20 replies to this topic

[CSNK]

Dear,
I am rating one of the heat exchanger with following configuration
Shell side - cooling water
Tube side - HC gas is cooled.

Shell id and length is all fixed, it is single segmental baffle type exchanger. Existing shell velocity is very less i.e 0.065 m/s.
I dont want to change shell id, but i can change baffle cut / arrangement.

My query is that i want to increase shell side velocity what are the options .
I am using HTRI for rating the exchnager.

Any idea on this shall be useful.

[Bobby Strain]

It is unusual to have cooling water on the exchanger shell side.

[breizh]

Hi,

Agree with Bobby ! 

Cooling water in a shell , with low velocity means Fouling .

I've attached a document for your consideration .

 

engineering tips underneath :

http://www.gulleyass...eeringtips.html

 

Good luck

Breizh 

Edited by breizh, 16 September 2020 - 10:59 PM.

[CSNK]

Thanks bobby/breizh.
Exchanger is in service since long and also hc gas is having impurities, so must have been allocated in tubes.

Thanks for path, yes now we are planning to change CW to tube and HC to shell ...
But my major concern is what are the options to increase CW velocity...

Thanks

[Chemitofreak]

Thanks bobby/breizh.
Exchanger is in service since long and also hc gas is having impurities, so must have been allocated in tubes.

Thanks for path, yes now we are planning to change CW to tube and HC to shell ...
But my major concern is what are the options to increase CW velocity...

Thanks

 

Once you change the CW to tube you can increase the velocity by changing the diameter of the tubes (most standard dia. are 3/4" or 1"), but you have to keep an eye on the pressure drop across the exchanger, high velocity will lead to high pressure drop, this might lead less flow of cooling water to the exchanger and in turn fouling. 

[IGC]

What is the gas pressure?
I have seen cooling water on shellside for high pressure cases for instance, or when gas on the shellside will cause vibration / noise issues.

[Bobby Strain]

To get your target water velocity in tubes, you can increase the number of passes.

 

Bobby

[breizh]

Hi,

What is the problem with the current set up?

Thanks 

Breizh

Edited by breizh, 18 September 2020 - 10:56 PM.

[CSNK]

@chemitofreak:will chk that once
@IGC: operating pressure of gas is 0.09 kg/cm2.
@bobby:will chk once.
@breizh: CW velocity is very low as it is on shell side and needs to be increased. Either in shell side or by changing configuration.

[breizh]

Hi ,

My question is about current set up and heat transfer , what is the issue ? Why cannot you increase the  cooling water flow rate ? 

You need to tell us more to get support . 

my view. 

Breizh 

[CSNK]

Actual issue is failure of tubes and reduction of tube thickness near baffles.
So we r exploring option of changing fluid or may finally go for tube mettalurgy change...
Existing tubes are CS.

Edited by CSNK, 18 September 2020 - 12:17 AM.

[latexman]

What is metallurgy of baffles?

[IGC]

Reduction of tube thickness near baffles is usually due to vibration induced fretting...  at 0.065 m/s though, there's something strange.  

It's just guesswork at the moment though. 

 

Provide:

 

Process details (flow, pressure, inlet / outlet temperatures)

Current unit details (TEMA Type, Shell ID, length, number of tubes, OD/WT, pitch / arrangement, baffle count, pass count, material)

Easiest way would be to create the as-built unit in HTRI and upload the TEMA from that.

[CSNK]

@ latexman : existing metallurgy is Carbon steel for shell / Tube / baffles .

@ IGC: i have developed As-built HTRI model also will current configuratoin - but  no vibration warnings are there(TEMA-BEM).

 

Now main concern is how can CW velociiy be increased. Because, we think due to lower velocity there could be concern.

I tried changing CW to tube instead of shell. but no major benefits.

 

Thanks.

[Bobby Strain]

So, why not replace the exchanger with a new one properly designed?

 

Bobby

[Jiten_process]

@CSNK

 

I think you are facing two different issues 1) failure of tubes. However, it is not clear what sort of failure you are experiencing 2) reduction of tube thickness particularly near baffles. But could you explain, what makes you think that increasing the cooling water velocity would solve the issue? Reduction in thickness indicates a localised corrosion issue which is less likely to occur on outside tube surface as the service is cooling water.  Do you experience reduction in thickness at every baffle and for all tubes ? I think, you should engage with metallurgy engineer on this. 

 

Nevertheless, what general way of understanding the heat exchanger issue is. Input your actual operating parameters and exact (as built) geometry in the HTRI and run it on rating mode. Analyse thermal design output and vibration analysis generated by HTRI. This will flag the concerned area. But make sure, you input every detail of your geometry in HTRI e.g. number of sealing strips, baffle spacing, baffle cut, etc. which i think you can get from exchanger fabrication and GA drawing only. if there is a really vibrational issue, you should see it in HTRI vibration analysis report. 

 

Now, addressing your question of cooling water velocity, since the exchanger is already built and you dont want to change any geomatry of built exchanger then the only way is increasing the flowrate of cooling water as already indicated by Breizh. Increasing baffle cut or increasing shell side passes by increasing number of baffles will increase the velocity too if you can change the tube bundle construction. If you could have cooling water on tube side, then you could block some tubes to increase the velocity. This is done many a times. But on shell side, unfortunately, you have limited option without changing the geomatry. Additionally, from HTRI run outputs you can check, if your leakage streams (B-stream, E-stream, etc) is higher, consider adding sealing strips (only if it warrants any benefit on HTRI run).

 

Good luck. 

Edited by Jiten_process, 19 September 2020 - 04:44 AM.

[CSNK]

thanks all for reply,

And sorry for delay in reply.

 

@bobby - we are finally designing new exchanger with CW on tube side and to meet process conditions.

 

For quick calculation, what is the best way to design a new exchanger and meeting all the norms of TEMA (i.e velocity of CW should be around 1m/s or high, Pressure drop under limit and best design).

Any suggestion / options are welcomed.

[CSNK]

Hi all,
I have designed new exchanger with rod type baffle and 10 tube pass..
Could rod type baffle and 10 tube pass be a problem?

[IGC]

You have been asked to provide the full process details in order for people to guide you.  Any advice that can be given will just be general and will not give the optimal solution.  

 

Personally I'd look at smaller tube ID if I really needed to go up to 10 tube passes or even a different technology like a plate and shell...  Again, more info needed.

[CSNK]

@IGC: On process parameters,

there is HC gas (in shell side - no phase change flow 10 TPH) getting cooled from 82.1 to 40 Deg C.

CW on tube side with inlet temperature of 38 Deg. C (flow and outlet of CW can be varied to meet process constraints).

Main constraints include CW velocity to be more than 1 m/s and shell side Delta P not more than 0.014 kg/cm2.

I have considered tube ID as 19.05 mm with 2.11 mm thickness. Lower than this is not recommended and hence, presently minimum 19.05 mm is selected.

 

I can achieve all constraints by selecting rod type baffles and increasing number of tube passes.

Hence, wanted opinion / concern  on usage of rod baffles and higher of tube passes. Or any other option if possible.

 

Regards,

[sak@675]

To get your target water velocity in tubes, you can increase the number of passes.

 

Bobby

The fluid velocity in water systems should not exceed certain limits to avoid noise and damaging wear and tear of pipes and fitting. The table below can be used as a guide to maxumum veolicities.