Is a receiver required or Not?

[Xeon th MG Pony]

A/C's Heat pumps and other such "comfort" cooling you will have noticed no receiver, yet, they will use TXVs! How do they pull this off? They over size the condenser and flood 3-5 rows and use that as part of their receiver but the system becomes a critical charge system! Another reason why they can get away with it is relatively stable load once under operation.


Now, technically, any system can be don this way, a Receiver just makes charging far easier and is more forgiving to error in charging. How ever it demands a higher system charge, and is an added risk of leakage or failure point, and drives cost of production up!

A critical charge system on the other hand, reduces net system charge, less potential failure points. How ever, not forgiving to charging errors, and must be care fully charged which can be time consuming and can at times perform poorly on low ambient environments, which a receiver would other wise benefit on good sub-cooling!

So, In the end, with care full thought for design, and in a situation where evap load remains relatively stable a receiver is not required, but rather, is an option

[Big SturL]

But since we're cooling computer processors, we will never have that option, right? Considering the the major heat difference between idle and load...

[n00b 0f l337]

Theres not that huge of a difference between idle and load. Two to three hundred watts is nothing, and as for the first stage of cascade, its always got a fairly decent load on it with the latent heat of evaporation from the 2nd stage.

[.p4nzer]

Theres not that huge of a difference between idle and load. Two to three hundred watts is nothing, and as for the first stage of cascade, its always got a fairly decent load on it with the latent heat of evaporation from the 2nd stage.

Correct my statement if I'm in error.
When the processor goes from idle to load is required a major fluyd so the tev open itself (reduce the open pressure). So the receiver is needed to avoid that the pressure go too high when the load is very low and the tev close the circuit.
If u tune the gas quantity for the load in thermal watt of an idle processor (an average 130/150 watt) u have a functionally system... 'cause when the load goes up the tev opens itself.

[n00b 0f l337]

Yes but theres one more step. We're talking on cascades after all. Load goes up on 2nd stage, pressures rise and heat is transfered to the first stage, the load goes up on first stage, so TEV opens a bit more and so on and so forth.
But theres quite a bit of load on the cascade first stage at all times, load on 2nd only adds more, not a 0 to 300 statement anymore.

[Xeon th MG Pony]

But since we're cooling computer processors, we will never have that option, right? Considering the the major heat difference between idle and load...

But the system is so tiny you have the option by defualt. No offense these systems are cute toys, no serious power in them.

[.p4nzer]

But the system is so tiny you have the option by defualt. No offense these systems are cute toys, no serious power in them.

But I think that if u have the money needed for a good cascade (the "old school" r402a/r1150 -100 and over Celsius) u can afford the expense for a receiver.
If u try a low cost cascade or u wanna build a custom unit with a personal touch u can do this choice. But I think receiver makes your life easyer, and it's cheap if u think about of a cascade cost.

All IMHO.

[n00b 0f l337]

Its not just about the money, also talking about simply not wasting refrigernat, if you have skill we're saying, and work and tune and build properly, its not needed. If you want to spend the extra cash, install a sight glass post condenser.

[Xeon th MG Pony]

by refining your fabrication and taking a bit more time the benefits of lower system charge, less components and cheaper production cost all are pluses that can be shared to the customer.

With a properly designed system there is no need when operating window is reliably constant, such as the systems here the operating window is tight, they are indoor so condenser temp can usually be well assured as constant, evap load is relatively constant. Thus with a properly designed system with a properly sized condenser there is no reason to have a receiver other then laziness or wide load fluxes.

I think it is time to start working on tightening up methodology here, to strive for the best design.

You have not met the perfect design when there is nothing more to add, but when there is nothing more to take away!

[SoddemFX]

Nice to start a new thread Xeon :)

As far as i can see there are two purposes to a reciever in commercial refrigeration. The first is to ensure a solid liquid feed to the metering device, the second is to hold the entire system refrigerant charge in pump down to allow servicing without recovery and recharge.

As far as i know, no one here has fitted a ball valve to the reciever outlet to allow pump down to be performed but it is viable and would be very usefull. If the system is designed to be pumped down, the volume of the reciever should be chosen based on 125% of the volume occupied by the entire system charge in liquid state at the maximum opperating temperature.

Depending on the diameter of the reciever, pipework and distance between the end of the dip tube and reciever base the system charge might increase between a little and a lot. I think the length of the liquid lines and evaporator type could potentially have a much larger influence.

Filling the bottom rows of a condensor with liquid refrigerant will increase the subcooling of liquid to the TEV compared to a reciever slightly and if the condensor is oversized to provide the required condensing capacity in the reduced space available then compressor performance shouldn't suffer either.

Temperature variation can be quite a lot, at least for me it is. Working in an unheated cow shed the ambient temperature can change from almost sub-zero in the winter to +30C in the summer. Assuming the liquid is at this temperature the change in density for r22 is about 9.5%. With a long liquid line and a "critical charge" taking up only a couple of rows in the condensor this ambient change could cause problems for a reciever-less TEV.

One of the things i realy don't like about this idea is that the evaporating capacity of a TEV with a given orifice depends quite a lot on the pressure drop accross the TEV, with no reciever the discharge pressure will depend on the available area in the condensor which with this you've made charge sensitive in addition to all the other variations you've got to deal with anyway.

A TEV is a brilliant metering device. I think that what you guys are saying will work but i also think it's kind of like running a Bugatti Veron on chip oil.

In the end this is your system and you can do whatever you want with it :)

Tom

[Xeon th MG Pony]

Ah yes, I seemed to forget about pump down cycles, but how ever again, this can be don within the condencer, it is generaly how minisplits are sold with the charge stored into the condencer.

This is not an untested thing, maney manufacturors have used no receiver for over 30 years now with air con systems and now heat pumps. It has been a well proven thing.

[SoddemFX]

If the system charge is high the condensor would have to be very large to pump down a non reciever system. I wouldn't want most people here to try this, i can see condensors exploding...

This isn't domestic air con designed by professionals who have the knowledge and responsibility to get it right :)

Tom

[[XC] 2long4u]

Couldnt you get a receiver drier to take place of the filter drier? It seems like you would want one. Besides who is trying to save money with phase? Everyone leaves the schraders on.

[Pansuu]

No need, just take a big drier :D

[n00b 0f l337]

Couldnt you get a receiver drier to take place of the filter drier? It seems like you would want one. Besides who is trying to save money with phase? Everyone leaves the schraders on.
I belevie your referring to processing the tubes at the end? Crimping and brazing shut and all that? Dont really want to do that as the unit then loses all access. i have other ways to know if the units been tampered with (the system charge and such) so I don't worry about the access valves. Not to mention there like $0.38 a peice.

[Xeon th MG Pony]

I think, you're 100% right.
We have a Carrier running of 600 kW cooling capacity without receiver.

You can go even further in your way of thinking: if the load on the Evaporator decreases, then you will inject less refrigerant.
If you inject less refrigerant in the Evaporator, the remaining refrigerant can then be stored in the condensor because with this smaller load, you need less condensor area. So, you then have spare Volume in the condensor to store this 'excess' refrigerant.
But... as you said, the system needs to be charged very critically and must work with a more or less stable condensing pressure.

If you store some liquid in the lower part, you then have a perfect subcooler.

This is a Mod from RE As he states he's working on a 600Kw system !

http://www.refrigeration-engineer.com/forums/showthread.php?t=9224

[Xeon th MG Pony]

Couldnt you get a receiver drier to take place of the filter drier? It seems like you would want one. Besides who is trying to save money with phase? Everyone leaves the schraders on.

Actualy I'm going to be using process tubes ;)

[n00b 0f l337]

Pressure switch on the fans on the condenser and ta da ;)

[Xeon th MG Pony]

If the system charge is high the condensor would have to be very large to pump down a non reciever system. I wouldn't want most people here to try this, i can see condensors exploding...

This isn't domestic air con designed by professionals who have the knowledge and responsibility to get it right :)

Tom


I was simply arguing that making TXV systems with out a receiver is very possible and quite oft don.

Durring pump down there is no load on the condenser, as you are simply removing the latent heat of the refrigerant and nothing more, so you need in the end only 1 or 2 rows to actualy condense the refrigerant!

Any system built here in Xs should never be capable of filling half the condenser, the systems are just much to small for that amount of refrigerant.

[SoddemFX]

I was simply arguing that making TXV systems with out a receiver is very possible and quite oft don.

Durring pump down there is no load on the condenser, as you are simply removing the latent heat of the refrigerant and nothing more, so you need in the end only 1 or 2 rows to actualy condense the refrigerant!

Any system built here in Xs should never be capable of filling half the condenser, the systems are just much to small for that amount of refrigerant.

But if you don't care about cost and the reciever / liquid lines are small enough not to greatly increase charge, why bother?

That's all the condensor is ever doing, if it's removing latent heat from the refrigerant then it has load. I think i get what you are saying tho: As the system pumps down the condensing load will initially be very high but as the charge is backed up into the condensor the suction pressure will fall and the mass moved by the compressor will exponentially fall. This causes an exponential load decrease on the condensor whereas i think the condensor capacity will decrease linearly with available volume, so i think i agree that it sounds safe in theory although at least 20% of the condensor volume should be available after pump down to allow liquid expansion.

The real danger in this is that if the condensor is undersized and the volume occupied by the system charge mass is greater than the space in the condensor - kaboom. A large CPU cascades first stage can have a charge ranging from 200 - 400g, with the density of r404a in liquid at +30C just over 1Kg/L the liquid refrigerant will occupy between just over 200ml to just over 400ml. Looking at the 500ml Coke bottle on my desk that's quite a lot of space and i'm not so sure a normal 2KW condensor like are used on a large cascade has this.

Luckilly it's not a problem at the moment anyway because as far as i know, no cascade first stages here are capable of pump down...

Tom

[Xeon th MG Pony]

That's all the condensor is ever doing, if it's removing latent heat from the refrigerant then it has load. I think i get what you are saying tho: As the system pumps down the condensing load will initially be very high but as the charge is backed up into the condensor the suction pressure will fall and the mass moved by the compressor will exponentially fall. This causes an exponential load decrease on the condensor whereas i think the condensor capacity will decrease linearly with available volume, so i think i agree that it sounds safe in theory although at least 20% of the condensor volume should be available after pump down to allow liquid expansion.

Thats correct, exactly the same as it pulls down to opperating box temp, colder you go less load on condenser and as such your condenser can be smaller as you get closer to opperating temp.

This is well knowen and practiced, remember the proper definition of theory too, a theory is a hypothisis with strong evidence, after all it is the theory of gravity and the fact of gravity ;) The theory tells how it works, the fact is you fall if you trip ;)