Pump Vibration Problem!

This is an absolute shot-in-the dark, as it's way out of my field, but could there have been any ground movement causing a change in the angular orientation of the motor/pump rotors? I imagine that something with that kind of inertia rotating at that speed could object quite strongly (gyroscope effect).

It's just a thought.

The fact that it came and went led me first to think of a change somehow in the process, but you say process parameters were unchanged. It is difficult to consider this further without seeing a system drawing.

I assume the pump & motor are mounted on a substantial baseplate, and the pump probably centre-line mounted?

You specifically mention that it was raining heavily at the time, and this is a fairly high temperature service. Could the rain have been cooling a certain area of the pump or system enough to cause differential expansion?

Pump and the motor are mounted on a baseplate 3.7 ton wt. Yes it is centreline mounted. There are actually 2 pumps in identical. The other one is not having this problem.

Then it is clearly a pump problem not a system problem.

With 38.1m NPSHa cavitation shouldn't be a problem.

Interesting that the problem came and went. Does it keep doing this, or did it do it just once?

If this kind of vibration is happening from time to time on the same pump without or with heavy rain, then the rain might not have anything to do with it.

Then, maybe a careful check of the coupling to the motor, the fixing of the impeller on the shaft for any play and the fixing for the motor cooling fan(if applicable) should be done when possible.

Might not be a bad idea to check impellar condition due to cavitation. Not an easy thing with the size of that pump.

What is the fluid? (308C - almost hot enough to melt lead)

The rain seems to be the only clue.

Could water or moisture enter the suction line?

Is there any way it could have been a slug of fluid with entrained gases passing through the pump?

Keep everything dry.

Total shot in the dark#1:

Moisture is lubricating some portion of the mounting hardware, allowing more travel when wet.

OK, I know it's not much.

Is there any chance that the pump is operating in it's critical speed range?

Is any of the pump piping exposed to the rain? If so the cooling effect of rain on piping carrying 308C fluid could be thermal contraction resulting in some excess and asymmetrical piping strain and a resultant pump casing distortion. Such distortion could cause internal rubbing of close clearances or cyclic loadings from slightly excess coupling misalignment. The amount of structural distortion that can create a problem like this is well within the range of tolerance variations between two "identical" installations. Thus a problem with just one pump. Carefully check for the presence of excessive piping loads. This stuff needs to be carefully engineered and controlled on pumps as large as you are dealing with.

Ed Weldon

You have pretty acceptable velocity readings of 1.3 to 1.7 mm/s. During the incident, what were the dominant peeks in the spectra?

How are you measuring the "online" shaft movement of 75 to 135 um, do you have proximity probes? If so, what does the orbit look like during the incident and what are the dominant frequencies?

Is this a between bearing pump with double entry impeller or a single stage overhung pump?

As Ed has said, the only clue that you have given is the "rain". Could you be distorting the pump due to uneven cooling of pump or piping due to the rain? Is the pump fully insulated and the insulation in good condition?

It is not clear whether pump is installed outdoors or indoor? Whether both the pumps were in operation at that time? The only difference caused is due to rains.

If the abnormility was existing during the rains only and other running pump is not affected, then the cause may be only due to difference in piping configurations at suction and discharge. The sudden change in temperature may cause angular twist at suction or discharge flange depending on length and bends in piping due to thermal contraction. This force is high enough to cause skewness in bearing journal contact sufaces resulting high vibrations.

I would once again thank all of you for putting up your valuable comments in response to my querry. I would make a list of these and visit the pumps shortly.

Some other details required are as given below:

The pump is installed outdoors and is pumping a hydrocarbon having 0.667 cP viscosity at 308 deg C.

It is a double volute, between bearing pump with double suction 1st stage.

Online vibrations at Non Drive End found rising suddenly from 75 microns pk to pk to 135 mircons pk to pk and that time it was raining heavily. After that vibration stabilised upto normal 67 microns pk to pk.

The other similar pump having identical piping and placed is similar situation didn't get this problem. Piping is insulated and so is the pump.

We do not know details of the pump, but I suggest that you check whether the pump is free to expand and contract. It may be that you are cooling the casing due to the rain and introducing stresses (axial or radial) due to differential expansion.

While replying in my last post I presumed that both pumps were in operation. You have not confirmed. Still I feel that you should criticlly check the piping configuration, whether they can cause thermal stress.

Normally in such machines cold misalignment is there to take care differencial thermal expantion between the pump and drive motor, so that it will get aligned at steady state operating temperature. It appears that heavy rains might have brought down the operating temp causing temporary misalignment between pump and drive shafts. in such cases a vibration spectrum if taken at that time could reveal more.

Thanks again.

Yes the other pump was operational at that time and no variation in vibration found in its operation.

As you have said there may be cold misalignment and the root cause could be moisture entering insulation either at pump or in the connected piping.

Pls suggest forward path for this case.

I think cold misalignment is misunderstood.

Cold misalignment is a deleberate act while aligning the machine with the drive in cold condition, so that at operating temp in steady state they get aligned. It is normally in radial vertical direction. Value of misalignment depends on ambient temp at the time of alignment and following parameters for machine & drive:

- Hight of shaft C/L from bed

- Body temperatures

- Cofficient of linear expansion

For your case machine may be around 2 mm up compared to drive while aligning in cold condition.

You have mentioned that "the root cause could be moisture entering insulation either at pump or in the connected piping". Cooling effect on pump body may cause radial misalignment and cooling effect on connected piping may cause twist and skewness resulting axial misalignment.

You might have missed the point that was specifically questioned above, that the pumps are centre-line mounted! This means that the pump is always aligned axially with the driver (within the required parameters of the driver and the coupling). In extreme cases, the pedestals apon which the pump casing wings are mounted, are also cooled, so I would disconsider this type of misalignment in this instance. Centre-line mounting precludes this type of misalignment.

If there is any misalignment problem, it is down to differential expansion and/or pipe stresses within the system, possibly due to localised cooling caused by the heavy rain, or maybe something else.

On this matter better Mr.Visal should have reacted after checking the alignment norms as per pump manufacturer. For this application, I strongly feel that the 'cold misalignment' must be there. OR else, pump should be heated in stationary condition before starting for suffient time by passing hot fluid through it or some other arrangement.

How long have these pumps been in service. do you have an hour run meter on them.

It may be because one of your pumps is coming to the end of its operatioal life, and needs to be rebuilt. (wear in the bearings, etc.).