In this research, power reactors' failures are analyzed and their root cause is systematically investigated in Red de Energía del Perú (REP) an energy transmission company. Failure analysis is conducted in two phases: artificial neural network (ANN) diagnosis and secondly the visual inspections with a using root cause analysis (RCA). Finally, current maintenance practice is reviewed and an effective maintenance planning has been identified and executed. This study provides a practical guidance to help maintenance personal for the best identification of problems, for design error.