Skip to main content


Could use some advice...

I am troubleshooting a Circulating Water Pump in a Power Plant... this pump is rated for 90k GPM and operates in parallel with a sister pump (also rated for 90k GPM)....both pumps operate in a closed loop and are used for condenser cooling.

This pump operates at 7.46 Hz.

I am reviewing the data I collected today,and I wanted to post this to ask some questions.

When I got to site... I found this pump had an amplitude of 5 mils peak to peak in the MOH direction (90° from flow) and nearly 15 mils peak to peak in the MOV direction (inline with flow)...highly directional (dominant 1x). I took cross channel phase data (1x) at the top of the motor and found that there is a 90.48° phase shift from horizontal to vertical.... looks like a text book imbalance.

I was however a little concerned that the highest amplitude is in the most stiff direction on this installation.

If you look at what I have attached.... you can see my ODS model that details the installation (havent done the ODS yet....I am doing it tomorrow), a linear scale spectrum from the MOH position and a log scale spectrum from the MOH position (90° from flow), as well as the same from the MOV position (inline w/ flow).

You can see in the MOH data that there is a resonance above 1x (7.46 Hz).... it's at about 10 Hz.... there is a ~25% separation margin there. 

However.... if you look at the MOV location... there is a resonance at about 8.23 Hz...this is about a 9.4% separation margin.

I cant shut this pump down to impact test it.... I am going to try to do some negative averaging impact testing with my 2 channel tomorrow but I haven't liked the results I have gotten in the past with this. I am going to try to take 50+ averages tomorrow to see if it will clean the signal up some.

Here is what I am thinking is going on.... they have a certain degree of imbalance that is being amplified by resonance....they are borderline on the separation margin and could stand to apply some stiffness to increase that margin.... they should also address the imbalance as well. 

My question is this...(1) please let me know what you think on the above (2) Assuming you agree with my train of thought.... could the proximity to resonance give me this 90° phase shift and maybe be making things "look" like imbalance when it's really not? My gut says it's a true imbalance that is being amplified by resonance.... but wanted to run this by the group and get some feed back.

Please let me know your thoughts? Thanks so much as always for your help.

Kind Regards,





Files (1)
CWP Docs
Original Post

Example: same pump and flow as yours, due to large amount of water movement lots of silt tends to come with all this flow. The silt deposited itself in front of the suction of the pump, like 50' sandbar, this in turn affected the 1x rpm and flow of the pump. Both of these examples, pump has a resonance right at 1x rpm!

Example: same pump and flow as yours, pump was rebuilt during major overhaul, put back in service and within a few months high 1x develops, ODS showed a clean circular orbit, imbalance, diver hired to inspect impeller, found a large chunk of repaired impeller missing

Directional vibration leads me to believe resonance, but what is the history of this pump, does the grout look compromised? Did the 1x just show up or been there for awhile, how many pumps are showing this? Yes I agree there is some imbalance, but where did it come from.....what changed, stiffness or damaged impeller?

What about phase from motor to top of pump? Is the motor rocking on top of pump head or is the whole machine moving?


Have you checked the bolts on the ground and motor bolts?

You are saying the highest stifness is at motor's vertical direction. That is not true when you deal with console type motor attachment. Or am I seeing the machine in wrong perspective?

How do you know those frequencies you mentioned are resonant frequencies?

Hi Dave - Thanks... the pump was repaired in the spring time and has vibrated ever since... there is only monthly route data that the plant takes which shows the elevated vibration amplitudes since day one... I am not sure if the imbalance was always there or something internal has occurred to create the imbalance. I have done an ODS, I have just not had time to animate the model.

Hi Becar - As I mentioned above, I have done an ODS, but have not had a chance to animate yet.. I will share those results. I am able to identify the resonance frequencies by looking at the spectrums in log scale. They are very clear.

Thank you both for your time reviewing this.


Add Reply

Link copied to your clipboard.