View Full Version : Compact AWHP charge determination
This is a parallel thread to "AWHP superheat & sub-cooling", in order to allow logical progression to continue in that thread, without creating confusion with a new case.
I've started it in order to investigate the correct charge determination for a compact air-to-water heat-pump.
This system essentially follows the format as that for the previous test AWHP, with a few design optimisations, as the main modifications.
The condenser is also of a different family to the tube-in-tube condenser used in the "AWHP superheat & sub-cooling" thread.
http://www.imagechicken.com/uploads/1250074231075327000.png (http://www.imagechicken.com)
Experimental results after 2nd day's set-up. Refrigerant R-134a.
1st run - set starting charge mass - based on calculated prediction, set to 'hot charge condition' (minimum mass).
2nd run - added refrigerant - now up to mid charge between 'hot charge condition' & 'startup condition' (maximum mass).
The computed range between min & max charge is 220g.
Total compressor run time at Run 2, approximately 5 h.
TXV still at factory setting.
Discharge line insulation to be applied before next run.
I have an idea of how I'd like to proceed on this unit, for the next tuning point, but would really value the applied wisdom of forum members.
PHE condenser?
Smaller evaporator?
Same compressor?
The condenser is working very well (5.32K approach), which is not too surprising given the light load from the evaporator.
The evaporator is working not so well (20.76K approach). (VIC to the rescue?)
The condenser outlet subcooling is 3.05K, while the TXV inlet subcooling is 5.62K. This is sufficient to feed the evap. Note that the cooling of the liquid line has added just enough subcooling to properly feed the evap. Is the liquid line insulated?
I would add a little more refrigerant to bring it up to 5.5K at the condenser outlet to see what effect this has on the condenser.
I would leave the TXV setting as is for now and focus on improving the evap appr.
1st run - set starting charge mass - based on calculated prediction, set to 'hot charge condition' (minimum mass).
2nd run - added refrigerant - now up to mid charge between 'hot charge condition' & 'startup condition' (maximum mass).
The computed range between min & max charge is 220g.
I take it the initial charge was insufficient to feed the evap at the higher temp conditions?
I take it the initial charge was insufficient to feed the evap at the higher temp conditions?
Yes, as I read the situation at the moment.
The initial charge was set to the calculated minimum system requirement, based on the 'hot system' operating condition.
PHE condenser?
Touche'. :)
Smaller evaporator?
Yes, different supplier - compact, upfeed distributor, not down-feed as in previous test-rig.
Same compressor?
Same manufacturer - one model size larger - scroll.
The condenser is working very well (5.32K approach), which is not too surprising given the light load from the evaporator.
This is the one selection I had under tight design & selection control. I'm pretty pleased it's running as per design.
The condenser outlet subcooling is 3.05K, while the TXV inlet subcooling is 5.62K. This is sufficient to feed the evap.
Good, then we are on the way there.
Note that the cooling of the liquid line has added just enough subcooling to properly feed the evap. Is the liquid line insulated?
Liquid line un-insulated. :)
I would add a little more refrigerant to bring it up to 5.5K at the condenser outlet to see what effect this has on the condenser.
Great. I agree with this. I'll add the charge this morning. I thought that this may be the case. Thanks so much.
I would leave the TXV setting as is for now
Agreed on the TXV setting.
and focus on improving the evap appr.
In your view, what could be the reasons for the evap problems? Under-size? Under-performance e.g. low heat-transfer (if so, I can get this licked pretty easily).
The evaporator has an inlet filter, which will be creating some restriction as well. I can pull it off to check performance change - after initial setup.
There are also a few casing air leaks, which may be causing pirate air issues & evap bypass. I'll get these blocked up. The fan may be marginal here, as well. I'll look into this.
The evaporator is working not so well (20.76K approach). (VIC to the rescue?)
I'm going to have to work on this side. I'll bring the machine up to my lab & test run it from every which way. To see where it is all going. The next machine development will kick off after these tests - this may very well be the time to introduce VIC. Good thought. :D
Important note:
The air outlet temp you will be using to calculate your approach, will include heating loss from the compressor, for this machine!!! :eek:
I will have a new probe inserted just behind the evap itself today, so that we can measure the true approach off the evap.
Thanks so much, Gary, for your stunning analysis. I am very grateful.
Yes, as I read the situation at the moment.
You read the situation correctly.
In your view, what could be the reasons for the evap problems? Under-size? Under-performance e.g. low heat-transfer (if so, I can get this licked pretty easily).
I think it comes down to evap surface area. The airflow is not at all marginal. That said, an increase in airflow will increase heat transfer. As usual, there will be a point of diminishing returns.
The evaporator has an inlet filter, which will be creating some restriction as well. I can pull it off to check performance change - after initial setup.
Should be interesting to see how much difference it makes.
We will only be able to continue the testing tomorrow, for various logistical reasons. Various components were insulated to conserve heat within compressor/discharge-line section.
One good thing to report is that the fan speed control, via compressor discharge temp strategy, will be implemented tomorrow - if all goes to plan.
I have already determined the initial control function profile & hope to report back on initial trial either tomorrow, or Saturday.
This will give us a real feel for the system response to this strategy. We can then fine-tune the control curve during the coming weeks & I'll implement the same unit on the lab test machine as well.
You read the situation correctly.
Thanks, Gary.
A large proportion of this experience can be attributed to your very kind tutoring on the lab test machine. Your methodology makes absolute sense & allows a logical appraisal of the system operation to be performed.
:)
Originally Posted by desA
In your view, what could be the reasons for the evap problems? Under-size? Under-performance e.g. low heat-transfer (if so, I can get this licked pretty easily).
Gary:
I think it comes down to evap surface area. The airflow is not at all marginal. That said, an increase in airflow will increase heat transfer. As usual, there will be a point of diminishing returns.
I measured the air inlet velocity to the filter - it is as per design. So, this is a good sign. We have selected air-speed carefully, to not be too large.
I also added a thermocouple directly behind the evap - to accurately measure the evap exit approach temp. We should have some experimental details tomorrow.
Originally Posted by desA
The evaporator has an inlet filter, which will be creating some restriction as well. I can pull it off to check performance change - after initial setup.
Gary:
Should be interesting to see how much difference it makes.
I'll let the system settle in tomorrow - onto the fan control strategy & perhaps remove the filter at a later point, once the system is settled. With the increased refrigerant mass charge, I'm expecting slightly more load on the evap - this should also close the approach a little more.
I have no practical experience with PHE's, but given the small micro-channels I would be wary of them possibly plugging more easily on the water side, thus requiring frequent maintenance. But I could be wrong on this.
Perhaps others have some hands on experience to share.
We may be able to gain a little more performance from this evap, but I'm thinking the gains will be limited. It is a smaller evap coupled with a larger compressor. I'm hoping the VIC will close the gap.
Any progress on the water regulating valve?
I have no practical experience with PHE's, but given the small micro-channels I would be wary of them possibly plugging more easily on the water side, thus requiring frequent maintenance. But I could be wrong on this.
Perhaps others have some hands on experience to share.
Fair comment.
There are four ways to manage this fouling challenge:
1. Reverse flush through unit from time-to-time;
2. Inlet filter/strainer;
3. Secondary loop;
4. Chemical cleaning when fouling builds up.
We may be able to gain a little more performance from this evap, but I'm thinking the gains will be limited. It is a smaller evap coupled with a larger compressor. I'm hoping the VIC will close the gap.
Good logic. Let's see just how far we can push this particular machine in terms of evap. We have a little more room to play, once the pirate air-leaks are closed & dust filter modified.
For the next machine (production level), I will watch this evaporator like a hawk & get involved with the designer in pushing the evap design limit. I have a few tricks up my sleeve here that will boost the performance limit up some 15-20% odd. This is my core expertise... ;)
I'll begin specifying the VIC for this unit, next week, on my return to my lab. We can then chat through the specification & refine it until it makes sense.
Will this be a PHE, or tube-in-tube type, as you envisage it?
Any progress on the water regulating valve?
I looked into this item, but, so far, my main component supplier has come up short.
I'll get onto my other suppliers early next week - I'll be able to find something. Actually, I'll chat to my heat-pump builder tomorrow - he'll know where to source it, I'm sure.
I looked into this item, but, so far, my main component supplier has come up short.
This is a very common component used to control the flow on water cooled condensers everywhere and they have been around forever. I'm surprised that your supplier would have any problem at all in finding them.
Will this be a PHE, or tube-in-tube type, as you envisage it?
I would go with tube in tube. The PHE would offer very little advantage here.
A further point to factor in on this evap:
The inlet air is now ~ 35'C upwards. If I take the previous (well oversize) evap figures from my lab machine & offset these upwards by the additional air temp difference - with the same dT,air across the evap, the exit approach works out to be in the region of around 12-15K.
In the current design, I have no need for cold off-air temps, as this will just dump excessive moisture into the drip tray. Our relative humidity over here is around 64% at 35'C. The system will fill up a 25L bucket in a few hours of run-time.
The designers watch very carefully the moisture blinding in the fins, as well as inlet air velocity - it's a real issue. So for our system, with a Te,sat of in the region of 12-13'C, I'd not be overly unhappy seeing an exit approach of 21-22K, for instance. We are under this.
I'd value your thoughts on this.
To minimize/avoid condensation you will need a delta-T of about 8K or less. This is primarily a matter of airflow, not approach. We can deal with this after we have maximized performance in other areas.
That said, when we reduce airflow to reduce load (fan control) this will increase the possibility of condensation.
Was there much condensation during these test runs?
In the second run the wetbulb was about 25C with the leaving air well above this. Unless the coil has a very high bypass factor, there should have been no condensation.
To minimize/avoid condensation you will need a delta-T of about 8K or less. This is primarily a matter of airflow, not approach. We can deal with this after we have maximized performance in other areas.
Very fair comment. So, the upper dT,air limit is around 8K. Thanks for this.
The airflow also has an upper velocity limit in terms of noise & moisture droplet entrainment.
That said, when we reduce airflow to reduce load (fan control) this will increase the possibility of condensation.
Condensation control is a real issue over here. Plain fins often need to be coated in order to allow the condensate to run off (well wetted), rather than hang in place, then blow off as droplets.
Large drip-trays are essential.
Originally Posted by desA
I looked into this item, but, so far, my main component supplier has come up short.
This is a very common component used to control the flow on water cooled condensers everywhere and they have been around forever. I'm surprised that your supplier would have any problem at all in finding them.
Sometimes it's a language issue over here in SE Asia. I'll re-word carefully & work through their product catalogs. I'm sure the 'ahah' moment will come. :)
Originally Posted by desA
Will this be a PHE, or tube-in-tube type, as you envisage it?
I would go with tube in tube. The PHE would offer very little advantage here.
Ok - this makes it very cost-effective & pretty easy to install. Let me work on this a little more.
Was there much condensation during these test runs?
In the second run the wetbulb was about 25C with the leaving air well above this. Unless the coil has a very high bypass factor, there should have been no condensation.
Actually, now that you mention it - for this particular unit, the condensation is indeed very much lower than the lab unit. The folks have a small drip bucket, which has not overflowed to my knowledge. I'll check this more carefully during tomorrow's runs.
Actually, now that you mention it - for this particular unit, the condensation is indeed very much lower than the lab unit. The folks have a small drip bucket, which has not overflowed to my knowledge. I'll check this more carefully during tomorrow's runs.
Given the low dT on the lab unit, I am surprised that it has condensate problems. Possibly the coil has a high bypass factor... or perhaps a higher dT in the early part of the run... or maybe just an unusually humid day.
Seems I recall there being a formula for calculating mean coil surface temp and this may be more definitive. If the surface temp is above dewpoint, then there should be no condensation.
Given the low dT on the lab unit, I am surprised that it has condensate problems. Possibly the coil has a high bypass factor... or perhaps a higher dT in the early part of the run... or maybe just an unusually humid day.
The lab is situated around 2km from one of the largest rivers in the region - the Mee Nam Khong (MeeKong). You can imagine the RH% up there. I'll buy in a RH meter & test local humidity, for future runs.
Seems I recall there being a formula for calculating mean coil surface temp and this may be more definitive. If the surface temp is above dewpoint, then there should be no condensation.
I'll look into this. Practically, though, with a Te,sat of around 10-13'C, the coil tubes will see condensation - you can see this if you look directly into the coil, from the inlet side.
I'll feed back on the latest charge settings, when I've worked through today's information.
The rest of the ideas, that essentially flowed from the previous lab-test machine, will be continued over there - so that we can keep everything in its place.
I've told my builder to box up the current machine & ship it up to my lab so that I can run my own tests in peace & quiet.
Powered by vBulletin® Version 4.2.1 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.