Vibration & Oil Analysis Trouble Shooting Case Study - Steam Turbine/Exhauster Bearing Oil Whirl

Applied Diagnostics Ltd - Precision with Pride


This data was collected from a small land based steam turbine that drives a critical gas exhauster within a chemical plant. Both the turbine and exhauster rotors are supported on sleeve bearings, which rotate up to 5200rpm. Routine condition monitoring surveys picked up increased vibrations that prompted this trouble shoot to be conducted.

Turbine Vibration Trend Monitoring


Frequency Spectrum Cascade Plot

To investigate the cause of the increasing vibration trend the following frequency spectrum cascade plot was collected while the turbine was run up from barring speed to 5200rpm.



Analysis of the collected cascade plot shows the normal run speed peak, however a sub-harmonic peak develops when the turbine run speed reaches 4500rpm. The frequency of this peak varied between 42 to 45% of the Run Speed and increased dramatically as full turbine run speed was reached. This vibration signature is indicative of instability within the sleeve bearing, more commonly known as Oil Whirl. This phenomenon can be caused by increased bearing clearances, or oil that is the wrong grade or contaminated. The first thing to check was the results of the oil sample collected during this troubleshoot.

Oil Analysis Tri-Vector Results

Tri-vector plots provide a quick and easy way of viewing the overall condition of the oil. The oil sample on the left was collected while the vibration problem exists. This highlights a severe contamination problem and also some evidence of wear, prompting a closer inspection of the oil analysis results and associated trends.


Tabulated Oil Analysis Results & Trends

The following table breaks down the oil analysis results into 3 discreet groups, Wear, Contamination and Chemistry. By looking at the individual test results within these groups a better diagnosis can be made. In this instance the severe Contamination is caused by Water within the sample, there is also evidence of early stage Iron wear, most likely from the sleeve bearing. The associated trends also show increases that coincide with the onset of the vibration problem and subsequent reductions following the oil change.



The oil results show that the presence of water within the sample was causing instability within the sleeve bearing, which caused the Oil Whirl vibration problem. A suspected break down in the steam turbine seal labyrinth was suspected as this would allow condensate to mix with the oil, thereby increasing the water content. While the strong vibration existed the onset of wear within the bearing also became evident, which if left would have caused bearing clearances to open up. The oil was replaced and the water/wear trends returned to acceptable levels, further more the vibration problem was also resolved.