Hello All,

We have a couple or three pumps that, from time to time, are plaqued with a peak and harmonics that is just above running speed of a 2 pole induction motor. The peaks, sometimes as high as 4.0 in/sec, seem fairly close to appearing in the pumps at about the same time even though they are several hundred feet apart and several pieces of equipment between them.
The peaks are not at the same frequency so I am leaning toward a resonance on each pump system/piping. And what I am also leaning toward is there is a change in the viscosity of the gas or oil (not sure what they actually pump but they are in a refinery) causing the piping to be excited which in-turn is vibrating the pumps.

At times the peak is not present and at times it is present but very low and then BANG at times it is very high. All this is observed from the monthly surveys.

I do not know for how long the peaks remain in the system since this is in a refinery we visit only once a month. I may need to spend a little time revisiting them during the 2 day survey time of the refinery to see if the peaks subside while I am there.

The motors are not variable speed.

Anyone ever seen this type problem before?

I have attached a plot of them here.

Mar's_pumps_June_2007.doc (215 Kb, 62 downloads) Strange Peaks in Pumps

Hey Ralph
Could this be vane pass or are these going over relief at certain times? I have some pumps that act weird sometimes on one visit they will be fairly quiet and the next month vibe levels are a lot higher. I have 7 of these in two locations and they all behave the same.I have not been able to figure out why so I will follow this post closely.

Lee

quote:

Could this be vane pass or are these going over relief at certain times?

Don't think it is vanepass, the orders are about 1.2 to 1.4.

What do you mean by "going over relief"? Can't recall ever hearing that expression.

could be pump recirculation from excessive wear ring clearances. Check the suction and discharge piping one of them will be pulsing at those frequency. This type of vibration will appear in one plane like a resonance. If they are turbine driven you change the speed a little and see the vibration drop instantly.
This is very common in hot condensate service and light hydrocarbon with high delta suction to discharge pressures. Also change in viscosity as you mention will cause recirculation if wear rings are marginal.

Ralph
As Erik put it pump recirc. Just a nother way of putting it. I know our pumps do this a lot.Not a good thing but it does happen. I am working with water and not familiar with refinery applicatons. But I think your idea about viscosities might be something.I on the other hand have a fairly steady viscosity it is a mixture of a synthetic lube and water. Which sometime is diluted when we have a big leak but never gave it that much thought.Maybe pipe strain the viscosity and resonace could be the mix that causes this.

Erik,

I've always associated recirculation pulsation to be frequencies below 1xSS. Is there any technical paper or other references that describes frequencies above 1xSS caused by recirculation? My expectation would be that for unidirectional vibrations that a resonant structure would have to be present. What other vibration or pump performance effects have you observed for this type of non-synchronous vibration? I am just trying to learn more about this.

Walt

Walt
Most recirculation or flow induced vibration is directional when it involves a volute and impeller. The cutwater area in the volute has the highest pressure profile compared to the rest of the volute. Worn wear rings increase slippage from the highest pressure back through the impeller suction eye wear ring. I can see this happening radialy somehow. Most of the time I see it around 1x but I have a pump that has it around 4x to 5x. The pump was overhauled and problem went away. I will post it next week.

quote:

could be pump recirculation from excessive wear ring clearances.

Erik,

It might be wear ring clearance, we rebuilt one of thees pumps and the vibration went away for about a couple of months but quickly returned.
The thing I noticed is the vibration will appear and disappear on its own. So this rebuild job may not have caused the amplitude to decrease since the peak has returned and left again since the rebuild.

Walt, I have not observed anything other than what I have mentioned here. I think I have some data on the suction pipe, but it is on the flange where it bolts to the pump and not on the piping itself. The next trip I will test the piping, even impact, if I can catch the pump down.

quote:

Originally posted by Ralph Stewart:
And what I am also leaning toward is there is a change in the viscosity of the gas or oil (not sure what they actually pump but they are in a refinery) causing the piping to be excited which in-turn is vibrating the pumps.

Pretty rare to get up to 4 ips without some kind of resonance involved, mechanical, acoustical or both. If you pursue this line of thinking further, suggest you (or owner) track the effect of process / product conditions, particularly temperature. Generally, temperature will affect sound speed and viscosity. Variable sound speed will influence acoustical frequencies, and variable viscosity will influence the sharpness of the peaks.