Centrifugal Pump Air Release

[jph90]

Hi guys,

Currently we are installing a process plant that features a closed hot water loop. There is effectively an atmospheric holding tank, that loads out directly to the suction of a centrifugal pump discharging to the various process applications. The flow in the supply line not used/returning from a heat exchange application then returns to the holding tank and the process repeats etc etc. The tank is complete with low level switches etc to ensure pump doesn't run dry.

A question popped up today around the installation of an air release valve on the discharge of the centrifugal pump. This is highlighted in a rev cloud on the attached sketch. I am curious as to why this would be required, given there is one on the return header (ultimately the same line) which is also the highest point in the system. A few things i want to run by you are:
- is an additional air release valve required on the immediate pump discharge or is the current air release adequate to protect the pump?
- if we were to add an additional air release, is this the best place for it?

Effectively i dont think it is required and the current set-up is adequate. Just want to check if i am missing something

Thanks in advance

 

[ashtree]

I am with you. Unless there is something that i am missing i can't see any reason to have an air valve installed. I assume that if you don't fit an air valve you will not have the upward loop in the pipework either. There should be no air to entrain or otherwise trap so there should be no need for the air release. The only possible reason was if the return streams caused massive aeration and turbulence in the tank such that the pump picked up air bubbles which then caused problems downstream, but i would think that is unlikely based on your sketch.
My advice would be to not do it.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[electricpete]

From the sketch it looks like you have a "local" high point. I think that is the reason someone thought air release valve is required. Just a random thought thrown out on a subject I don't know much about.

=====================================
(2B)+(2B)' ?

 

[LittleInch]

I realize its a sketch, but you show the ARV as a pressure relief?? and does the return line enter below low liquid level or free fall into the tank??

If it is an ARV, it does seem a bit OTT I must admit.

Is the flow velocity very low? So long as you have >1m/sec in pipework you should be able to flush out air pockets and then only need to put vents or an air release valve on the HX.



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

 

[dbill74]

I can see the air relief being an effort to remove as much air from the system as possible before it reaches the process equipment. If this is a local high point in the system, it could also be used to help facilitate drainage.

I don't see any harm with having a second ARV. It's a cheap preventative measure to minimize damage to the process equipment and extend the life of said equipment.

Personally, I'd rather see an air separator upstream of the pump to protect the pump and process equipment. Especially when dealing with an atmospheric tank as described (lots of potential for entrained air in that tank).

 

[bimr]

The symbol that you are referring to is a safety relief valve symbol, not an air release symbol.

 

[jph90]

My mistake on the PID symbol, definitely mean air release and not pressure relief. Realised after I posted

The return line is top entry to the buffer tank as shown, at approx 4m and thus higher than the high level switch/level in the tank. Velocity in the line is 1m/s. I don't think aeration from the tank is likely from the return line injection. The water in the tank should have time to settle a bit. I assumed the current air release prior to entry back into the tank was primarily for start-up, with minimal air being able to re enter the system during operation.

If protection of pump/heat exchanger should be considered then shouldn't the air relief be on the pump suction in theory

 

[electricpete]

I think it should be at the local high spot (if there is not sufficient flow available to flush air out of that high spot)

=====================================
(2B)+(2B)' ?

 

[jph90]

On second thought, is there a chance of air lock between the pump and heat exchanger? In this instance, i think we should fit the ARV

 

[dbill74]

I don't think aeration from the tank is likely from the return line injection. The water in the tank should have time to settle a bit. I assumed the current air release prior to entry back into the tank was primarily for start-up, with minimal air being able to re enter the system during operation.

I have to respectfully disagree with this. I believe you are underestimating the ability of water to absorb and transport dissolved air. Especially when you have a flow of water falling through an air gap as you have detailed.

 

[bimr]

On these types of systems, the air is removed during the startup procedure. Once the air is removed, there is little likelihood of the air re-entering.

The heat exchanger will probably have a drain on it to purge the air.

Your system is pumping upward to a tank that you show as being 40 meters high. Air binding is more of a problem if you are pumping downward, not upward. Design for a velocity of 1 m/s and you will not have to worry about the air. A velocity of 0.8 m/s will clear the air out of a 2-Inch diameter pipe installed at 90[sup]0[/sup] slope.

Air release valves in small pipes may be of little or no value. You do not need any air release valves.

 

[georgeverghese]

Not much detail in the posting, but I would guess this is a centrifugal pump which has a high point within the casing - hence the auto air release on the pump discharge. But this would work only if the take off for the air release is above the elevation of the pump casing high point and that there are no low points from pump discharge to the auto air release take off point.

This trapped gas venting can be done manually also everytime this pump is started up. An auto release may be justified only if the standby recirc pump is auto started.

Troubleshooting the source of this air accumulation in the standby pump is another topic altogether, from what little we know so far.

 

[lilliput1]

All high points of piping should be provided with means to remove air. This can be manual vent or automatic air vent plus isolation valve. The air in piping would interfere with heat transfer. Also all low points of piping should have drain valve to allow removal of sediments and to relieve pressure in piping when equipment is valved off for removal or servicing. All strainers should have blowdown valve to clean out strainer screen.

 

[bimr]

One would think that many of the posters have overlooked the pipe size of DN50 on your sketch. Air release valves in small pipes may be of little or no value

 

[georgeverghese]

If the pump casing is indeed not self venting (and hence the air release valve), the pump would not develop pressure on startup, or its capacity would be much lower than predicted.

 

[lilliput1]

Does the water tank have vent to atmosphere? Vent size should match size of inlet piping. Vent would prevent pump from creating vacuum in tank. The air release valve in the piping would be too small.