Rolling element bearings in paper mill machines are considered in this paper due to their critical role in the machine function. In general, the cost-effectiveness of a vibration-based maintenance policy mainly depends on the technical and economic effectiveness of the vibration monitoring system, i.e. software, hardware, and personnel, being used. In this paper, a model for predicting the vibration level of rolling element bearings in paper mills is developed. In the model, a combination of bearing vibration history, current condition and data from operating conditions is used to predict the vibration level of a bearing during a near future.

The factors influencing the condition of rolling element bearings during operation are discussed. These factors are related partly to operating and environmental conditions and partly to previous changes in the condition of the bearing. The operating life of a bearing is divided into four phases characterised by their vibration behaviour. One of the important conclusions is that a mechanistic model would be more appropriate than other models for predicting the vibration level of a rolling element bearing and describing the deterioration process, which are needed to achieve better understanding of the bearing vibration behaviour.