Sub Synchronous fault

Hello:

Hope all are fine.

I am facing the high vibration issue of Vertical VS4 type pump (as per API-610). Pump is mounted on the vessel for draining the acid water. Here are some spectrums for your reference, needs your comments/suggestions.

Recently the pump is overhauled and bearing(bush & rolling) was changed.

Motor is in perfect condition gives 0.3mm/sec vibration in solo run.

Attachments

Files (1)
Original Post

Vertical pumps on vessels can have vibrations from other machines, so check into that first.

Typically vertical pump can have one or more natural frequencies (vibration modes) below 1xSS. An impact test may verify this. Typically an increase in vibration amplitude indicates impact excitation (assuming first item is checked) caused by shaft bearing or pump wear ring increased clearance. Also pump impeller may be impacting case if Lift setting is incorrect.

Walt

I agree. I would also add the possibility of turbulence.

 

It might be my first step to take data over several  minutes or so to understand if the patterns are really stable or not. A CSI unit can do this using spectrum and overall monitor modes.

Before touching any internal parts, it's important to eliminate external causes. The data may show beating while the pump is running. Take background monitor mode data to evaluate this possible issue. In the same time, do the impact test.

 

Regards- Ali M. Al-Shurafa

I really appreciate your responses walt & shurafa....I'll must check for natural frequency. Whereas as per said of shurafa regarding monitor mode, I have done this type of activity but one time only suddenly vibration drops to 5mm/sec..but now i continously monitor in live mode and not observed any decrease in vibration...here you both might be correct for turbulency and wear ring clearnces. As i have reported that the wear ring clearnces increase intentionally(but it is in under limit of manufacturer) also maintenance team put mesh type net of S.S at suction side might be the holes are too small which create some turbulency. Lets see what happened.....any suggestions after the scenario mentioned.

Rohit02 posted:

Dear Mohiuddin

Is this machine by any way a VFD drive? As the RPM mentioned on the spectrum is varying from 3288 to 2988 rpm.

Secondly is the RPM of pump is same as the Motor or there is any gear unit in between? 

Regards

Rohit

Dear rohit..VFD is not installed and motor synchronous speed of 3000 rpm....and pump is directly coupled,,same rpm..might be rpm varying by electric frequency change or any parameter issue in spectrum.

John from PA posted:
mohiuddin posted:
might be rpm varying by electric frequency change or any parameter issue in spectrum.

You line frequency varies that much?

I also observed first time..I am just surprising to see the comment by Rohit, well noticed by him. I will check what the issue is...

mohiuddin posted:
John from PA posted:
mohiuddin posted:
might be rpm varying by electric frequency change or any parameter issue in spectrum.

You line frequency varies that much?

I also observed first time..I am just surprising to see the comment by Rohit, well noticed by him. I will check what the issue is...

I doubt the speeds mentioned on the right side of the plots were actually measured values. By default CSI plots are designed to show the speed with each spectrum based on certain "best guesstimation". The speeds in the attachment could be actual in this particular case but this would be, if true, very odd for a synchronous motor. The speed figures are calculated by the software and presented with the spectra even if the machine does not have speed pickups. The software does its calculations based on the pre-configured speed of each point.

Correctly mentioned that CSI plots are designed to show there RPM values. Adding to it the order calculation in CSI is also done based on this RPM. and hence if the RPM mentioned is not the actual RPM then there will be fault indication of order of frequency. 

Calculating with the provided RPM say 3000, 0.75 order corresponds to 2250. This could not be the running speed as then there would be a slip of 750 cpm. And that would be too much.

possibly the vibrations seems externally excited.

Rotating Guy posted:

hi,

Does the equipment run at its BEP? can you adjust the fmax to 800 and no of lines, and revert again. if you could get the phase for additional inputs is better.

regards, 

 

Yes, we adjust the flow at different values in ''Prefered operating region''. also at nearly about BEP. but no significant change has been observed.

Phase also conducted, but magnitude is too low around 1.2 mm/sec,cwhich is satisafactory.

I will take data next time at Fmax:800 and at 1600 L.O.R..and share you 

Rohit02 posted:

Correctly mentioned that CSI plots are designed to show there RPM values. Adding to it the order calculation in CSI is also done based on this RPM. and hence if the RPM mentioned is not the actual RPM then there will be fault indication of order of frequency. 

Calculating with the provided RPM say 3000, 0.75 order corresponds to 2250. This could not be the running speed as then there would be a slip of 750 cpm. And that would be too much.

possibly the vibrations seems externally excited.

Right!! 2250 cpm is not possible.....

Can you give e any hint about external vibration source: the equipment is mounted on the vessel separately from other pump. 

Rotating Guy posted:

if you could get the phase for additional inputs is better.

regards, 

 

I missed to comment on this in my last post. 

The common meaning of phase is associated with a shaft position at certain conditions such a specific speed. In this particular case, the phase angle might be unreliable of the key frequencies. The phase of 1X is associated with very low amplitude (less important and less reliable). They key amplitude is assumed at 0.7 X which has not stable phase and is not measured by CSI.

Maybe measuring the actual speed by an independent sensor (not through the vibration spectrum) is a necessity in this case for many reasons.

Regards- Ali M. Al-Shurafa

Shurafa posted:
Rotating Guy posted:

if you could get the phase for additional inputs is better.

regards, 

 

I missed to comment on this in my last post. 

The common meaning of phase is associated with a shaft position at certain conditions such a specific speed. In this particular case, the phase angle might be unreliable of the key frequencies. The phase of 1X is associated with very low amplitude (less important and less reliable). They key amplitude is assumed at 0.7 X which has not stable phase and is not measured by CSI.

Maybe measuring the actual speed by an independent sensor (not through the vibration spectrum) is a necessity in this case for many reasons.

Regards- Ali M. Al-Shurafa

Yes shurafa is right, measured phase didnot show any fault relating identification(for 0.75X), rather I saw the phase shown on CSI is only about 1X which has low aplitude(description: Motor DE and NDE are in phase horizontaly and vertically with low amplitude of 1.2-1.5mm/sec., also at Motor NDE phase between horizontal and vertical is 90º with low amplitude of around 1.3mm/sec, Coherrence is 0.99, which is satisfactory according to me.) BUT ISSUE ABOUT 0.75X is not solved through phase relationship.

mohiuddin posted:
Rotating Guy posted:

hi,

Does the equipment run at its BEP? can you adjust the fmax to 800 and no of lines, and revert again. if you could get the phase for additional inputs is better.

regards, 

 

Yes, we adjust the flow at different values in ''Prefered operating region''. also at nearly about BEP. but no significant change has been observed.

Phase also conducted, but magnitude is too low around 1.2 mm/sec,cwhich is satisafactory.

I will take data next time at Fmax:800 and at 1600 L.O.R..and share you 

Can you share the picture of the pump? i want to verify the piping and the structural condition. 

What i mean in getting the phase is for you to start checking the ODS. since the dominant frequency is subsync or maybe this subsync is the primary freq of other machine and hence it can contribute as dominant freq in this equipment. 

Another thing is looseness also contribute this subsyc, most probably in bushing or any related. 

 

regards,

 

Rotating Guy posted:
mohiuddin posted:
Rotating Guy posted:

hi,

Does the equipment run at its BEP? can you adjust the fmax to 800 and no of lines, and revert again. if you could get the phase for additional inputs is better.

regards, 

 

Yes, we adjust the flow at different values in ''Prefered operating region''. also at nearly about BEP. but no significant change has been observed.

Phase also conducted, but magnitude is too low around 1.2 mm/sec,cwhich is satisafactory.

I will take data next time at Fmax:800 and at 1600 L.O.R..and share you 

Can you share the picture of the pump? i want to verify the piping and the structural condition. 

What i mean in getting the phase is for you to start checking the ODS. since the dominant frequency is subsync or maybe this subsync is the primary freq of other machine and hence it can contribute as dominant freq in this equipment. 

Another thing is looseness also contribute this subsyc, most probably in bushing or any related. 

 

regards,

 

Ok.I will share to you.....Pump bushes are replaced 4 days before the high vibration observed, so less chance of it.

I notice at motor ob vertical the waveform is looks like modulating for me, the reason why i want to see the picture of pump is to check the piping support if any or connection to the pump. Also, looseness will also contribute to your high vibration, can you also tighten the foundation bolt. 

regards,

Dear, at All.
I had similar experience before with vertical motor, subharmonics peaks appeared when the shaft was moving up and this behavioer was related with flow changing, we checked the nut of thrust rolling bearing and we found was untighten.
this kind of motors have angular rolling bearings in tandem arrangment for control of axial load (only one direction restriction) , and if the shaft move up can generate loossneses in this kinds of bearings.
If the shaft move up is abnormal behavior, it is necesary find the cause why it happend, (design problem or bad axial rotor setting).

Hello

we have found the cause of vibration, after maint. significant reduction is observed..we found bolt loose of buffer vessel (buffer seal pot), after tighten the bolt the vibration drops from 21-23mm/sec to 12-14 mm/sec (still at  higher side)..attached find the file to see pictures.

WE WANT FURTHER REDUCTION, NOW OUR WORKING IS GOING ON, TO LIMIT THE VIBRATION AT COMPROMISING LEVELS.

Attachments

Files (1)
mohiuddin posted:

Hello

we have found the cause of vibration, after maint. significant reduction is observed..we found bolt loose of buffer vessel (buffer seal pot), after tighten the bolt the vibration drops from 21-23mm/sec to 12-14 mm/sec (still at  higher side)..attached find the file to see pictures.

WE WANT FURTHER REDUCTION, NOW OUR WORKING IS GOING ON, TO LIMIT THE VIBRATION AT COMPROMISING LEVELS.

You could do cross phase when detecting loose bolt part. 

regards,

Interesting. Is this vertical vessel a part of a mechanical seal piping plan?

I don't recall seeing a bracket like this to support accessories to a pump's structure. It could be completely perfect, simply it's not common in my world (which is very small).

Regards- Ali M. Al-Shurafa

mohiuddin posted:

Hello

we have found the cause of vibration, after maint. significant reduction is observed..we found bolt loose of buffer vessel (buffer seal pot), after tighten the bolt the vibration drops from 21-23mm/sec to 12-14 mm/sec (still at  higher side)..attached find the file to see pictures.

WE WANT FURTHER REDUCTION, NOW OUR WORKING IS GOING ON, TO LIMIT THE VIBRATION AT COMPROMISING LEVELS.

You can try some rubber packings. It might help in further vibration reduction.

Shurafa posted:

Interesting. Is this vertical vessel a part of a mechanical seal piping plan?

I don't recall seeing a bracket like this to support accessories to a pump's structure. It could be completely perfect, simply it's not common in my world (which is very small).

Regards- Ali M. Al-Shurafa

YES!!! This is a seal pot vessel for buffer liquid for the M.seal plan

ramonruiz posted:

Dear MOHIUDDIN

Are the same amplitude, radials vibration readings in all directions?
for example, same direction to pipe line and orthogonal direction to this (90 degrees each other).

Best Regards.

 

 

 

No vibration are different at radial position...Refer to the attached file in previous session shows image with arraow sign.

Dear Mohiuddin,

First, for sharing data and simplifying nomenclature, I suggest you use a different point naming convention for verticals as opposed to horizontals. We use Motor Top Inline (inline with discharge piping) or MTI and MTP (p for "perpendicular to discharge) and MTA (axial) then MBI, MBP, and MBA. 

To address the sub synchronous vibration, with verticals, by far the most common sub-synchronous excitation is too much freedom in the line shaft bushings allowing shaft whip (and this often coinciding with shaft resonance). This is so common, it is hard for me not to be fixated on it. Do you have any extensive history on this machine? Is this new? Or has there always been a fairly large sub-synchronous component at around the same frequency?. 

Add Reply

×
×
×
×