Showing papers by "Sivaji Chakravorti published in 2017"

Assessment of interfacial charge accumulation in oil-paper interface in transformer insulation from polarization-depolarization current measurements

Abstract: Accumulation of interfacial space charge in oil-paper interface is a critical issue in insulation diagnostics of transformers. This interfacial charge mainly accumulates due to the conductivity difference of oil and paper. Accumulation of interfacial charge leads to localized field enhancement, which further leads to partial discharges and acceleration in the aging of insulation. Therefore, from the point of view of transformer insulation diagnostics, assessment of interfacial charge is very important. However, it is not easy to estimate interfacial space charge behavior from the transformer diagnostics methods currently in use. In case of Polarization-Depolarization Current (PDC) measurement, a well known method for transformer condition monitoring, the effect of interfacial charge is reflected in the non-linearity of current response during polarization and de-polarization. During de-polarization process, a part of the interfacial charge accumulated during polarization period is absorbed by the electrodes producing a current, which is difficult to separate using conventional linear dielectric theory. In this paper, an attempt has been made to separate this current component from de-polarization current through considering charge de-trapping mechanism. Terming this current component as de-trapping current, its relationship with other parameters of transformer insulation is discussed. The developed methodology has been applied on several practical transformers. It was observed that the time constant of de-trapping current is related to the paper conductivity, oil conductivity, dissipation factor and age of the insulation.

A deep learning framework using convolution neural network for classification of impulse fault patterns in transformers with increased accuracy

Abstract: The paper presents a method using deep learning framework based on convolution neural network (CNN), for identification and localization of faults of transformer winding under impulse test. The results show that the proposed method outperforms the existing methods significantly. The present scheme eliminates the requirement of separate feature extraction and classification algorithms for the analysis of fault current patterns. A part of the proposed network performs feature learning and the other part classifies the features in a supervised manner. The method is computation intensive but capable of achieving very high degree of accuracy; on an average a margin of more than 7% compared to other published literature till date.

Effect of measurement temperature on power transformer insulation diagnosis using frequency-domain spectroscopy

Abstract: Frequency-domain spectroscopy (FDS) is a widely accepted method for the estimation of moisture content in oil-paper insulation of power transformer. Researchers have shown that measurement temperature is an important factor that not only shifts the tan δ curve both horizontally and vertically with respect to frequency axis but also affects the value of paper-moisture content of transformer insulation. This implies FDS data measured at different temperatures yields different results even if the insulation under consideration remains more or less unaffected. Availability and construction of master curve for in-service real-life units might not always be available. This study proposes a method that is capable of predicting the profile of tan δ curve at different measurement temperatures. Existing expressions available for predicting paper-moisture does not consider the effect of change in measurement temperature. This study also proposes modification of existing expression to predict the moisture content of insulation paper at any measurement temperature. In addition, this study outlines a method to evaluate the value of activation energy ( E a ) from the insulation response. The discussed technique is first tested successfully on data recorded from a laboratory sample. Thereafter, the method is applied on data collected from a real-life power transformer.

Investigations on charge trapping and de-trapping properties of polymeric insulators through discharge current measurements

Abstract: It is generally accepted that formation and accumulation of space charge in polymeric insulation is a critical issue in the design of high voltage electric systems, as it leads to insulation degradation and premature failure under high voltage surges and polarity reversals. Even at service stress level, accumulation of space charge may cause electrothermal aging. The insulation degradation and aging is reflected in the trapping and detrapping characteristics of the insulation, making it a useful aging marker. Space charge measurement methods are frequently used for trapping and de-trapping analysis. However, estimation of charge trapping and de-trapping behavior is very difficult due to the involvement of material properties and also partly due to the fact that proper measurement of space charge is itself a very complicated issue, requiring very sophisticated experimental arrangement and signal processing techniques. Moreover, application of these methods is limited to polymeric insulations of certain geometries (planar or radial). In this paper, the charge trapping and de-trapping behavior of low density polyethylene (LDPE), a widely used polymeric material for electrical insulation purpose, have been investigated employing a direct method based on the discharge characteristics obtained after the sample was stressed for short durations. A relationship is established between the de-trapped charge and the extracted charge measured from the discharge current. From this relationship a relative distribution of trapped charges across different trap depths was obtained. It was noticed that the trapped charge distribution is of exponential nature in the trap depth range of 0.93–1.11 eV. The presented method facilitates direct estimation of trap distribution independent of the insulation geometry. It was also observed that the trapped charge is highly dependent on the applied field.

Evaluating the effects of lower molecular weight acids in oil-paper insulated transformer

Abstract: Paper insulation breakdown of a transformer will directly result in the failure of its operation. Temperature variation in transformer initiates cellulose chain scission. Studies indicate that carboxylic acids are formed due to this scission in an aged transformer. Organic acids that are produced as byproduct is one of the key factors for insulation degradation in an oil-paper insulated transformer. Low molecular weight acids that are water soluble, prefer to stay in paper, which eventually affects the paper insulation, whereas the higher molecular weight acids are known to be fat soluble and affect the oil insulation. Along with the traditional methods which are used for condition monitoring, initiatives must be taken to understand the influence of acids generated inside the transformer. With a view to look into the detrimental effects of acids on the insulation status, organic acids are mixed in a definite proportion with transformer oil and the experimental results on an insulation sample are reported. An attempt has been made to correlate the dielectric test results with type and concentration of organic acids present in the transformer insulation system.

Cole-cole representation of transformer oil-paper insulation dielectric response

Abstract: Information about the condition of oil-paper insulation is of extreme importance for evaluating ageing status and maintenance requirement of transformers Dielectric testing techniques in both time domain and frequency domain are being assessed in recent years as supplements to traditional techniques by researchers and utilities Several attempts have thus been made to understand the physical processes of dielectric relaxation taking place during such tests by correlating conventional tests, time domain and frequency domain dielectric measurements, and chemical techniques In practical dielectric systems, the dipoles are found to relax according to a distribution of elementary Debye relaxation frequencies during the dielectric response measurements Researchers have proposed models that can simulate such many-body interaction processes with the help of suitable distribution density functions Frequency domain spectroscopy (FDS) data can further be computed from the distribution density function thus formulated The present contribution discusses mathematical formulations used for transforming time domain data to distribution domain and further to frequency domain Findings about the influence on operating temperature, moisture content, charging voltage etc on the relaxation process of oil-paper insulation have been reported in this contribution Results of tests on field transformers are also presented This paper also highlights the possible use of Cole-Cole models and employs the features of Cole-Cole diagrams for analyzing condition of dielectric materials

Effect of charge accumulated at oil-paper interface on zero of transfer function formulated using classical debye model parameters

Abstract: PDC measurement and analysis is one of widely used tool for reliable diagnosis of power transformer insulation. In different reported methods of analysis it is considered that polarization current is composed of the current due to dipole movement and conduction current. Similarly the depolarization current is assumed to be composed of relaxation of dipoles. However when the dc conduction effect is removed from polarization current it is found the resulting current is not similar to the measured depolarization current. This deviation in both the currents show the presence of nonlinearity in the system. This nonlinearity arises due to migration of trapped charges that reside at the interfacial region of oil-paper insulation. The present paper shows the effect of such free charge on some important performance parameters like paper moisture and zero of Transfer Function of Classical Debye Model.