Partial discharge (PD) is a naturally occurring phenomenon in electrical insulation systems. The phenomenon is characterized by small arches inside the insulation or adjacent components. The progressive development of partial discharge activity mark the start of almost all the failure mechanisms of the insulation of high voltage equipment. When the detection and repair of such mechanisms is not performed it is expected that the PD magnitude and repetition increases, leading to the complete breakdown of the insulating equipment, which is accompanied by subsequent defects and loss of revenue due to unavailability of the HV equipment. Even when the presence of PD is detected it is vital to identify its source (internal, surface, corona, etc.) and given a technical advice as to dealing with the failure mechanism. Therefore, monitoring and diagnosis PD phenomenon is deemed as necessary to HV equipment. In this work this task was performed with analysis of some absolute values and a series of graphs generated in a computer program written in the MATLAB language syntax. The program reads .tsv files and reproduces 2D statistical maps (φ-qn), magnitude histograms (pulse height distributions) and enable us to know Qm value of a measurement. This work analyzes the files of a test on 2 hydrogenerators of Tucuruí Hydropower Plant, using national monitoring system IMA-DP. In the results examples of the graphics are gathered and diagnosis of Roebel bars insulation condition is given for two machines to discriminate the source of discharge. The work is based on technical standards and other works on the physical phenomenon and the measurement of partial discharges.
|
