6 replies to this topic

[krishna106738]

Dear Team,

I have query related to BEU type  Shell and tube condenser.I am working on a project in which top vapor of distillation column is around 105 C , we are using cooling water to condense it.

 We were initially using BEM type for this ,but static person mentioned as per his analysis need to give  expansion joints on shell side .Hence we thought to  study and design on BEU type so that expansion joins can be avoided.When I started simulating it in HTRI,I was getting  "B" stream of shell is very low , "F" fraction stream of shell is greater by 30 % than max . hence heat exchanger cannot perform well.Following are conditions and criteria

 

Hot Side (shell)

Alcohol vapor:5700 kg/hr(93% wt conc)

Inler temp :105 C

Pressure :~2.9 bar a

Cold Side (tube)

Cooling water in :28 C

Cooling Water out :36 C

Pressure:3.5  bar a

 

Tube OD:25.4 mm

Baffle : Single Segmental, orientation : Parallel(vertical)

Tube layout :90

Pass partition horizontal

 

I studied  in U tube the "F" stream is high generally due to central lane formed due to U bend.Initially I have attempted to change baffle cut and space but it seems they are not much relevant to this situation.

 

I attempted following

1.perpendicular baffle orientation to reduce "F" stream fractions,but i came across  that there will be issue of condensate drain.

2.Used sideways vapor nozzle entry  in U bend and found "F" stream fraction decreased and B fraction of stream became high but  few seniors said it is not recommended generally.

 Then I attempted  vertical pass partition and ran simulation with same, HTRI didnt showed  any issue.But want to check whether it is good to go with this vertical pass partition  in BEU shell side condenser or not.  I am unable to get source/literature for same.

 

Please help me out in this issue,also let me know if i need to furnish any details.

 

Thank you team

 

 

[Pilesar]

Why use a 90 degree tube layout? Is the shell side a fouling service? I think shell side condensing alcohol would not need to be mechanically cleaned. I suggest trying a 60 degree layout to see if it gives better results.

 

I would not be concerned about F stream fraction in a total condenser. Check how much this matters by changing your input Geometry -> Tube Layout -> Bundle Clearances -> Block Bypass Stream -> F stream. 

How does this affect the shell side heat transfer coefficient and the overall heat transfer? If the F stream bypass is a real problem, you can use dummy tubes or seal plates in the pass partition area. Usually having low pressure drop in condensers is more important than B stream flow fraction. The pass partition gap can help distribute the vapor and reduce pressure drop. So in general, keep generous baffle cuts and spacing. 

 

Your large temperature difference is large. Are you trying to subcool the alcohol extensively in this exchanger?

[krishna106738]

Thanks Pilesar,

We are not going to subcool this , we want to condense only. 

 

"  I would not be concerned about F stream fraction in a total condenser. Check how much this matters by changing your input Geometry -> Tube Layout -> Bundle Clearances -> Block Bypass Stream -> F stream. "

I did this option it improved B stream fraction and F stream fraction became zero . Geometry wise no change observed. HTRI Simulation converged but in practical also there will be no issue due to F stream fraction.

[IGC]

Going back to the selection at the start, considering the flowrate, pressures and temperatures, would a plate and frame unit not be more suitable?

[Pilesar]

  If this were my design, I would start with AEU exchanger with 19 mm tubes and vertical baffle cut. The vapor entry nozzle after the U-Bend would be from the top and I would have an adequate liquid drain. With such a large temperature difference, I would evaluate using more than two shell passes.

  I would not consider a plate and frame unit for this service because: 1) I am not familiar with designing them. 2) Our maintenance and operations are familiar with shell and tube. 3) Even if my shell and tube design were not the best optimum solution, it would not be 'wrong'. Implementing newer technology is a risk to to my company and a risk to my career. Sometimes the risk is worthwhile, but in this case there seems to be limited potential reward to me or my company. Others are in different circumstances with different experience. This is just my thought process.

[breizh]

Hi ,

About Plates and Frame HX I've attached 2 documents to support .

Good luck .

Breizh

Edited by breizh, 12 September 2019 - 07:25 PM.

[IGC]

  If this were my design, I would start with AEU exchanger with 19 mm tubes and vertical baffle cut. The vapor entry nozzle after the U-Bend would be from the top and I would have an adequate liquid drain. With such a large temperature difference, I would evaluate using more than two shell passes.

  I would not consider a plate and frame unit for this service because: 1) I am not familiar with designing them. 2) Our maintenance and operations are familiar with shell and tube. 3) Even if my shell and tube design were not the best optimum solution, it would not be 'wrong'. Implementing newer technology is a risk to to my company and a risk to my career. Sometimes the risk is worthwhile, but in this case there seems to be limited potential reward to me or my company. Others are in different circumstances with different experience. This is just my thought process.

 

That's fair enough.  I work for a manufacturer so considering maintenance aspects isn't really up my street other than providing instructions to do so!

 

I would probably design the unit so that the vapour entrance is closest to the tubesheet rather than the U-bend so you can take advantage of the large temp difference.

 

@krishna106738, I believe you are kind of solved now?

 

Are you using plain tubing only or have you looked at using low-fin tubes as well?

 

You've mentioned that no subcooling so the other thing you can check are condensate outlet nozzle size (Fr < 0.3, for this criteria), to ensure there would be little liquid hold up in the shell.

 

The other thing to check is the height under the inlet nozzle as you could have bypassing there due to the positioning of the baffles / allowances for impingement protection.  If you check the HTRI TechTip-1 for this.