Sistema de alerta para detecção de falhas em equipamentos elétricos

Abstract

In a substation, equipment failures are often associated with thermal anomalies, which is why thermographic inspection is part of the maintenance and monitoring requirements for installations. In addition, thermography has the advantages of being a non-invasive technique, nowadays, with the technologies available, it can be carried out remotely and in real time. However, the technique has its limitations, especially when carried out in open environments, where the results obtained can be considerably influenced by environmental conditions. This paper presents the development of a realtime warning system for detecting faults in electrical equipment. To this end, thermographic images and data from a solarimetric station, installed in the same location as the thermographic images, were first captured. Using autoregressive vector models, the main atmospheric factors influencing the operating temperature of the equipment in an open substation were determined. With knowledge of these parameters, control chart models (EWMA) were used to define acceptable upper temperature limits for each operating time of the substation’s equipment. This information formed a database with all the data relating to thermal acquisitions by time of day, carried out five times a day, creating an operating history for each piece of equipment that was compared to the real-time capture value, initiating a classification process. This alert was calculated based on the probability of possible failures, ranging from 0.0 (lowest risk of failure) to 1.0 (highest risk of failure). The classifier begins its procedure by defining whether the equipment processes current or not, since its thermal values are strongly related to the current level at the time its temperatures are recorded (in addition to the influence of environmental conditions). Five classification levels have been defined for each piece of equipment, given that the last two levels triggering alerts for the operator. With this system, the thermography specialist does not need to dislocate to distant substations, nor does he need to spend hours analyzing all the images taken during an inspection in order to later produce specific risk reports. With this SMT-SE system, the specialist will focus solely on analyzing risk situations, which will speed up the maintenance process and identify faults at an early stage. In this way, this system allows equipment with possible faults to be identified before they can compromise the substation’s operation and impact on the energy supply provided by the utility company.