Cascade: Refrigerant blending and Oil Management

[mytekcontrols]

I dont think r50 and r1150 is going to give you much but MyTekControls is probably the best one to answer

Scott, Adam, have you guys ever experimented with putting a small SLHX (evaporator suction return -exchanging with- liquid line) on a given stage of a cascade prior to feeding the cap tube? This added sub-cooling would most likely be very beneficial when working with mixtures (much more so than simply wrapping or bonding the cap tube around the evaporator suction line). I know it works wonders on my CryoBUG design.

Keep in mind that the more sub-cooling that occurs to the R1150, the more R14 or methane can dissolve into it, thus creating a lower evaporating temperature than R1150 alone. This is the whole idea of utilizing a mixture, to create a non-existent refrigerant of a desired boiling point. Obviously if you could find a single refrigerant having a boiling point of -110 C, that would be ideal.

[n00b 0f l337]

Yeah I've done it on alot of first stages, but as people pestered for closer and closer to -100C and cared a bit less for load as long as it was decent I stopped SLHX on 2nd stage. Also found it made tuning a larger pain in the bum. Never tried actually throwing some r14 in there as I felt it would be mostly cosmetic on temperatures though.

Man if only I had the time these days

[Patrickclouds]

i think it depends of the usage of the unit and the load you have to deal with.

when we talk of two stage cascades for computer cooling with r404a/r507 in the first and r1150 in the second stage, we tune them for a load of around 350watts.

350 watts load causes a high pressure of 17 to 18 barg and very hot discharge temps.

if you now add some r50 or r14 the discharge temp and high pressure will increase.

i know a load test with 350 watts for one hour is tough and maybe not comparable to the use on a computer chip where you have the max load only for minutes while running a benchmark.

maybe you can get rid of the pressures when tuning the mixture properly, but it's not as easy as filling the unit with ethene and add afterwards some methane.

who will be the first that gives it a try ?

[DFI pit bull]

Can't figure out how the r50 or r14 would condense, please explain, how does it work or at least in theory, has anyone tried, have not seen it done before.

[n00b 0f l337]

It's not going to condense. Simply it will dissolve into the already condensed r1150 (primary) refrigerant. And lower the average boiling temperature.

[mytekcontrols]

First of all to answer your question DFI pit bull, and as Adam pointed out, I am not talking about condensing either the R14 or Methane into the R1150, but instead creating the conditions where some of the gas can dissolve into the R1150. Doing this will modify the normal boiling point of R1150 towards the colder side of things (not as cold as either -128 C or -161 C, but colder than R1150 alone). The other effect of adding either the R14 or Methane, is known as partial pressure, where it can create what looks to be a lower pressure environment in the evaporator, thus causing the refrigerant to boil off colder. This is basically how propane heated refrigerators work, with the hydrogen gas part of the mixture creating a partial pressure environment for the room temperature liquid ammonia to evaporate in (see: How 3-way Fridges Work).

Man if only I had the time these days

Yeah me too

who will be the first that gives it a try ?

Well I kinda of do this all the time whenever I build an AutoC (a working one that is ).

[n00b 0f l337]

'Before' has blended r50 and r1150 with good success in the past as well

[sdumper]

MyTekcontrols: Do you have multi stage cascades running for 24 hours a day 7 days a week?

Without shutting down...

[mytekcontrols]

MyTekcontrols: Do you have multi stage cascades running for 24 hours a day 7 days a week?

Without shutting down...

Not Standard Cascades, but plenty of Autocascades in all sizes ranging from below 1 HP on up to 10 HP. Many of these systems are run non-stop for months at a time.

'Before' has blended r50 and r1150 with good success in the past as well

Perhaps he can chime in on this topic for verification. Like I said earlier, I think with some sub-cooling via an SLHX, I don't see why this wouldn't be of benefit. It's not really that much different than what is being done in a single stage of an AutoC.

Check out the final stage of my CryoBUG for an example...



As can be seen, I am getting about 30 degrees of additional sub-cooling (-105 C versus -135 C) in the last HX stage which is nothing more than a SLHX or SUBCOOLER as I call it in the diagram. It is here that I do the bulk of the work to concentrate the condensed R14 with Argon gas, thus giving me the -156 C temperature at the evaporator inlet. R14 alone at 20.5+ psig evaporation pressure would only yield about -115 C. And as I also theorized, the perceived lower partial pressure produced by the Argon gas in the evaporator, also plays a role in lowering the evaporation temperature.

[kang-China]

'Before' has blended r50 and r1150 with good success in the past as well

i think ,if you use R50 and R1150 blended,you can do that ,use a subcooler in the system,,

if you use the subcooler ,the suction will become high .

but ,you can do that ,R1150 and R14, is better ,

[kang-China]

mike, welcome

[sdumper]

Not Standard Cascades, but plenty of Autocascades in all sizes ranging from below 1 HP on up to 10 HP. Many of these systems are run non-stop for months at a time.

Any reason why a normal two stage cascade wouldn't be able to run a true 24x7 at various loads?

[n00b 0f l337]

Just oil return really. At some point enough oil will escape past the oil separator to end up trapped in the evaporator. Its why 24x7 systems often have a defrost cycle.

[mytekcontrols]

Just oil return really. At some point enough oil will escape past the oil separator to end up trapped in the evaporator. Its why 24x7 systems often have a defrost cycle.

I agree

On some of our big AutoC's we use up to 3 oil separators in series to totally alleviate the oil migration problem. And keep in mind that in an AutoC the phase separators also act as oil seps. So to only use one oil separator on a cascade you're probably sacrificing reliability, even with a hydrocarbon refrigerant. I know Temprite makes claims of 99% oil separation for their coalescing units, but I doubt that that is really completely true unless everything is ideal. Although placing a muffler between the oil sep and compressor is suppose to help, because gas pulsations tend to cause more oil carry over.

Hello Kang

[n00b 0f l337]

Yeah I've seen some of your (and other tech's) work with serial Temprites. It adds a lot of cost for a cascade which will probably never see 24x7 usage. And its so simple to have a defrost or hot gas bypass style solenoid on a switch.
I think its also why we've always liked autocascades in that sense, the need for an oil separator seems far less essential.

[mytekcontrols]

Yeah I've seen some of your (and other tech's) work with serial Temprites. It adds a lot of cost for a cascade which will probably never see 24x7 usage. And its so simple to have a defrost or hot gas bypass style solenoid on a switch.
I think its also why we've always liked autocascades in that sense, the need for an oil separator seems far less essential.

Yep AutoC's tend to play nicer with oil. My CryoBUG is a good example of this, using nothing more than a large bullet strainer as it's one and only phase separator (and no oil separator). I've been testing out a charge with R600a, R23, R14, and Argon and so far so good. Although I haven't run for more then 2 days thus far between shut-downs. I switched over to R23 from ethane due to having difficulty staying below 57 grams of flammables.

Back to the original topic; I would still be interested in someone trying a bit of R14 or methane with the r1150 in the final stage of a cascade. Does anyone have a link to Before's test?

[n00b 0f l337]

I believe it was this project...
http://www.xtremesystems.org/forums/......&highlight=

But in his 3 stager he tried R50/r1150 and has used it since, and with decent success.
There's no reason why it shouldn't work with a good SLHX down towards ~ -70 to -80C, but in the -35C leaving the HX I don't expect anything.
I think much of the time our systems run close to our "limit" of 250-300psi on the high side anyway, so we've never seen any room for the r50.

[mytekcontrols]

I believe it was this project...
http://www.xtremesystems.org/forums/......&highlight=

But in his 3 stager he tried R50/r1150 and has used it since, and with decent success.
There's no reason why it shouldn't work with a good SLHX down towards ~ -70 to -80C, but in the -35C leaving the HX I don't expect anything.
I think much of the time our systems run close to our "limit" of 250-300psi on the high side anyway, so we've never seen any room for the r50.

That posting stopped at page 4 for some reason, and just when it was getting interesting. He was at -130C with R14, and still waiting to do the methane test.

[n00b 0f l337]

Damn I didn't check that far, must have been the three stager of his.
The old archives are getting dusty

[sdumper]

Anybody tried a small amount of r23 with r402a?
Is it worth trying?

[n00b 0f l337]

With enough subcooling it'll work, thats the whole idea here though. R402a is already a blend, its already pretty much r125 into r22. Except that the r125 is also condensable at the temperature range. If your willing to run a proper SLHX (not just a little tube or wrap) then yeah theres no real reason why not past that r402a is already very high pressure.

[n00b 0f l337]

The idea here that I think your missing Scott is that part of the goal is to use the extra headroom in a system.
With r402a you dont have much headroom.

Now if you used r134a or r290, and your condensing at 200 psi. Then you've got headroom in the condenser and head pressure.
That's when if you have the subcooling potential (SLHX or HX) you can add a non-condensable like r23 until you run out of headroom, say at ~265psi or so.

I've done co2 and r290 back in the old school autoc times. That works fine actually although dry ice was always a problem.

[sdumper]

Good points adam but head room should be relative to compressor size somewhat I would think. Regardless, I cant try it now because I don't have the right adapter for the r23 tank they sent me


Still not happy with the load testing...not one bit but I have time to tweak.

[n00b 0f l337]

It's not so much compressor size as system goal.
Give it a shot! Just stop into an air gas for the adapter to order it 2 day. Easy enough!

[mytekcontrols]

Now if you used r134a or r290, and your condensing at 200 psi. Then you've got headroom in the condenser and head pressure.
That's when if you have the subcooling potential (SLHX or HX) you can add a non-condensable like r23 until you run out of headroom, say at ~265psi or so.

Couldn't of said it better myself. Thanks Adam.

Another good possibility in a pure hydrocarbon blend is R600a (iso-butane) and R170 (ethane). But if I properly recall the start of this thread was about getting colder in the 2nd stage of a cascade then what could be accomplished with either R23 or R1150 alone. So this brings us back to the idea of using an SLHX and dissolving a normally non-condensible (something you would use in the 3rd stage of a 3-stage cascade) by sub-cooling into either the R23 or R1150 such as R14 or R50. So with that in mind, the R23/R14 combination seems like the best bet to me. Also adding a tiny amount of R600 or R600a would help with the oil return.