Jump to content

Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog


Design Of A Shell And Tube Evaporator/condenser

4 replies to this topic
Share this topic:
| More

#1 andrea.r


    Brand New Member

  • Members
  • 2 posts

Posted 20 December 2018 - 05:08 AM


I designed a Single Effect Mechanical Vapor Compressor desalination system. Now I want to size the shell and tube heat exchanger, which is the main component, used for heating brine with superheated steam in order to be able to calculate the pressure losses in the evaporator/condenser.

The sensible heat released from desuperheating the steam is low, so the driving force considered in sizing the heat exchanger is only the latent heat.    

Shell side: saturation pressure for water at 0.1995 bar is 60 C, but because of the boiling point elevation (which is extimated through the Pitzer model to of 5.3767 C), the evaporating brine  temperature is 65.37 C; flow rate 2.7 kg/s.

Tube side: steam at 0.3921 bar it will be cooled and condensed (saturation temperature of water at 0.3921 bar is 70 C) from almost 131.56 to 75.37 Celsius; flow rate 7.4414 kg/s.

When I decide the tube OD, ID (respectively outside and inside diameter), length and number of tubes, I can check and adjust my choice, knowing that the two phase flow velocity in the tubes has to be around 2-2.5 m/s. When I calculate velocity, the density of the two phase flow is needed, due to the following formula v = (4 * m * N_p)/(density * pi * ID^2 * N_tt), where:

- m: mass flow rate

- N_p: number of tubes passages

- pi: is 3.14..

- ID: internal diameter

- N_tt: number of tubes

The vapor is supposed entering the tubes with a vapor quality x_in = 0.01, and leave the tubes at x_out = 0.99.

The question is which value of density should I use for this calculation? Just an average one or the homogeneous one? I search in lots of books but I did't find anything.

Thank you.


#2 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,721 posts

Posted 20 December 2018 - 12:01 PM



To get a clearer and accurate picture of what you describe, are you stating:

  • You are using superheated steam to heat a brine?
  • You are using a conventional shell & tube heat exchanger - probably with a removable tube bundle?
  • You are flowing the brine through the exchanger's shell side - with one pass?
  • You are flowing the superheated steam through the tube side - with one pass?

My initial comments are that I've always seen brine flow in the tube side of an exchanger - for obvious operational and maintenance reasons.

I have never relied on using the sensible heat in superheated steam.  I know of no one that has.  The size of the exchanger, the calculations, and the cost have never been justified.  All superheated steam slated for use in a steam heater has always been de-superheated prior to use.  All you can exploit is the latent heat - in a practical sense.

A steam heater should be a very easy heat exchanger calculation when only the steam's latent heat is heat source.   Can you please furnish a copy of your calculations for our member's review and comments?   This may greatly simplify your concerns.

#3 Pilesar


    Gold Member

  • Members
  • 562 posts

Posted 20 December 2018 - 12:12 PM

Condensing steam in the tubes is a more complicated calc than shellside condensing because the pressure drop for the steam/condensate in the tubes is significant. As the distance from the tube inlet increases, the saturation temperature drops further. With your close temperature approach, you cannot assume constant condensing temperature. 

#4 andrea.r


    Brand New Member

  • Members
  • 2 posts

Posted 20 December 2018 - 02:26 PM


In first, there isn't a real plant, it's a theorical thesis; so I'm talking about a model (a simulation) which predicts the area of the evaporator/condenser, in order to estimate the price of the system. I just want to size the components at the design point for my work.

For what I've learnt, a  lot of thermal desalination systems I've seen in papers in my initial research, present shell and tube heat exchangers which are of this configuration (i.e. brine in the shell side, and condensing steam in the tubes - both one pass) - for example, the book from which I started developing the model is "Fundamentals of Salt Water desalination" by El Dessouky and Ettouney (https://www.scienced...er-desalination ), and all the models have been developed in this configuration (horizontal shell and tube heat exchangers with steam in tubes and brine outside).

How the system work:

The steam heater is only needed for the start-up process; because after that, in steady state conditions, the steam that is used to heat the brine (sprayed over the tube bundle) is the one you get from the solvent evaporation. The formed vapor passes through a demister and it is intaken from a centrifugal compressor which is used to enhance the saturation pressure of the steam and further increasing its temperature giving it an additional quantity of superheat. The superheated steam coming out of the compressor flows inside the tubes of the condenser. And the cycle restart. There are also two preheaters that use the heat of the discharged distillate and brine to heat up the incoming brine. If this explanation can help.

The calculations I've done till now gives me the mass flow rate of brine and steam and the related inlet/outlet temparatures. So, with these input data I can use the K-deltaT method to size the heat exchanger, am I wrong? (where K is the overall heat transfer coefficient and deltaT is the supposed costant -it is one of the assumptions of the model- deltaT between the evaporating brine and condensing steam).

My calculation are based on the "Fundamentals of Desalination" ones, with a different resolution algorythm (in MATLAB, because the final brine concentration is higher and the correlation for NaCl thermophysical properties I used are different, non-linear and therefore they needed iterations in order to be solved) and I haven't them in paper.

But, I can't understand which calculation should I post here in relation with the first question, because all I have (in order to size the shell and tube heat exchanger) by know is the input data for the K - deltaT method, that I have obtained from the model simulation. I have anyway the procedure I'm going to follow for sizing it; if you meant that I can post it.

For a clearer understanding of the previous explanation I have posted a simplified scheme of the system. It was normally used for seawater desalination, but due to environmental problems regarding for brine discharge (based on the brine volume) it is now used for brine volume reduction. I hope the problem it is know clearer.


http://slideplayer.c...sion (MVC)..jpg

I'm sorry for my English (I'm not a native speaker); and for the length of the post.

I hope the system is clearer now. Thank you for your time.

#5 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,721 posts

Posted 20 December 2018 - 05:07 PM



The scientific paper you refer to doesn’t make any detailed explanation of how the process really is designed and fabricated.  Kindly refer to the attached paper on the subject of dealing with the design of a mechanical vapor compression desalination process.  Look at how the author deals with the subject of steam heat transfer in this type of application and how the energy employed in a vapor compressor is obtained.  When dealing with a fluid such as sea water concentration due to evaporation, conventional heat exchangers are avoided for obvious reasons of efficiency, corrosion, maintenance, dependability, and economy.


The illustration you show of a mechanical vapor compression process is very deficient.  It fails to denote what is driving the centrifugal compressor.  It also does not show mechanical details - especially as to type of heat transfer apparatus used.  This is typical of a purely academic paper.  Scientists visualize and theorize.  Engineers design and build - with an economic incentive.  If we are to believe the sketch, the compressor is using electrical energy to compress the low pressure steam.  This doesn't make economical sense.


Pilesar, for example, points to the simple design basis an experienced engineer always follows: keep the design simple.  In this case, Pilesar rightfully points out to the avoidance of 2-phase flow in a heat transfer conduit.  The simplest and most practical design for this application is to avoid the 2-phase flow to interfere with the vital heat transfer required.  That makes common sense and stands as a prime reason to place the condensing steam outside the tubes (if tubes are used) and rely on drop-wise or film-wise condensation with gravity assisting to remove the desired condensate.  In a purely scientific discussion, these engineering details are of little or no worth.  However, if one is to dedicate complex and costly engineering manhours to a successful design, these details are of prime importance and make the difference in a credible, dependable, and successful process.


By selecting the proper heat transfer method, you don’t have to confront the complex dilemma of determining the density of a differentially condensing vapor traveling inside a heat transfer tube.  There is more than one way to skin a cat.


Attached File  An Advanced Vapor-Compression Desalination System - TAMU2005.pdf   1.52MB   29 downloads

Similar Topics