Condenser Puzzle

[Gary]

Here's an odd problem:

On a water cooled condenser, the temperature of the refrigerant leaving the condenser is lower than the temperature of the water entering the condenser.

What could be the problem?

[Tycho]

is there a pilot receiver on the system?

if so, possibly pilot valve stuck in open position

[Gary]

is there a pilot receiver on the system?

if so, possibly pilot valve stuck in open position

Nope... no pilot receiver. Assume a simple water cooled A/C system with TXV.

But if there were a pilot receiver, how could that cause the refrigerant exit temp to drop below the water inlet temp?

[chemi-cool]

Some restriction inside the condenser makes part of it work like evaporator.

[Gary]

Some restriction inside the condenser makes part of it work like evaporator.

I agree.

Are there any other possible causes?

[monkey spanners]

Faulty thermometer

Jon

[Gary]

Faulty thermometer

Jon

Nope... sensors have been checked and are accurate.

Any other possibilities?

[mafoi]

Some restriction inside the condenser makes part of it work like evaporator.

I concur this restriction inside condenser answer. Otherwise, it doesn't make thermodynamics sense.

[Gary]

The condenser is a PHE. Does that change anyone's opinion?

[nike123]

Is this question related with Alfa Laval book?

[Gary]

Is this question related with Alfa Laval book?

Could be...

[Gary]

Let's take this a step further:

If there were a restriction in the body of the condenser, the reduced flow of chilled refrigerant would quickly equalize with the incoming water temp, making the exit refrigerant temp equal to the incoming water temp... therefore, the restriction is at (or very near) the exit.

[sedgy]

hey gary. give us a clue, whats the ambiant ?

[Gary]

hey gary. give us a clue, whats the ambiant ?

An excellent question, to which I do not know the answer, however the PHE is insulated as is the piping and sensors.

[monkey spanners]

Its piped up backwards, so its feeding in the bottom and condensing, and has mostly filled with liquid, which is now boiling off to get out again...

[Gary]

It would still take a sudden pressure drop (restriction) to cause it to boil off and drop the temp.

[Magoo]

Orifice tube installed as for an evaporator, and PHE installed as a condenser by mistake.
magoo

[Gary]

Orifice tube installed as for an evaporator, and PHE installed as a condenser by mistake.
magoo

Hmmm... orifice at the exit... I like it... seems unlikely... but imaginative. I've seen stranger things.

[Magoo]

Hi Gary.
You can see if orifice tube is installed by the reducer as it is not the regular size, and some manufactures dimple to tube to orientate the position of the linear slot relative to the feed to PHE when applied as an evaporator.
Talking about strange things, I had a water chiller that would not perform at all, and the problem was the TX valves two of, were welded in the wrong way around, and that was from the factory after all the quality control garbage.
Weird things do happen

[desA]

Is this question related with Alfa Laval book?

Hypothetically: What if the condenser exit piping were looped upwards, then over, down through a filter drier? (This is common received knowledge in regards to filter-drier orientation & placement, it seems).

Let's assume that, for argument, the level of the filter-drier were around 20% from the top of the PHE?

Does this filter-drier operate like a 'through liquid receiver'?

Where does the liquid level sit in the condenser, under such an arrangement?

[Magoo]

Des A
you are in the wrong post.
magoo

[desA]

Des A
you are in the wrong post.
magoo

No, I'm in the correct thread.

The question stands. I'm following on from Nike's lead regarding the potential link between receiver location and condenser discharge temperature (the alfa laval book).

This is the correct status of the application mentioned in the OP - not all facts were provided in order for the members to determine the cause-&-effect process. This led everyone to the 'restriction' conclusion - perhaps a little prematurely, in my view.

[Gary]

As Nike correctly guessed, this is the same system we were discussing in the other thread.

The refrigerant exits the condenser at the bottom, loops upwards towards the top of the condenser, then down through the filter/drier.

So this bring up two questions:

1. Can this arrangement cause liquid to back up into the condenser?

2. Can such a liquid backup cause the condenser exit temp to drop below the incoming water temp?

[desA]

I see Chef lurking on the thread.

I'd really value your insights on the filter-drier as 'gas bubbler' theory you proposed a few months back. What would its effect be on the condenser performance?

At the root of this is the conventional logic which says that a filter-drier needs to oriented vertically downwards to act as a small accumulator, rather than lie horizontally. This comment has been raised many times on RE. Experimental results would tend to show that this orientation can have unconsidered consequences.

I'd like to know more about the gas/liquid dynamics between condenser & filter-drier - what is really going on in that part of the system?

As part of this, the condensation dynamics in a vertically-oriented PHE need to be well understood, especially in regards to the effect of excessive back-pressure, or imposed liquid trap level.

[Magoo]

DesA
I hummmmmmmmmbly apologise. Oh great and wise person.
magoo

[desA]

DesA
I hummmmmmmmmbly apologise. Oh great and wise person.
magoo

Hahah...

[Magoo]

Hi Gary.
so now the posts are intergrated, that is a new twist.

[Magoo]

DesA,
speaking to the receiver above the condenser discussion. The delta P to drive liquid up to receiver could be an advantage, so as to introduce more heat to system, penalty being the COP. From memory 2.3 foot head equas 1 Psig, and then apply the specific density factor for refrigerant. So long as the piping is trapped out of condenser and before the receiver.
Generally with ammonia plants the discharge pressure is applied to top of receiver as well, with the condenser above the receiver. Mainly because of the specifc density of ammonia being lighter than anything else, and to get the gravity transfer factor back to the receiver.

[desA]

Thanks Magoo.

DesA,
speaking to the receiver above the condenser discussion. The delta P to drive liquid up to receiver could be an advantage, so as to introduce more heat to system, penalty being the COP.

Could you perhaps expand this a bit further?

From memory 2.3 foot head equas 1 Psig, and then apply the specific density factor for refrigerant. So long as the piping is trapped out of condenser and before the receiver.

What would this trap look like?

Generally with ammonia plants the discharge pressure is applied to top of receiver as well, with the condenser above the receiver. Mainly because of the specifc density of ammonia being lighter than anything else, and to get the gravity transfer factor back to the receiver.

Ok, understand this. Thanks very much.

[nike123]

Hypothetically: What if the condenser exit piping were looped upwards, then over, down through a filter drier? (This is common received knowledge in regards to filter-drier orientation & placement, it seems).

Let's assume that, for argument, the level of the filter-drier were around 20% from the top of the PHE?

Does this filter-drier operate like a 'through liquid receiver'?

Where does the liquid level sit in the condenser, under such an arrangement?

I am not that good to answer your questions. I simply added 2 an2 and concluded that Gary possibly throw us bite to catch on this interesting matter, after he little "surfed" that Alfa Laval book.

[desA]

Wondered about your detective work.

[Chef]

It may be that the velocity in a PHE is quite high so the refrigerant cools to close to the liquid coolant temperature but has a high kinetic energy. When this reforms to normal pipe flow the kinetic energy is converted to static pressure. If it is an adiabatic process it would manifest itself as subcooling but if it were non adiabatic it may show up as colder refrigerant.

However there does not seem to be a process where the refrigerant is colder than the liquid coolant unless there is a Joule Thompson expansion just before the pressure conversion back to static pressure.

Chef

[desA]

Thanks very much for your thoughts, Chef - much appreciated.

I'll tell you my thought processes:
1. Filter-drier located high up condenser;
2. Condensate backs up inside condenser;
3. This raises operating pressure (Tc,sat);
4. Condensate flow out discharge, up & over into filter-drier;
5. Vapour-liquid separation, some level of vapour locking in upper bend;
6. Expect process to be slightly erratic, with a small level of surging (nothing undue noticed in pressure gauges);
7. Sub-cooled boiling occurs in condensate - bubbles rise upwards.

The amount of excess sub-cooling begins very small at startup, but rises to around 6.5K towards the hot part of the cycle, as the liquid line heats up. This would correspond to more reactivity of the refrigerant - tendency to flash.

We have to always bear in mind that we are dealing with 2-phase phenomena here, & that kinetic energy to static pressure conversion will be a little less simple.

Physically, an exact replica of the same cycle, except for filter-drier position, runs as per expectation. The condenser is manufacturer-supplied with refrigerant stubs brazed & tested by them. It is one of the leading brands. I do not expect an exit blockage - although, it could always be argued so. (I will cut this line open & rebuild the pipework, to test the standard case.)

I have strong suspicions that the 'pat' answer we often provide with respect to willy-nilly having a filter drier set vertically with flow downwards, without taking care with respect to location & relative height, may not be entirely valid. It is perhaps something to think more deeply about.

[Gary]

DesA,
speaking to the receiver above the condenser discussion. The delta P to drive liquid up to receiver could be an advantage, so as to introduce more heat to system, penalty being the COP. From memory 2.3 foot head equas 1 Psig, and then apply the specific density factor for refrigerant.

The lift being far less than 2.3ft, the effect upon head pressure would be far less than 1 psig... and this assumes lifting a solid column of liquid which is not necessarily the case. A vapor/liquid mixture would further reduce the effect upon head pressure... perhaps to the point where it is relatively inconsequential.

[Gary]

However there does not seem to be a process where the refrigerant is colder than the liquid coolant unless there is a Joule Thompson expansion just before the pressure conversion back to static pressure.

Chef

If I understand you correctly:

Joule Thompson expansion just before the pressure conversion back to static pressure = restriction at the exit

[desA]

The lift being far less than 2.3ft, the effect upon head pressure would be far less than 1 psig... and this assumes lifting a solid column of liquid which is not necessarily the case. A vapor/liquid mixture would further reduce the effect upon head pressure... perhaps to the point where it is relatively inconsequential.

How sure are we of this, in practice though? Are there practical examples in the field where we can draw comparisons? Has anyone examples of a vertical filter-drier positioned at a level above the condenser outlet?

A condenser, as I understand things, can actually develop a reduced pressure as the vapour is condensed internally. This would be compounded as the liquid rains down. The effect of an induced back-pressure, with possible vapour-liquid pulsing/slugging is not well understood, at least in my head.

I'd love to be able to see inside the condenser-pipe-filter-drier combo.

[Gary]

How sure are we of this, in practice though? Are there practical examples in the field where we can draw comparisons? Has anyone examples of a vertical filter-drier positioned at a level above the condenser outlet?

The more common example would be the receiver inlet above the condenser outlet... and in fact this would seem to be the normal configuration for most smaller condensing units.

[desA]

http://www.compactheatpumps.org/docs...flow_x_0.2.m1v

http://www.compactheatpumps.org/docs...flow_x_0.2.m1v

Take a look at the above videos of flow within compact plate heat-exchangers. You will observe that a fair amount of wavelike motion & thrashing about takes place in the flow streams. Add to this the interference from back-pressure, condensate levels etc & I'd believe that the dynamics will be anything short of smooth, uninterrupted flow.

Take a look & think about it. Plate HE's are not the same at tube-in-tube systems, I'll bet

[desA]

The more common example would be the receiver inlet above the condenser outlet... and in fact this would seem to be the normal configuration for most smaller condensing units.

If anyone has some pictures of this idea applied to PHE's, this would be useful.

[Gary]

Interesting videos, Des.

The transition from liquid to vapor and vice versa might correctly be viewed as contained explosion/implosion. Nothing smooth about it.

[desA]

I thought they were downright scary when I first saw them.

I think that the compressor pulsations also have a lot to do with the flow instabilities. The pulsing swings back & forth at entrance/exits are something I've seen in flow simulations, with definite instabilities - a lot of whipping back & forth. The short contact length of these beasts makes for some very interesting system dynamics.

I think there's a lot to still be learned about how these HE's operate. Very challenging.

[desA]

Now, take a look at the downward flow video, again. Concentrate on the liquid outflow through the small pipe.

Now, imagine a certain amount of condensate filling the bottom of the PHE, then back it up a bit further, say due to back-pressure in the discharge line. Combine this further with bubbles of vapour trying to climb backwards back into the HE.

What would the outlet nozzle flow dynamics & pressure drop be like? Could this lead to a 'blockage' situation, resulting from flow-induced phenomena alone?

Perhaps under these conditions, oversizing of the discharge exit nozzle could be an advantage?

[Chef]

What would the outlet nozzle flow dynamics & pressure drop be like? Could this lead to a 'blockage' situation, resulting from flow-induced phenomena alone?

Perhaps under these conditions, oversizing of the discharge exit nozzle could be an advantage?

Seems like your thinking along the same lines as I was, some fairly dynamic action and as the dynamic part slows it adds to the static pressure, but at the outlet of the plate the vena contractor may look like a restriction to flow and so a possible temp drop?

The JT effect can be caused by a blockage or flow induced pressure drops.

Chef

[desA]

Thanks very much, Chef.

The PHE flow dynamics are fairly challenging.

[Magoo]

Hi DesA.
Watched the videos, weird stuff going on inside PHEs. Hence the failure rate from internal failures from rapid and continual temp changes.
I think the low pressure drop through PHEs has alot to do with the problems. I have had that opinion for years, and the videos confirm that.
magoo

[desA]

Hi DesA.
Watched the videos, weird stuff going on inside PHEs. Hence the failure rate from internal failures from rapid and continual temp changes.
I think the low pressure drop through PHEs has alot to do with the problems. I have had that opinion for years, and the videos confirm that.
magoo

PHE's are also very thin-walled. With pressure fluctuations, the units may very well fatigue over time.

I wonder if some sort of inlet pulsation damper, or equaliser, to damp out the compressor pulsations, would help here?

[Tycho]

Nope... no pilot receiver. Assume a simple water cooled A/C system with TXV.

But if there were a pilot receiver, how could that cause the refrigerant exit temp to drop below the water inlet temp?

Had a system where the pilot valve was stuck in a semi open state and when the compressor was on low capacity the refrigerant was evaporating almost all the way back to the condenser

[Chef]

So riddle me this and riddle me that - oh thats Batman. Off Topic.

So are we closer to a real solution to this riddle?

[desA]

So are we closer to a real solution to this riddle?

Hi Chef,
I'm currently running charge trials on an equivalent system & am watching for similar behaviour, if any, to the raised exit filter-drier.

I'll report back when sensible findings occur.

The forward plan is to re-pipe the offending machine, & re-test that one, to see if a direct cause can be found. Still a work-in-progress, I'm afraid.

Either way, this comparison can be used to benefit the greater public learning.