Gravity flow and self-venting design - how important is it ?

[Mike4chemic]

Hello all
I have a question regarding design of pipes for gravity flow and self-venting flow regime.
There is a steam condenser (not a direct contact. A shell and tube ) ,which its condensate flowing from the bottom of the condenser head and to a condensate tank downwards.
At that condensate tank, a vertical pump will be installed, and also a level transmitter
I wish to avoid adding another level transmitter at the condenser head bottom, and to be sure that no condensate is accumulated there (so it won’t flood the tubes and reduce heat exchanging efficiency).
My question is: how important is it to design the condensate pipes from the condenser to the tank to be self-vented ( meaning :Fr<0.3 ) ? (To me it will dictate a difference between a 4/6” line to a 10” line )
A few figures ( See also an attached schem ) :
The elevation differences between bottom of condenser head to top of tank : 6.5 ft
The pipe length including equivalent length for valves and fittings: 50 ft
The expected condensate flow: 300-350 GPM
The tank will have a vapor equalization line between it and the condenser head.
The pipe will be slightly submerged in the tank liquid ( to avoid steam escaping )

I have read threads : 378-136655, and others, but I still debate with this issue .

I will appreciate your comments

 

[bimr]

At the condenser, you should have some type of vortex action as the condensate drains out of the condenser into the pipe. You should expect some type of backup as the water enters the pipe.

One would think that you would also have some surging and bubbling at the transition point where the drain line changes from partially full to full.

Would think that if the intent is no condensate backup, then you would be better served with the larger pipe.

 

[katmar]

I have seen too many condenser return lines surging and pulsing to ever consider not using a self venting design. The way you have sketched it, if you use a 6" line, you will continuously draw steam down from the condenser into the tank. It will have to flow back up the equalization line, which isn't really a problem since that is what the line is there for, but how will you size the equalization line? It is difficult to predict how much steam will be entrained with the falling condensate.

You should also avoid any horizontal piping that is not flooded with liquid. The pressure drop through a 6" line for the flow you have given would be about 4 inch of water. This means that the horizontal section as you have shown it will run part-full.

It would be better to prevent steam flowing back up the condensate pipe by using a goose-neck (U-bend) rather than just having the line discharging below the liquid surface. The condensate line from the condenser should come down vertically to the level of the base of the tank, and this vertical section should be sized for Fr < 0.3. From that point onwards you can use smaller pipe if the cost is important. Run the smaller line horizontally to the tank and then rise up to an entry nozzle on the side of the tank. If it would be a tripping hazard you can run the horizontal section under the floor.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

Mike,

I would listen to katmar if I was you.

For me your only alternative is to locate a control valve at your tank controlling on say 6 feet of back pressure to always keep your line full on the basis that you have some sort of slope on your horizontal sections so that it will actually fill. That way your line will stay full, but still be just below the condenser drain point, which you may need to make bigger for the initial 1' of fall, like a boot.

depends how much the different sized pipe means to you.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[ione]

Mike4chemic,

Have you taken into account the amount of steam is going to be flashing in the condensate line? There's a quite high likelihood to have two-phase flow in the line. The amount of flash steam is definitely going to affect the fluid velocity in the line. We don't know upstream and downstream pressures involved as you have not mentioned them, but a 6" pipe for such a flow rate seems a bit undersized to me.

 

[Mike4chemic]

Thanks you all for your valuable and quick answers.

I believe I will design it for self-venting as Katmar proposed ( 10" vertical section )

A few more details to add and a question:

The horizontal section here is very small ( the condensate tank is right beneath the condenser and will be buried ) .

Regarding using the U-bend instead of the seal inside the tank - Can you elaborate why is it preffered,and how critical is it ?

I am asking becasue the level transmitter inside the tank will allow me the keep the maximum and minimum operational heights.
The minimum height will be of-course -according to the NPSHr, and the maximal - was intended by me so that the condensate entrnace pipe will be submerged with the required level to avoid that steam entrainment.
Isn't it basically the same thing as long as I keep the required level which is equivalent to the pressure difference I need to avoid steam entrainment ?(the same pressure loss the U-bend will create? )

I am asking because in burried pipes it will be harder for the piping designer to add that U-bend.

Now, just to see I understood it correctly,let me summarize :The vertical section must be designed for self-venting (Fr<0.3 ).
the horizontal section (if exist ) must( or can ? ) be kept totaly full from 2 reasons :
1) for long horizontal section it will save cost if we minimize line sizes. So,if it is not necessary to create a self-venting section - we should design it for full flow.
( becasue to avoid gas entrainment,it needs to be either totaly full or self-venting. anyway in between will create sluggs and vibrations etc.)
2) To create enough pressure drop to avoid steam from entering the tank (In this case, this pressure drop can be replaced by other means such as :valve,seal height in the tank etc. )

Ione,I agree with your comment regarding flashing in general, but here the temperature & pressure are: T =176 F, P=17 psia ,so it is rather subcooled

 

[katmar]

There are 2 reasons for wanting to avoid coming in from the top and sealing with a dip pipe. The first is that you would be relying on the level controller of the tank to ensure the seal. This is not a major issue.

The main reason is to avoid 2-phase flow in the horizontal section. If the 10" line were running full the pressure drop for 350 gpm over 50 equivalent feet would be less than 1" of water column. The equilibrium level in the 10" line is therefore going to be very close to the level in the tank, and all the piping above it is going to be part-full.

A few bad past experiences has made me overly sensitive to horizontal 2-phase piping. A 10" line is probably big enough to be safe in the horizontal section because it appears to be quite short. If you can install it with a slope of 1:30 then it is going to run (theoretically) around 1/4 full. This should be more than safe, and if there are good reasons (which you do have) to not install the U-bend then sloping the horizontal section and making it the same size as the vertical section would be OK. It is when horizontal sections are assumed to be running full of liquid and the diameter is based on a rule-of-thumb of 4-6 ft/s that things go wrong.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"