How to calculate surge drum height in flooded evaporator ?

[Tom16]

We are trying to assemble the package chiller ourselves. It is ammonia flooded type evaporation system. Application is to cool glycol from 48 to 26 deg.F in plate heat exchanger.

My question is what should be the distance between surge drum outlet to plate heat exchanger inlet and plate heat exchanger outlet to surge drum inlet ? Is there any formula we need to use ?

Please help.

Thanks

 

[StoneCold]

Tom 16
You are trying to get the ammonia to thermosiphon through a plate heat exchanger? I don't think that will work. You need to have a sinle pass shell and tube with the ammonia on the tube side if it is a verticle thermosiphon or on the shell side if it is a horizontal design. If you are pumping the ammonia around then it will not make much difference how far it is to the exchanger, other than cost of course.

Regards
StoneCold

 

[sterl]

This done through Plate and Frames or even brazed plates all the time. In fact one of the Brazed Plate selection programs gives a required surge drum height based on the refrigerant side pressure drop, the evaporating temperature selected, the total duty and so on.

Who did you buy the plate type from? Or have you bought it already?

The answers in terms of formulae: see International Journal of Refrigeration; Rahmatollah Khodabandeh (author) .....based on work of Russian named Kutateldaze; MIT professor named Rohsenow; and Klein et al. As a collection of math, a PhD Dissertation on the Web at

http://www.me.umn.edu/~kgeisler/Geisler_PhD_Dissertation.pdf

gives you 400 pages of sho has done what with what and expands the science somewhat.

Without going through all that: the H Ex manufacturer should be able to fill in the blanks and you do need it because I've had quotes come back that siad to make this function you needed to put the Surge Drum 104 feet above the heat exch. Obviously, quote was rejected but the point is: The heat exchanger did not suit the duty, a very different geometry was required.

 

[Montemayor]

I've done what Stonecold describes - flowing the liquid ammonia through both the shell and the tube side - and with success each time. In both applications the evaporator was in the vertical orientation. I prefer the refrigerant through the shell side, mainly when there are non-condensables that have to separated from a condensed process fluid flowing downward in the tubes.

I designed and constructed my units overseas, away from disturbing and nagging U.S. engineering departments, so I had free rein and empowerment to make practical decisions. I designed my vertical orientation such that the expected shell side pressure drop was negliglible. I did not use a thermosyphon effect. I employed what a normal, flooded refrigeration evaporator uses: negligible pressure drop while evaporating the refrigerant. Therefore, my expansion drum (what you call a "surge" drum) which:

1. Generates the required low-pressure, cold ammonia liquid;
2. Generates the expansion ammonia vapor resulting from the generation of the low-pressure ammonia liquid and routes the same to the refrigeration compressor;
3. Receives the ammonia vapors generated in the evaporator and routes them to the refrigeration compressor.

was located alongside and its ammonia liquid level was controlled slightly higher than the tube sheet of the vertical evaporator - approximately 3 - 6 inches - such that the hydrostatic head made up for the shell side pressure drop. The liquid from my expansion drum fed the bottom of the vertical condenser and flowed upwards to displace the evaporated ammonia liquid that went to the expansion drum as vapor. There is no formula to use - just common sense and common pressure drop considerations.

If you intend to use a cheaper construction, such as plates, then you might have to compensate for a higher pressure drop accordingly. I know it can be done; it's just a matter of what the trade-offs are and how expensive.

 

[sterl]

Not to argue, but the principles Monte describes are those of a thermosyphon; as is the steam drum on a water tube boiler or the oil coolers on many industrial compressors and even rudimentary forms of heat pipes.

There are a million or more finned coil R-717 evaporators cooling air spaces from temperatures of 40-deg. down to minus 40 or so, all essentially installed with the SD operating level 6" above the topmost tube on the finned coil. Any liquid head required is in fact a penalty to heat transfer as the saturation pressure is higher and the benefits of vapor generation to convection are diluted...

Course they are all designed with 10 to 15 deg. gross TD and adjustments on the Op Liq Level....and to a level of 90% and higher effectiveness, they all work.

By contrast: the folks who put Surge Drums right on top of redwater chillers or CIP brewery chillers have risers between barrel and SD of as much as 6-feet...and liquid at 7-ft over the topmost tube. And these things are only targeting 35 to 40-deg Outlet Temps....

And earler thermosyphon oil coolers relied on some large liquid heads to drive liquid through a S & T:
The major manufactures indicate 7 to 8 feet to support a recirc rate of 3:1. Difference was: smaller refrigerant passages and a whole lot bigger temperature differences.


Only with the incorporation of plate and frames or brazed plates or drum and plates has this whole thing become a topic. Largely that because these devices were originally considered sensible heat exchangers, and their passages are nearly symetrical by nature. With 4-connections and a stack of rectangular plates, the only thing that determines the available flow area is the spacers (often single layer gaskets in a P & F) between adjacent plates...

Consider that, for R-717: The vapor volume at Minus 40 & Sat is over 20-times as great as it is at plus 40: Op Temp and Temp Difference is a big topic to those trying to fit a 50-tr HTX in a briefcase; whereas for space cooling, a similar capacity device consists of (50)-parallel 7/8" OD tubes formed into serpentines, occupying something like 350 Cu Ft. and its own vertical dimension is 6-ft or better...

Well, its just not quite comparable. And there's currently lots of R-717 brazed plates on glycol out there that make ice in a diagonal line across maybe 30% of their exposed surface; the other 70% barely even sweats.