Application of Predictive Maintenance Technique with Vibration Monitoring Method for Analysis of Damage Trend and Damage Diagnosis at Make Up Water Pump on PLTGU Installation

摘要

In a plant installation there are equipment that have their respective roles. In the Combined Gas and Steam Power Plant installation there are two main power components, namely gas turbine and steam turbine. Each of these turbines must have a supporting component. One of the important supporting components for steam turbines in the UPJP Priok Installation is the make-up water pump which plays a role for the distribution of water to steam turbines. If this distribution stops then the steam turbine cannot operate. The steam turbine itself can generate power by 40% of each gas turbine. In one block there is one steam turbine with three gas turbines. The capacity of each gas turbine is 125 MW, so that each steam turbine block alone can produce a maximum power of 150 MW. The significance of the make-up water pump function requires that the component is always in top condition. To maintain the prime condition itself needs to be implemented maintenance method. One of the most effective and efficient maintenance methods is the predictive maintenance method. Relying on vibration as an indicator can be generated an analysis of trends adjusted to the prevailing standards for predicting the most efficient maintenance time, but also a diagnosis of the frequency spectrum to determine the source of the vibration that also indicates the source of the damage. The research was conducted by taking vibration data using transducer at certain point which then analyzed the tendency of growth of excessive vibration based on ISO 10816 then made the spectrum diagnosis at the point with the excessive vibration so that the source of damage can be known. Trend analysis is performed to predict the time of varying damage each point. While the spectrum diagnoses show the damage occurred due to the unbalance on the pump shaft, turbulence, and damage at the pump impeller.