PDA

View Full Version : Ammonia/water pump suction line velocity



virk@gmx.net
09-04-2008, 09:18 AM
Hello to all!

My name is virk and this is my first post:-)

We have to design a system in which we need a pump for 200m3/h of a mixture of ammonia and water. This liquid is in equilibrium (boiling).
We will make use of KSB pump ETABLOC with NPSH of about 2,5m.
Now we are talking about suction line design. If we did this as we do it usually a diameter of DN300 (12") would be the result. The pump suction side is DN150 (6").
Now we want to rethink our design criteria for this suction line diameter.
Of course vertical distance between vessel and suction inlet of pump can be made large enough.

What diameter/velocity would you suggest AND WHY?

Greetings virk

US Iceman
09-04-2008, 03:23 PM
Hi virk,

Welcome to the RE forums.

Since you have provided the NPSH (required) of 2,5 meters for the pump I am assuming you are taking this into consideration in the design of the system. Any of the friction losses that could occur in the pump suction line will affect the NPSH (available) to the pump.

As you are already probably aware, the NPSH (available) must always be greater then the NPSH (required) by the pump curve for the operating conditions encountered.

If the pumping system is a closed system and not hydraulically decoupled from the pump suction this is a better situation. If the system consists of a closed system with a tank on the suction side of the pump (hydraulically decoupled) the NPSH calculation become more involved.

In either event, the NPSH calculations should be performed for available NPSH due to the piping scheme you select.

virk@gmx.net
10-04-2008, 05:40 PM
Hi USiceman,

Thank you for your answer. Of course we consider NPSH but we want to go deeper into detail. For example, concerning NPSH calculations we could allow a velocity of f.e. 3m/s in the suction line of this boiling liquid; but nobody is doing that! What we come across is about approx. 1m/s. Are there perhaps some other rules besides stationary pressure losses which forbid the design for higher velocities in boiling liquid suction lines. What velocity should/could we design it for and why? This is what we want to clarify.

Regards virk

US Iceman
10-04-2008, 08:56 PM
Hi virk,

The problem is due to the saturated condition that is encountered on the pump suction. Any pressure loss causes a resulting increase in possible vapor formation (bubbles).

The 1 m/s rule of thumb is in my opinion one way to allow for buoyancy of vapor bubbles in a liquid stream. If the liquid moves slow enough it is hoped the bubbles will disengage from the flow stream and vent back to any vapor space above the pump suction.

So, if you wanted to go into a lot of detail... the drag force of the liquid on the bubble (and gravity) would have to be less than the buoyancy of the vapor bubble.

There are other considerations also such as; minimum loss in valves and fittings or other fluid behavior such as fluid turning when flowing through elbows or the need to provide sufficient distance between a turn in flow direction and the pump suction to allow the flow streams to be less turbulent. The rule of thumb for this is usually 10-12 pipe diameters.

You might also want to use a pipe from the pump suction up to the vapor space above the liquid. This would help to vent any bubbles in the pump suction line.

Of course, if you can provide sufficient static head on the suction side to overcome the friction losses of the pump suction pipe (and fittings, valves, or other strange behavior) the liquid pressure should always be greater than the vapor pressure. Then the liquid would not boil at all!

I hope this helps.