Power Transformer Fault Characterization Through Oil Contaminants Evaluation

Dissolved gas analysis (DGA) in insulating oil is performed periodically to access operating conditions of power transformers. DGA results suggest possible incipient faults and maintenance actions for transformer repair. The more information the process gives, the better and cheaper corrective actions can be. To optimize asset management, determine the need for repair in a transformer failure, or choose options for managing operating conditions, an additional tool that allows maintenance engineering to previously identify failure type would be interesting. When an incipient fault occurs, all construction materials involved will be affected and, therefore leave traces in insulating fluid. This paper is a first step in evaluating potential analytical methods to be used in determining the presence of transformers construction materials in insulating fluids due to incipient faults. Insulating oil samples were first contaminated with different transformer construction materials in order to test different analytical techniques to determine the presence of those materials in oil and/or their influence in oil properties. Samples were thermally aged and then tested by different methods to find out contamination effect on physical chemistry properties, such as total acid number, interfacial tension, dielectric loss and others. The presence of added contaminants both soluble as well as particulate material is also investigated. Soluble contamination materials are determined by use chemical analysis techniques such as Fourier Transform Infrared Spectroscopy (FTIR), Gas Chromatography Mass Spectrometry (GC-MS), Inductively Coupled Plasma (ICP), and enhanced DGA. Some nonconventional analysis techniques were also used to evaluate particulate contaminants related added construction materials, such as particles counting, particle quantification index (PQI), and Analytical Ferrography that gives particles nature, shape and size. This work is part of the R&D research project ANEEL PD-222.2017.