Plate heat exchanger -- Dimensioning

[Claudio156]

Hello,
I am busy with a feasibility analysis of a sea water air conditioning system. This is a system that provides cooling with deep sea water. Therefore, I need a heat exchanger, especially a plate heat exchanger. In my research, they are always saying that the overall heat transfer coefficient for this kind of heat exchange is around 4000 W/(m^2K).

The problem is that I never come to this value (only when I am playing with some values, but that is not the way it has to go). Mostly I have a value around 1500 W/(m^2K) without taking into account some fouling factors.

The general purpose is to dimension a plate heat exchanger starting from temperature values across the heat exchanger. I assume in the beginning that the geometrical information is known and that I just have to choose how many plates I need.

I am using the following technique:
First I assume that I know the temperatures in the in- and outlet of my heat exchanger at both side. Thus, I can calculate my LMTD value. Next, I want to calculate the Reynolds number of both fluid: sea water and district cool water. Therefore I make an assumption: I assume the plate heat exchanger consists only 1 plate for the sea water and for the district cool water. The Reynolds number are calculated and then the heat transfer coefficients are calculated with a Dittus Boelter relationship.

With the formula 1/U= 1/Hcold + 1/HHot + PLThickness/KMaterial

I can calculate my U-value. This new U-value, I am using for calculating my heat transfer area:

A=Q/(U*LMTD)

With information about the plate area, I can calculate how many plates I would need. This value will be compared with the previous value and if the difference is larger than 5%, the new value is assumed and the calculation are redone.

The iteration gives me always a value around 1500 W/(m^2K).

*If the explanation is not clear, I can give you a site where the procedure is explained:

https://books.google.be/books?id=hm...EIJzAB#v=onepage&q=heat transfer area&f=false

On page 422

Two important remarks:
• I have a very high cooling duty, in the order of 30-120 MW. Thus, I am considering that the amount of plates will be split over a few heat exchangers.
• The assumed plate geometry is:
Plate area= 3 m^2
Plate Height: 3.567 m
Plate Thickness= 0.5*10^-3 m
Plate gap=2.8*10^-3
Plate width=1.172 m
Plate diameter= 0.5 m

It is maybe because of the choice of plate heat exchanger that I am using. If I increase the area of my plate heat exchanger to 4, I am in the good range. But then, I am not working with a real plate heat exchanger because above values are taken from some data sheet.

Can anyone help me? I already investigated a lot of time in this topic and need some feedback in how to solve this problem.

Kind regards,
Claudio

 

[Drazen]

that calculation is actually very complex and few formulas will not help.

for 120 MW you will need to contact some manufacturers and ask them to tailor-design unit for you.

 

[HVACMech]

I would encourage you to contact a manufacturers rep, and give them your conditions and allow them to select the size etc. of your heat exchanger. They have special programs to run all the variables through, and will be able to tell you if your conditions can be met or not. While you could try to keep doing the manual calculations yourself, they will not necessarily match what would get purchased in the end.