Direct expansion evaporator liquid volume

[SoddemFX]

Hi guys,

In a direct expansion evaporator with healthy outlet superheat (6K - 10K), what is the percentage of evaporator volume occupied by liquid?

I've seen a few different views ranging from 25% to 66%, i would expect it to be closer to the 66%

I've used the "Refrigerant charge calculator" tool available here:

http://www.defra.gov.uk/environment/climatechange/uk/fgas/index.htm

To predict the a systems total charge (including evaporator) for a r404a reciprocating compressor system with a reciever, plate heat exchanger and 400W evaporating capacity at -40C. The excel macros predicted high and low refrigerant masses for the system (0.4Kg to 0.5Kg) appear ball park right.

In itself this tool doesn't answer the question i want to know and i think it's making generalised assumptions on evaporator volume based on system evaporating capacity and evaporator type.

I would like to know the typical percentage of evaporator (type) volume occupied by liquid, understand how this figure is reached and what factors affect it.

Many thanks in advance :)

Tom

[Freddie123]

I think I remember Chilly1 used to say he tried to evap around 30% before it hit the base which would lead to your 66% theory but i'm not sure what he based this on. I'm sure DetroitAC may be able to help here, he seems the knowledgable

[Unknown_road]

depends on the evap 's volume and how you charge it ;)

theoretically it would be possibele to have only a few % liquid. If all the liquid out of the cap tube almost instantly evaporates but is still able to cope with the heat load.

[SoddemFX]

Hi Jurgen,

Thanks for the reply :)

I realise this but i want to be able to know the percentage of internal volume occupied by liquid which will give 6 - 10K superheat on the outlet.

I realise i can decrease the amount of liquid in the evaporator resulting in higher superheat at the outlet but generally lower capacity.

For a specific reference in am thinking of plate interstage heat exchangers as the evaporator, in this case the internal volume of the "evaporator" is approx 100ml.

Hope this makes the question clearer although i'm probably still missing a few things out...?

Tom

[[XC] 2long4u]

Tom you should make a clear evap.

[Unknown_road]

Oh, I forgot the superheat part :D sorry I can't the answer that question.

[DetroitAC]

That's a tough question Tom,

It's really not an analytical problem, it's one of those physical phenomenon that is far easier to quantify experimentally.

The problem is that some liquid sort of "hides out" in the evap.

Since you're talking about a plate/plate I'm quite qualified to say this since I've worked on a lot of plate/fin evaps. I've built evaps with clear end tanks to see where the liquid was going, where the liquid was pooling and just sitting there. I can tell you that these things cannot be predicted on paper, or in excel, or in CFD for that matter. Given enough time and resources a skilled CFD jockey might create a model that matches reality (they love doing THAT!), but will not work for any other situations.

If not for the pooling of liquid in evaps there would be only a very tiny volume occupied, since there is so much vapor by volume flowing and so little liquid, but again the liquid that is "hiding out" throws all of that theory out the window.

I would think copper tube evaps would have less pooling since they don't have nooks and crannies and dead spots, but I don't know...

AFAIK charge is always determined experimentally. You could ballpark it like you've done, I've done a bit of that years ago to understand how much mass exists in each component to try to reduce the charge levels. At the end of the day you're going to have real hardware anyway, so usually you just wait until then to get the charge number.

You must have a good reason for wanting to do this, maybe you could explain the problem further?

-Erik

[SoddemFX]

Hi Erik,

Thanks for the reply.

I'd hoped it might be relatively straight forward, wishfull thinking i guess :(

The goal of the question was to roughly size expansion tanks needed for a cascaded system.

For parts of the system where you know only gas or liquid are present it's relatively simple to work out the mass of refrigerant contained from design temperatures and pressures.

I think the condensing interstage heat exchanger will only contain a small amount of liquid, possibly < 10% of it's volume since the majority of the refrigerant will migrate to the evaporating interstage heat exchanger - would you agree or disagree?

The evaporator liquid content has been the thing that's been bugging me for the past few days because the liquid content could potentially be very large in order to avoid a high suction superheat, which would be pretty catastrophic in this case.

If i bring the system to absoloute vacuum and then fill the evacuated system with a known volume at a known pressure of nitrogen (i.e. 2L at 400PSI), from the resulting pressure i can determine the system volume.

If i have a figure for the mass of refrigerant needed and set myself a target static pressure (maybe 150PSI) i can convert this refrigerant mass into a required volume in gas state at +30C. I then subtract the system volume from the required volume and i have the volume of expansion tanks needed.

Thats the idea anyway :)

Tom

[Pete]

Tom you should make a clear evap.

Someone did didn't they...some odd shaped thing on a 3/4 stage cascade! Think it was made from glass. The Aura project i think it was

[DetroitAC]

Ah, what you're doing makes sense to me now.

I think the condensing interstage heat exchanger will only contain a small amount of liquid, possibly < 10% of it's volume since the majority of the refrigerant will migrate to the evaporating interstage heat exchanger - would you agree or disagree?

I don't get what you're saying with this part. Are you talking about the system being off? When you say "condensing interstage" I picture the condenser of the coldest stage of a cascade. When you say "evaporating interstage" I picture the evaporator of the warmest stage of a cascade. These are two separate loops, so I'm confused (obviously) when you say it'll migrate.

[[XC] 2long4u]

Someone did didn't they...some odd shaped thing on a 3/4 stage cascade! Think it was made from glass. The Aura project i think it was

No linky?

This is the only thing I could find. http://www.p-a-hilton.co.uk/R633_Edition.2_GREY.pdf

[SoddemFX]

Ah, what you're doing makes sense to me now.


I don't get what you're saying with this part. Are you talking about the system being off? When you say "condensing interstage" I picture the condenser of the coldest stage of a cascade. When you say "evaporating interstage" I picture the evaporator of the warmest stage of a cascade. These are two separate loops, so I'm confused (obviously) when you say it'll migrate.

As an example the second stage of a three stage cascaded system.

The refrigerant condenses in the stage 1/2 interstage heat exchanger and evaporates into the stage 2/3 interstage heat exchanger.

What i mean is that when the metering device working well, the volume of liquid refrigerant held in the condensor (stage 1/2 interstage heat exchanger) will be small, much smaller than the volume of liquid refrigerant contained in the evaporator (stage 2/3 interstage heat exchanger)....?

Migrate was the wrong word to use, the refrigerant is just "there" :D

Tom

[DetroitAC]

As an example the second stage of a three stage cascaded system.

The refrigerant condenses in the stage 1/2 interstage heat exchanger and evaporates into the stage 2/3 interstage heat exchanger.

What i mean is that when the metering device working well, the volume of liquid refrigerant held in the condensor (stage 1/2 interstage heat exchanger) will be small, much smaller than the volume of liquid refrigerant contained in the evaporator (stage 2/3 interstage heat exchanger)....?

Migrate was the wrong word to use, the refrigerant is just "there" :D

Tom
I'm not trying to be difficult or argumentative, but I think the opposite. :D

In automotive systems, the majority of the entire system charge (apart from the receiver) resides in the headers of the condenser. I think a plate/plate condenser would hold a lot of charge.

I suppose the amount that the evap holds would be greatly influenced by how it's mounted and circuited. If it were not possible for the liquid to pool, I think the charge could be quite small (possibly at the expense of performance). If the evap were circuited such that inlet is at the bottom, and the flow goes upwards and through many passes, I think a much larger amount would be held up by the evap.

I don't know how much I'm helping, it's really a bunch of arm waving at this point...

-Erik

[SoddemFX]

No, not at all, it's great to have your experience :yepp:

If the condensing plate heat exchanger was the same type, orientation and volume (basically identical) to the evaporating plate heat exchanger, do you think the majority of the charge would still be held in the condensing plate heat exchanger? Why?

What do you use in your work do you determine the volume occupied by liquid in a plate heat exchanger?

I realise that the answer to my original question is sounding like "try it and see" but if it's ok with you (?) i have a lot more associated random cr*p in my head i want to ask you or anyone else who can answer :)

Thanks Erik,

Tom

[DetroitAC]

If the condensing plate heat exchanger was the same type, orientation and volume (basically identical) to the evaporating plate heat exchanger, do you think the majority of the charge would still be held in the condensing plate heat exchanger? Why?

I would run a plate/plate condenser with inlet at the top, and passes serpentine towards the bottom, or multiple plate passes serpentine towards the bottom. I would circuit an evap the opposite, flowing bottom to top, except if the plate passages were wide, and I was worried about oil return.

On the other hand in the case where I wanted to minimize the total system charge, I think I'd circuit it the same as the condenser, flowing downwards.

If it is a condenser flowing downwards, it is going to be easy for the liquid to pool in the bottom few plates, and the sides of the plates. I picture condensers in my head as having a high velocity vapor inlet and no liquid flow out, the liquid is just such a trickle compared to the vapor.

If an evap were circuited that way, downwards, the increasing velocity of the vapor would tend to wash the remaining liquid over the plate surfaces, preventing pooling. I would think also at the expense of performance, since it'd be easy to dry out sections of plate. Now that I think of it, flowing downwards with an evap, I'd circuit the plates with a single pass serpentine downwards to keep from missing any plates.

What do you use in your work do you determine the volume occupied by liquid in a plate heat exchanger?

For a rough number, I just estimate volumes and add them up. Like a plate interior is 48mm x 230mm x 1.5mm x 0.95 for lost volume to beads and corner radiuses x number of plates. Then separate estimates for tanks or cups or whatever.

Refrigeration oil, and a scale works somewhat, air bubbles are a giant pain and it takes forever to be satisfied you've got it with a complex exchanger.

Hot liquid refrigerant into a room temperature capped heat exchanger works much better, but there is the setup time of building fittings to cap the ports with an access valve. I use a heated charging cylinder, heat the refrigerant to about 30C, charge it into the capped evacuated exchanger. Since the heat exchanger is room temp (25C), the refrigerant is going to both flow in (because of thermosyphon) and contract once it's in since it will be cooled off (otherwise it makes a boom). oh, forgot use a scale with that one too.