12 replies to this topic

[emmanuelv2012]

Good day all, please does the hot fluid always flow through the tube side of the heat exchanger always? and also the cold fluid from the shell side

[fallah]

Hi,

 

In general, yes...because looking at thermal standpoint, it's more effective than the contrary...

 

But there might be exceptions to force doing vice versa...

[emmanuelv2012]

thank you my regards

[CheAmine]

Emmanual,

Sometimes it is prefered to put the high pressure stream in the tubes. This will minimize the amount of material required to withstand this high pressure if compared to putting the high pressure stream in the shell side. Consequently, this will lead to cost reduction.

CheAmine

[emmanuelv2012]

I have been trying to learn how to design a shell and tube heat exchanger(theoretically) using the kerns method but I havent been able to make much progress....Pls I seriously need help.

[IGC]

No, not always.  It depends on the process, what fluids you are considering, if the fluids are corrosive, operating & design pressures etc. etc.

 

For example, an LNG shell and tube heat exchanger would have LNG on the tubeside (cold), with typically hot water/glycol on the shellside.

 

As to what you are looking to design:

 

What fluids are you considering?
Operating pressures?

Flowrates?

Inlet temperatures / required outlet temperatures?

 

Where are you getting the Kerns method from?  E.g. in Coulson & Richardsons there is a step by step example of it?

[breizh]

Hi ,

A few documents for you and others interested with the subject .

Enjoy.

 

Breizh

[emmanuelv2012]

A big thanks to Mr Breizh for those files they've truly proven to be very helpful....
Yes IGC am actually using C AND R vol6, my confusion all started in the very first example (12.1) I understood it up to a certain point bt my confusion started where they said U=600w/m2k (they quoted fig 12.1 to get this value bt i dnt get hw it was choosen).
Also my next confusion was the basis on which they choosed the outer and inner diameter to be 20 and 16mm respectively from there downwards was were I stopped understanding what they were doing...
I know my questions may seem elementary am just a Young chemical engineering student with a zeal to learn...ur prompt responses would be really appreciated...Tanx

Edited by emmanuelv2012, 02 June 2016 - 09:15 AM.

[IGC]

The Kern method is iterative, that is that you make initial assumptions and then based on your calculations, you go back and repeat until you have a more reasonable answer.  The U = 600 w/m2k comes from figure 12.1.  Your process fluid is the condensing organic vapour, while the service fluid is the river/well/sea water.  Drawing a straight line between them gives you an estimate for the overall heat transfer coefficient - in this case, around 600 W/m2k.  This is just a guess to get you started.  

 

Tubing is based on OD, with a certain wall thickness.  In this case 20 mm OD with 2 mm wall thickness, which gives you 16 mm ID.  The process fluid has to flow through something!  Again, it is just a guess for the example.  They could have easily chosen 1" or 0.5" tubing.  It all becomes a bit of a balancing act trying to get the most efficient cost / performance balance.  If you were to make up the required surface area with 1" tubing, you'll most likely end up with a larger shell diameter -> more material -> more expensive.

[Ankit Malhotra]

The choice of shell and tube-side fluids for an heat exchanger are often governed by more demanding requirements which have implications on the safety, cost, maintenance time and feasibility of the heat exchanger and cannot be generalized as cold fluid on shell side and hot on tube-side. 

 

Some examples on how this choice is made might be - 

1-Frequent mechanical Cleaning- Tubeside to more fouling fluid, if mechanical cleaning is required. Note that chemical cleaning can always be done even on shell side. Mechanical cleaning on shells side however would require access to tube outer surface which in turn would require design features such bundle pullout possible / tube-tube gap/tube layout. It is more difficult than tubeside cleaning as it will require dismantling the whole exchanger. eg most Cooling water exchangers in process industry, which are frequently cleaned using hydrojetting.

 

2-Pressure - Tubeside to high pressure fluids (Shell side high pressure increases the thickness of shell and can have great implications on cost of the exchanger). Remember, shell cost forms the major cost of the exchanger in most cases.

 

3-Viscosity - Shell side to viscous fluids as turbulence, which can be more easily induced on shell side might be difficult (note viscous and fouling fluids are not essentially the same thing and are often confused as high viscosity implying high fouling). And then there are tube inserts which can give you that effect on tube-side as well.

 

4-Corrosive - Tubeside to corrosive fluid - Less components for it to see and eat up

 

5-Metallurgy - Expensive metallurgy on tubeside- Process requirements such as compatibility of your fluid or high or low temperatures might require you to use a particular metallurgy. It is always cheaper to use the more expensive metallurgy on the tubeside as it will require less of that metal.

 

This is not an exhaustive list and you can find more things in any chemical engineering text on heat exchangers for more info.

 

It gets tougher than this in real life where you might have competing requirements and this is often a choice between lesser of the two evils. To give you an example would be an  HF - Cooling water exchanger, where you will have to choose between a corrosive or fouling fluid to be kept on the shell side. 

 

You therefore need much more information for this simple decision if you are designing an actual exchanger.

[emmanuelv2012]

Thanks Mr Malhotra....
bt pls which is process fluid and which one is service fluid??
And please how did they arrive at that 4.88m used as d length?

Edited by emmanuelv2012, 03 June 2016 - 09:05 AM.

[Ankit Malhotra]

Hi Emmanuel

 

I could not understand your question. What do you mean by process fluid and service fluid ? Also are you referring to some specific case with 4.88m as the length ? 

[emmanuelv2012]

Yes Mr Malhotra the very first example in heat exchanger design in coulson and richardson vol 6. example 12.1, page 675