Magnetic Flux Entropy as a Tool to Predict Transformer's Failures

This paper proposes magnetic flux entropy as another technique to assess electric transformer failures. The obtained tool based on entropy may reveal some irregularities, distortions, aging, overloading, and harmonics, all of them reflected over the electrical sine wave that supplies power to dwelling and industry. Some entropy concepts are reviewed and the basic equations are applied to data obtained from magnetic and infrared waves radiated by the (18 watts 110-20 V) prototype. Several experiments were made through non-invasive and indirect measurements of magnetic and thermal fluxes. Entropy, as an accepted indicator of physical complexity, is proportional to the logarithm of the number of states in a thermodynamic system. Failure prediction based on entropy of radiated magnetic waves could be easier, faster, and cheaper.

I. Introduction

The transformer is an essential component of the transmission and distribution system. Transformers represent one of the key components in utility systems requiring a constant monitoring. For this reason, it is necessary to improve the techniques and methods for keeping good function of the network that helps to prevent transformer faults. The machine electric life is a decisive factor for managing the risk associated with the network reliability and of course provides a quality server directly impacting the availability. Currently are executed various preventive tests in energized transformers which use technics as ultrasound, power factor, physics and chemical analysis.