Morphing Air Conditioner into Autocascade System

[mytekcontrols]

So in your opinion, propane to ethylene. Thats -42C to -104C. Or a 62C boiling point delta. Do you think a partial phase separator would be necesary? R600 to R170 is a 89C boiling point delta.

I wish I could give you a definitive answer, but as you can imagine, it depends on a lot of different factors (eg; phase separator efficiency for one). Since I don't have ethylene or anything else with this same boiling point, I can't run a test. But you can

[[XC] gomeler]

So in your opinion, propane to ethylene. Thats -42C to -104C. Or a 62C boiling point delta. Do you think a partial phase separator would be necesary? R600 to R170 is a 89C boiling point delta.

On a similar note, could you get away with strictly R290/R1150 or should you keep the R600 loop for compressor cooling and as a way to help purify the R1150 by having 2 loops to pull the R600/R290 components out of the mix?

Basically R290 -> R1150 or R600/R290 -> R290/R1150 -> R1150.

ninja-edit: Thinking about it, I'm betting the 2nd option would be better as that 1st mixture will boil somewhere close to propane, say -30C. Hopefully the 2nd mixture would boil around ethane's BP, and the final stage would be nearly pure R1150. I'm in the planning stages of my autocascade which is why I ask and I don't want to screw it up

[n00b 0f l337]

Hahaha Michael ya should get ethylene, might be able to help you guys replace r14

[Clemmaster]

What about blending 2 HC? Don't you think they will blend to form an azeotropic gas in the end? Some of you use blend on cascade (methan+propan), so why would it be different here?

You're expecting to condense ethan or ethylen with respectively R600a or R290? Good luck

Considering they won't blend (I'm not sure they won't...) :
R170 is the hardest mid range second stage refrigerant (~-85°C @ 1bar like R23 and R13) to condense because of very high latent heat of vaporisation. A R600a compressor must have 65% greater displacement than needed with R12 to get the same capacity, so don't expect to remove the heat of R170 to condense it at decent pressures.

That's worse with ethylen and propan

[n00b 0f l337]

They will be blending, and theres also going to be saturation. And not even isobutane, we're talking about normal butane.

[mytekcontrols]

You're expecting to condense ethan or ethylen with respectively R600a or R290? Good luck
Considering they won't blend (I'm not sure they won't...) :
R170 is the hardest mid range second stage refrigerant (~-85°C @ 1bar like R23 and R13) to condense because of very high latent heat of vaporisation. A R600a compressor must have 65% greater displacement than needed with R12 to get the same capacity, so don't expect to remove the heat of R170 to condense it at decent pressures.

Actually I think I can condense it at reasonable pressures with only R-600 (43 suction/240 discharge):

First Test Results (-102.6C @ 125 Watts)

And with R-123+R-22 (49 suction/275 discharge):

-101C @ 150 Watts Applied Load

On a similar note, could you get away with strictly R290/R1150 or should you keep the R600 loop for compressor cooling and as a way to help purify the R1150 by having 2 loops to pull the R600/R290 components out of the mix?

I think it best to use both the R600 and the R290 for the reasons that you stated in the end.

Edit: I'll be running my next test with R600 and R290.

[n00b 0f l337]

Oh cool made a r290 adaptor?

[mytekcontrols]

Oh cool made a r290 adaptor?

Yes I did. Thanks for the idea

Still waiting on my extra parts to arrive first, so I can implement the buffer valve feature. But soon.

[n00b 0f l337]

No problem and not my idea, been used here for years

[Clemmaster]

Actually I think I can condense it at reasonable pressures with only R-600 (43 suction/240 discharge):

You're talking about condensing R170 with R600 right?

[n00b 0f l337]

Yes he is, butane will boil much lower then its normal BP as it absorbs and saturates with ethane.

[Clemmaster]

How do you know where it will actually saturate?

Ethane involve a very high condensing rejection so you must have a good displacement to carry such load on first stage...

[n00b 0f l337]

It will saturate everywhere... Dunno what you mean by that? The auxillary condenser before the first phase separator will lower refrigerant temps and boost saturation as well.

[mytekcontrols]

Clemmaster, When I made the statement:

Actually I think I can condense it at reasonable pressures with only R-600 (43 suction/240 discharge)

I really meant to say "I know" instead of "I think". The reason I say this is, that I know beyond a shadow of a doubt, that the only way I could have obtained -100C or colder with a charge based on R-600, R-170, R-14 -and- with a 240 psig condensing pressure -while- evaporating into a 43 psig suction, would be by condensing at least the R-14, which would have minimally required condensing and evaporating the R-170.

My questions to you are: Have you studied, or have been working with autocascades? Do you understand that an autocascade is not the same as a standard multi-compressor cascade, and that in an autocascade a mixture of refrigerants are fractionated out and used to cool successive stages? If not, then this may be the source of your dilemma about why I was able to condense R-170 (and R-14 which has a much lower boiling point), while using only one other refrigerant (R-600).

One last question for you: Have you read through this entire thread? If not, then I would highly recommend that you do so, since much has been covered on the topic of autocascades and how we think they work.

I'll be the first one to admit that there has been much theorizing on the subject, and I'm sure you'll find that some of the later posts will be in conflict with what had been thought to be fact in earlier ones. But overall I believe it has been a very productive effort, and hopefully is allowing others to have the knowledge to build their own autocascades, and have their own successes at condensing R-170, and/or R-1150, and/or R-14.

[n00b 0f l337]

Mytek if you need a rotary cascade as you were thinking from the r22 problem thread, send me an email I have a large inventory at the moment.

[mytekcontrols]

I just read over my response to Clemmaster's questions, and although I didn't mean to come off as and over bearing know-it-to-all, I think it could possibly be perceived that way. My apologies go out if it felt this way to you as well Clemmaster

Anyway the point I was really trying to make is, that there is a tremendous amount of information on autocascades already covered by this thread. So to keep from being redundant (by going over what has already been discussed), it is best to start at the beginning, and read it through to the end. That way we can try to keep this from getting into a 1000+ postings, which to say the least, gets a bit unwieldy.

However, I am always happy to try and answer the questions, and try to do so as accurately as I can.

Also... to this thread.

[Clemmaster]

Clemmaster, When I made the statement:

I really meant to say "I know" instead of "I think". The reason I say this is, that I know beyond a shadow of a doubt, that the only way I could have obtained -100C or colder with a charge based on R-600, R-170, R-14 -and- with a 240 psig condensing pressure -while- evaporating into a 43 psig suction, would be by condensing at least the R-14, which would have minimally required condensing and evaporating the R-170.

My questions to you are: Have you studied, or have been working with autocascades? Do you understand that an autocascade is not the same as a standard multi-compressor cascade, and that in an autocascade a mixture of refrigerants are fractionated out and used to cool successive stages? If not, then this may be the source of your dilemma about why I was able to condense R-170 (and R-14 which has a much lower boiling point), while using only one other refrigerant (R-600).

One last question for you: Have you read through this entire thread? If not, then I would highly recommend that you do so, since much has been covered on the topic of autocascades and how we think they work.

I'll be the first one to admit that there has been much theorizing on the subject, and I'm sure you'll find that some of the later posts will be in conflict with what had been thought to be fact in earlier ones. But overall I believe it has been a very productive effort, and hopefully is allowing others to have the knowledge to build their own autocascades, and have their own successes at condensing R-170, and/or R-1150, and/or R-14.

Don't worry for me

I didn't look at the pdf, and I now know the purpose of this cooling, nothing to be afraid of indeed

[Clemmaster]

I just read over my response to Clemmaster's questions, and although I didn't mean to come off as and over bearing know-it-to-all, I think it could possibly be perceived that way. My apologies go out if it felt this way to you as well Clemmaster

No problem, that's the way I talk too I understand that it isn't personnal

I apologize not having read the entire topic, I will, I'm sure I will find lots of information. I didn't know where the blend would saturate but as you said, It will saturate somewhere between butan and ethan point and that's the way it's supposed to be, that now make sense . I've always been thinking of "quasi-non blending" refrigerant (R290+R744 or R290+R23) not on HC mixtures and that's a part I'll we be working on!

Thank you

[mytekcontrols]

I've always been thinking of "quasi-non blending" refrigerant (R290+R744 or R290+R23) not on HC mixtures and that's a part I'll we be working on!

I like the term "quasi-non blending". In other words; a virtual refrigerant, one that exists only under certain conditions

This is probably as good of a description as any, for what happens to the combination of refrigerants used within an autocascade as they are subcooled, and then go into solution with other nearby boiling point neighbors. Thus creating a condensate with a unique boiling point.

I also think that the other affect that we see to lowering temperatures, is that brought on by evaporating a condensate into an environment which consists of gases that have a lower boiling point then the condensate itself (eg; R-14 evaporating into Argon rich environment). This could be compared to the effectiveness of an evaporative water cooler in a high humidity environment vs. a dry low humidity one. The low humidity environment, being of a large part composed of lower boiling point gases, allows for an increased flow of heat from the evaporation of the water. Where as the high humidity environment is less effective due to saturation.

This is all part of the fun and magic to getting an autocascade to work properly

[Clemmaster]

I like the term "quasi-non blending". In other words; a virtual refrigerant, one that exists only under certain conditions

Yes I don't know how to call these mixtures because two gases always start a reaction (even if it stops really quick, that's the way it is ).

[mytekcontrols]

Yes I don't know how to call these mixtures because two gases always start a reaction (even if it stops really quick, that's the way it is ).

This has been a problem for the engineers at Polycold for years. There was only one guy from Russia that had come up with a program that could accurately simulate a refrigerant mix for an autocascade. He had developed the program himself, and never shared it with anyone else. Unfortunately for Polycold he left, and went back to Russia, taking the program with him.

[mytekcontrols]

I received the high pressure switch yesterday (see photo below). It looks like it'll do the job just fine, and as it turns out, has the correct switch logic (Close on Pressure Rise). The Ranco Part Number is: 010-2054, and it is called a Condenser Fan Cycling Control. Range: 100 to 400 psig, Differential: 40 to 150 psig.

I will most likely leave the differential setting at 40 psig (default), and set the cut-in to 350 psig.

[Marvin]

Great, what about this ?
http://www.fullgauge.com/PDF/PCT-1600.pdf
http://www.fullgauge.com/PDF/PCT-420.pdf
http://www.fullgauge.com/PDF/PCT-100Ri.pdf
http://www.fullgauge.com.br/
regards,
Marvin

[mytekcontrols]

Marvin --- those are pretty sophisticated controllers, but I think a bit more then what is really required for pressure control on the AC-2 unit.

Below you will see an updated piping diagram showing the Buffer Valve and Pressure Control.

[n00b 0f l337]

Yep thats the switch
HPCO will cut out, a fan cycler will cut in. And they can be cheaper then that, but thats a rather high quality one