Induced air? Cavitation? Pumping problem 2

[misterd]

We are starting a new wastewater unit using ceramic membranes. The problem with those membranes is that they are "allergic" to air. If the water stream contains air, the membranes will collapse.

After running for a month, one membrane exploded...After analyzing the possible causes, we are now down to the pumps. Can air go through a centrifugal pump (65 gpm @ 41 psi)with mecanical seals?

The unit is equipped with air vents, intakes are immerged, piping is straight and welded, feed tank is open to atm ...

Any ideas?

 

[PUMPDESIGNER]

Large area of discussion.
Centrifugal pumps can sure handle quite a lot of air, passes it right through. Some people claim to know amounts of air a centrifugal can pass, but any value given would not be fully trustworthy.
McNally claims .5% entrained air by volume is definitely not a problem while 6% can be a big problem.

Air can come out of solution if NPSH drops low (even if NPSHa is greater than NPSHr), because the lower the pressure the lower the saturation point. In other words, pressure is reduced in the pump intake, water that was not saturated before entering the intake can then become saturated or even super-saturated as pressure drops both leading to the pump and inside the pump as fluid velocity increases and pressure drops.

I have to say this however, it is hard to believe that a little bit of air coming out of solution could be considered harmful to your process. Is the total dissolved gas content of your pumpage really that high? You could do an easy test, just put some of the pumpage in a glass jar and reduce the pressure slow to see when the gas starts to come out of solution. Remember thought that when pressure drops low enough the pumpage itself will begin to vaporize once you drop pressure below vapor pressure.





PUMPDESIGNER

 

[saxon]

misterd, Vortexing in the suction well? Suction line flanges or piping conn. loose? Improper start up after shutdown? These are just a few of the things to check.

saxon

 

[misterd]

We took special care in order to avoid vortexing and triple checked our connections...hummm.

The pump inlet is 2". Half of the inlet piping is 2" and the other half is 1½" (don't ask why...). Is it possible that the increase in velocity in the 1½" piping creates a depression lower than vapor pressure? The fluid is wastewater coming from an industrial laundry.

 

[imok2]

As I am not a pump engineer I'll put in my two cents for what it's worth
get the npsh from the manfg then measure the inlet pressure also I would look at possibly installing a clear plastic spool maybe 10 dia at the discharge of the pump to see how much bubbles you have and finally you may have to repipe the 1-1/2" line to a full 2"

 

[Latexman]

misterd,

An industrial laundry has lots of agitators that whip air into the discharge water. The presence of surfactants may or may not stabilize the dissolved air; I don't know for sure. At the 2" to 1.5" transition, do you have an eccentric reducer with the flat side on top of the line? This will minimize the accumulation of an air pocket that can burp over to the pump periodically and feed the dissolved air more evenly and continuously. Maybe this transient behavior is what caused the problem?


Good luck,
Latexman

 

[uncleal52]

Maybe put some sort of reservior or high point vent in the process feed where you can trap air and vent it?

Are there any vacuum breakers in the lines?

Seems like an odd problem being allergic to air???? you sure your not overpressurizing the system, pressure spikes, water hammer, etc?

 

[misterd]

We installed an air vent on top of the recirculation loop (highest point) and on top of our filter bags. We suspect, that small bubbles are not captured by the vents. Eventually, those small bubbles become big bubbles!

From what I can understand, bubbles of air passing trough the membranes creates water hammering. The membrane then collapses.

Vacuum breakers...here is something we didn't think about...

 

[Montemayor]

misterd:
If included or entrained air in your wastewate is really what is causing the problem, then uncleal52 has brought up some valid and proven solutions. High point vents may not be resolving the problem.

If the air threat is real, then a serious, engineered solution should be done: install a feed tank immediately before your ceramic membranes. This tank should have the ability to segregate and automatically vent out any existing collection of air at that point, before entering the membranes. The selective purge (or venting) of the included air is done with a conventional air purge valve mechanism (not a "vacuum breaker") that is nothing more than a ball-float valve working in reverse: as air accumulates in the feed tank, the top water level lowers and the ball float opens to purge the trapped air. I have used this mechanism on many installations with success in order to positively purge high spot air/inert accumulations in water systems. A vacuum breaker does just the opposite: it allows atmospheric air to enter the process - an effect that you say you don't want.

I hope this helps.


Art Montemayor
Spring, TX

 

[unclesyd]

Depending on the type of mechanical seal you can get air introduced into the system through the seal.

A quick check is there any leak at the seal, I'm talking about drops of liquid?

You are pumping a dirty stream that could cause the seal to leak by several means. Get with the tech rep that supplied the seal and see if the have the best seal for your application.

 

[misterd]

All seals have been checked earlier and we don't have any leaks or pressure lost.

While investigating on the pump, we realized that we had microbubbles when we were choking a valve. At first, we tought about the pump because that valve has a direct effect on it. When the valve is completely open, there is no problem. If it was pump cavitation, it would only get worst.

But now : How do you know you're cavitating in a valve? We are using 1½" SS Crane #9502-2 ball valve. Flowrate can vary from 65-190 gpm @ dp = 20-40 psi...