Arijit Baral

An expert system approach for transformer insulation diagnosis combining conventional diagnostic tests and PDC, RVM data

Search for a reliable and efficient insulation diagnostic tool has always been the interest of power utilities. Today a large number of methods are available that can be used for insulation condition monitoring. These methods include both traditional and newer techniques. However due to complex aging process of oil paper insulation under the influence of different types of stresses, insulation condition assessment is generally performed by experts after carefully evaluating different measurement data. Furthermore, measurement data are influenced by various factors (like conductive aging byproducts, furanic compounds, paper and oil-moisture) in addition to measurement error (if any). This makes prediction of insulation condition based on single type of measurement rather difficult. This paper presents an Expert System designed to perform insulation diagnosis. The Expert System considers measurement data obtained using both traditional and newer techniques in order to come to a definitive conclusion. The Expert System extracts insulation condition sensitive information from data obtained using different techniques and then uses these to devise an optimized insulation model. This optimized model is used to predict paper-moisture content and other insulation condition sensitive parameters. Since these values are predicted using optimized model, they are not dependent on a single type of measurement and hence are less likely to be affected by error of any specific measurement. The performance of the developed Expert System is first tested on a laboratory sample and then on several real life power transformers belonging to NTPC Ltd.

Prediction of moisture present in cellulosic part of power transformer insulation using transfer function of modified debye model

Polarization-Depolarization Current (PDC) analysis is a well known technique for condition monitoring of oil-paper insulation in power transformers. However, due to variation in insulation geometry, information obtained through analysis of PDC measured for one transformer cannot be used for predicting condition of another transformer even if the loading history of the two units are similar. Furthermore, parts of solid insulation closer to winding are exposed to much higher temperature than the parts away from it. Prolonged exposure to this temperature variation leads to nonuniform aging in cellulosic parts. It is reported that Modified Debye Model (MDM) is capable of modeling this non-uniform aging. In the present work it is shown that a parameter sensitive to insulation condition can be obtained from Transfer Function of MDM. Several laboratory samples having different physical dimensions has been constructed. PDC data recorded from these samples are used to obtain the performance parameter (which is less influenced by insulation geometry) and its relation to paper moisture content. The results obtained from the proposed method have been compared with that obtained using Frequency Domain Spectroscopy (FDS) based commercial instrument (IDAX 300) in the case of real-life power transformer.

Condition assessment of cellulosic part in power transformer insulation using transfer function zero of modified debye model

Analysis of Polarization-Depolarization Current, recorded from high voltage oil-paper insulation using insulation model is common among researchers. It is reported that paper insulation of power transformers undergoes non-uniform aging. Unlike Conventional Debye Model (CDM), Modified Debye Model (MDM) is capable of modeling such non-uniform aging. However, factors like insulation geometry affect the values of the MDM branch parameters. Therefore, model parameterized using data obtained from one insulation system finds limited use in assessing the condition of a different transformer, even with similar loading history and power rating. The present paper shows that a performance parameter, which is less sensitive to insulation geometry, can be evaluated from Transfer Function, TFM(s) of MDM. The parameter is the zero Z1 of TFM(s) which is located farthest away from the origin in the Left Half Plane of s-plane. The capability of Z1 as an insulation condition sensitive parameter is first tested on laboratory samples and then on data recorded from several real life power transformers. Results obtained for these transformers show that there is a good correlation between magnitude of Z1 and paper moisture content obtained from Frequency Domain Spectroscopy (FDS) using IDAX 300.

A new approach for determination of moisture in paper insulation of in-situ power transformers by combining polarization-depolarization current and return voltage measurement results

Moisture content in an oil-paper insulation system of power transformer accelerates the aging process of cellulose that degrades insulation properties. As a result, dielectric response measurement data is also affected by moisture present in insulation. In time domain these measurements include Return Voltage Measurement and Polarization-Depolarization Current measurement. In addition it is a known fact that power transformer insulation consists of complex structures which is difficult to model in laboratory. Therefore, results obtained from samples prepared under controlled environment may not be applicable to real life power transformer. The present paper discusses a new approach to combine the results obtained from these two methods namely, Return Voltage Measurement and Polarization-Depolarization Current measurement for several power transformers, to find the relation between oil and paper moisture content that is relevant to insitu power transformers.

Application of S-transform for removing baseline drift from ECG

The three major sources of noise in an ECG signal are baseline drift, powerline interference and other electromagnetic interferences. With the recent success of Wavelet transform in identifying and eliminating noise from a wide variety of signals, frequency transformation techniques are gaining more and more importance for eliminating noise from ECG signals. This paper applies S-transform in eliminating baseline wander from ECG signals and the proposed approach is subsequently compared with the method based on wavelet transform.

Denoising neutral current of a power transformer measured during impulse test by framelet technique

Condition monitoring of the insulation of the windings of a power transformer depends on the accuracy with which the neutral current is measured during impulse test both at full and at reduced voltage. The most common problem faced while measuring neutral current is that it contains noise that causes ambiguous fault signatures. This paper presents an effective means for filtering out the noise from the neutral current using framelet technique. A fast and yet simple algorithm has been used here to calculate the framelet coefficients. Results presented are for an electromagnetic transient program (EMTP) simulated model of a 3 MVA transformer mostly used up to 33 kV. Studies have been carried out with both high and low frequency synthetic noises. Comparison of the results with that obtained from wavelet technique based method has also been reported in the paper.

Applicability of Stockwell transform in diagnosing impulse faults of a transformer

Lightning impulse test is one of the mandatory tests that every transformer has to undergo after assembly. This is done for assessing the condition of insulation of its windings. In the case of any fault, the neutral current and/or tank current contain(s) typical signatures depending on the nature and location of the fault. Since these current signals are non-stationary in nature hence each of these faults can be classified from such a transformation technique that has the time-frequency localization property. This paper investigates the performance of Stockwell Transform, popularly known as S-transform, for identification of insulation failures of a transformer winding during impulse test. The present study is restricted only to identify failure of winding insulation due to some defects that are already present in the winding and have not been developed due to the application or during the propagation of impulse voltage applied during the test. The results presented in this paper are for simulated models of a 3 MVA and a 5 MVA transformer. From the results, the prospect of S-transform in identifying the nature and the location of winding faults in distribution transformers is reflected.

An integrated method for numerical computation of electric field using parametric CAD module

The purpose of this paper is to discuss on the performance of parametric CAD module that has been developed for illustrating electric field distribution over high voltage (HV) systems having either a 2-dimensional or an axi-symmetric or an asymmetric configuration. The results of the electric field computation that are presented here are for an asymmetric gas insulated system (GIS) configuration. The electric field is computed by indirect boundary element method (BEM).

A novel methodology for on-site validation of RV measurement data

Large number of power transformers that are operating today have already outlived their designed life. Such equipments, which are now prone to failures, are still in operation in order to maximize asset utilization. Extending the life of such equipments by policies based on proper condition assessment can mean substantial saving for the utilities. Modern techniques based on dielectric response measurement and analysis provides valuable information regarding the condition of the HV insulation. However, commercial equipments available for recording dielectric response are extremely costly. The paper reports the design of a low-cost instrument developed at High Tension Laboratory of Jadavpur University for measurement of return voltage of high voltage oil-paper insulation. In addition a method is also discussed that facilitate user to carry out on-site validation of RV measurements recorded using the above instrument.