SUMMARY
Maintenance on machines is a mandatory asset management activity to maintain asset reliability in order to reduce losses due to failure. 89% of defects have random failure mode, the proper maintenance method is predictive maintenance. Predictive maintenance object in this research is Steam Generator Amurang Unit 1, which is predictive maintenance is done through condition monitoring in the form of vibration analysis. The conducting vibration analysis on Amurang Unit 1 Steam Generator is because vibration analysis is very effective on rotating objects. Vibration analysis is predicting the damage based on the vibration spectrum, where the vibration spectrum is the result of separating time-based vibrations and simplifying them into vibrations based on their frequency domain. The transformation of time-domain-wave into frequency-domain-wave is using the application of FFT, namely AMS Machinery. The measurement of vibration value is done on turbine bearings and steam generator of Unit 1 Amurang using Turbine Supervisory Instrument and CSI 2600 instrument. The result of this research indicates that vibration spectrum from Unit 1 Amurang Power Plant indicating that there is rotating looseness, even though the vibration value does not require the Unit 1 Amurang Power Plant to stop operating (shut down). This rotating looseness, at some point, can produce some indications that similar with the unbalance. In order to avoid more severe vibrations, it is necessary to do inspection on the bearings in the Amurang Unit 1 Power Plant.