Avaliação de cavitação em turbinas hidráulicas por emissão acústica

Abstract

The production of electricity in Brazil is predominantly Hydraulics and the potential in the coming years will continue to be widely used hydraulic, whether through great works like Jirau and Santo Antônio plants on the Madeira River and Belo Monte plant on the Xingu, in the construction of power plants smaller dams or through Small Hydropower around thecountry. Until these plants come into operation and contribute to the generation of electricity for the country, generating units in operation need to meet growing energy demand. To meet this energy demand, many plants are operating already installed in your ceilings or evenoverloaded. As a consequence many plants have operated their turbines with some level of cavitation. The phenomenon of cavitation causes a variety of problems in power plants, such as erosion of turbine blades, vibration of the structure, loss of income, noises, etc. The erosioncaused by cavitation alters the original profile of the blades compromising the efficiency of the turbine, being necessary to conduct periodic inspections on turbine blades in order to map the damage caused by cavitation, for later retrieval. The frequency of these inspections is defined in terms of hours of rounds machines and do not consider the load regime whereby the machine worked this time interval. This fact contributes to machines with the same hours rounds have very different stages of degradation due to having worked longer in charge ofarrangements that favor the onset of cavitation. Thus inspection rounds based only in times of machines can occur when the damage is already well advanced in, and with this downtime, repair time and cost of corrective maintenance are much higher when compared to that whenmaintenance occurs on time appropriate. Given this, many researchers have studied indirect ways of measuring the levels of erosion on turbine blades analyzing vibration signals or their own internally generated sounds. The acoustic emission technique has been widely used inresearch as a tool to solve this problem. This paper evaluates the application of acoustic emission technique for monitoring and diagnosis of erosion caused by cavitation in turbines. The results show that there is a clear correlation between the acoustic emission signals with the state of wear of the blades of these turbines analyzed.