Air entrainment/cavitation/vortexing issue

[CheemaJ]

Hi, I'm having some issues with a centrifugal pump which is being used on a CIP tank.

Pump is 18.5 kW and is on a frequency inverter drive.

[ul]
[li]Tank capacity is 2,500L[/li]
[li]Pump suction pipework diameter is 3" NB[/li]
[li]Pump suction pipework length is ~ 1m[/li]
[li]There is a full bore ball valve in the suction line and a Y-strainer[/li]
[li]Liquid level is minimum of 1m above suction of pump[/li]
[li]Tank is vented to atmosphere[/li]
[li]Liquid is water at 20 degC[/li]
[/ul]

It's possible to re-circulate back into the tank, so I've been testing the pump using this route. I'm finding that at low speeds the pump runs quiet and no vibration. As I inch the speed up and it gets to around 35 hZ the noise and vibration starts. At 45 hZ it is extremely loud and vibrates a lot.

I've tried throttling the discharge by closing a valve - the noise seems to get worse.

When I throttle the inlet ball valve, the noise improves.

Have removed the filter basket from the strainer and didn't really make any difference.

I am wondering could this be an issue with vortexing in the tank? Is there any way to check this or remedy it?

The pump has several CIP duties, one of which is to supply liquid to a rotary spray jet at 5 barg. The flow rate at this pressure is 9,200 Litres/hour. So relatively low flow. The pump has to run at 45 hZ to generate this head and the issue of noise and vibration persists - despite the low flow. So then I wonder is it not vortexing and something else?

Thanks in advance.

 

[Artisi]

Is the return into the tank just freely discharging into the tank, ie free-fall, if yes you are probably driving air into the tank which is then entrained into the pump 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.)

 

[CheemaJ]

When returning into the tank there is definitely free-fall which would lead to air entrainment as you say.

But when the pump is delivering flow to the rotary spray jet, there is no return back to the tank and yet the issue persists.

I've just tried closing valves on both suction and discharge pipework around pump in an attempt to stop any air from getting into the pump case. Ran the pump and it still sounds rough.

 

[Artisi]

Is there any evidence of air entrainment or vortex from the surface with only 1 metre of coverage over the inlet?
Can you post the pump curve and a bit more detail on the hydraulic duty.

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.)

 

[CheemaJ]

I'd have to remove a flange on the top of the tank to get a good look but will try this.

Also I've realised that the inlet strainer is less than 5 diameters from the suction flange. Considering getting a spool piece made up to put in it's place. With the spool piece in place there'd be about 10 equivalent diameters of straihgt pipe. Could that improve things do you think?

Attached a copy of the pump curve. Just waiting for NPSHr information from the supplier.

Duty details are as follows:

Flow rate required m3/h
Duty 1 – 33
Duty 2 – 51
Duty 3 – 9.2
Duty 4 – 51

Total pressure drop
Duty 1 – 32.1 m
Duty 2 – 38.6 m
Duty 3 – 57.4 m
Duty 4 – 49.0 m

 

[Strong]

Sounds like a guessing game about vibrations with no vibration data! Yes, a hydraulics cause is possible, but a resonant structure (pump and piping) is also possible given the variable speed. A quality vibration analysis can pinpoint the cause of the vibrations and lead to a practical solution.

Walt

 

[LittleInch]

The attachment won't open as it has a strange symbol in the file name - can you repost please.

The flow and head look high range.

I think you're running either on the extreme left or right of your pump curve.

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

 

[danschwind]

Are you able to record a video to show us this vibration?

 

[georgeverghese]

What configuration do you have for min flow recirculation for this pump ? A min flow recycle loop with FCV or a PCV or a crude RO? What is the set flow for this min flow line at the speed corresponding to 50Hz and is it the same setting as that recommended by the pump vendor? Can we see a sketch of this setup or a P&ID?

 

[Artisi]

The problem isn't vortexing, cavitation, or air entrainment, it's simply operating too far left on the curve at 9.2m3 & 57.4 m at near full speed.
It's blade pass frequency, shaft deflection, internal recirculation just to point out a couple of areas causing the problem.
if you need to operate at this duty - you will need to by-pass some of the flow back to the tank or pump 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.)

 

[CheemaJ]

Thank you all for your replies.

There is not currently a minimum flow bypass installed around the pump however I think this would be beneficial.

Based on the curve, could you recommend the best method for sizing this and type of valve for controlling the flow?

Also, it would be easier from an installation point of view to pipe the re-circulation line back to the pump suction pipework - is this acceptable?

Given the relatively small size of tank I am thinking of installing a vortex breaker as well.

Out of interest, for future pump sizing, when presented with a pump curve like that above - how would you decide whether a minimum flow bypass is necessary?

 

[LittleInch]

If pushed the vendor can determine a Min continuous flow line.

Your curves listed don't have efficiency on, but if they did a figure of 30% from the Best Efficiency Pont (BEP) is a good starting point as a MCF figure.

but it is a very pump and situation dependent thing. Some pumps won't like anything below 50% of BEP, others are Ok down to 10 or 15%.

ARV (automatic recirculation valves) are good but might be a bit OTT if you don't have it happening very often.

A cheap and easy way is simply a manual valve with an orifice suitably sized which you open at the same time as your duty 3 operation

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

 

[georgeverghese]

The pump vendor should tell you the min flow for the 50Hz operation straight off the bat. Another simple alternative for the min flow is to use a self operated backpressure CV.

 

[Artisi]

As LI has noted minimum flow is to too variable to put a number on, only the manufacturer can give real data or your own site testing will soon tell you, as you have ready discovered.
Yes you can recirc. back to the inlet, again seek advise from the manufacturer of someone with a similar arrangement.

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.)