Early Warning Anomaly Detection of an Increased Bearing Vibration at an Aluminum Rolling Site

Industry: Metals, Steel & Aluminum
Region: S. America
Asset Type: Hot Finishing Mill

A deviation in a hot finishing mill at an aluminum rolling site was found in the beginning of November. GE Vernova's SmartSignal Predictive Analytics AI/ML digital twin of the hot finishing mill revealed an increase in the mill motor Non-Drive End (NDE) bearing vibration X (velocity RMS) and motor NDE bearing vibration. These values are near the alarm thresholds. The GE Vernova Engineering team, Industrial Managed Services (IMS) issued a case, included this item in the weekly report for discussion with the customer, and recommended inspecting the bearing.

The IMS team promptly initiated a case to address this deviation and included it for discussion in the weekly customer report. They recommended an immediate inspection of the motor bearing to determine the cause of the increased vibration. Upon detailed inspection, the customer discovered significant wear marks on the lower half of the bearing, with minor wear evident on the upper half as well. Additionally, the presence of contaminants within the bearing housing was noted, likely contributing to the increased vibration levels. The customer subsequently undertook repairs, which included polishing and reassembling the bearing. Following these corrective actions, the vibration levels successfully returned to the normal ranges predicted by the digital twin model.

The early alert and guidance provided by the IMS team offered substantial benefits to the customer by enabling timely intervention during a planned shutdown in December. This proactive measure allowed the customer to address the bearing wear and potential shaft issues before they could escalate into more severe problems, thereby preventing the need for more extensive repairs or replacements.

The foresight provided by the IMS team helped the customer avert forced outages and major equipment failures that could have had significant operational and financial implications. By tackling the issue at an early stage, the customer not only maintained operational continuity but also avoided the high costs associated with unscheduled downtime and major repairs. The estimated savings from this proactive maintenance approach are approximately $136,200, calculated based on the average production loss avoided in North America. This case highlights the critical role of predictive analytics and early intervention in maintaining the reliability and efficiency of industrial operations.