Pump Knocking - Apparently not cavitation, so what is it?

[BronYrAur]

I have a centrifugal pump that is causing a lot of vibration and knocking in the adjacent piping. The pump draws ambient water from a tank, sends it through a weir-type diaphragm valve (wide open) and then through a UV "exchanger". The water leaves this area and goes out to process branches which consume the water. There is a bypass valve out in the system that maintains a differential and send unused water back to the tank. The tank also has a makeup valve that maintains a level.

The pump has at least 10' of static water head in the tank, so I should be fine for NPSH. If you listen closely to the video, you can hear knocking that sounds like I am pumping marbles. The manufacturer has apparently ruled out cavitation (not sure how, other than impeller looks good).

I am wrong with some of my pipe size comments on the video. I got interrupted and didn't have a chance to take another one. The connection on the bottom of the tank is 3". Suction pipe increases to 4", which is the suction connection size. Pump discharge is 2-1/2", and pipe expands to 3". Wier valve is 3". I am not familiar with the UV device, but its nameplate says that it is rated for 1000 GPM. I am told that the system pumps around 200 GPM. I am trying to get a pump curve. They run the pump around 75% on the drive. If they crank it up, the vibration gets crazy to the point where they are breaking the UV lamps.

There are no flex connections where the piping connects to the pump. I think this would eliminate the vibration that I can feel when touching the pipe. But, what about the knocking? That seems worse than the vibration. Any thoughts?

https://youtu.be/Tz0QKc1lhf0

 

[LittleInch]

Sounds very much like excess flow to me but you need more information/ data. What are those pressures? Will need a pump curve and a flow diagram to make some sense of this.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[hacksaw]

The UV unit is using ultraviolet light to sterilize the water. Must be a bio-pharma unit, with s lot of thin walled piping.

You may be dealing with pulsations in the pump discharge, so with the rather complex piping layout and plenty of opportunity for trapped gas bubbles in the pipe, you may be dealing with resonance effects in the discharge piping that will get drastically worse with increased flows.

You definitely need a piping design engineer involved, and high point vents to insure liquid full piping.

Good luck!

 

[Compositepro]

It sounds like there is lots of gas bubbles in the flow causing water hammer when going through a restriction. Where would the gas be coming from? Perhaps a seal leak where the pressure in the pipe is sub-ambient. Is there a filter somewhere in the pump suction?

 

[georgeverghese]

At 200gpm, velocity is 2.8m/sec(9fps) for 3inch (about acceptable), and would be unacceptably high at 14m/sec at 1000gpm, which would explain the high vibration / breaking UV lamps in the steriliser unit when trying to ramp up beyond 200gpm. Lines seem to be undersized for 1000gpm too. Sounds like a high speed pump, so NPSHr may be high.
Could there be air in the make up water line?
Could the pump be operating at below min flow for the developed dp? - they ought to be returning water based on the dp between pump immediate discharge and pump suction.

 

[Artisi]

So far all guesswork crystal-ball gazing, need a lot more detailed data.
Just one point for quick check, how is returned / by-past water being returned to the inlet side - you could have highly aerated water entering the inlet.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[Latexman]

How old is the installation? Has it always done this? Or did it start all at once after some time?

Good luck,
Latexman

Engineers helping Engineers

 

[danschwind]

If you leave that discharge valve barely open (i.e. high pressure drop), the noise and vibration persists?