air bubbles in pump discharge

[ports394]

i'm using a centrifugal pump to pump RO water. The RO water originally has some fine air bubbles. But they should leave over time.

I have a flooded suction with a 2.5" suction, and about 1ft of straight pipe before the suction.

I'm recircing the water right back into the tank, and it just continues to create bubbles and sometimes airy-foam. The recirc back into the tank is under the liquid level and you can just see all the air bubbles discharging.

I dont have any leaks in the system, pump doesnt vortex or appear to suck air in any of the fittings before the inlet. Any ideas?

 

[1gibson]

Packing?

 

[ports394]

I was wondering about that.

It's a direct coupled pump, and there isn't a gearbox etc. It has a mechanical seal.

None of the pump leaks, and I've had it up to 100 psi on the discharge. I would think that if someone were lose/worn and it was sucking air, that when the pump was off, or under high pressure, it would leak water out..

 

[ports394]

1gibson, I tightened the bolts on the housing. 8 bolts. eahc turned about 1/8th of a turn. And the bubbles went away.

I guess i don't understand why it didn't leak under high pressure. But it's fine now.

I spent all morning trying different combinations with the pump and the system to see if i could isolate where the air was coming from, finally decided it was the pump and tightened the bolts.. viola.

thanks

 

[hydtools]

Seals may not leak under pressure but will under vacuum. Kind of unidirectional that way. High pressure may be enough to force the seal tight, low pressure below atmospheric may not. At rest the seal has enough self-pressure, squeeze, to not leak.

Ted

 

[1gibson]

What kind of pump? Is it a between bearing (where packing would be located in suction region) as opposed to a vertical pump (where packing would be located in discharge region.)

Probably the former based on symptoms?

 

[keyepitts]

It's also not unusual for some air to come out of solution in the impeller inlet eye and not fully re-dissolve into the liquid. This can happen well before the onset of cavitation due to small disturbances in the flow path.

 

[Clvet]

I've seen something similar with the suction piping. Low pressure in the suction piping(from the running pump) was actually drawing small amounts of air into the suction line at one of the connections, it caused a few performance headaches and the visual of small bubbles going through the flow meter.

 

[ports394]

It's just a normal flange connection centrifugal pump from Flowserve. The work I'm doing requires me to run the feed tank very low, and with the low NPSH I vortex at times. Finally figured out a great new vortex breaker.
And with tightening the bolts on the pump casing, there's no more air in the system. Uber happy.

What i've noticed, since I work with tri-clamp, is on some of the discharge piping, if I have a high enough velocity, it works kinda like a venturi or jet pump. It will suck air there instead of leak. So many strange things.

 

[Pumpsonly]

ports394,
It is very proper that you gave your feed back on the problem and how it was solved.There are many differing opinions in each and every thread.Some correct, some not so correct and some totally irrelevant. This allowed fellow readers to learn the correct solutions to the problems encounter.

Keep it up.

 

[ports394]

Pumpsonly, thank you.

I'll post a picture of the vortex breaker tonight. You guys will laugh, but it's crazy how well it works.

 

[alseng]

dude, maybe you are facing cavitation

- Prior Planning Prevents Poor Performance -

 

[ports394]

alseng, lol no. it's not cavitation. The air bubbling was occuring even with a full tank and very low flow. I'm talking 100 RPM on a 3500 RPM pump/motor.

 

[ports394]

FYI.

Eventually found the problem. The return line goes up, alongside the tank, then takes a 180 into the tank, and there is a drop down discharge inside the tank to reduce splashing. The down leg was creating a siphon... I found about 0.5 psi vacuum. Just enough to create some bubbles.. put a siphon break hole in the return line, inside the tank... problem solved.

 

[MikeHalloran]

We love a happy ending.


Mike Halloran
Pembroke Pines, FL, USA

 

[bimr]

What do you mean when you say RO water? Are you referring to ultrapure water?

Are you aware that RO product water that has passed through an RO membrane will typically be super saturated with carbon dioxide gas?

A 0.5 psi water column should not ordinarily release significant volumes of air bubbles. That's only 1.2 feet of water.

 

[ports394]

bimr, I'm not being a smartass. What else could I mean when I say RO water, besides reverse osmosis product water, ie ultrapure water?

0.5 psi of vacuum in this instance was enough to create a fine stream of bubbles in the return line. I have a 2 foot section of clear PVC that I can see it in.

As far as the supersaturated CO2, that should be released pretty early on with the pumping and heating of the RO water.

 

[bimr]

With all due respect, RO water is not descriptive and can mean many different things. RO water is not ultrapure water. RO water passes through multiple stages of treatment before it is termed ultrapure water.

First you said you had no leaks.

Then you said you had an air leak.

Then you said you fixed the problem by tightening pump bolts.

Then you said you fixed the problem by removing the siphon.

Regarding, "As far as the supersaturated CO2, that should be released pretty early on with the pumping and heating of the RO water."

Pumping does not remove CO2. Unless you have a deaerator or chemical treatment, all of the CO2 will not be removed.

 

[ports394]

I had no liquid leaks.

There was a twisted gasket on the pump discharge, covered by a clamp. The next startup, I saw no bubbling, and thought it was solved. - I had major bubbles are times in the system.

Then on a later run, we found a fine mist of bubbles in the sight tube. - Siphon break solved this.

As far at the casing bolts, I was wrong. I just hadn't turned up the pressure high enough to see the bubbling.