View Full Version : Which refrigerant to use for low temp apps?

24-08-2002, 12:53 PM
I am asking this for another Dutch guy who also wants to build a refrigeration system for his processor. He cannot ask it himself over here since his English is even worse than mine.

He plans on building a system where the refrigerant evaporates in an evaporator directly on the CPU die. Depending on how my experiments with a TEV in such small system work out, the evaporator will be fed by a TEV, captube, or fixed evaporator pressure valve.

The question is: what is the best refrigerant to use for an evaporation temperature of approximately -40 °C, preferrably -50 °C? He says he can choose between R22, R404a, R407, R600a and R134a.

To me, R600a and R134a seem unsuitable since compressors are not optimized for those low suction pressures. Another problem is the rather low evaporator surface (he did not yet finish the design of his evaporator, but I guess the refrigerant->metal contact area is not larger than 0.05 m<sup>2</sup>), so a refrigerant with a high heat transfer coefficient and high latent heat of vaporization seems to be more optimal.

Based on this, I would say either R22 or R407, but my thinking proved to be wrong many times earlier.

24-08-2002, 05:39 PM
Of the refrigerants listed, my choice would be R404A, mainly because I am familiar with it. But I am certainly not an expert on refrigerants. :)

26-08-2002, 01:40 PM
History has proven R22 is not a good low temp refrigerant. The high heat of compression will cause the compressor to overheat. You could run less superheat, but an experimental system with R22 would be one more risk. I would go with 404 also.

26-08-2002, 03:35 PM
If you are talking about R407c, this also is not a good refrigerant for low temperatures. Compared with R22 it's efficiency drops off dramatically when the evaporating temperature drops below zero and superheats are above 6k.

With the small evap surface your friend is contemplating I would assume that he will experience very high superheats.

26-08-2002, 05:43 PM
You would get high superheat only if your load exceeds the flow rate of refrigerant. If you have high superheat, increase flow rate of *****. With a small surface area the DT will be high though. (I always get those backwards, difference in temp between load and ***** is what I mean)

I would say a pure refrigerant would give better performance on a small surface. A blend boils at different temps (glide). I would quess minimum glide would be best for smaller surface area.

27-08-2002, 09:10 AM
OK, so 404a seems to have the best cards?

Another possibility is the use of R290 (propane). This should work using a captube and an R22 compressor. Would this perform better than R404a? I don't think so, but asking does not harm.

27-08-2002, 02:16 PM
Does Prof Sporlan have any gems of wisdom to share on this subject? I think the rest of us are speculating. Some with a refrigerant company please come to the podium.

Prof Sporlan
27-08-2002, 07:25 PM
For single stage refrigeration applications operating down to -50C, your best bets will likely be R-404A or R-507. At least you'll find a compressor manufacturer willing to talk to you about this application... :)

No way on R-600a (isobutane), as your pressure ratio would be out of sight. R-134a isn't much better in this respect. R-22, as previously noted, has problems with high discharge temps at these operating conditions, and R-407C is largely a refrigerant for a/c applications.

The Prof has in the past sized many TEVs for R-290 (propane), but not with applications operating at -50C evaporator. And he sees no benefit using R-290 with R-404A or R-507 being readily available, and non-flammable :)

28-08-2002, 10:36 AM
Since R507 is not in his list, R404a seems to be the most suitable refrigerant. Also, parts for R404a are readily available, and there is plenty of iinformation about R404a.

So, I translated the usefil tips in this thread to Dutch and sent him an E-mail.

Thanks guys!

11-09-2002, 01:20 AM
R-22 is an excelent low temp refrigerent, Things just get more complicated. We have single stage systems that use liquid injectors that run as low as -25 F, -30F. We also custom built a compound cooling system that runs 22" suction for a special application. I belive he runs a -60F end product temp.

11-09-2002, 08:51 AM
But would R22 be a better choice than R404a?

11-09-2002, 11:54 PM
Depends on unit size, in the application above it would probly not be cost effective. When telling a custom the price of 2000# R-507 vs R-22 plus the cost of 25 gal POE oil vs Min. R-22 usually wins. I should have been more specific in my responce. R-22 will work when set up right, but depending on the specifics of the facility may or may not be the best option. When I read the thread it sounded as though R-22 was ruled out for the wrong resons. If he has 20 other machines running R-22 it might be a good idea to use it and he sounded like it was his gas of choice.

20-09-2002, 06:51 AM
I recently worked on a deep freezer for a sperm bank.
It had 2 compressors.

On one of them, it used R95

The other was R290 and R22 (Can't remember for sure)

I still can't find out what R95 is.........

The freezer is rated to -88deg C

21-09-2002, 12:41 AM
R-95 might have been Suva-95.

Prof Sporlan
21-09-2002, 03:17 PM
The Prof has heard the term "R-95" used for Suva 95, a.k.a. R-508B, a R-23/R-116 blend, 46/54 percent by weight