Metodologia para prognosticar a reducao de temperatura do enrolamento de campo de um hidrogerador

Abstract

The medium and large hydrogenerators are tailor-made in a few identical units for each particular hydroelectric power plant. So, it is not expected that we can find identical generators in different power plants. Because these hydrogenerators are not previously tested, they onlypresent design or erecting problems during the field acceptance tests or in the beginning of their operation. Among this kind of problems we have the erroneous calculation of the reluctance of generators magnetic circuit, that leads to high losses and consequently to high temperatures in the rotor winding than expected. This problem is generally detected in the generator heating test at rated output, during commissioning. In this dissertation a semi-empirical model is proposed to evaluate how effective are the air gap diminution and the air volume flow increase foreliminating the rotors winding overheat. This model is based on the commissioning test results of the generator itself. The model was applied to the unit n° 2 of Nova Ponte Hydroelectric Power Plant By, that belongs to Companhia Energetica de Minas Gerais CEMIG. Confronting the calculated values with the experimental ones we can conclude that the model is able to give good predictions. In particular, the rotor winding predicted temperature (94,4°C) lied very close of the measured one (95,2°C). The model can be used to prescribe the availability of hydrogenerators performance upgrade to generate even more power.