BKU Type Chiller - Propane (Shell) level below top of tube bundle

[RicoB]

Hello everybody

I have a BKU (Kettle style) chiller in a gas refrig. plant. Tube site is NGL, shell site is propane as refrigeration medium.
I just learned that in a BKU (Kettle style) chiller the propane (shell) level is controlled to just below the top of the tube bundle.

I would like to know why?

Thanks
Rico

 

[EmmanuelTop]

There are several reasons I can think of:

1) Provide sufficient space for disengagement of liquid droplets in the vapor space above the tube bundle (avoid carryover of liquid with Propane vapor to equipment downstream);
2) Avoid excessive superheating of Propane vapors as they leave the chiller, assuming more than one row of tubes would be above the liquid level (superheating of refrigerant contributes to heat transfer in a minor extent, hence it it just a waste of surface area);
3) Avoid potential tube material damage by exposing naked tubes to the hot stream only, without being cooled by colder Propane;
4) Maintaining the level above tube bundle makes no additional heat transfer effects, since there is no contact with warmer stream - it just increases carryover of liquid phase from the chiller.



Dejan IVANOVIC
Process Engineer, MSChE

 

[RicoB]

Hello Emmanuel

I do understand number 1 as reason to keep the liquid level below the upper tube rows.

Number 2 would rather speak against it. My understanding is, if the upper tube row is not covered by liquid propane, the propane vapor phase will be super heated by the upper tube rows. So how does a not liquid covered upper tube row avoid superheating of the vapor phase? It does actually the opposite, what makes somehow sense to me.
Same applies to number 3. How does the naked tube avoid material damage? Wouldn't it be the opposite?

The only reason to keep some rows uncovered might be to purposely superheat the vapor phase, to avoid caring over liquid propane into the compressor.

 

[EmmanuelTop]

Yes, I was writing about reasons for avoiding both higher and lower level of liquid, relative to the top of the tube bundle. So the bullet points (2) and (3) explain why the level should not be kept excessively low. The point (4) about higher level is self-explanatory.

The process of boiling inside a kettle exchanger is quite turbulent and surface of the boiling liquid pool is anything but flat. This pool of liquid "moves" and keeps wetting the top tubes even if - physically - the top tube row is at some degree higher then the overflow weir that sets the nominal height of the liquid pool surrounding the tubes. You have probably seen water boiling at home (e.g. when cooking eggs) so you get a picture how the boiling process manifests in reality.

Dejan IVANOVIC
Process Engineer, MSChE

 

[RicoB]

Thanks Emmanuel

As my question was, why the level has to be below, So I could not relate (2) and (3) to my question. The chiller I have here does not have a weir, it is level controlled by an LIT. However, out of your explanation, for me the most reasonable explanation would be to slightly superheat the vapor phase just to avoid liquid drop out in the line to the compressor.

 

[gcastan]

The bundle should be submerged entirely to use all of the tubes for heat transfer.

Gabriel Castaneda, P.E.

http://www.gabcheminc.com

 

[RicoB]

Thanks Gcastan

I just got told the upper row of the bundle has to be free of liquid. I could not make any sense out of it and tired to find and answer here. What is the standard for BKU chiller? How are they operated normally? It seems there are different opinion about it.

 

[RicoB]

Thanks Gcastan

I just got told the upper row of the bundle has to be free of liquid. I could not make any sense out of it and tired to find and answer here. What is the standard for BKU chiller? How are they operated normally? It seems there are different opinion about it.

 

[georgeverghese]

Strange you dont have an outlet weir to maitain level in the tube bundle.
One reason for this may be because you have horizontally cut baffles on the tube bundle ?

 

[StraighOuttaND]

Emmanuel Nailed it with the first answer. Insufficient space in the Chiller will result in droplets/mist being carried out of the Tube/Shell into the suction header.

I have also been told that leaving some tubes exposed can increase efficiency and save horsepower on compressors? Can someone please explain?

thanks!

 

[RicoB]

Hello StraighOuttaND

Less horsepower makes somehow sense as you would need to compress less volume. You simply decrease the flow but keep the same retention time. Subsequently the efficiency would increase. I would not say you save HP as you design the chiller at this level anyway. Maybe someone correct me please, if I am wrong?

 

[georgeverghese]

Typically, BKU units are operated with the tube bundle flooded. As suggested, there may be some vapor space demisting limitations. Else I cant see why the top row is left exposed.

But so far you havent said how temp control on the process (tube)side of this chiller is enabled, since propane level is fixed ?

Level control issues ? Where is the top tap into the shellside for this LT - would it be too close to the propane level if the top row were to be flooded?

 

[StraighOuttaND]

@Georgeverghese

Process temp. control would be controlled by the chiller pressure. A back pressure control valve is utilized to keep desired pressure on the chiller and control process outlet temps.



Thank you for the response RicoB.

 

[don1980]

There's a lot of different applications for kettle exchangers. The process control strategy depends on the specific application, and for some applications there are different options for control.

For the application discussed here, the purpose of the kettle is to chill the NGL (control the NGL at a specified temperature), and that's being done by vaporizing propane in a refrigeration loop. There are two basic process control strategies for the propane side of this exchanger. You can maintain a fixed liquid level and vary the pressure (mentioned above by StraightOuttaND), or you can vary the liquid level and maintain a fixed pressure. Either one will work.

The first option (fixed level) appears to be the one being used by the OP, and he asked why this fixed level is set at a point below the top of the tube bundle. I don't know why that's being done, but I seriously doubt that it's really necessary. If the kettle was designed properly you should be able to operate with the bundle submerged. Operating at a fixed level below the top of the bundle is a waste of surface area. You wouldn't want to design a kettle with that plan in mind - it's wasteful. Perhaps the plant experienced some entrainment of liquid propane out of the vapor outlet nozzle, and they decided to drop the liquid level to reduce the entrainment. However, it's almost certain that this propane vapor is passing through a liquid KO drum before re-entering the refrigeration compressor. Another possible explanation is that the exchanger is operating at a heat load which is much higher than design. If so, the propane flowrate is higher, and maybe this was causing excessive entrainment, overloading the KO drum.

If none of these explanations are true (no liquid entrainment problem on the propane side experienced) then there's no need to maintain the level below the top row of tubes.

 

[RicoB]

Hello Don

The chiller is designed with 10% excess duty.
The argument what was brought up to me was to operate at a lower level was and to slightly super heat the propane vapor so that no condensation in the suction line occurs. However, there is a suction scrubber too, which makes this argument kind of pointless. Anyway, I was just curious if this is a typical way to operate a chiller and yes I am 100% with you, that designing a chiller with more duty than you need, is a waste of money.

Thanks

 

[georgeverghese]

With variable kettle pressure control to manage process temp, propane pool temp would vary on the shellside, and so will density. That will mean the LT dp cell will report different levels even if actual level is fixed (unless you have some algorithm to correct the reference density with temp feedback). I dont see much superheating of propane on just one top row of tubes exposed.

 

[georgeverghese]

NGL exit temp control via propane backpressure control at the chiller would be sluggish, due to the large volume of propane on kettle side. You could get tighter control on exit temp with propane level control or hot NGL bypass - this latter option may have some drawbacks depending on the actual application.