Degradation Of Electrical Insulation Research Articles

Estimation of the Degradation of Epoxy Resin due to Void Discharge using Weighted SVM and Partial Discharge Characteristics

The performance and reliability of power apparatus are affected by the degradation of electrical insulation, often caused by voids in resin‐molded components. To address this issue, we investigated the use of machine learning to estimate the degree of degradation due to void discharges in epoxy resin. We prepared samples with artificial void defects and collected partial discharge data up to the dielectric breakdown, applying long‐term AC voltage. Utilizing this data, a machine learning model was developed to estimate degradation levels. The results showed that the accuracy was 55% and the false estimation rate which underestimates the degradation level was as high as 31%, which was a significant problem for practical use. Therefore, for regions of data that were difficult to classify, we constructed a weighted SVM model that predicted higher levels of degradation by weighting the data according to their degradation labels. With this approach, the accuracy remained at 57%, but the rate of underestimation of the degree of degradation was reduced to only 4%. Consequently, we were able to develop an effective estimation model that is practical for the maintenance of power apparatus. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.

Graphene Nanoplatelets Effect on Electrical Tree Inhibition in Epoxy Resin Insulator

Epoxy resin is widely used in high voltage insulator material because of its high dielectric strength. Electrical tree phenomena are one of an electrical insulation degradation. It is easily initiated due to the distorted high electrical field. In this study, graphene nanoplatelets (GNPs) with the contents of 0.003-0.010 wt.% (2.6-8.8 mg) were filled into epoxy resin to improve electrical tree inhibiting ability. Treeing experiments were conducted at a fixed voltage of 15 kV, 50 Hz. LCD digital microscope was used to observe the electrical tree inhibition of GNPs in epoxy resin. The digital image of insulators sample was then analyzed by MATLAB program. The percentage of electrical tree damages of GNPs content of 0.007 wt.% was found at 2.75 %. Experimental results showed that electrical tree has propagated from main branch tree to many small branches. The distributed electrical tree around the main branches occurred after the addition of GNPs. Filling GNPs content of 0.005 wt.% will result to a complex branch tree structure with much longer branches than others. Nevertheless, the average tree length was inhibited at the GNPs filling contents of 0.003 to 0.007 wt.%.

НАДЁЖНОСТЬ ИЗОЛЯЦИИ СТАТОРНЫХ ОБМОТОК ЧАСТОТНО-УПРАВЛЯЕМЫХ ЭЛЕКТРОДВИГАТЕЛЕЙ

Relevance Widespread use of semiconductor frequency converters is due to the significant advantages of this technology: smoothness and accuracy of controlling the speed and torque of the electric motor, smooth start and stop; increased energy efficiency; the possibility of remote monitoring, etc. The accumulated operational experience has shown an increased failure rate of stator winding insulation of asynchronous motors operating with pulsed power. The presence of high-frequency overvoltages in the inverter signal can lead to the appearance of partial discharges even in the low-voltage motors windings. It’s leading to accelerated degradation of electrical insulation and premature breakdown. Over the past three decades, a significant amount of information has been accumulated in the technical literature: high-frequency transient models have been created; partialdischarge and corona-resistant test methods have been developed; the most important voltage parameters determining the intensity of insulation aging have been determined; regulatory documentation for testing and selection of insulation systems have been developed. Analysis and systematization of existing results are necessary to determine the directions of further research in the subject. Аim of research Identification of new solutions and trends of insulation reliability windings of lowvoltage electric machines when powered by modulated voltage. Research methods Bibliographic analysis, methods of statistical data processing. Results Analysis of the most relevant works on the reliability of low-voltage winding insulation systems in frequency electric drive systems is carried out, a bibliographic list of these works is given. The main results concerning the overvoltages simulation, insulation testing methods, studying the aging mechanisms of winding polymer insulation with pulsed power supply, determining the characteristics of emerging discharges are shown. The ways of further development aimed at modeling overvoltages, taking into account high-frequency effects, increasing the volume of experiments to estimate the insulation lifetime, studying the conditions for the occurrence of partial discharges, creating a common approach to insulation testing are determined.

Measurement of Activation Energy and Accelerated Degradation Time by Thermal Analysis Methods for Polymeric Insulating Materials

With the development of HVDC systems, there has been much research regarding insulating materials applied in the DC environment. Of particular concern is the degradation of electrical insulation due to external environmental factors caused by the long-term use of the insulation material of a power device. If such degradation continues, a leakage current may flow on the surface of the insulating material, eventually causing a partial discharge and potentially resulting in a major accident. Therefore, it is necessary to understand the long-term reliability of insulation materials. To do this, accelerated degradation tests and lifetime analyses should be performed by measuring the activation energy of insulation materials. There are currently various analytical methods to determine the activation energy, but the values obtained according to the different analysis methods vary. Because there are no established measurement criteria, it is necessary to accurately measure the activation energy of insulation materials. Therefore, in this paper, six analysis methods were performed through thermogravimetric analysis (TGA) to measure the activation energy, which was used to select the degradation time for the accelerated degradation test for which a reliable degradation time for acceleration degradation could be calculated. As a result of the experiment, deterioration acceleration time appeared in the order of PP > PE > PTFE > UPE, and accelerated deterioration of the insulating material through the deterioration time range measured in a future study will be performed to evaluate the life of each insulating materials.

Classification of Partial Discharge Images Using Deep Convolutional Neural Networks

Artificial intelligence-based solutions and applications have great potential in various fields of electrical power engineering. The problem of the electrical reliability of power equipment directly refers to the immunity of high-voltage (HV) insulation systems to operating stresses, overvoltages and other stresses—in particular, those involving strong electric fields. Therefore, tracing material degradation processes in insulation systems requires dedicated diagnostics; one of the most reliable quality indicators of high-voltage insulation systems is partial discharge (PD) measurement. In this paper, an example of the application of a neural network to partial discharge images is presented, which is based on the convolutional neural network (CNN) architecture, and used to recognize the stages of the aging of high-voltage electrical insulation based on PD images. Partial discharge images refer to phase-resolved patterns revealing various discharge stages and forms. The test specimens were aged under high electric stress, and the measurement results were saved continuously within a predefined time period. The four distinguishable classes of the electrical insulation degradation process were defined, mimicking the changes that occurred within the electrical insulation in the specimens (i.e., start, middle, end and noise/disturbance), with the goal of properly recognizing these stages in the untrained image samples. The results reflect the exemplary performance of the CNN and its resilience to manipulations of the network architecture and values of the hyperparameters. Convolutional neural networks seem to be a promising component of future autonomous PD expert systems.

News from Japan

The necessity of diagnosis and monitoring of electrical insulation integrity has been steadily growing since many industries want to make the operational period of power apparatus as long as possible. Furthermore, natural renewable energy resources such as solar power and wind power are recently growing rapidly. However, the outputs from such renewable energy resources tend to fluctuate significantly. To balance such fluctuating outputs with the demands, thermal power generation is subjected to rapid load changes and frequent start-and-stop operations more often than before. This would cause severe degradation of electrical insulation in steam turbine generators. From these backgrounds, the importance of insulation monitoring has been increased significantly.

Multishot Damage of Insulator in a Medium-Caliber and High Linear Current Density Launcher

As an important part of pulsed high-current launcher, insulator takes the functions of electrical insulation and support. Surveys were conducted after multishot to analysis matrix damage and insulation performance of insulator based on a medium-caliber and high linear current density launcher. Mechanical damage of matrix is mainly related to local current. The most serious damage located at zone of flattop and initial decline stage. In addition, serious damage may be caused by sprayed molten aluminum rather than high-temperature plasma. Metal contamination is dominated by local current and contact condition of rail-armature. There are more metal deposits on surface of high current and high muzzle voltage zones, and corresponding flashover voltages decrease significantly. Considering the effect of different damage forms and the operation conditions, the main damage form is matrix ablation located at the contact position of rail-insulator rather than the degradation of electrical insulation. Moreover, the main damage form of bore insulators is different for variety of caliber and linear current density launchers. Insulation damage may be considered mainly in small-caliber and low linear current density launcher, while mechanical damage may be considered mainly in large-caliber launcher and high linear current density launcher.

Estimation of the Hottest Spot Temperature (HST) in Power Transformers Considering Thermal Inhomogeniety of the Windings

The degradation of electrical insulation in transformers is traced to thermoelectric processes. Existence of localized hot regions due to thermal insulating properties of electrical insulation would cause thermal runaway around these regions. In an earlier paper , the authors presented analytical methods for estimating the temperature and its distribution at different points of the transformer based on a closed-form mathematical technique using a generalized heat conduction (GHC) model. Certain aspects of the inhomogeniety of the several components of the winding and incorporation of distributed heat source were not addressed. Also, a rigorous treatment involving the changes in winding resistance, which was built into the thermal model as an empirical correction factor, has now been modified and incorporated in the GHC. These considerations have now improved the accuracy of estimation of hottest spot temperature.

Prediction of hottest spot temperature (hst) in power and station transformers

The degradation of electrical insulation in transformers is more often due to thermal stress. In this paper, the authors attempted to work out a method of assessing the magnitude of temperature and the hottest region in the body of a large transformer with a reasonable degree of accuracy. A theoretical model has been developed based on the boundary value problem of heat conduction in transformer winding using finite integral transform techniques. The model requires, in addition to electrical parameters of the transformer, information on the actual design data. The authors believe that this model predicts HST with precision and also indicates a possibility of online data acquisition.

Diagnosis and monitoring of high-voltage insulation using computerized analysis of partial discharge measurements

Transmission of electrical power requires various types of apparatus. All the apparatus and transmission-line conductors must be insulated from the high and extra-high voltages arising from the need of transmitting a large quantum of energy. As the need grows, the quality and reliability of the insulator must also grow to handle extra stresses. Hence an attempt is made to evaluate insulation in power apparatus based on the computerized analysis of partial discharge (PD) testing measurements and observations, leading to the identification of PD fault-sources of electrical insulation degradation. The database offers a decision-making aid for product development and supports condition monitoring, assessing the state of insulation. By conducting suitable experiments on two-electrode models, one does partial discharge characterization by considering one PD defect at a time. Using the database extracted from the various PD distributions attained from physical models fabricated, PD sources or defects in simple power apparatus such as surge arrestors, current transformers and artificially created defects such as slot models and rigid terminal connectors are identified successfully and evaluated. Successful recognition rate is achieved with the developed software package even in the presence of two PD sources; say, surface and internal discharges contributing to electrical insulation degradation.

Long term degradation of electrical insulation of Al 2O 3 under high flux fission reactor irradiation

Long term increase of electrical conductivity under irradiation was observed in some Al 2O 3 (polycrystal and single crystal alumina) specimens in in situ electrical measurements in the High Flux Isotope Reactor in the framework of the Japan/USA collaboration called JUPITER project. Especially, chromium doped single crystal alumina showed substantial increase of electrical conductivity at moderate doses of ∼0.1 dpa, which had similarity to the phenomenon called radiation induced electrical degradation (RIED). Irradiation to doses above 0.1 dpa did not cause a further degradation of electrical conductivity, but instead resulted in an improvement in insulating ability. Behavior of the observed increase of electrical conductivity is described in conjunction with the reported phenomena in RIED and some mechanisms responsible for the observed increase are speculated from the results obtained.

Lifetime assessment by intermittent inspection under the mixture Weibull power law model with application to XLPE cables.

This paper proposes a new treatment for electrical insulation degradation. Some types of insulation which have been used under various circumstances are considered to degrade at various rates in accordance with their stress circumstances. The cross-linked polyethylene (XLPE) insulated cables inspected by major Japanese electric companies clearly indicate such phenomena. By assuming that the inspected specimen is sampled from one of the clustered groups, a mixed degradation model can be constructed. Since the degradation of the insulation under common circumstances is considered to follow a Weibull distribution, a mixture model and a Weibull power law can be combined. This is called The mixture Weibull power law model. By using the maximum likelihood estimation for the newly proposed model to Japanese 22 and 33 kV insulation class cables, they are clustered into a certain number of groups by using the AIC and the generalized likelihood ratio test method. The reliability of the cables at specified years are assessed.

Review of chemical indicators of degradation of cellulosic electrical paper insulation in oil-filled transformers

Chemical and thermal degradation of electrical insulation in oil-filled transformers, mechanisms of failure and methods of condition monitoring and of life prediction are reviewed. An insulation life prediction model is developed and estimates of insulation life are presented under a variety of oil/paper conditions. Systematic measurements of the concentration of chemical indicators in transformer oil are used to demonstrate that furaldehyde and related products can be used to detect high rates of paper degradation. Similarly, phenol and related products indicate electrical degradation in phenol-formaldehyde resins.

Temperature considerations in electric motor noise reduction

Induction motors represent a major noise problem in the petrochemical industry. Retrofit noise reduction frequently requires the use of inlet and outlet mufflers. Be they commercially available units or custom designs, the most severe constraint is often the maximum pressure loss the motor can support without serious reduction of operating lifetime due to excess temperature. Typically, a 10°C increase will reduce the lifetime by half, due to degradation of electrical insulation. One cannot base the true cost of a noise reduction treatment on its capital cost alone if it may reduce the operating life of the machine. This paper presents a technique to simulate the effect of a muffler in order to determine the motor flow characteristics, and at the same time, determine the motor internal temperature versus flow characteristic by electrical resistance measurements. Actual mufflers whose designs were based upon such measurements will be described. The influence of mufflers on the explosion-proof rating of motors will also be discussed.

The Contribution of Differential Thermal Analysis to the Estimation of Thermal Endurance of Insulation

The equations governing the thermal degradation of electrical insulation and those governing the shape of a thermogram have been combined to give that of the required life line. The numerical solution has been shown to require an empirical correlation between the physical properties involved. In the case of thermogravimetric analysis (TGA), weight loss can be correlated with electric strength (the failure criterion) in an independent experiment. However, in differential thermal analysis (DTA) (at least with normal DTA technique) the heat evolved cannot be correlated with electric strength. By means of a numerical integration of a DTA thermogram, together with the determination of the cell cooling constant, an activation energy can nevertheless be obtained and the thermal rating estimated. DTA constitutes an ideal supporting technique, since if the activation energy agrees with that given by TGA this is good confirmation that the value is appropriate to life-line slope. A worked example of the method (using a formvar enamel) is described in detail.