Air in a cooling tower pump 2

[leo31]

Hello there, I went to a learning visit to a plant where there's a problem with air getting trapped into a pump moving cooling tower water towards an heat exchanger. They can't quite figurer out why air could get there since the pump is positioned 2,5 m (8 ft.) under the cooling tower basin, so the NPSH available should always be greater than the one required. The cooling tower is used in a batch production, so it's often shut down to be turned on again later.

Could it be because of a pan-drain down during the start up phase ? I noticed how the basin volume looks very small compared to the pipeline circuit and heat exchanger that would have to be filled before sending back the water to the top of the cooling tower. I also noticed there's no check valve installed so I suppose the pipeline circuit doesn't stay flooded after each shut down of the cooling tower. (could this be determined by looking if Volume_pipeline + V_heat exchanger > V_tower basin ?)

Or should they not overlook the possibility of having air getting stuck in the vertical drop after the cooling tower basin ?

Could a solution be to place an additional tank to provide the additional volume needed to fill the whole circuit ? Or should they place valves to prevent piping to tower drainage (pic below)

Sorry if something I said might sound silly, I'm just trying to learn and my professor didn't have an answer for me. Many thanks in advance ! (I attached a sketch of the circuit)

 

[katmar]

Air is often a problem with cooling tower pumps. If you do a web search using the term "cooling tower pump suction piping design" you will find recommendations on the required submergence of the suction piping, how to avoid vortex formation and a host of other potential problems.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

I'm sure your sketch means something to you, but it doesn't to me....

That schematic is Ok but really a bit too simplified/

My guess is that your NRV D/S the pump is passing over time so that the condensor and the pipework in the "water leg" drain back and then there is a time gap between the pump turning on and enough water getting back through the cooling tower to provide water flow into the pump.

Only be observing the level in the tower pan can you determine if this is the cause of air getting into the system.

A start would be to replace or add into the pump discharge line an isolation valve so that it closes on pump stop to prevent drain back of the condensor.

Then before pump start, ensure that the level in the tower pan is high.

You still might need a bigger tower pan to provide enough water to prevent air inclusion and to allow sufficient time for any entrained air coming off the tower to bubble back out. These unts are notorious for not allowing enough time for the air entrained during the cooling process to dissipate and hence they get stuck in the pump and pipes.

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

 

[leo31]

Thanks for your feedback, I asked to give a look at what happens in the cooling tower basin and apparently the water level does get pretty low at the start up, but more importantly swirls are formed ! I calculated the minimum submergence required with the Froude number and it is in fact higher than what they have now. Could a simple vortex breaker help or should I look into an intermediate tank with a higher level ? Thanks again

 

[georgeverghese]

The cold water collection cell just upstream of the pumps may be too small. Not only should there be adequate submergence of the exit nozzle from this cell, but cold water needs to be well degassed. For good degassing, typically upward terminal flow velocity for a 300micron air bubble should be more than the bulk velocity of water flowing downward to the exit nozzle. See also Hydraulic Institute Standards for other requirements for cold water cells in open loop cooling towers.

No check valve installed on pump discharge ? So almost the entire circuit must be draining out into the cold cell in the reverse direction on pump stop.

 

[LittleInch]

Vortex breakers would help or at least not do any harm, but very low level means that the water in the basin doesn't have time or space to de gas.

Add a non return valve if there really isn't one, think about adding an isolation valve to close when it stops and add a vortex brakes are easy things. A bigger pond is more difficult.

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

 

[pierreick]

Hi,
Vortex breaker is a must to avoid swirls. Pay attention to the solubility of air in water (probably 2nd order)
Note: The level in the basin must be monitored, float, radar or delta P to compensate evaporation and drifts and adjust make up water. In any case the level must be stable and high enough to prevent cavitation.
Good luck
Pierre

 

[pierreick]

Hi,
I've added a better document, for you to assess your refrigeration system.
Pierre

 

[Artisi]

You say the air is getting trapped in the pump, what does this mean?
Is the pump going off-prime and not pumping or is the flow surging?
What size is the pump / flow rate, head required.

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.)

 

[shvet]

@leo31
See appendix A ANSI/HI 9.8

 

[pierreick]

Hi,
I'm adding a great resource about Cooling towers and fundamentals.
Enjoy the reading
Pierre