A Tool for estimating remaining life of power transformers

Abstract

Part 1- Modelling the Lumped Parameter Network of a Power Transformer using Sweep Frequency Response Analysis Accurate modeling of power transformers is beneficial for fault diagnosis and the health condition assessment of the power transformers. This project proposed a novel approach of synthesizing the lumped parameter equivalent circuit of a power transformer using its frequency response by utilizing artificial neural networks and the genetic algorithms. A physically realizable transfer function derived using the sweep frequency response data of a power transformer is used to construct its lumped parameter equivalent circuit model. Then, the parameters of the model are estimated using an artificial neural network and the genetic algorithm. All parameters, objectives and constraints of the genetic algorithm and the artificial neural network are set appropriately to ensure accurate results within the shortest time span. The entire approach is validated by means of an error analysis. The proposed method is easy to implement and can be used effectively to improve the reliability of power system though accurate fault identification and health assessment of power transformers. In addition, by replacing general equivalent circuit with the proposed model, the accuracy of transient analysis results can be significantly enhanced as the high frequency behavior is simulated using lumped parameter network. Part 2 - Fractal Characteristics of Creeping Discharges Propagating on Solid/Liquid Dielectric Interfaces A solid/liquid dielectric interface is considered a weak point in a composite insulation system, as it facilitates creeping discharges on the interface. This project proposes the use of nano-dielectrics to minimize the effect of damages which occur due to creeping discharge activity on such interfaces. A point-plane electrode arrangement-based test apparatus, energized by a high voltage supply, is used for analyzing the propagation of creeping discharges over solid/liquid interfaces using visual observation. An algorithm is used to determine the fractal dimension of creeping discharges propagating over various solid/liquid insulating interfaces. The results show that the use of nanofillers can increase the dielectric breakdown strength of epoxy by 7%.