Direct expansion ammonia evaporator/condenser

[virk]

Hello!

Perhaps somebody of you has little information and/or experience about the following cascade heat exchanger application:

Because of low space we want to use a plate heat exchanger for condensing ammonia at -38°C (-36.4F) by evaporating ammonia at about 4K colder on the other side of the heat exchanger. Here are my questions:

1) What type of plate heat exchanger should we use and why? How should we install it? We think about direct expansion installation in order not to have to install a separator. (There are no problems with droplets because of ammonia absorption refrigeration system)

2) What additional problems we should take into account and consider concerning this application? (In general we are familiar with this application)

Of course we have some information of potential suppliers but I would like to learn something from you.

Kind regards

virk

 

[gepman]

virk, I have not used a plate and frame heat exchanger as a condenser and my two-stage ammonia experience has been compound not cascade. I don't have any experience with absorption systems either.

My experience with plate and frame heat exchangers in ammonia applications is limited to using them as either flooded or liquid overfeed evaporators for chilling glycol. In this case many people want to use the completely welded type with no gaskets since ammonia leaks now have to be reported. I have not used a plate and frame on a DX system and I think that it would be a waste of the advantages of the plate and frame system that you are trying to obtain (limited space and heat transfer area) since the superheat in a DX system reduces your heat transfer.

Sorry that I couldn't be much help but I don't know much about the application only the heat exchanger.

 

[TXiceman]

Generally you should not use DX on ammonia below about 0 dF. The correct method is to use a flooded design or a pumped liquid overfeed design

As for the exchanger, I'd look at something in a nickel brazed plate type over a gasketed plate and frame. Try

http://www.flatplate.com

for a US supplied.

If you do go with a DX design, use an electronic expansion valve.

Ken
KE5DFR

 

[virk]

Thank you gepman and Ken for your contribution.
Ken, why do you suggest not to use DX-evaporation below 0 dF and why should we use an electronic expansion valve.
(We plan to use AlfaNova from company AlfaLaval) In case of DX-system we can do without a separator, which saves money and space.

Kind regards

virk

 

[gepman]

I have looked at the AlfaNova units and they appear to be very good (although new). I would see if Alfa Laval has any references in your application. Although I cannot speak for TXiceman I almost never see any DX systems in ammonia and certainly not at low temperatures. In mechanical refrigeration DX systems produce superheat which reduces the efficiency of the compressors (lower density and therefore less capacity for a given swept volume for the compressor). However I haven't studied absorption systems enough to know how superheat affects the efficiency of the absorption cycle. I also believe that his recommendation of an electronic expansion valve is because you can control the superheat much more accurately with an electronic expansion valve than a thermostatic expansion valve. DX systems just aren't seen much in larger refrigeration systems because of the penalty that you pay in evaporator and compressor performance. This may not be an issue (obviously the compressor isn't) in an absorption system.

 

[TXiceman]

The way it was explained to me by folks that design DX exchangers was that it is a problem with predicting two phase flow through the HX as it progresses through the exchanger. You need to keep the flow within proper ranges to get good heat transfer.

I do know that i had to work on one DX NH3 glycol chiller at a -15 dF ET and it would not work well. We finally wound up converting the DX chiller with new heads and a thermosiphon drum. Never changed the surface area, just a single pass thermosiphon design and it worked fine.

Ken



Ken
KE5DFR

 

[gepman]

There will be two phase flow in all evaporators including DX, flooded, and liquid recirculation. If you look at evaporator coil ratings in catalogs the same coil is always rated less BTU/H for DX than for flooded or liquid recirculation. This is because in a DX coil, by the very nature of the thermostatic expansion valve, the coil is always starved for liquid refrigerant so that there is no liquid refrigerant carryover. Since approximately 20% of the coil surface is not wetted with liquid refrigerant there is less heat transfer since the largest amount of heat transfer takes place with the change in phase of the refrigerant (heat of vaporization) and not with the increase in sensible heat of the vapor. There are all kinds of reasons that your DX coil may not have worked, it is always difficult to get good distribution of refrigerant from the expansion valves to the various coils. Large coils usually have special distributors from the expansion valve to the individual coils.