Failure Monitoring in Pelletizing Process Fans Using Statistical Methods

Pelletizing furnaces require the use of high-power fans to circulate air in different regions of the furnace, ensuring the efficiency of the pellet burning process and optimizing energy consumption. The occurrence of a failure in one of these fans results in the interruption of the production process, leading to economic losses. This work proposes a significant contribution to strengthening routine maintenance practices for these equipment. The signals from the installed instruments are sensitive to a variety of faults and are widely used in the precise diagnosis of problems. Through these signals, it is possible to create monitoring and failure prediction models that will assist the entire maintenance and operations team in their necessary actions and maneuvers. This approach not only minimizes unplanned interruptions, increasing plant uptime, but also contributes to the overall operational efficiency of the system. Two algorithms will be used: Three Sigma and Cumulative Sum Control Chart (CUSUM). The results indicate that these statistical methods can accurately identify failures in pelletizing process fans.