the PD control criteria must be studied taking such influence into consideration.

The lightning impulse withstand voltage for an oil-immersed power transformer is determined by the value of the lightning surge overvoltage generated at the transformer terminal. This overvoltage value has been conventionally obtained through lightning surge analysis using the electromagnetic transients program (EMTP), where the transformer is often simulated by a single lumped capacitance. However, since high frequency surge overvoltages ranging from several kHz to several MHz are generated in an actual system, a transformer circuit model capable of simulating the range up to this high frequency must be developed for further accurate analysis. In this paper, a high frequency circuit model for an oil-immersed transformer was developed and its validity was verified through comparison with the measurement results on the model winding actually produced. Consequently, it emerged that a high frequency model with three serially connected LC parallel circuits could adequately simulate the impedance characteristics of the winding up to a high frequency range of several MHz. Following lightning surge analysis for a 500 kV substation using this high frequency model, the peak value of the waveform was evaluated as lower than that simulated by conventional lumped capacitance even though the front rising was steeper. This phenomenon can be explained by the charging process of the capacitance circuit inside the transformer. Furthermore, the waveform analyzed by each model was converted into an equivalent standard lightning impulse waveform and the respective peak values were compared. As a result, the peak value obtained by the lumped capacitance simulation was evaluated as relatively higher under the present analysis conditions.

In many cases, the vacuum circuit breaker is used as the single-break, and investigation is not conducted so much about the double-break vacuum circuit breaker. So, we are investigating the relation between the post arc current and the voltage distribution ratio in the double-break vacuum circuit breaker. In this investigation, the post arc current and the voltage distribution ratio were measured in order to clarify the relation between the post arc current and voltage distribution ratio. Experiments are measurement of the post arc current which flows into one vacuum interrupter, and measurement of the voltage distribution ratio and post arc current of a double-break vacuum circuit breaker. The results of investigation showed that the bias of voltage distribution ratio occurs when the post arc current value which flows into each vacuum interrupter in a double-break vacuum circuit breaker became imbalanced. In addition, it turned out that the post arc current of spiral electrode is very scattering as compared with axial magnetic field electrode. Therefore, it turned out that it is much to use the axial magnetic field electrode for a double-break vacuum circuit breaker.

It is important to comprehend the interruption phenomenon of contact materials for aiming at adaptation to high voltage of vacuum circuit breakers. Then, we carried out measurement of arc energy and anode surface temperature and observation of vacuum arc by using the vacuum chamber. Moreover, we made prototype of new contact materials that was based on these tests result. And these contact materials were made for adaptation to high voltage of vacuum circuit breakers. Additionally, we carried out above tests and interruption test about these contact materials.