On-line Surveillance Monitoring in Continuous and Batch Process Applications

Background

Successful on-line surveillance based condition monitoring and evaluation rely on acquiring data at known, repeatable machine operating states and reporting detected abnormalities as quickly as possible. Most industrial machinery operates at reasonably steady states for extended periods. These machines are relatively easy to diagnose, as changes in their “condition indicating characteristics” can normally be attributed to changes in their physical, mechanical health. Identifying the operating state, beyond the simple verification that a machine is running or not, to control when data should be collected is usually not required. Unfortunately, not all industrial machinery can be expected to always be running at known,
repeatable conditions.

The objective is not to alter or suspend production to put a machine into a known, repeatable state to make measurements; rather, it is to maintain production and make measurements when, in the cycle of normal operation, the machine is operating in an identifiable, repeatable state.  This application note goes into detail about the challenges and benefits of such systems.  Although the document makes reference to older legacy SKF systems, the general practical knowledge can still be applied today with current SKF on-line systems.