intermittent high 1x vibration on steam turbine

[gallaz]

I have a steam turbine/generator that undergoes huge escalations in 1x radial vibration at the front turbine bearing. The odd part is that the vibration increases greatly and then reduces to normal running conditions over about a 10-20 minute period. Has anyone seen this before?

 

[Strong]

Galaz,

The symptoms seem like Spiral Vibration. If you have a 1xSS tach pulse (Key Phasor), then measure filtered vibration level and phase. A polar plot of the 1xSS vector would take a spiral path, or circular path if vibratoin amplitude does not keep growing. The cause is related to a shaft/rotor rub in combination with a critical speed (balance resonance) near operating speed.

Walt

 

[magicme]

Is this a startup phenomenon? I mean are you simply passing through a critical speed on the way to operating speed?

If this is a steady operating speed phenomenon, how far away from the T-G lateral critical speed do you operate.

Possibly is it a thermally-bowed rotor due to thermal gradients throughout the rotor that "blend out" after some period of running? Is it possibly a bearing oil film issue? Changes in bearing oil film flow (during startup) could change the bearing spring rate and thus the critical speed. Are the oil pumps started first (before the turbine)?

 

[gallaz]

It is a steady operating speed ocurrence. There is no indication of 1/2x running speed vibration usually associated with a rub. I agree it must be some kind of excitation but can't find the cause.

 

[MachineryWatch]

Do you have proximity probes installed on the turbine, or are you looking at accelerometer data? If you have prox probes it might be revealing to look at the orbits and waveforms. Rubbing does not always produce 1/2X RPM frequencies, so do not rule out a rub on that basis.

Skip Hartman

http://www.machinerywatch.com

 

[hacksaw]

what is happening on the steam side when you experience this?

 

[gallaz]

yes, the vibration is seen by proximity probes... there are also accelerometers on the bearing casing which do not show much action.

On the steam side, there is normally a rise in steam flow when the vibration event occurs. This happens after vibration begins... seemingly a result not a cause.

I have posted more info in thread666-82790. Your input has been muchly appreciated.

 

[JCBerggren]

We have experienced this same kind of problem on a 27,000 HP centrifugal air compressor in our nitric acid plant. The vibration would rise and fall on a regular cyclic pattern over approx a 230 minute period. Amplitude variation was 1.2 mils max with over a 200 deg phase shift. The pattern would repeat day and night for weeks on end between runs. Frequency was primarily at 1xRPM, with a significant subsynch vibration at the rotor natural frequency. Vibration could be decreased with speed reductions, axial thrust position shifts, and discharge temperature changes.
This is a two stage compressor running at 11,100 to 12,600 RPM with overhung steam turbine and hot gas expander at opposite ends. Rotor components are connected through highly tensioned curvic couplings.
Problem was corrected on two rotors by opening up axial clearance on 2nd stage, converting to honeycomb shaft seals, and mininmizing mechanical runout in shaft segments.

 

[sms]

The symptoms are consistent with a rub. In the class of machinery you are talking about, the most common rub characteristic is high 1X, or a 1X that cycles from high to low. I actually would be very supprised to see a 1/2X in the presence of a rub on a large steam turbine generator. What does the axial position do in relation to the onset of the high 1X vibration?

 

[McVibe]

Check the control system, it could also be feedback from the grid. Might not be mechanical at all...

 

[MachineryWatch]

And, if it is feedback from the grid, it may also be exciting torsional natural frequencies. There are several well documented cases of catastrophic failures resulting from grid feedback exciting torsional natural frequencies, so it is important to consider and eliminate this as a possibility.

Skip Hartman

http://www.machinerywatch.com

 

[gallaz]

Thanks for the useful replies everyone. We have investigated control system feedback... the reason I beleieve that the machine is not being externally excited(or excited by control system)is that the peak in amplitude occurs at running speed of 50Hz. The machine's critical speeds/natural freq are not close to this at all. I would expect an external or feedback excitation to only materialise this severely if it caused a resonance, ie. at natural frequency of machinery or its components. Any thoughts on this?

 

[Franko]

In your other thread you say that bearing temp. also increases. Normally that does not occur - you may have an oil supply problem. Try increasing oil pressure - and check for pressure drop from gauge to front bearing housing. If increase in temperature is actual metal temperature, there can be non-uniform radial load if a partial admission steam turbine - perhaps water in steam that can occur from picking up condensate in pipes.

 

[gallaz]

I expected that bearing oil temperature would increase due to the increased friction associated with vibration in a journal bearing. Let me know if this is wrong, we have considered oil flow problems but as yet have not run any tests for it.

The problem has not occured for some time as the machine is running on load limit at less than full load.

 

[Franko]

My experience is that bearing temperature does not change much unless there are the two circumstances given above. If it's not an oil supply problem, the cause of the vibration would also have to simultaneously affect loading of the bearing. Also check oil drain sight glass before and after next vibration excursion - should not be running full with minimal foaming (do you vent bearing caps with a small breather?). Thre may be tips in:

http://www.lube-tips.com/BackIssues/2002-02-20.htm

 

[chjura]

I have understand that the peak occurs during the start up of the turbine just when the turbine have reached nominal speed and after a 10-20 min period reduces to normal running conditions. Is this correct?

I agree with DAVELEO that this is due to thermal gradients that "blend out" after some period of running, but not necesary throughout the rotor, but throughout the turbine casing.

After the turbine is shut-down the rotor barring system is rotating the rotor at about 200-250rpm and therefore it is not very likely that the rotor will be thermally-bowed. But on the other hand the turbine casing could egxert thermal-distorsion and that could cause large bearing loads or tip-clearance exitation. After same time (10-20 min), the thermal gradinets reduce to normal running conditions together with the vibration level.

 

[gallaz]

Sorry, I didn't explain the situation well enough. The turbine/generator is running continuously at 3000rpm. The vibration events come and go. The vibration increases to a peak and then decreases with the same gradient and returns to normal vibration levels. This process takes 10-20 minutes.

 

[hacksaw]

You are describing load changes on your machine.

 

[gallaz]

There are sometimes load changes associated with these events, but not always.

 

[hacksaw]

the vibration levels routinely change with load with most machines. your turbine manufacturer can advise.

ultimately you have to open the machine for a detailed inspection.