Lately we have been getting a spike on radial bearing of the steam turbine. Every day I go to the field, the operators would tell that they had a spick on the turbine last night. It usually happens around 11pm. It has been like this for a week.

The machine runs smoothly. Even the spike that comes on the DSC doesn’t not reach the alarm setting. But I am wondering what would cause the vibration to jump for sometime and gets back to normal.
Turbine Details
Speed: 6500RPM
Vendor: Nuovo Pignone

tell me if you want more details

Get some data unless you believe in turbine fairies.

I thought Spike was a Gremlin and not a fairy. They are far worse. :0

well the turbine runs with little excessive axial displacement,0.052&0.47mm, which is close to the alarm setting,0.6mm. but this case it is not the issue, because it has been for a while. However, the spike that starts to show up every night could have some related to this axial movement! i am not sure.
on the trend chart, the radial vibration is around 24um which is normal, and the spike reads up to 50um.

hope that would help?

Define spike in terms of duration. Are we talking something that is a fraction of a second, or something that lasts for a minute or two?

John from PA

At a certain time and 60 or 120 Hz?

Jawal,
You need to find out if the vibration is real (mechanical, electrical or electro-mechanical)or just an electrical/controls issue. If it is real vibration, then you need to find the cause of it and determine if it is an early warning of a developing fault. Look at your DSC system trends for any other parameters (load, temperatures, flow, pressures) that may have changed at same time or just before/after.

There are simply too many possibilities that could cause the indicated vibration events, so more data is much better than guessing, or worst just ignoring these events. I would set up a temporary multi-channel vibration analyzer-monitor to continuously measure vibrations independently of and more completely than the DCS system. I use an IOtech ZonicBook with ezTOMAS software. Other multi-channel systems are available from GE-Bently or Maars could be used. The system would be connected to the buffered signal output of the proximity probes and key phasor (speed sensor). You could rent the system or hire a consultant (like me) to make the measurements.

Walt
w_f_strong (at) msn (dot) com

Check your vacuum system. See if there is a change in vacuum reading at the same time. Could be a gland condenser issue.

Sam
there is no certain time for each event. like for the past 10 days, the spike would come up like 2am then it wont show up again till next night at10 or 2 in the morning. its 60HZ.

Walt and Steve: im going to do what you recommended then i will get back to you guys
thanks a lot

Steve do you mean the LP steam?
can you tell me more how the vacum could casue a vibration on the radial side

Again jawal, please describe spike in terms of duration. Are we talking something that is a fraction of a second, or something that lasts for a minute or two?

What Steve describes for instance is something that will typically occur over a duration of minutes, not a few seconds.

John from PA

This message has been edited. Last edited by: John from PA,

it lasts for seconds.

If the generator is a 3600 rpm (60 Hz) machine, then the "vibration" at 60 Hz on the turbine would have to be caused by the generator. This is possible for a turbine-gear-generator machine train. You should also see an increase in 60 Hz vibration on the generator as well, since it is the source; unless it is torsional. Otherwise the "vibration" is probably caused by an electrical problem, such an an intermittent ground loop.

Walt

maybe you should check out if there are other machines stopping or starting in the plant at these particular times, just thinking outside the square. we had a set of pumps that would spike when a particular fan started up elsewhere in the plant

Wall
we have agreement with Emerson to do some check if needed. and they are coming today to connect the multi-channel vibration monitor.

i still have little concern. Connecting the multi-channel is there any thing to do with the DSC? because we still not sure if that spike is a real or not yet and I thought the monitor analyzer is something is plug to the machine while its running and its independent of the DSC.

If the tachometer and prox probes have a buffered signal output, then you should be able to connect while machine is running. If you and your I&C guy are not sure and Emerson is not sure, then don't do it.

Walt

I survived the spike gremlins in my steam turbine and an air compressor. The reasons were totally different.

1. The air compressor vibration cables were routed through a common conduit. Soome stray voltage was being p

Any disgruntled employees taking a break at that time with a little hammer. Believe me, stranger things have happened.

We had some weird radial vibration spikes on a turbine. The weeper holes were plugged in the lower oil seal at 1 of the bearings. Crud would build up in the seal, the oil wedge would change shape and the vibration would increase to 7-11 mils until it was violent enough to break up the crud and remove the wedge, then it would rapildy drop back down to normal. We lived with it for over a year. Mostly it happen once a day, sometimes twice, no real pattern. The event took 20 - 40 minutes.

If the unit was shut down/restarted it would take 1-2 weeks before the vibration started acting up again. During the last outage the seals in question were removed and replaced.

Just an interesting addition to the remarks area - probably not related your problem.