View Full Version : Refrigerant to water heat exchangers

17-02-2005, 01:17 AM
Hi. Now at the outset I'll admit I know very little. Apart from a short course and a Safe-Handling cert, there's not much stands between me and total ignorance. I have a little courage (stupidity), that permits me to jump into a project with both feet, and sometimes land on them! My interest in A/C is accidental. I am working on a project to do my bit to reduce the amount of energy I use, which means an integrated system of solar/wind/woodgas etc etc. At the start it all seemed so easy. In reality it is like an onion: Peel off one layer and guess what's underneath? Anyway, to make it more difficult the whole thing has to be done on a shoestring, preferably using recycled stuff from the scrapyard. The A/C comes in with the ground-coupled heat recovery system. Initially I believed this could be done with old fridges, as I convinced myself that the back of a fridge pushed out a lot more heat than the electricity that runs the motor could provide. Ok, I've now come across the idea of COP and realise that A/C units are better than old fridges, but they are also dear, particularly if the COP is high. So the inevitible compromise is to use compressors from discarded but servicable A/C units or large fridges/freezers etc, add a TEV if the unit didn't have one, run the liquid and vapour lines through respective water cooling and water heating heat exchangers, and retire to an ever-warm house!? Not being good at the academic stuff, I have no idea how long or big these exchangers must be. Here's what I know (guessed) so far: Water in the ground loop has antifreeze in it and comes back at about 7 deg. The loop is 400 metres of 25mm poly pipe about six feet down (by comparing with stuff I have found on the net this should be ok for about 12-15KW input to the A/C unit. The hot side obviously wants to be as high as possible. Assuming I'm lucky if I get a COP of 2.5, anything around there would make me happy. Water circulation is by standard central heating pumps (120W). My choice of A/C pumps are; Toshiba PG330X3F.4LS (R407c), Nett 18,000BTU (R22), various car A/C pumps, Fosters chiller 9,000BTU (R134A). To cut to the chase, my big problem is how long should I make the heat exchangers? Should they be the same length on both the hot and cold sides? Bear in mind they must be simple to make, so I was assuming tube within a tube with counter flow and no fins on the inner tube. Does it matter if they are longer than necessary, will that mess up performance of the A/C system? Having read some threads about pipe sizing I realise it is a complicated subject and there probably is no way I'm going to get it right first time, but if anyone has any experience of this sort of system and can tell me definitely not to do this or that, it could save buckets of time for me and would be much appreciated. So far I have experimented with the Fosters chiller, I've got two 5m coils of 3/8" in parallel inside an old imersion tank connected to the ground loop. The two coils don't have identical flow rates and one has more frost exiting the tank than the other. I have put the bulb from the TEV on the frosty coil. It seems to cool the water down fine but not a lot of heat seems to be going along the line into the condenser. I would have expected more, although the air is warmish coming from the fan, fine for a pre-heater I suppose. The fear of over-charging has probably made me undercharge the unit by way too much, the liquid line sight glass is not completely full all the time. I tried to get some info from Fosters web site about charge amounts, pressures etc but so far nothing found. The unit is an 8H, and the pressures I'm getting at the moment are 1ATM and 5ATM which seems very low to me. Well I suppose I'm expecting too much when there are so many variables, but maybe someone can point me in the right direction on some of it, and then I can trial and error the rest. Thanks for reading, kind regards

17-02-2005, 06:46 AM
you don't fancy cutting that up a tad do you?

personally for me a huge slab of letters like that just blurs and I give up before the first sentence



17-02-2005, 09:36 AM
Fair comment. Here's the reduced version:
Can anyone give me a rough idea how to size refrigerant to water heat exchangers for a 15KW ground coupled heating system. A simple formula that could get me to within 30% would be brilliant compared with where I am at present.
For simplicity and cost I want to make the exchangers from copper pipe, one within the other with counter flow. Thanks

750 Valve
17-02-2005, 10:51 AM
Not sure after all that if I understand what you wanna do. Here's a basic formula that applies for evaporators and condensers

Q = U x A x LMTD
where Q is cond/evap capacity in kW
U is the heat transfer coefficient of the cond/evap material (copper pipe) in watts/m2 K
A is the surface area of the pipe in m2
LMTD is the logarithmic mean temp difference between the refrigerant in the exchanger and the heat exchange medium (air,water,earth?) in degrees kelvin.

If you intend burying coils in the ground LMTD may or may not be hard to find, if the heat dissipates in the earth well, I imagine just a simple TD between ground and refrig would do, but if you intend harnessing the built up heat in the earth through reverse cycle and an insulated ground (ie reject heat into ground during cooling season and remove same energy from ground well for heating season) well then this may be harder to size.

Hope it helps, good luck. :)

17-02-2005, 05:17 PM
Now thats good timing.
Saw my uncle for first time in ages last week, he has a farm in Shropshire where he has 12 large Log cabins.
The conversation started off by harnesing wind to charge batteries or power none essentials.

After talking for a while we also came round to the idea of using a 'heat-pump' type system to heat domestic water and a generator to power/drive the compressor.
Be interested to get some pictures of what you have running already.

17-02-2005, 11:36 PM
Thanks for the formula 705 Valve. I think I can almost handle that apart from the logorithm bit. Am I finding the mean difference and then just looking up the log value for that? I know I can be dumb, but is that base 10 logs or the natural ones, my book has both?

Reeferjon, always glad to share ideas with fellow energy fanatics! PM me your email and I'll see if I can get some pics together for you. Most of what I am doing seems to be perpetual 'projects in progress', it's amazing how much pipe and cable has to be put in the ground for this sort of thing, I spend more time in a JCB than doing the interesting stuff. My biggest success to date is the woodgas project (by the way, I am told you can run an ordinary car engine on woodgas for nothing!). I'll be pleased if I can save you some time avoiding the pitfalls that I have climbed out of!!

18-02-2005, 06:10 AM
Thanks for the formula 705 Valve. I think I can almost handle that apart from the logorithm bit. Am I finding the mean difference and then just looking up the log value for that? I know I can be dumb, but is that base 10 logs or the natural ones, my book has both?



For refrigerant side, temperature is usually considered constant.
Use natural log.



We astigmatics find small paragraphs most helpful. One idea or question per, would make specific answers more easily supplied.


18-02-2005, 08:41 AM
Useful links, thank you.

I've done a Google on Heat Transfer Coefficiant. I now know about 'fat splitting', and how to refrigerate a computer, but no value for copper heat transfer coefficiant drops out. Could a kind someone with a table look it up for me please?

18-02-2005, 08:57 AM
Try this link http://www.engineeringtoolbox.com/36_429.html

18-02-2005, 01:26 PM
And another useful link,


I like their products and they are very helpful.
You can contact them at:

Alfa Laval Ltd
7 Doman Road, Camberley
Surrey GU15 3DN
Telephone: 01276 63383
Fax: 01276 685035
E-mail: general.uk@alfalaval.com

Chemi :)

19-02-2005, 12:08 AM
By the way, apologies to all for the block of words at the start of this thread. Comments noted. I won't do it again.

Thanks for the extra links, interesting and useful.

One answer throws up another question though. In the formula, the Heat Transfer Coefficiant is in W/m2K. The link gives me the Thermal Coefficiant in W/mK. The units of the formula won't balance unless there is m2 on the bottom. Is there a conversion for Thermal Coefficiant to Heat Transfer Coefficiant to give the extra m on the bottom line?

19-02-2005, 04:37 AM
One answer throws up another question though. In the formula, the Heat Transfer Coefficiant is in W/m2K. The link gives me the Thermal Coefficiant in W/mK. The units of the formula won't balance unless there is m2 on the bottom. Is there a conversion for Thermal Coefficiant to Heat Transfer Coefficiant to give the extra m on the bottom line?
I almost remember this.

U (overall Heat Transfer Coefficient is made up of k, thermal coefficient of material, and film factors which depend on viscosity and flow regime (turbulent, not-turbulent, pool boiling, etc.) k is given per meter of thickness of material, U per square meter of surface area. Your missing dimension is the actual thickness of the material.

In practice, thermal conductivity of copper, taken as resistance to heat flow, is insignificant compared to surface film factors.

Best route might be to research data from mfgs of tube-in-tube heat exchangers who would give values for different dimensions and flow rates.


19-02-2005, 10:44 AM
Ta for that. I'm on the trail...

18-09-2005, 03:04 AM
We have developed a new Patent that can improve the operating efficiency of heat pump water heaters. It is significantly more energy efficient than other Heat Pump Water Heaters and savings above 75% of hot water heating costs have been recorded at night!

The Patent design in question relates to what is called a condenser/tank.

We have been field trailing 340 litre condenser/tanks connected to air conditioners with a 1.3 kW compressors and the results are extremely impressive. As Len Taylor says concerning his own test results:

“This is really an excellent performance for an air source heat-pump of this capacity on such a cold day and would facilitate efficient operation with almost any combination of on-demand or off-peak modes...Examples: ambient 8C to 10C, run time 3hrs 36 mins, C.O.P 2.98; ambient 16C, run time 2hrs 2mins, C.O.P 4.80; and so on.” L.E. Taylor M.Eng Sc: CPEng. August 2005

The water heater operates at night to take advantage of cheap off-peak electrical power. Because it heats up in two to three hours, on off-peak power and the heater does not use heating elements, hot water costs are also 75% cheaper than conventional hot water heating costs.

The only other product like it is the Quantum Energy System and thsi new condenser/tank heat pump water heaters far exceeds the performance of the Quantum.

Carlo Hansen
18-09-2005, 12:09 PM
Hi Sundwind

It is always difficult to calculate, how to do the right thing in a refrigeration system.
If someone is interested, i do have a PDF file about 186 pages, wich describe nearly all you have to know about
the calculation. ( It is realy hard stuff )

Just send me a mail, and i will send you the file.

Best regards
Carlo Hansen

18-09-2005, 07:09 PM
Hi Carlo , I would like to receive a copy of that pdf file , if you don't mind . . . . :)
My email is : send2pierre@hotmail.com

19-09-2005, 09:22 PM
Thanks Carlo , i received the file . . . and it's good


23-04-2006, 11:20 PM

We have been making and developing Heat Pump Water Heaters for 20 years. See siddonsolarstream.com


22-05-2008, 12:43 PM
It will be very kind of you if you can send me the 186 pages file to my mail box at
zahid.noor@salemjci.com or zahid.noormohammed@yahoo.com

22-05-2008, 07:08 PM
If this file is still possible to get, I also kindly ask, that you send it to me to!
My e-mail will be sent to you by PM.

24-06-2008, 01:40 PM
Hi,I'm a student,my major is co2 heat pump water heater .So I want to find some info on it.If you can e-mail me some files ,I will study more.my e-mail is dongdong4848488@163.com,Wait for your reply,thank you!