Development of a reduced-model laboratory for testing predictive fault system in internal combustion engines

Early diagnosis of a fault is relevant in the industrial environment to increase the availability of critical equipment. In a scenario where the thermal plants have fully functioning for long periods of time, to monitor the condition of generators and their primary machines becomes increasingly imperative. This paper presents the development of a reduced-model laboratory to test a predictive system for diagnosing thermos-mechanical faults in internal combustion engines through electrical signature analysis of voltage and current signals of the generator. This laboratory serves for the identification of patterns of failure in combustion engines and the verification of the system developed for installation in SUAPE II thermal power plant, in Brazil.