Secondary Chilled Water Pump standby vibration problem

[abdulmdg]

Greetings!

I work at a Process Cooling Plant where we serve chilled water and we are experiencing some problems with our secondary pumps. We have a total of 10 secondary pumps in the plant that are connected in parallel. Two pumps are running and the other 8 are standby. The standing pressure of the standby pumps is 2.16 bar. The problem is, the pressure of some standby pumps drop to 1.5 to 1.9 and start vibrating out of nowhere. We would "fix" it by turning the vibrating pump on and turning another off. It's more like a temporary fix. But as time passes, the vibrations happen more frequently.

What might be the problem here? Is this a cause of cavitation? Some of our technicians suspect it is a problem on the check valves.

[beenthere]

Moved to Tech to Tech commercial forum.

[lzenglish]

Sounds like Turbulent Water, causing Cavitation, to me. How long has this situation gone on? What has changed in your process plant, as of late? I would also look close at the check valve/valves.

[pecmsg]

we need pressures in and out?

Speed drives or fixed speed?

[servicefitter]

37.7 psi= 2.6 bar is that pressure correct? What is the suction pressure and have anyone looked at a pump curve? How are they.piped off the same header?

[abdulmdg]

It's been happening long before I started working here (January 2019). We supply chilled water to a Turbine Air Inlet for a Gas Turbine plant. I think what changes often is the chilled water return from that plant. It goes way off the designed chilled water return for the chillers.

[abdulmdg]

we need pressures in and out?

Speed drives or fixed speed?

Pressures in are usually 2.1 bar. Pressures out are 5 to 6 bar. They are fixed speed pumps.

[abdulmdg]

37.7 psi= 2.6 bar is that pressure correct? What is the suction pressure and have anyone looked at a pump curve? How are they.piped off the same header?

Suction pressure is at around 2.1 bar. They are piped off the same header in parallel. How do you use the pump curve? Am I going to need a pump curve for the parallel connection or is one pump curve of all the pumps enough?

[lzenglish]

There are 2 ways to go after this problem. The first is to continue to throw guesses at it, and hope one of them lands, OR, you can start at the beginning, and work your way thru it. I would do the second, as it will not only be a challenge for you, but, you will learn so much more in the process. Step one, locate the mechanical design blueprints and make copies to mark up. Step 2, Interview all past and present engineers who were/are operators, and quiz them concerning this problem. Step 3, Locate all start up, and daily logs, which will indicate past and current pump readings, pump change outs, modifications and or repairs.
Step 4, Find all pump literature pertaining to these pumps, to include installation baseline frequency values, performance curves, and design and operating parameter's. Step 5, Armed with all of that information, start at the beginning, and systematically "Wring the Entire Chilled Water System out".
This will not be done in one day, or even a couple of days, but will take as long as it takes. As you are wringing the system out, make engineering sketches, and mark up original copy prints to show all your findings. When you get to the root of the problem, you will know it ! Good Luck to You !

[Scoobie]

You say the pumps NOT running are vibrating. Could water be going backwards through the discharge check valves on these pumps? Try closing a valve on one of the standbys when it starts vibrating.

[servicefitter]

Model # brand of pumps, impeller size and a pic s of the pumps would be nice. Lzenglish is correct need to look at the prints all the pump info should be on them and piping. Working in some process plants never have seen 8 pumps 6 are back ups. Tring to picture having 8 pumps on the same header and are they piped reverse return?

[abdulmdg]

There are 2 ways to go after this problem. The first is to continue to throw guesses at it, and hope one of them lands, OR, you can start at the beginning, and work your way thru it. I would do the second, as it will not only be a challenge for you, but, you will learn so much more in the process. Step one, locate the mechanical design blueprints and make copies to mark up. Step 2, Interview all past and present engineers who were/are operators, and quiz them concerning this problem. Step 3, Locate all start up, and daily logs, which will indicate past and current pump readings, pump change outs, modifications and or repairs.
Step 4, Find all pump literature pertaining to these pumps, to include installation baseline frequency values, performance curves, and design and operating parameter's. Step 5, Armed with all of that information, start at the beginning, and systematically "Wring the Entire Chilled Water System out".
This will not be done in one day, or even a couple of days, but will take as long as it takes. As you are wringing the system out, make engineering sketches, and mark up original copy prints to show all your findings. When you get to the root of the problem, you will know it ! Good Luck to You !

Thanks a lot! These steps are really helpful! I'll try and do all these as best as I can. Thank you!

[abdulmdg]

Model # brand of pumps, impeller size and a pic s of the pumps would be nice. Lzenglish is correct need to look at the prints all the pump info should be on them and piping. Working in some process plants never have seen 8 pumps 6 are back ups. Tring to picture having 8 pumps on the same header and are they piped reverse return?

I'm new to this site and from what I understand, I'm still not allowed to post pictures.

Right now, because of low ambient temperatures and on-going preventive maintenance, our client only requires a few chillers and pumps running.

[abdulmdg]

Model # brand of pumps, impeller size and a pic s of the pumps would be nice. Lzenglish is correct need to look at the prints all the pump info should be on them and piping. Working in some process plants never have seen 8 pumps 6 are back ups. Tring to picture having 8 pumps on the same header and are they piped reverse return?

The pumps are Bell & Gossett VSH 12x14x17.5A. Impeller diameter = 445 mm.

[lzenglish]

Thanks a lot! These steps are really helpful! I'll try and do all these as best as I can. Thank you!

You are Welcome, and Please let us know of your findings.

[servicefitter]

http://bellgossett.com/literature-library/#search/Vsh
Just page up to find your pump.
So this is a 700 hp 1780 rpm double suction pump
The 2.1 & 6.0 bar running 2 pumps?
Are you using a know good gauge with the gauge valves flushed. Should be some documentation from the start up of the plant like the water balance report.

[Deux1]

Sounds like you are getting some short-circuiting through the standby pumps, which would suggest a problem with the check valves.

[71CHOPS]

If there’s is a problem with the check valve, how is the standing suction pressure dropping?

With roughly 80 psi on the discharge and 40 on the suction, there should be an increase in standing suction with back flow, should there not?

Sounds more like a cavitation problem, causing a loss of flow.

[WAYNE3298]

Cavitation in a chilled water system has a lot of energy. If you have cavitation the pump impellers will be damaged in short order. Ay least some of the secondary pumps should be VFD controlled or at least have some sort of pressure control.

[Poodle Head Mikey]

The only thing I can quickly think of that would make a not-running pump vibrate is water flow through it. Especially air filled water. Do the shafts of the vibrating-non-running pumps turn when the vibration occurs? If so; is it normal or reverse rotation?

Are there discharge checks on all the pumps?

AS switching the pump on stops the vibration - I would suspect reverse flow through the vibrating pump(s).

I think I would first try loosing the packing and closing an isolation valve on a vibrating-non-running pumps. If that stops the problem I would take-a-look-inside the check valve / replace the check valve.

Another approach would be to close a valve on each standby pump - every time it is shutdown - opening it only to start that particular pumps.

Why are there 8 standby pumps supporting only 2 operation pumps? What controls the pump operation? Are they supposed to be linked to the chiller loading? And here I am wondering why the chiller RWT spikes.

PHM
-----------

Greetings!

I work at a Process Cooling Plant where we serve chilled water and we are experiencing some problems with our secondary pumps. We have a total of 10 secondary pumps in the plant that are connected in parallel. Two pumps are running and the other 8 are standby. The standing pressure of the standby pumps is 2.16 bar. The problem is, the pressure of some standby pumps drop to 1.5 to 1.9 and start vibrating out of nowhere. We would "fix" it by turning the vibrating pump on and turning another off. It's more like a temporary fix. But as time passes, the vibrations happen more frequently.

What might be the problem here? Is this a cause of cavitation? Some of our technicians suspect it is a problem on the check valves.