Andy Schoen's Receiver Subcooling Article

[DTLarca]

This matter of subcooling and receivers is a good example of the nonsense that pervades our industry and probably because historically those who write the books and design the courses just do not properly know what they are talking about. I know because I have scrutinised nearly all the books and have presented some of the key courses myself.

Andy Shoen is correct - and I want to make more salient some of the key principles. But I will get a little more technical in the article version in a few months time. I just want to give at least Mad Fridgie something to think about since he has shown a little skepticism toward Andy's arguments.

Here's the article in question...

http://www.sporlanonline.com/Februar...Cold%20WAR.pdf

I'm just going to start off with the main and very basic points - in case no one is interested anyway.

Consider water in a pot on the stove. It has a vapour pressure always. When that vapour pressure equals atmospheric pressure the water boils. The water boils at 100°C at atmospheric pressure. At atmospheric pressure it cannot properly boil at any temperature above 100°C. This is to say that when adding heat the water rises and rises but eventually reaches a certain temperature above which it cannot go - this highest temperature is a function of the prevailing pressure.

What is important to notice here is that I am talking about things as they happen when we are ADDING heat. Not removing heat. ADDING heat. We generally expect that no matter how fast we add the heat the water will not rise above 100°C - this is a reasonable expectation to have when thinking of how things are within the confines of refrigeration systems.

Because the vapour at 100°C is leaving the water body the vapour just above is also at 100°C. This vapour rising above the water will have a pretty constant density.

Now here is a key point overlooked relevant to scenarios where we are going in the opposite direction - its oversight is the source of confusion - which is a principle not brought into salience by Andy.

Consider a rigid container with 1 bar of nitrogen gas at 20°C where the weight of gas is 1kg. I can double the pressure of the gas by either halving the volume of the cylinder slowly allowing the gas to cool thus also maintaining a gas temperature of 20°C. I can also double the pressure of the gas by raising its thermodynamic (absolute) temperature - say 290K to 580K. But I can also double the pressure of the gas by simply adding another 1kg Nitrogen or by adding 1bar of any other gas. If I added another 1kg of Nitrogen and thus another 1 bar of nitrogen then I would have doubled the density of the gas. So I can have double the pressure of the same gas while maintaining the same low temperature - just that the density of the gas has increased (doubled in this case). This is very important to picture mentally.

When we are removing heat from a liquid vapour mixture at a constant pressure, as happens to say R22 in the condenser, the liquid and vapour temperature is by no means prevented from dropping below the saturation temperature. Who says just because when you are adding heat the liquid cannot reach a temperature above saturation that when you are removing heat the temperature cannot drop below saturation? It just tends not to in the early stages of the condenser. Hell, we regularly see 4K subcool with vapour presence in liquid lines - even slow liquid lines where there is minimal vapour carry-over occurring on account of sweeping.

All that happens then - toward the end of the condenser - is that we have a cooler but higher density vapour above the subcooled liquid where it simply is the case that both the liquid and the vapour are subcooled for the existing pressure. The higher pressure in the condenser/receiver is being caused by the hotter, higher pressure, lower density gas coming in the from the compressor and also that at saturated condensing temperature where all the condensing is taking place. But the vapour with the liquid toward the end of the condenser and in the receiver is equal in temperature with the liquid - its density is just greater having been compressed by the higher temperature, less dense, hotter vapour entering the condenser. Any heat that tries to get to the liquid at the end of the condenser from the hot gas at the beginning of the condenser is quicker removed along its way by the cooler heat exchanger walls than it is conducted by the separating liquid vapour bridging beginning to end.

[mad fridgie]

In reverse, i have not stated that sub-cooling does not occur in a condensor (different to my skeptism in a reciever), I also except that you can get bubble entrapment in a sub cooled liquid (temp and time limited).
I also except that you can have sub cooled liquid in a vessel (more than just its static head), this due to stratification (changes of density) heavy at the bottom, light at the top.
To have flow you must have a pressure drop, or flow will not occur.
Thus your reciever will be at lower pressure than your compressor discharger pressure or even SCT depending i which point this measured.
So lets look at the diagram.
1: The sub cooled liquid is injected directly into the vapour. Example on a cooling tower do we just have fan blowing over the surface of the water to achieve our cooling effect or do we spray the water. We spray the water to give optimum heat transfer between the water and the air.
The bigger the temp gradient the bigger the flow of energy.
What about pressure, i hear you say, it will drop. come to this later.
The liquid leaves the pipe and free falls until it hits the surface of the liquid within the reciever. The statement clearly says that the interface between the vapour and liquid will be at the higher temperture.
Depending upon how much liquid/vapour is the reciever, how far the sub cooled liquid falls, will determine how much of the boundry surface is broken, and how much recoil is rejected back upwards into the vapour and how big the waves and ripples are. basically is the reciever in a mixed state or not.
After hitting the boundry layer, will it sink, like a lead shot straight down or does it disperse slightly, moving in a downward direction, this being the case then (forgeting the vapour at this stage) the sub cooled liquid must be in direct contact with warm liquid and heat transfer will occur, warming the sub cooled, coolling the saturated.
rate flow decreases as density become closer, increasing thermal length. Are we not now producing a internal current within the vessel aid mixing further.
Let look at the practical application.
What the purpose of the reciever, in simple terms to allow for varying mass flows of refrigerant (no need to go into this area) or as storage vessel for pump down.
We will use pump down as it will show best effect (oversized reciever in running conditions)
The main object if possible is to keep a liquid seal on the reciever outlet (yes or no), the object is not to keep the reciever full of liquid. (yes or no)
Basically you charge the system until you have achieved this liquid seal (when the sight glass is full, I want to make a note that a sight glass is an aid to charging not the be all and end all)
This being the case, by volume the reciever does not have a great deal of liquid refrigerant in it.
So all the effects above will take effect, and no/little actual sub cooling.

[mad fridgie]

HOW TO MAKE THE THEORY FIT.
looking at the drawing, If the liquid feed was fitted with a perforated tube which went close to the bottom of the reciever and was surround another tube with open ends towords the top and the bottom, then your sub cooled liquid would remain at the bottom, the vapour could escape from the top , no mixing/boundry breaking would occur, you would then have genuine sub cooled liquid leaving the reciever. (temp, time, and sized based)

[DTLarca]

After hitting the boundry layer, will it sink, like a lead shot straight down or does it disperse slightly, moving in a downward direction, this being the case then (forgeting the vapour at this stage) the sub cooled liquid must be in direct contact with warm liquid and heat transfer will occur, warming the sub cooled, coolling the saturated.

The liquid arriving at the receiver from the condenser is subcooled to the same temperature as the equally subcooled liquid in the receiver.

The vapour arriving in the receiver is not really arriving - it is hanging around there above the liquid in the receiver and along the last runs of the condenser including the drain leg and this vapour in the area is supercooled. It wants to condense but it does not -it just a lot more dense than we would normally expect.

[mad fridgie]

The liquid arriving at the receiver from the condenser is subcooled to the same temperature as the equally subcooled liquid in the receiver.

The vapour arriving in the receiver is not really arriving - it is hanging around there above the liquid in the receiver and along the last runs of the condenser including the drain leg and this vapour in the area is supercooled. It wants to condense but it does not -it just a lot more dense than we would normally expect.

The articles clearly states that the liquid at the boundry layer is at saturation point, super cooling is an effect , and i have covered this in the "time temp base"
The vapour can actually travel up and down the drain leg or equalisation leg if it has one, it does not just hang around

[DTLarca]

The articles clearly states that the liquid at the boundry layer is at saturation point, super cooling is an effect , and i have covered this in the "time temp base"
The vapour can actually travel up and down the drain leg or equalisation leg if it has one, it does not just hang around

Up and down is hanging around - as in if the condenser fans speed up or the condenser fans slow down or the TEV opens up or the TEV closes down - during these sorts of events the supercooled vapour will be drifting up and down the drain line tending to equalise pressure differences between the receiver and the condenser - I call this hanging around because it is not flowing through - it can't get past the liquid seal.

Here's a video of supercooled water.
http://www.youtube.com/watch?v=fSPzMva9_CE

[mad fridgie]

Per article

It is simple: we will have 110°F at the liquid and vapor interface, but the refrigerant liquid immediately below the
interface will be at 100°F. The 110°F saturation temperature will only be found at the interface, and with the vapor

above the interface.

[DTLarca]

Per article

It is simple: we will have 110°F at the liquid and vapor interface, but the refrigerant liquid immediately below the
interface will be at 100°F. The 110°F saturation temperature will only be found at the interface, and with the vapor

above the interface.

In simplistic thinking - as Andy says in his article - he is assuming a simplistic model to impart the fundamentals.

In reality the liquid arrives at the receiver already subcooled and on the way there through the last leg passes supercooled, not saturated, vapour.

[mad fridgie]

The article does not bring to the attention pressure drop, most liquid sub cooling is actually pressure drop. hence not actual sub cooling

[DTLarca]

The article does not bring to the attention pressure drop, most liquid sub cooling is actually pressure drop. hence not actual sub cooling

This is only if there is a pressure drop between where you measure pressure and where you measure temperature.

The article assumes negligible pressure drop as is typically the case in low horse power commercial fridge condensing units.

[mad fridgie]

This is only if there is a pressure drop between where you measure pressure and where you measure temperature.

The article assumes negligible pressure drop as is typically the case in low horse power commercial fridge condensing units.

There lies practical reality. when is it sub cooled and when is not.
Having $100,000s of test rigs and measuring many (not all by a long way) units the actual amount of liquid sub cooling leaving a reciever is very small, in comparrision what folks like to believe is sub cooling.
It is not as simple as saying you or I are right or wrong. A bumble bee can not fly, but no one has told the bee that!

[desA]

The article does not bring to the attention pressure drop, most liquid sub cooling is actually pressure drop. hence not actual sub cooling

No, not necessarily the case - at all.

The condenser (non receiver type) is capable of producing sub-cooled condensate in the range of 4-5K (plate), or down to 8.5K (tube-in-tube), or more, if required - dependent on system mass charge.

[mad fridgie]

No, not necessarily the case - at all.

The condenser (non receiver type) is capable of producing sub-cooled condensate in the range of 4-5K (plate), or down to 8.5K (tube-in-tube), or more, if required - dependent on system mass charge.

As stated earlier, i have no problem with sub cooling and when or where it is produced, my issue is that you need to measure correctly liquid pressure and liquid temperature.
The nature of our industry, does allow for a wide range figures, rules and practical application.
As i am most familar with air cooled condensing units, this where I base my observations, my observations indicated pressure drops between comp discharge and liquid pressure out of the reciever.
On a shell tube, your pressure drop would be minimal (in comparrrision)
The point of this thread was around the article, and the article from Marcs point of veiw was incorrect, as proved, But his theory of course is correct, when it can happen.
Your examples are not free draining,
Like most things in life, finding the balance between academic excellence and practical application.
How many times (even on this forum) have we seen engineers give discharge pressure and liquid line temps and give some massive level of liquid sub cooling. " Measure you liquid pressure and your liquid temperature"

[nike123]

http://www.tutorvista.com/content/ph...-animation.php

[mad fridgie]

Great video, also called solid rain in Canada

[mad fridgie]

Just one point, if the vapour does not get past the liquid seal, how do you see bubbles in the sight glass, and if the the liquid is highly sub cooled, how could bubble form if there is no pressure drop.

[DTLarca]

Just one point, if the vapour does not get past the liquid seal, how do you see bubbles in the sight glass, and if the the liquid is highly sub cooled, how could bubble form if there is no pressure drop.

These are the results of constant but slight variations in receiver and liquid line pressure - both static and dynamic.

Ask yourself also why could you possibly have vapour in the receiver when only subcooled liquid is falling into there via the condenser drain where there is a liquid seal leaving the condenser in that drain leg giving positively subcooled liquid to the receiver.

The vapour in the receiver is supercooled - it is quasi-stable - it is not saturated vapour and it is not saturated liquid - it is in between - it is denser than a saturated vapour would be but it is no where enough dense to be a liquid.

[mad fridgie]

These are the results of constant but slight variations in receiver and liquid line pressure - both static and dynamic.

Ask yourself also why could you possibly have vapour in the receiver when only subcooled liquid is falling into there via the condenser drain where there is a liquid seal leaving the condenser in that drain leg giving positively subcooled liquid to the receiver.

The vapour in the receiver is supercooled - it is quasi-stable - it is not saturated vapour and it is not saturated liquid - it is in between - it is denser than a saturated vapour would be but it is no where enough dense to be a liquid.

Correct, so the article is then incorrect, and the article is the reference point, hence my points are valid, moving the goal posts does not count.

[DTLarca]

Correct, so the article is then incorrect, and the article is the reference point, hence my points are valid, moving the goal posts does not count.

I'm missing something - I'm using the article as a foundation of agreeables and going further. What in particular do you say you do not agree with the article?

[mad fridgie]

I'm missing something - I'm using the article as a foundation of agreeables and going further. What in particular do you say you do not agree with the article?

Hi Marc,
"

It is simple: we will have 110°F at the liquid and vapor interface, but the refrigerant liquid immediately below the
interface will be at 100°F. The 110°F saturation temperature will only be found at the interface, and with the vapor

above the interface."

This clearly indicates that the vapour is at saturation and is not supercooled, would you not agree.
Thus this being the case, then the rest argument is incorrect.

[mad fridgie]

Knowing that you are not going to ask a question without first knowing the answer, why is it the vapour remains supercooled, when all around it is shock energy and seed particles

[DTLarca]

Knowing that you are not going to ask a question without first knowing the answer, why is it the vapour remains supercooled, when all around it is shock energy and seed particles

This is something to think about

It's almost like the vapour in the receiver wants to be a fog

The vapour makes up for its lower temperature, for the given pressure, without condensing, by instead holding more molecules in suspension - that is the greater density.

[mad fridgie]

This is something to think about

It's almost like the vapour in the receiver wants to be a fog

The vapour makes up for its lower temperature, for the given pressure, without condensing, by instead holding more molecules in suspension - that is the greater density.

This question was pure curiosity! no counter arguments on this one

[mad fridgie]

I do think that this article should be re written, clearly indicating correct temperatures (removing doubt and counter acting my arguments, which would be true as stated at the moment) It should also include a pressure drop between the discharge and liquid line, to indicate true measurement of sub cooling (which practically is my biggest bone of contention). Technical information should be correct or stated that it is not, but has been written to aid understanding.
The term "supercooling" should be used, with an ananolgy of cloads are supercooled water vapour, something that most of us understand.

[mad fridgie]

The next question, should be?
Why do we need know what the "actual liquid sub cooling" is leaving the reciever?

[lawrence1]

stay tuned for round 2

[mad fridgie]

stay tuned for round 2

According to my score card, that was my round, even though he did have the technical moves, dancing and jabbing, he just could see the large upper cut coming, even though I had try to throw it once during the middle of the round, that floored him at the end of the round, he managed to get himself up, or was it the end of the round that saved him.
Marc only picks fight he is sure he can win, he does have technical skill to bash most around the ring, you can only beat him with cunning and pure practical brute force.
Mad F, does not mind loosing a fight, as long as he can learn from the fight, he already knows in this game there is not a single master, many weight catagories, and fighting styles. Nothing like a good "scrap" to keep us fight fit

[desA]

Some useful academic links.

http://www.wlv.com/products/thermal-...databooks.html

Collier G.C., Thome J.R., Convective boiling and condensation, 3 ed, Oxford Science Publications, 1994.

[DTLarca]

Not much chance of adding more comments just yet but I do have a moment enough to post two images taken recently measuring subcool (3.5K) on a water cooled pack system with a receiver. The pressure and temperature readings are at the same location -that is on the liquid line leaving the receiver.

[monkey spanners]

Monkey at work measuring condenser u bends.

http://www.youtube.com/watch?v=UUpwH-NHkuY


Jon

[Quality]

Monkey at work measuring condenser u bends.

http://www.youtube.com/watch?v=UUpwH-NHkuY


Jon

Now thats what I call subcooling

ps DTlarca needs a new manifold those digi refco are carp

Testo all the way

[mad fridgie]

Monkey at work measuring condenser u bends.

http://www.youtube.com/watch?v=UUpwH-NHkuY


Jon

Hi MS, what refrigerant, and what was your pressure (a bit hard to see) It looked like you were measuring liquid pressure, is that correct.

[monkey spanners]

Hi MS, what refrigerant, and what was your pressure (a bit hard to see) It looked like you were measuring liquid pressure, is that correct.

Hi, its R437A or MO49 plus, looks to be about 220psi, the gauge is on the reciever as there is no compressor high side service valve. Its only a 1/2 hp ish tecumseh unit.

Jon

[mad fridgie]

Hi, its R437A or MO49 plus, looks to be about 220psi, the gauge is on the reciever as there is no compressor high side service valve. Its only a 1/2 hp ish tecumseh unit.

Jon

Hi Jon,

thanks, no data on that refrigerant, you are reading at the correct place for liquid sub cooling.
It would be intersting to see the temperature across the height of the reciever

[monkey spanners]

Hi Jon,

thanks, no data on that refrigerant, you are reading at the correct place for liquid sub cooling.
It would be intersting to see the temperature across the height of the reciever

Its an R12 drop in, very similar to R417A but with some extra propane or something! I believe the pipe clamp thermometer is on the liquid line after the reciever. A few months ago now, job to remember Was just playing about and thought it would be interesting to see the temperature measurments and put on my youtube channel

[DTLarca]

Mad F, sorry, texting from blackberry, but have you never heard of Berno
ull's therum? Liquid and gas pendulums

[mad fridgie]

Mad F, sorry, texting from blackberry, but have you never heard of Berno
ull's therum? Liquid and gas pendulums

Yes i do understand the theory of a venturi, also how flow is attracted to a surface, can not remember the name of the principle, begins with "C", eastern european sceintist.

[DTLarca]

Yes i do understand the theory of a venturi, also how flow is attracted to a surface, can not remember the name of the principle, begins with "C", eastern european sceintist.

Not sure how this will come out. I can hardly see the text on my blackberry screen.

But, my question then, how would you describe the efficiency losses, in nature, when considering liquid pendulums?

[mad fridgie]

Not sure how this will come out. I can hardly see the text on my blackberry screen.

But, my question then, how would you describe the efficiency losses, in nature, when considering liquid pendulums?

I am a simpleton, please explain as if i was an idiot, because I am!
Slightly of subject, why is at Christmas, they have to show the Wizard of OZ? struggling with whos law this belongs to?
It is present 1.29pm xmas day, under the influence of russian based working fluid!

[desA]

Yes i do understand the theory of a venturi, also how flow is attracted to a surface, can not remember the name of the principle, begins with "C", eastern european sceintist.

Coanda effect.

[mad fridgie]

Coanda effect.

Can always rely on old desA to know his facts, good onya!
For the next round, against marc, can I have desA in my corner, braun and brains, must be a winning combination

[desA]

Mad F, sorry, texting from blackberry, but have you never heard of Berno
ull's therum? Liquid and gas pendulums

Marc, you will need to expound further on this. Please be advised that the Bernoulli equation is applied to flow along a streamline only, by definition & for a single-phase fluid.

It cannot be used across a streamline, for instance.

[desA]

LOL... You are most welcome...

[desA]

Perhaps it's time to begin defining the problem a little more clearly?

Incoming fluid properties? Pressure, temp, refrigerant.
Receiver vented, or not?
Heat-transfer to/from receiver, or adiabatic?
Receiver internal pressure?

Actually, working through my process simulator, I suspect that Andy's temperature explanation at the fluid interface is not correct at all.

[mad fridgie]

Perhaps it's time to begin defining the problem a little more clearly?

Incoming fluid properties? Pressure, temp, refrigerant.
Receiver vented, or not?
Heat-transfer to/from receiver, or adiabatic?
Receiver internal pressure?

Actually, working through my process simulator, I suspect that Andy's temperature explanation at the fluid interface is not correct at all.

As a rule,but not limited to, non vented is the most common.
Andy figures are incorrect, this is has been agreed by Marc, but I also believe he has amore than valid point about the super cooling.
A do see vertical and horizontal recievers acting slightly different (practical) and how these could be effected by external forces, so I would for this excersise call it adiabatic.
You pick a refrigerant (R134a would be OK as this one you are familar with, and has no glide, so no deviations for other reasons)

[DTLarca]

Perhaps it's time to begin defining the problem a little more clearly?

Patients - one thing at a time - things are made more clear by a slow process of eliminating those blury edges stunting our progress here.

Marc, you will need to expound further on this. Please be advised that the Bernoulli equation is applied to flow along a streamline only, by definition & for a single-phase fluid.

You'll have to wait a few days till I get back home to the comfort of my own desk.

However - let me put it this way for starters. Are you sure you cannot see any similarility in the principles of physics between a pendulum and, say, an orifice plate - in either air or water?

[desA]

LOL... I tend to make my simulated theoretical pendulums display chaos - even though the gurus say it can't be done...

[DTLarca]

LOL... I tend to make my simulated theoretical pendulums display chaos - even though the gurus say it can't be done...

Come on Des. Because the pendulum is a paradigmic icon - since back with even Galileo - but still especially central to the structure of our current scientific paradigm it is one of the first tools of description and analysis used in any field of physics including chaos theory. Did you read James Gleick?

The pendulum is a prime tool used in the demonstrations of chaos theory.

Watch this video to the end

http://www.youtube.com/watch?v=Qe5Enm96MFQ

Are you sure you are not just stalling on answering my last question

[desA]

Come on Des. Because the pendulum is a paradigmic icon - since back with even Galileo - but still especially central to the structure of our current scientific paradigm it is one of the first tools of description and analysis used in any field of physics including chaos theory. Did you read James Gleick?

The pendulum is a prime tool used in the demonstrations of chaos theory.

Watch this video to the end

http://www.youtube.com/watch?v=Qe5Enm96MFQ

Are you sure you are not just stalling on answering my last question

You are confusing the classical 2D pendulum, with a 3D pendulum. Chaos theory requires an additional degree of freedom over that which a 2D pendulum can provide... however... it is a trivial matter to introduce chaos into the 2D classical nonlinear pendulum system.

James Gleick's book is rather ancient, by now...

You should attempt to further qualify your assertion of the link between the simple 2D pendulum & whatever you are attempting to prove. The proverbial ball is now in your court... Let your infinite wisdom proceed to shine...

Personally, though, I would prefer you to set the process conditions for your receiver problem, so that detailed process modeling & simulation can proceed. I suspect that we'll be able to provide a few insights into your earlier dilemma.

[DTLarca]

You are confusing the classical 2D pendulum, with a 3D pendulum.

It doesn't really matter - how many components the pendulum has - does it?

The fact is there actually is no chaos in any of them - as the Gleick example illustrates - you will always find, from some perspective, a pattern in the chaos - chaos is only apparent. Put it this way - pendulums - all types - obey the laws we use to describe how nature goes - they cannot go any other way.

Chaos theory requires an additional degree of freedom over that which a 2D pendulum can provide... however... it is a trivial matter to introduce chaos into the 2D classical nonlinear pendulum system.

Which was demonstrated in the video I linked to

James Gleick's book is rather ancient, by now...

But not refuted

You should attempt to further qualify your assertion of the link between the simple 2D pendulum & whatever you are attempting to prove. The proverbial ball is now in your court... Let your infinite wisdom proceed to shine...

Yes - I would like to discuss gas and liquid pendulums (Newton and Bernoulli) - especially liquid to eliminate a concern that Mad Fridgie has.

Personally, though, I would prefer you to set the process conditions for your receiver problem, so that detailed process modeling & simulation can proceed. I suspect that we'll be able to provide a few insights into your earlier dilemma.

We know we cannot achieve subcool in say a shell and tube condenser on, for instance, a water cooled centrif which has no liquid capturing tube at the bottom of the condenser still exposed to the condenser water - we have instead a low side float valve or even a high side fload valve.

We are talking serpentine condenser coils per the article being examined here.

Certainly 4K Subcool tends to be quite the normal thing on a serpentine condenser with receiver and site glass. 6K is achievable at full design load. Although with electronic expansion devices the subcool remains pretty constant through all load conditions because they do not have the capacity/superheat curve that TEV's do.