Hi Experts,

We are observing high frequency vibration in the steam turbine at 69,938 cpm. This individual frequency is dominant with amplitude of 5 mm/sec RMS. The same frequency is observed in hor, ver, axial directions.

Rated speed: 1500 RPM. MFR : Dresser Rand, turbine directly coupled to fan.

For vibration measurement we are using SKF microlog and accelerometer with magnet base mounting.

Fan vibration is normal and no presence of 69,938 cpm.

New equipment , we don't have previous readings...

Any suggestion shall be of help in identifying the problem..

Note : we have two identical equipments and both are having the same high frequency in turbine spectrum.

Thanks in advance

 

Original Post

1500 rpm is a very low speed for a steam turbine!! Verify speed at nameplate, specifications, and measure exact speed with tachometer or strobe.

Is the amplitude and frequency load sensitive?

Is the frequency an integer multiple of shaft speed?

Does the frequency change/increase as turbine speed increases during run-up?

I suggest using a microphone and quickly scan close to the surface of the entire turbine from main/extraction steam lines, admission valves, HP and LP casings, and at each shaft seal. Sound pressure near 1000 Hz is very easy to measure, even with a low cost microphone or meter connected to your vibration analyzer.

Walt

Dear Walt,

We shall try to do the following as suggested:

"I suggest using a microphone and quickly scan close to the surface of the entire turbine from main/extraction steam lines, admission valves, HP and LP casings, and at each shaft seal. Sound pressure near 1000 Hz is very easy to measure, even with a low cost microphone or meter connected to your vibration analyzer."

Can you please explain what problem we can identify by doing the above?

Regards

Interesting case and comments.

 

To verify some of the possibilities, it's worth to spend time collecting high quality vibration data. I mean high frequency with high resolution. You may want to have also high number of linear averages or synchronous averaging.

Also, I would make sure if the speed is really stable or not. At this high frequency, a small speed change can mean a lot of frequency change at VPF (if the problem is associated with VPF).

Regards- Ali M. Al-Shurafa 

Thanks shurafa for your reply.

At present I am out of office. I remember we have increased the Lines of resolution, to get quality data..with 8 averages. I shall confirm by next week.

Please let me know your suggested settings then we shall collect one more reading.

Thanks in advance

Add Reply

Post
×
×
×
×