Confiabilidade metrológica em termografia aplicada em sistemas elétricos

Abstract

The diagnosis of thermal anomalies in electrical devices power and utility companies is crucial for the maintenance of theirs plants and power plants and supply systems. In this context, thermography is a nondestructive inspection technique widely used for the diagnosis of thermal anomalies without intervene on its operation providing securityfor operators and high-yield for the of maintenance sector. However, thermography has been addressed only as a qualitative tool for diagnosis of thermal anomalies, limiting the reliability of results and the potentiability of this technique. In this study, we investigated the measurement uncertainty of thermography in the diagnosis ofabnormalities in electrical connections used in urban energy supply systems. The measurement uncertainty was calculated by GUM (Guide to the Expression of Uncertainty in Measurement) and Monte Carlo methods by using the IMT software - Measurement Uncertainty Thermographic developed in MatLab ®. The studies were conducted in LabTermo, Thermometry Laboratory School of Mechanical Engineering,Universidade Federal de Minas Gerais. Tests were performed in electrical connections typically used in urban energy supply systems. The samples were supplied by Companhia Energetica de Minas Gerais, Cemig.. In some of the samples were introduced defects during its assembly to reproduce the deterioration of faulty connections. In order to validate the methodology, was also carried out a thermographicinspection at a substation where temperature measurements were taken on devices with a tendency to failure. The results shows the need to address the measurement uncertainty as a way to ensure the quality and reliability of thermography apply to diagnosis of anomalies in electrical connections.