Cascade ammonia condenser/evaporator

[virk@gmx.net]

Hello!

There is need for a heat exchanger which evaporates ammonia at -50°C/-58F at 0.408 bara/5,9 PSI and where ammonia condenses at -48°C/-54.4F at 0.459 bara/6,6 PSI. Capacity will be approx. 1.500 kW/426 TR.
The purpose is to extend the capacity of an ammonia absorption system by a smaller compression refrigeration system.
The cascade heat exchanger is provided for to separate both ammonia circuits. My questions are as follows:

1) How would you design this heat exchanger?
2) What k-values do you expect in your design?
3) How would you solve the "non condensable gases"-problem?
4) What other aspects will we have to consider?

Any hints appreciated and if you require additional info kindly let me know!

Regards

virk

[lowtempman]

Some engineers younger than me will suggest plate and frame I am sure. Me, an old timer used to design cascade condensers with steel and tube design and they worked fine. Being below -40 F you will need to consider impact resistant materials such as SS, nickel steel, and or impact tested carbon steel. Steel material is suitable for ammonia tho there have been recent concerns about hydrogen embrittlement you will have to investigate. As far as air and purging are concerned a small ammonia compressor connected to a conventional thermal type purge unit will work. There are a lot of design considerations in this whole matter so don't proceed without some good educated design help. Oil recovery is another matter you need to look at.

[virk@gmx.net]

I received first designs from manufacturers of plate heat exchangers. One company gave an overall heat transfer coefficient of approx. 400W/m2K, the other company gave 4200W/m2K. I will think about this after the week-end.
Concerning materials we know what to do. It is our daily business. Only with this application: cascade heat exchanger with very low temperature difference and running at these low pressures we do not have experience.
Until now I haven't got a design for a shell&tube exchanger. I personally think a horizontal spray chiller with ammonia condensing in the tubes to be the "best" solution. Not only for me it is difficult to size this apparatus. I would believe overall heat transfer coefficients for this applications of something below 1000W/m2K, if the system is designed correctly.
Any further suggestions kindly appreciated!

(I will immediately discuss your hint "using a normal purge unit with a small ammonia compressor". Seems to be a possible solution if there is no other:-))

[lowtempman]

You are on the right track with your spray type heat exchanger. Since you have such a small temp difference, 3.4 F, spray type avoids the static head penalty with a plate type exchanger and a flooded shell and tube exchanger. Of the many cascade condenser designs I did years ago with a shell and tube exchanger I used a flooded design. Looking at my old engineering data I would have calculated an overall heat transfer coeficient of about 200 Btu per hour per square foot of external tube surface per degree F temp difference based on about 0 F ammonia condensing with no superheat and +10 F ***** boiling on the outside of the bare 3/4" steel 18 BWG wall tubes. Sorry I do not think in metric terms. Many of these were 500 TR exchangers used in breweries. [We wanted ammonia on the low side for good heat transfer and small pumps and ***** on the high side so I could use a multi stage centrifugal compressor on the ***** with water coled condenser.] Spray type exchangers came along later after I was no longer active in heat transfer design but I know spray type has better heat transfer coef than flooded and as noted above avoids the staic head. You have relativley small temp diff and may need lots of surface. Also do not overlook penalties that may occur if you have scale or other impurities on tube or shell side which can easily add more heat transfer resistance than the boiling or condensing coeficients, like oil. Hope this helps. Let me know if you need more info.