Search

The performance of a high-voltage switching impulse (SI) measurement system is highly dependent on the characteristics of the used voltage divider. Self-built or commercial voltage dividers do not always fulfill the requirements set for reference-level devices. Software corrections to improve the non-ideal divider response are easy to implement. This article describes the traceability and performance of the SI measurement systems of VTT MIKES up to 400 kV. Traceability is based on a calculable impulse voltage calibrator, which is used for calibration of a reference digitizer and a 10-kV divider. The 10-kV divider is then used to calibrate the 400-kV modular divider. Responses of both dividers are corrected using two different software methods. Implementation of these software methods allows state-of-the-art impulse parameter uncertainties. The most significant uncertainty component of time to peak is related to its definition and evaluation according to IEC 60060-1:2010 and IEC 61083-2:2013.

The breakdown and streamer behavior in double-layer oil-paper insulation under impulses of varying front time were investigated in this study to determine the prebreakdown process in oil-paper insulation. The relationships between front times and breakdown characteristics were investigated and impulse-polarity-dependent variations were observed. The complete impulse breakdown process and precise role of front time were assessed as per streamer initiation and streamer propagation based on emitted light signals and shadow images. The re-initiation of streamers in non-first oil gaps was determined to be the key factor in breakdown as-dominated by applied voltage and space charge accumulation. The initiation and propagation times of streamers were observed to find that the streamers tend to propagate at phases with higher voltage amplitude as front time increases, which facilitates streamer re-initiation and thus decreases the breakdown voltage. Equivalent applied voltages during streamer propagation were further calculated to quantitatively investigate the effect of front time on streamer propagation.

Partial discharge inception voltage measurements under repetitive impulsive voltages play an essential role in evaluating the reliability of inverter-fed low voltage rotating machines. Using impulsive voltages, partial discharge (PD) signals and the disturbance induced by power electronics switches have often similar energy distribution in the frequency domain. Thus, PD test technologies used with sinusoidal voltages are no longer effective. Under repetitive impulsive voltages, ultra-high frequency detection is recommended to couple high-frequency spectral components from the PD and remove the low-frequency disturbance coming from power electronics. However, to obtain high signal to noise (S/N) ratio, the parameters of the antenna used for PD detection should be designed considering the PD and the disturbance energy distributions in the frequency domain. In this paper, we discuss the design and show the properties of an antenna with Archimedes plane spiral structure intended for PD detection under repetitive impulse voltages. The simulation shows that the antenna has low voltage standing wave ratio (

There has been a vast vacuum in the literature on the importance of the location of remote earth carried out during impulse test on grounding systems using field measurements. Some published work show that the remote earth is installed surrounding the electrode under tests, whilst some have used separate remote earth, away from it. This work studies the impact of return electrode location on transient behaviour of the earthing electrode. The electrode under test is a single rod. Two types of remote earth configurations are used, a circular electrode with multiple rods surrounding the single rod, and one that is 80 m away from it. Impulse tests are conducted under both impulse polarities. Voltage and current traces are captured for a single rod, using two different types of remote earth configurations, with the impulse resistance measured as well. It has been established that the arrangement of the remote earth poses no influence on the test results. That said, higher oscillations of current traces and higher resistance values have been observed in negative impulse in comparison with positive impulse polarities.

This paper presents the results of the application of a novel method to estimate the parameters of the Disruptive Effect Model for the prediction of the volt–time curves of 15 kV insulators. Besides the standard lightning impulse voltage waveshape, four non-standard impulse waveshapes (namely 1.2/4 μs, 1.2/10 μs, 3/10 μs, and 7.5/30 μs), of both polarities, were adopted in the tests. The measured volt–time characteristics of a 15 kV class pin-type insulator were compared with those predicted by three different procedures related to the Disruptive Effect Model: Darveniza and Vlastos, Hileman, and the one recently proposed by the authors. The results show that for some impulse waveshapes and polarities, the methods by Darveniza and Vlastos and by Hileman do not predict insulator breakdown for the lower voltage levels. On the other hand, a relative good agreement is found between theoretical and experimental results for the calculations performed using the method proposed by the authors.

The disconnector is the key electrical equipment in DC field. In China, the ±1100 kV DC project uses the indoor DC field, where the switching impulse withstand voltage of the disconnector influences the clearance a lot. To improve the switching impulse withstand level of the ±1100 kV disconnector, first of all, the switching impulse test was carried out on the disconnector with traditional double-balanced ring structure and the 50% switching impulse discharge voltage was obtained. By analysing the discharge path, it is found that the grading ring did not effectively shield the upper part of the disconnector. Therefore, the structure of the end fittings of the disconnector was changed to the structure of the mushroom head. The electric field simulation was performed on the optimised shield，it is shown that the maximum field strength is lower than the control field strength. Then the switching impulse test was carried out on the optimised disconnector, it was found that the 50% impact discharge voltage was above 2400 kV which met the engineering requirements.

The goal of the paper is testing the short-pulse dielectric strength of chip capacitors and lightning impulse testing with subsequent failure analysis. The experimental setups are developed in order to determine the short-pulse dielectric strength and lightning impulse withstand voltage. The distribution functions of these characteristics have been found to follow the normal distribution. The oscillograms of current and voltage across the capacitors during the tests are presented. A brief failure analysis is carried out. The main failure mechanisms are highlighted. The obtained results will be the basis for the method of preliminary estimation of short-pulse dielectric strength in chip capacitors.

The effect of the coaxial cable type and length on the lightning impulse parameters is often neglected. However, studies have shown that the coaxial cable length can affect the parameters that are calculated according to IEC 60060-1:2010. This study investigates how different coaxial cables influence the impulse waveforms by measuring the input and output of the cable under test with a 2-channel digitizer. Non-standard lightning impulses were generated using an impulse voltage calibrator. Results indicate that coaxial cables attenuate the test voltage and increase both the front time and time-to-half value.

This work presents an extension of our previous work on voltage stress analysis of a 3 MVA, 33/11 kV, 3-phase, 50 Hz, Dyn 11 transformer against the application of standard and non-standard impulse voltages. A preliminary comparative study on the variation of maximum voltage to ground and voltage across the coils under different impulse voltages indicated that non-standard impulse voltage waveforms develop higher voltage stress in the windings. They pose high risk and are critical to transformer insulation system. The obtained test results provide the basis for further study of non-standard impulse voltage waveforms and make necessary correction in the existing impulse testing standards. However, to take care of the stringent effect of actual lightning and other impulse voltages occurring in practical field a more detailed study considering a wide variety of non-standard impulse voltage waveforms is aimed to be done in future.

A concept to predict the Lightning Impulse Voltage (LIV) breakdown probability of medium voltage Vacuum Interrupter (VI) tubes has been developed and tested for the first time employing the Voltage Holding Prediction Model (VHPM) originally formulated at Consorzio RFX. The VHPM is capable to calculate the Weibull breakdown probability curve of any multi-electrode multi-voltage system insulated in vacuum under dc voltage. Even though the possibility to employ the VHPM in the prediction of voltage breakdown under LIV conditions is not straightforward, the potential benefits of the VHPM usage in VI tube technology with regard to development, design and testing at medium and high voltage levels, strongly recommend the assessment of VHPM under LIV conditions. The paper aims to describe the methodology adopted to identify experimentally the VHPM parameters to be employed in such a novel application to assess its prediction capability. The measurements have been done in the Siemens Berlin VI tube factory test stand for LIV, compliant with the IEC standard. Different vacuum tube types manufactured by Siemens were investigated. The main difficulty was to obtain, with such a test bed, voltage breakdown distributions well fitted by a Weibull Distribution — upon which the VHPM is based — primarily due to the difficulty to reach a well-defined end of the conditioning test sequence preceding the relevant voltage breakdown test sequence suitable to be analyzed by the VHPM. Nevertheless, the VHPM prediction resulted in reasonable agreement with the measured probability curves. Finally, investigating VI tubes with spiral-type Radial Magnetic Field (RMF) contacts, characterized by a geometry strongly deviating from axial 2-D symmetry, the first implementation of a full 3-D version of the VHPM has been tested.

The new IEC 60060-1 standard establishes the k-factor function to determine the test voltage value of lightning impulses with superimposed oscillations, but no rule is given to determine experimental k-factor values. Testing methods, testing setups, and formulas to determine k-factor values are some crucial issues to obtain good results. This paper describes the two procedures used in the investigation funded by the European community and referenced in the Annex D of the last edition of IEC 60060-1 standard to determine experimental k-factor for air gaps, one for nonhomogenous fields and other for quasihomogenous fields. Both procedures were implemented using a setup with only 1 generator and a setup with 2 generators. Compatibility of new experimental k-factor values determined for air gaps up to 900 kV in quasihomogenous field with the factor function stated in IEC 60060-1 is shown by means of several examples, but additional examples show a significant discrepancy for air clearances of a sharp rod-plate configuration in the voltage range 250 to 800 kV.

The goal of this work is to study the transient behavior of the cable against the application of standard and non-standard lightning impulse voltage waveforms. A 66 kV cable model has been developed in MATLAB Simulink and standard and non-standard impulse voltages are applied to it. A preliminary comparative study on the obtained voltages indicated that non-standard impulse voltage waveform developed higher voltage stress in the cable. Simulation results helped to investigate which impulses (standard or non-standard) represent the worst possible voltage stresses on the cables. This study provides the basis for further study of effects of non-standard impulse voltage waveforms and make necessary correction in the existing impulse testing standards.

Short circuit fault of a motor, due to breakdown of an insulating material of winding, is one of the most probable faults in motor drive systems. Establishment of an easy and effective fault diagnosis method has been strongly required to assure operation with high reliability. This paper proposes a novel diagnosis method for insulation failure of motor windings by a combination of impulse testing and pattern recognition based on Hidden Markov Model (HMM). A voltage waveform across two winding terminals is recorded under application of an impulse voltage. HMM is exploited to distinguish a small difference in voltage waveforms between healthy and faulty winding insulations. Usefulness of the proposed diagnostic method is verified through voltage waveforms experimentally obtained for motors with several kinds of turn-to-turn insulation failures.

Dielectric withstand voltage tests to verify oil-filled transformer insulation performance are in most cases carried out using two to three types of tests from lightning impulse (LI), switching impulse (SI) and power frequency (PF) withstand voltage tests based on economical factors and past achievements. Test conversion factors (TCF) are used to evaluate whether the tests cover overvoltages not directly dealt with in tests and the omitted tests. In this paper, insulation characteristics for LI, SI, and PF (AC) are first organized including the recent ones. Several sets of TCF are next calculated based on these characteristics, compared with TCF in present standards, and representative values are then derived. Finally, the dielectric withstand voltage tests specified in present standards are compared using the representative TCF, and the AC withstand voltage test was found to be the severest test overall.

Incipient faults are low magnitude and fast decaying transients, lasting only for a few microseconds during impulse tests of transformers. They are very difficult to detect, even for experts, and may cause severe problems leading to faults and consequent outage. Thus, simulation and diagnosis of incipient faults is very important requiring a valid model and advanced detection technique. In this paper, a new arc discharge model is proposed to simulate arcing phenomena between two discs of a power transformer winding, as one type of important incipient fault, during impulse test. Using wavelet transform techniques, a new methodology is developed. It is shown that winding input current, affected by arcing phenomena and obtained through impulse voltage test, provides important information for detection and analysis of this transient type of minor fault. A number of experiments and simulations were performed to evaluate the validity and accuracy of proposed model and diagnostic technique.

To develop a propagation model of a local discharge on a wet polluted insulator, the authors have examined the propagation characteristics and the propagation conditions of the local discharge on an electrolytic solution simulating a wet polluted insulator surface. To calculate the relation between the propagation length of,the local discharge and the time required for the propagation, an electric field distribution near the tip of the local discharge must be known. In this paper, the authors have examined a voltage distribution in an electrolytic solution to obtain the electric field. As a result, the shape of the contacting surface between the local discharge and the solution is obtained under impulse voltage. A calculation of the voltage distribution is performed with the shape. A comparison between calculated values and measured values of the distribution is given.

The main aim of this paper is to discuss important aspects with respect to the switching impulse voltage testing of insulator sets under laboratory conditions. Those factors include mainly rain parameters, clearances to the ground, laboratory walls and larger grounded objects (like a rain tower).

Live working on very old structures often poses unique challenges, such as dealing with the effects of arc rings and the behavior of old insulators. This paper presents the results of tests conducted for Public Service Electric and Gas (USA) to confirm that very old broken insulators can be replaced live on 138 kV double-circuit steel lattice structures with suspension I-strings. Each string had nine porcelain units and two large arc rings with a clear air ring-to-ring distance of 1.07 m (42"). Issues addressed in this paper include: the effects of broken and shorted insulators; the effects of live working tools and workers; and the possibility of selective barehanding. For the above-listed conditions, switching impulse tests were performed on a full-scale mockup of the structure. Actual hardware retrieved from service was used for the mockups. Workers were represented by mannequins. The paper gives switching impulse test results for the outlined conditions. Test data suggest that there is little difference between broken and shorted insulators in the tested cases. However, the CIGRE formula for calculation of the effects of defective insulators is shown to be sufficiently conservative.

We have investigated the influence of surface charges on the discharge development across particle contaminated spacer surfaces under SF/sub 6/ for lightning impulse (LI) voltages (1.2/50 /spl mu/s). Surface charges may be generated by dc, ac or lightning impulse stress. The discharge characteristic shows a strong reduction of the insulation strength if the applied voltage and the surface charge have opposite polarities. The investigations were performed with a needle protrusion attached to the bottom electrode to give severe field distortion. The bottom electrode was biased positively. The influence on the discharge process is observed by measuring the surface charge distribution and predischarge currents. The results reveal changes in streamer onset voltage, streamer to leader transition, and leader development.

Standard spark gaps have lower time limits of gap recovery when testing with high repetition rate impulses and in high-current high-voltage multiple impulse testing. More important, they are not suitable for the generation of opposite-polarity successive impulses using conventional capacitive-type generation techniques. In this article, an improved design incorporating two rotating spark gaps is described. The new spark gap was used to test ZnO surge arresters with two successive impulses of currents up to 5 kA and voltages up to 50 kV. Time intervals as low as 10 ms were achieved between the two impulses. Both similar and opposite polarity tests were carried out successfully.

New power-inverter technology has made sharp voltage rises increasingly common in today's motor and magnet wire insulation applications. That is why voltage pulse performance is an important property of any high-voltage insulation material.

One of the main characteristics of the distribution grid is the reliability of medium voltage overhead lines which can be improved by appropriate insulation coordination. This paper assesses common insulator types in Estonia. Old glass pin insulators and newly installed porcelain pin insulators were tested against different lightning overvoltages (impulse flashovers) in order to determine their volt-time characteristics. Test results show convincingly that the new porcelain pin insulators can be easily physically broken when stressed by lightning impulses. This can lead to the numeral increase of outages in medium voltage overhead lines in the distribution grid.

Power transformers must be evaluated under impulse tests. For that, it is necessary to compare voltage and current waveforms corresponding to reduced-level and full-level voltages. Any difference may correspond to a failure in the transformer under test. This test is based on the principle that there is no change in the generation of the impulse voltage. However, with chopped waveforms, many impulse generators have significant differences in the time of chopping that make difficult to get a conclusion on the test result. This paper presents a new algorithm that generates reduced-level chopped voltage and current waveforms, with arbitrary chopping time, from a couple of reduced voltage of full and chopped waveforms. In this way, it is possible to compare full-level voltage chopped waveforms with reduced-level ones with different chopping times.

Because of the increasing importance of the effects of the nonstandard lightning voltage waves on power apparatus, the Task Force was formed to (1) compile a bibliography of the research on the nonstandard lightning voltages, and (2) review the previous research. The bibliography and the review dealt with: (1) behavior of air gaps under nonstandard lightning voltages, (2) generation of fast-front voltage waves, and (3) measurement of fast-front voltage waves. The bibliography is presented in this paper. The bibliography contains 207 publications, published by IEEE and by other societies, in English and in other languages. >

There have been several cases of dielectric faults and failures in power transformers identified internationally that have been associated with system operation. This paper discusses UK examples and concludes the evidence indicates internal over- voltages arising from part winding resonance. These resonances are initiated by remote energisation, either intentionally or by the action of automatic re-energisation schemes after line trips. Some designs are more prone to the problem, and it is here that diagnostics such as sweep frequency response analysis may give some indication as to designs more at risk.

This paper deals with the interactive Web based interface used as a tool to explain the features and functioning of an impulse voltage generator. Impulse voltage generator is used for impulse testing of various power apparatus to assess the efficacy of their insulation system. Details of the simulator, developed using HTML and JavaScript along with the results, are also presented

The investigation was carried out with the purpose of obtaining more information about electric discharges in sandy media. A cylindrical setup was chosen for impulse testing of earth rods to simulate the voltage gradients in the field when lightning propagates to earth rod. The temperature images at the surface of the sand and rod were captured during the experiment. The surface temperature increased slightly higher with the increased of voltages. The CDEGS validates the increase of electric field higher at the edge of the cylindrical test cell. And also for the first time, the conduction current in dry sand was observed. It would be expected that the current produced in dry sand be due to soil ionization and also thermal effect due to charge movement in the soil, since there are air voids in the dry sand.

The development of an arc prolongation device for the secondary impulse method used in power cable fault location is reported in this paper. The faulty cable breaks down by a HV surge. Prior to the arc extinguishing, the device will supply, automatically, the energy to the faulty cable to sustain the arc sufficiently for the secondary impulse test. The device consists of an energy generator, wave blocker, impulse duration modulator, and signal coupler which are mounted in a compact case. This makes it easy to link the device with the surge generator and faulty cable on site. Since the output voltage is low, the device is safe for the cable's good insulation. Both simulations and laboratory tests show the device can prolong the arc for many milliseconds. The on-site secondary impulse method tests using this device produces fault location waveforms on the instrument's screen

Accelerated lifetime tests on EPR specimens were performed with the aim to investigate the endurance under impulsive electric stress at room and 90 /spl deg/C temperature. The results outline that relationship of the electric field on the number of applied pulses, at 63% of probability of breakdown, can be described by the well-known inverse power model.

Impulse tests are performed on power transformers to assess their insulation integrity. During these tests, impulse voltages are applied, and the resulting voltage and current waveforms are recorded. After postprocessing of these data, the presence of faults in the winding should be detected. Different methods and models have been developed for pattern recognition of transformer winding current and classification of impulse faults, . The drawback of these methods is the representation of a fault as a short circuit with constant conductance. In this paper, the breakdown and arc were modeled to represent a more realistic fault during impulse tests. Using this model, it is possible to generate more realistic data, which are necessary to develop more reliable fault-detecting algorithms and, as a result, better pattern recognition and classification of impulse test waveforms. The model has its roots in the classical Mayr equation, and it is simulated with TACS and MODELS modules in EMTP. The simulation results verify the ability of the suggested method to represent disk-to-disk breakdown and arc fault during impulse testing.

The effect of repetitive surges on the impulse strength of interturn insulation was studied for coils in three motor stators. In each stator, the coils were divided into several groups. Breakdown voltages of coils in one group were measured for 0.1- mu s risetime impulses and the average breakdown voltage of these unaged coils was determined. Coils in the other groups were first subjected to a number of surges with magnitudes 3 to 6.5 pu (1 pu is the peak-rated line-to-ground voltage of the machine), and then the impulse breakdown voltages for these aged coils were measured and compared with those for the unaged coils. There was an indication of surge aging of the turn insulation only for one stator. Surge magnitudes capable of causing detectable aging of turn insulation appear to be higher than those likely to occur at normally operating utility motors. It is suggested that there is a threshold value of surge magnitude for producing aging effects. >

Results are reported of investigations into the influence of the space charge associated with impulse corona on the growth and direction of the leader in a rod-plane of 1.72 m with a negative-polarity lightning impulse voltage applied to the rod. The charge oscillograms show three distinct steps of 2.5, 5.0, and 8.0 mu C, and these are attributed to the first corona and secondary discharges including leader development. In order to assess the effect of impulse corona space charge on the field distribution, field calculations with and without space charges were carried out using the charge simulation method. From the results of the calculations together with photographs taken of the discharge, attempts are made to explain the direction and growth of the leader in the gap. >

The work models the performance of mobile particles in compressed-gas-insulated substations under an intentionally applied impulse voltage of reduced magnitude. The subsequent microdischarges are used to detect the presence of particles and identify their approximate sizes. A comprehensive computer model is described. It describes and interrelates the potential and field distributions on and around the particle, particle lifting and movement, timing of the initial microdischarge, and the complete pattern of the microdischarges as might be detected by an external monitoring circuit. >

The authors describe a space charge measurement technique for a long XLPE cable specimen. The length of the cable specimen has been limited in conventional methods, because of the problem of pulse propagation through the cable. The problem was overcome by using the pulsed electroacoustic method, and also by applying the pulse voltage at the outer shield where the piezoelectric device was fixed. A preliminary experiment carried out using a 3.2-mm-thick XLPE cable confirmed the validity of the method. >

An electrode arrangement in compressed SF/sub 6/ was studied to verify its applicability as a precise voltage chopping device under standard lightning impulse for testing purposes. The effect of irradiation of the chopping gap and the application of a low DC sweeping voltage on the gap breakdown voltage dispersion under a standard lightning impulse was also studied. The results obtained show that the effect of irradiation and the DC sweeping voltage for the configuration and pressure tested was practically negligible, leaving the electrode conditioning as the most important factor to be taken into account when considering the utilization of this electrode arrangement as a chopping gap for standard lightning impulse chopping at the crest. >

In order to clarify the factors involved in breakdown under AC and lightning impulse voltages, the authors have developed a prebreakdown detection method, and have discovered a number of facts concerning the factors determining the breakdown of insulation in XLPE (cross-linked polyethylene) cables. Experimental results reveal that impurities in the insulation constitute the governing factor in breakdown under AC voltages, whereas breakdown under impulse voltages is predominantly governed by protrusions in the shielding layers. >

Full-scale impulse testing of various cable systems that were carried out under the DSTAR (Distribution System Testing, Application, and Research) program, of which the NRECA Rural Electric Research program has been a participant, is described. The primary objective of this phase of the DSTAR program is to evaluate the neutral performance of various cable types under impulse conditions. It also attempts to evaluate the efficacy of a number of neutral grounding schemes. >

The results of an experimental investigation of the lightning performance of selected electrode configurations as well as 115 kV and 230 kV substation structures are presented. The evaluation of the lightning performance of the air clearances for structures used in HV substations was based on the measured critical flashover voltage (CFO) values for standard 1.2/50 /spl mu/s lightning impulses with both positive and negative polarities. The measurements of CFO voltages were performed in the Mississippi State University High Voltage Laboratory. The method used for estimating the air clearance in HV substations is discussed. The results obtained enable the designer to optimize air clearances for a specific application. >

Impulse trees in insulators have been measured three-dimensionally to investigate their development mechanism. Since a usually observed tree is a projected one on the plane, the projected tree has pretend branches and the directions of its branches are not exact. In this paper, the impulse trees in the polymeric insulator (PMMA) have been measured three-dimensionally utilizing a CT (computer tomography) method. Successive impulse voltages have been applied. In each step, the tree has been measured and also the properties of the tree which cannot be obtained by the projected tree, such as the cross section. The volume and the distribution of the branches have been obtained for both polarities. >

This paper presents results from the experimental study of the current and charge behavior of predischarges occurring in insulating oil under impulse voltages. In order to minimize the influence of stray capacitances on the current measurement a specific device was developed and constructed. The streamer inception and propagation was visualized using high speed shadowgraph photography technique. To permit a chronological relation between voltage, camera monitor signals as well as the current, these characteristic quantities were simultaneously recorded with a digital storage oscilloscope. For both predischarge polarities the propagation must be differentiated in three successive stages: primary, secondary and tertiary streamer. During positive primary streamer propagation a continuously increasing charging could be observed. However, beyond a critical voltage, short current impulses could be detected discharging and charging the streamer structure in less than 200 ns. With the inception of the secondary streamer the charge increases dramatically. For the first negative predischarge stage a similar behavior could be observed. However, the charge curve changes from the linear to a step-wise increase leading to a step-wise streamer growth.

Newly introduced pulse width modulated type drives have an inherently fast rise-time resulting in transient voltages that are much higher than the operating voltage. A need has been created for corona detection in the motors used with these drives, preferably using a voltage impulse. A simple method has been developed to detect corona levels in a small coil or motor and experiments have been performed to verify the new method. This paper explains the theory and procedures used to determine corona levels experimentally. An automated system has been developed for corona detection and is also discussed.

As tree growth is dependent on discharges occurring in narrow and multiply branched tree channels, its mechanism is complicated and a clear understanding has not been obtained yet. In order to study discharge phenomena in a tree channel, gas discharges in tubules have previously been discussed, in which the tree channel was simulated using AC, DC and impulse voltages. In this paper, the discharge voltage in PMMA tubules was investigated, when an impulse voltage and the superposition voltage of a DC and an impulse voltage were applied.

This paper presents the study results conducted on the critical flashover (CFO) voltages of the fiberglass distribution line pole and its combination with polymer and porcelain insulators. Special attention was made to evaluate the added CFO voltages by the fiberglass pole to the insulators when pole served as a second component in the distribution line structure. A discussion of the results follows the data obtained.

Patterns of electrical trees in insulators are estimated numerically by fractal dimension. However they are not simple fractal but multifractal. They can be estimated by global spectrum and singularity, which do not have clear local information. A local fractal dimension and its spectrum are introduced. Local fractal dimension maps of impulse trees were calculated and the development considered by it and its spectrum. The change of the pattern of the impulse trees in applying the next impulse voltage can be predicted by considering the difference in the local fractal dimension spectrum in each stage.

Stress-induced leakage current (SILC) decreases when the time-between-pulses (T/sub bp/) of an AC-pulse waveform is increased. The amount of SILC reduction generally decreases for the same increase in T/sub bp/, with increasing stress voltage magnitude and stress pulse width. A model developed to describe the trap generation and relaxation processes occurring during transient high-field stress from unipolar and bipolar pulse waveforms is presented in this paper.

Corona-stabilization in SF/sub 6/-insulated systems is a known phenomenon, but up to now not fully understood. In case of slow-rising impulse voltage stress, a stabilization mechanism, mainly depending on the steepness of the applied test voltage, occurs. Essential for the phenomenon is a space-charge accumulation in front of the probe tip. The space charge reduces the field near the rod and thus an increased voltage level is needed to initiate the leader process. This paper presents experimental results. An imaging system consisting of 2 still-video CCD cameras has been used to record the paths of spark breakdowns across a positive point/plane gap. With subsequent computer image processing and analysis the dimension of the space-charge cloud could be estimated. Studies have been done for various probe tip lengths, gas pressures and steepnesses of the applied test voltage. The probability of corona-stabilization was determined for each test sequence. It was found that this probability decreases with the steepness of the applied test voltage for certain gas pressures.