AWHP superheat & sub-cooling

[desA]

This is a branch off Gary's Refrigeration 101 thread.

Coil outlet superheat varies with type of system. Generally speaking, a freezer should have 6-8F/3.5-4.5K superheat, a cooler should have 8-10F/4.5-5.5K superheat, and an A/C should have 12-16F/6.5-9K superheat.

What figures for evap outlet superheat & condenser sub-cooling should be used for an air-to-water heat-pump?

At what operating temperature would the refrigerant mass charge determination be most appropriate - startup Tc,sat 35'C; or hot end Tc,sat 70'C, or mid range?

I have no given mass charge to work off in this case & have to determine it in-house.

[Gary]

Cap tubes or TXV's?

[desA]

^ A TXV system.

[Gary]

Is this an existing system? Something you are designing?

[Gary]

I would charge it in cooling mode to obtain 12-16F/6.5-9K superheat with the evap entering temp at design and the evap out temp 20F/11K below design (dT can be altered by adjusting blower speed). And the subcooling at 15F/8.5K.

Then flip it over to heat mode and see if any alterations need to be made.

[desA]

I would charge it in cooling mode to obtain 12-16F/6.5-9K superheat with the evap entering temp at design and the evap out temp 20F/11K below design (dT can be altered by adjusting blower speed). And the subcooling at 15F/8.5K.

Thanks Gary.

Then flip it over to heat mode and see if any alterations need to be made.

It is an air-to-water heat pump, with the air being used as a heating source. There is no 4-way valve in this system.

[Gary]

So... this is a water cooled air conditioner?... or maybe a water cooled refrigeration system?

[Gary]

What would be the expected temp range of the air?

[desA]

Is this an existing system? Something you are designing?

It is a new machine. The design is complete & the machine is currently in the build stages.

The design was developed around the ARI standards of 11.1K evaporator superheat, 8.33K condenser sub-cooling, as the first pass.

The thing is, practically, my test rig (similar spec, but tube-in-tube condenser - not my design) seems to prefer operating with a slightly lower evap superheat & condenser sub-cooling.

My thoughts on the test rig is that is may be running as it is, due to its being a cobbled-together system I inherited from a now defunct heat-pump manufacturer - it was a suck-&-see system. The evap/compressor/condenser combinations were ill-matched in my view. The other issue was that the previous manufacturer merely loaded in refrigerant as he saw fit - until 'it looks right'. There was no science behind what went in.

My new designs are finely matched in terms of evap/compressor/condenser & I'm looking for a high COP system. To get there, I need to be able to finely-determine the refrigerant mass charge.

The condenser sub-cooling can be optmised a little via the design flow-rate, entry temp for the water stream, as this affects the HX performance.

[desA]

So... this is a water cooled air conditioner?... or maybe a water cooled refrigeration system?

This is a dedicated water heating system.

In Asia, our local ambient air temperature typically runs in the range of 25-35'C. This provides a useful heating source for heating water for home, hotels & guest-houses.

The unit currently in build can supply enough hot water for a 20 room guest-house, with a greatly reduced electrical consumption, compared to direct electrical heaters. Over here, the only viable heating source is hydro-produced electricity.

Most establishments use element-type heaters. The heat pumps can typically operate in the range of 14-25% of this electrical consumption. They have potential to be great power-savers.

What would be the expected temp range of the air?

Typically runs in the range of 25-35'C.

[Gary]

The higher temp range on both high side and low side makes for some interesting design challenges.

You say the superheat is set at 11.1K, but at what evap air in and air out temps?... and at what saturated suction temp (SST)?

[desA]

The higher temp range on both high side and low side makes for some interesting design challenges.

They certainly do. It seems to be a very fine design envelope to work in.

You say the superheat is set at 11.1K, but at what evap air in and air out temps?... and at what saturated suction temp (SST)?

The initial design selection of 11.1K for evap superheat comes from wanting to stick closely to industry standards, for reference purposes. Practically though, the evap, under a more heavily loaded refrigerant charge, at the upper end of the heating cycle, the superheat tends to reduce to something closer to around 5K. This tendency has been noted by other folks for heat-pumps.

The typical saturated suction temp (SST) range for these units is in the order of 10-17.5'C (R-134a) (start - end of heating cycle) - with air on temp around 28-33'C (measure in lab).

[Gary]

Is there some means of further subcooling the liquid before it enters the TXV? Warm incoming liquid will take up valuable coil area in flashing down to SST.

[desA]

^ Extremely valuable comment - indeed.

Thoughts:
1. Dedicated sub-coolers are something in the pipeline, based on the results of the early prototype machine. Definitely.
2. Sub-cooling of refrigerant too far closes the gap on useful air temp range before forcing defrost cycles.
3. Sub-cooling aids heat-pump COP,r.

This is a fine balancing act, in the aim for low-cost machines for the region.

[Gary]

The initial design selection of 11.1K for evap superheat comes from wanting to stick closely to industry standards, for reference purposes. Practically though, the evap, under a more heavily loaded refrigerant charge, at the upper end of the heating cycle, the superheat tends to reduce to something closer to around 5K. This tendency has been noted by other folks for heat-pumps.

Given the high pressure pushing the liquid through, I would expect the TXV to have a strong tendency to hunt at the upper end of the heating cycle?

[desA]

Given the high pressure pushing the liquid through, I would expect the TXV to have a strong tendency to hunt at the upper end of the heating cycle?

You are correct here. I've selected the TXV fairly tightly in terms of the compressor/TXV load curves.

Another driver for TXV disturbances is that the mass balances for the evap/condenser are different at start of cycle & end of cycle, from a thermodynamic & heat-exchanger design point of view. The system tries to re-establish this balance as the heating cycle progresses.

I use simple bulb-driven TXV's to smooth out any hunting as far as possible (slow, lazy response) & so what comes over are slow, low-amplitude, smooth waves towards the upper end of the heating cycle, rather than short, choppy swings. My test heat-up time is aggressive at 1.75h & so this tends to show up these issues fairly early.

At the top end of the heating cycle, the compressor is close to its upper performance envelope.

[Gary]

I'm thinking a compressor designed for R404A might be more appropriate than one designed for R134A. It doesn't know what it is pumping, it just reacts to the pressures.

[desA]

^ What's the critical pressure/temp for R404A?

For instance R-134a is ~101'C. The distance between the condenser Tc,sat & the critical temp (top of vapour dome), is what has me using R-134a at the moment. The present design heats water up to 65-70'C.

I'm very open to looking at other low-impact refrigerants which could perhaps provide more compact systems at lower pressure. A very good point - thank you.

[Gary]

^ Extremely valuable comment - indeed.

Thoughts:
1. Dedicated sub-coolers are something in the pipeline, based on the results of the early prototype machine. Definitely.
2. Sub-cooling of refrigerant too far closes the gap on useful air temp range before forcing defrost cycles.
3. Sub-cooling aids heat-pump COP,r.

This is a fine balancing act, in the aim for low-cost machines for the region.

Keeping in mind that design is not my forte, here is how I would handle it:

I would install a vertical suction/liquid heat exchanger (HX).

The suction leaving the evap would enter the bottom of the HX and exit the top. The TXV bulb would be mounted at the exit.

I would install a pressure reducing valve (PRV) in the liquid line to reduce the liquid to a pressure that is more appropriate to R134A, say about 100-110psi.

This reduced pressure liquid would enter the top of the HX wher it would be heavily subcooled and exit the bottom, then on to the TXV.

[desA]

^ That's very cunning.

So in essence:
1. Superheat the evap outlet vapour;
2. TXV controls on evap outlet temp;
3. Sub-cool the condenser exit liquid;
4. PRV reduces liquid pressure, before entry to TXV.

Basically an inter-cooler strapped across suction & liquid lines, plus pressure control pre-TXV.
With the HX, I could then reduce the evap footprint = good!
With the PRV, the TXV will control better.

Now, we're cooking...

May need to check on TXV at HX discharge - something I've read is niggling here - may need to use an electronic expansion valve due to an inverse relationship (homework for me).

[Gary]

^ What's the critical pressure/temp for R404A?

For instance R-134a is ~101'C. The distance between the condenser Tc,sat & the critical temp (top of vapour dome), is what has me using R-134a at the moment. The present design heats water up to 65-70'C.

I'm very open to looking at other low-impact refrigerants which could perhaps provide more compact systems at lower pressure. A very good point - thank you.

I would continue to use R134A, but in a compressor that is designed and rated for R404A.

[Gary]

1. Superheat the evap outlet vapour;
2. TXV controls on evap outlet temp;

No...

1. Superheat the HX outlet vapor.
2. TXV controls on HX outlet temp.

Superheat at compressor inlet remains as is.

[desA]

I would continue to use R134A, but in a compressor that is designed and rated for R404A.

Can you elaborate further on this idea?

[Gary]

4. PRV reduces liquid pressure, before entry to TXV.

When the pressure is reduced the liquid will want to flash. That's why it needs to then enter the HX to bring the temp down and return it to subcooled liquid state.

In addition to everything else, the HX, being vertical and bottom fed, will act as an accumulator in the off cycle.

[desA]

No...

1. Superheat the HX outlet vapor.
2. TXV controls on HX outlet temp.

Superheat at compressor inlet remains as is.

Ok... that's fine. (My head was thinking of HX outlet monitoring, but my fingers typed evap).

The compressor inlet is the one that has to be held - agreed.

[Gary]

Can you elaborate further on this idea?

Where a compressor that is designed for R134A under such high pressures might be close to kicking out on it's internal overload, a compressor that is designed for R404A would see these pressures as normal.

[desA]

^ That's a very interesting point. I'll look into it & feed back.

I've tested the current compressor range beyond its normal envelope conditions without the internal kick-out activating. This is done by gradually increasing the Hi/Lo pressure cutout switch settings. These compressors are usually set around 10-15% higher for the internal cut-out, than the current operating range I'm working in. So there is still some margin left.

[desA]

When the pressure is reduced the liquid will want to flash. That's why it needs to then enter the HX to bring the temp down and return it to subcooled liquid state.

In addition to everything else, the HX, being vertical and bottom fed, will act as an accumulator in the off cycle.

This idea is excellent. I like the pressure control & accumulator aspects. Thanks so much for your wisdom.

[Gary]

^ That's a very interesting point. I'll look into it & feed back.

I've tested the current compressor range beyond its normal envelope conditions without the internal kick-out activating. This is done by gradually increasing the Hi/Lo pressure cutout switch settings. These compressors are usually set around 10-15% higher for the internal cut-out, than the current operating range I'm working in. So there is still some margin left.

Given improvements in heat transfer in the evap (better TXV control, lower superheat, minimal flashing, maximized fan speed, etc.) the SST should rise and that margin could quickly disappear. You may need a compressor with a higher operating (pressure) range.

[desA]

^ Thanks Gary.

Is this where you'd be looking for a R404A compressor, but pumping R-134a?

The upper safe envelope limit for the current R-134a compressor is given as Tc,sat of 75'C, at Te,sat of 15'C - top right point.

This operating point is always a major top end issue to a standard heat-pump compressor & it would be nice to be able to run up a bit higher without problems. The discharge line temps & compressor base temps are currently well within the manufacturer's limits listed in their technical spec sheets - it's that upper temp limit of 75'C that concerns me.

I've taken precautions on the positioning of the compressor to keep the casing reasonably cool - looking to push the envelope margin a little, if necessary.

[Gary]

^ Thanks Gary.

Is this where you'd be looking for a R404A compressor, but pumping R-134a?

Yes.

The upper safe envelope limit for the current R-134a compressor is given as Tc,sat of 75'C, at Te,sat of 15'C - top right point.

This operating point is always a major top end issue to a standard heat-pump compressor & it would be nice to be able to run up a bit higher without problems. The discharge line temps & compressor base temps are currently well within the manufacturer's limits listed in their technical spec sheets - it's that upper temp limit of 75'C that concerns me.

I've taken precautions on the positioning of the compressor to keep the casing reasonably cool - looking to push the envelope margin a little, if necessary.

Improvements in evap efficiency should have little to no effect on the high side limits (this is primarily a matter of water temp and flow volume) and the lower more stable superheat should in fact give you a lower discharge temp.

However, improvements in evap efficiency should result in higher refrigerant mass flow which equates to higher amp draw. It is the amperage limit and consequent tripping of the internal overload which we will be running up against.

[Gary]

Te,sat of 15'C...

This is the limit that concerns me.

[Gary]

The typical saturated suction temp (SST) range for these units is in the order of 10-17.5'C (R-134a) (start - end of heating cycle) - with air on temp around 28-33'C (measure in lab).

And apparently you are already exceeding the Te,sat 15C (SST) limit at the end of the cycle.

[desA]

And apparently you are already exceeding the Te,sat 15C (SST) limit at the end of the cycle.

True & it does concern me. These results are from the test machine in my laboratory.

I originally considered the main reasons for this Te,sat value to be due to:
1. The evap is too large for the heat-pump service (by some 60% according to my calculations);
2. Local ambient temps are very high, pulling the Te,sat (SST) up towards the end of cycle;
3. Current refrigerant mass charge in test rig is not yet at optimal value. (With lower charge values Te,sat is reduced slightly);
4. The lab machine evap/compressor/condenser/piping are mis-matched (quite a large discrepancy, according to my calculations).

As I understand - from the compressor manufacturer's tech sheets - the right line of the compressor envelope may be crossed if compressor shell cooling is implemented.

For the record, the supplier of my lab machine, had supplied a number of units abroad, into hot climates, with some measure of success. (It was his business model that failed, with unscrupulous partners, rather than poor product.)

[desA]

These Asian-condition AWHP machines are a very fine design balance, at the outer limits of the compressor envelope. Of that, there is no doubt.

What some folks will do, is to limit the achievable hot-water upper temperature limit. This will move the Te,sat (SST) & Tc,sat values to just within the published compressor envelope.

In this game, oil control is a must. Pipes are carefully sized & laid to ensure correct oil return.

[Magoo]

Hi desA
Interesting post/topic, with evap superheat testing, setting the TEV is critical for system performance. Start by reading the air on temp., and the actual evap pressure converted to temp., this is system TD. Then read suction temp at TEV bulb versus the evap pressure/temp. The superheat of TEV should be 60 > 70 % of system TD. Can be set up during pull down or at design, any adjustments to TEV wait 15 minute for TEV to stabilize. Doing this method of checking gets rid of all the "rule of thumb " ideas
I would stick with R134a, but with high suction temps would consider a TEV injecting liquid before an accumulator to cool suction gas entry to compressor, reduces motor winding temps, and discharge temps.. to a controllable level. High discharge temps kill oil quality, and even consider fitting a head cooling fan and oil cooler coil above heads as well.

magoo

[desA]

Thanks so much Magoo for your post.

... with evap superheat testing, setting the TEV is critical for system performance. Start by reading the air on temp., and the actual evap pressure converted to temp., this is system TD. Then read suction temp at TEV bulb versus the evap pressure/temp. The superheat of TEV should be 60 > 70 % of system TD.

For example:
Air on temp (Ta,in) = 28.1'C
Evap sat temp (Te,sat) = 14.0'C
TD = Ta,in - Te,sat = 28.1 - 14.0 = 14.1'C
0.6*TD = 0.6*14.1=8.46'C
0.7*TD = 0.7*14.1=9.87'C

So, acceptable superheat in range 8.5-9.9'C.

Can be set up during pull down or at design, any adjustments to TEV wait 15 minute for TEV to stabilize. Doing this method of checking gets rid of all the "rule of thumb " ideas

At which operating point in the heat-up cycle would you optimise the superheat settings?

For example, for a hot-water heat-pump, the vapour compression cycle moves from a condenser Tc,sat of around 35'C up to 70-75'C, for instance. During this excursion, the whole vapour compression cycle has to continually re-position itself - everything is in a state of dynamic instability & has to continually re-adjust itself to track the hot water temperature.

Where in this range would be the best place to optimise the system settings, charge determination & so forth?

I would stick with R134a, but with high suction temps would consider a TEV injecting liquid before an accumulator to cool suction gas entry to compressor, reduces motor winding temps, and discharge temps.. to a controllable level. High discharge temps kill oil quality, and even consider fitting a head cooling fan and oil cooler coil above heads as well.

Excellent points - well taken. Thank you.

At the moment, in a high-load test, the base of the compressor casing runs at 45.3'C, in the test rig - this has not gone above 55'C - well under the compressor manufacturer's guidelines for safe operation. The new machine has the compressor temps well managed.

Currently, the maximum discharge line temps are well under the allowable temp limit suggested by the manufacturer - they are being very closely monitored.

[Gary]

Since the evap is not in fact being used for comfort cooling, I see no reason the compressor cannot be mounted in the path of the evap leaving air.

[desA]

^ Touche...

[desA]

What are the pro's and con's of introducing a dedicated de-superheater & sub-cooler into the system, instead of trying to do sub-cooling, condensing, sub-cooling in one single condenser unit?

[Gary]

If the compressor is close coupled (short suction line) there is no need for de-superheating as the compressor inlet superheat will be the same as the evap outlet superheat. De-superheating is for long suction lines.

[Gary]

Dedicated subcooling assumes an external source of cooling for the liquid. What exactly did you have in mind?

[Gary]

A possible source of cooling for the liquid would be the evap leaving air. A spiral of copper tubing in the path of the cool air could act as a subcooler and if sized properly could provide just enough restriction to drop the liquid pressure, thus eliminating the PRV.

[desA]

Dedicated subcooling assumes an external source of cooling for the liquid. What exactly did you have in mind?

Use the incoming water to first sub-cool in the sub-cooler, before passing into the condenser.

The thought process here is whether a stand-alone liquid sub-cooler/water pre-heater is more efficient/cost-effective than the sub-cooler integrated as part of the condenser itself.

Essentially, this same logic applies to a dedicated de-superheater.

So, in essence, which option is more efficient/cost-effective:
1. Integrated de-superheater/condenser/sub-cooler (single HX);
2. Separate de-superheater + condenser + sub-cooler (3 HX's)

[desA]

A possible source of cooling for the liquid would be the evap leaving air. A spiral of copper tubing in the path of the cool air could act as a subcooler and if sized properly could provide just enough restriction to drop the liquid pressure, thus eliminating the PRV.

That's brilliant! I love that idea...

Now, that has got to be looked into. Using a air 'waste stream' to do further duty. You've made my day. Thank you so much.

[Gary]

Use the incoming water to first sub-cool in the sub-cooler, before passing into the condenser.

What temperature is the incoming water?

[desA]

What temperature is the incoming water?

This can be tuned by adjusting the water flowrate to suit the system.

Water-heating systems trade water temp rise & flow-rate in using the condenser heat output. The flowrate selected will determine the number of water passes through the heat-pump (heating cycles).

In other words, on the one extreme, (1) you have water moving from 20'C to 65'C in one pass through the heat-pump i.e. low volume flowrate, high dT,water.

The other extreme is (2) water coming in at say 60'C & leaving at say 65'C, at high flowrate.

A sub-cooler would be useful in case (1) & not much use in case (2), for instance.

So, to answer your question, with a Tc,sat ~ 70'C, & typical sub-cooling of 8.33K, Tc,sub ~ 61.67'C. Allowing for an entry approach of 5K (good HX), this means that the entry water could be around 56.67'C, or slightly higher - even up to 60'C with an excellent HX.

If the water flowrate is now reduced, allowing for less passes through the condenser (less heating cycles), then a lower inlet temp can be selected. Say we use Tw,in=45-50'C for example, then we can further sub-cool the refrigerant by using the incoming water stream.

Heating circuits are a balancing act & this can be used to advantage.

[Gary]

Are the condensers counterflow or crossflow?

[Gary]

Water-heating systems trade water temp rise & flow-rate in using the condenser heat output.

This is not an even trade. Lower flowrate will give you higher temp water, but far less heat transfer. It is a mistake.

[desA]

Are the condensers counterflow or crossflow?

For most of these AWHP's, the condensers are either tube-in-tube coils, or plate heat-exchangers. Both can be treated in essence as counterflow units.

For a condensing phase unit, the cross-flow correction factor is taken as unity, in the phase-change region. It is only when you get to the sensible heat-transfer regions (de-superheating & sub-cooling), that the effect of cross-flow, counter-flow, or parallel flow, is noticed.

In general, the HX's are taken as counterflow.