[AC_Hacker]

Originally Posted by wsexton1

I have finally managed to read through the wealth of knowledge available on this thread, and I've got to mention how impressed I am.

William,

Thank you. It was a lot of work, but I did it in the hopes that people just like you could succeed in a project such as you are now undertaking.

Originally Posted by wsexton1

So to the build...As per your advice I analysed the heating load on the building (new build, 2 years old) and discovered a max heating load of 5kW for the room to be heated.

OK, so when I translate that into my antiquated US (AKA: British) system, I get 17060 BTU/hr or about a Ton and a half. Sounds reasonable.

Originally Posted by wsexton1

Luckily the underfloor heating circuit was installed recently in a 75mm thick insulated screed floor with 150mm spacing. Which I believe allows for a relatively low flow temp (40-50C).

Floor = 3" thick concrete, insulated below, with 6" PEX spacing. (perfect)

Input water temp = 104F to 122F (just the right range for GSHP)

Originally Posted by wsexton1

Onto the borehole (actually 80m deep), which after speaking to a local expert informed me of an extractable heat of around 59W per m giving around 16000 Btu/hr.

Bore hole = 80m = 262.5 feet deep

Local estimate = 59 watts per meter = 4720 watts = 16105 BTU/hr

So far, it seems right on target!

More bore hole is always better, but this sounds just about right.

Originally Posted by wsexton1

Progressing to the heat pump. I have decided to start off on a smaller scale to get some practise and teething errors out the way first time around and see what sort of COP is possible before purchasing a bigger compressor.

Great, just great!

Originally Posted by wsexton1

Potentially putting the second heat pump in series with the first to have the capability of having the system on 'half power' (which I hear may be more efficient anyway). Having purchased an old 14000 Btu/hr mini split unit and having discharged it, I'm now in the process of getting hold of some heat exchangers.

Your idea of multiple compressors is a good one, and your decision to build a smaller test unit for 'proof of concept' is really the only sane way to proceed.

With all that said, I would say that the 14000 BTU unit you now have is a perfect match to your heat load, floor, and bore hole. It will not provide 100% of the heat required for the worst case condition, but the remainder can be supplied by other means.

So, this first build may well be the only build you may need to do, and if you proceed carefully, you may be able to get it right on the first try.

Originally Posted by wsexton1

This is my question for you at current. Why is it essential for the heat exchangers to be sweat fittings? Only because there are much cheaper threaded versions available here in England.

What ever kind of junction you choose to go between your copper refrigerant tubing and your HX will be subjected to two different kinds of stress, mechanical stress from vibration and also thermal stress due to the large temperature excursion which is a result of off-on cycling. Sweat fittings (that I continue to insist should be called 'brazed fittings'), when properly done, will provide a junction that can easily survive both kinds of stress, because a brazed connection makes the HX and the copper refrigerant line become one continuous structure, so in effect you have eliminated the junction entirely.

As a lucky coincidence, you happen to have a fellow heat pump hacker in The Neatherlands who is also at this time, building a heat pump RIGHT HERE.

He is using a Brazed Plate HX which has compression fittings, and he is not going to use a sweat fitting.

It may be that the problem of stresses that a HX is subjected to have been solved through properly-engineered compression fittings. If that is the case, it makes life easier and less expensive.

If you are going to use compression fittings, be sure that you fully understand all of the details of implementing the compression fitting before you proceed. I am thinking of details like assuring that you have made a good flare that is clean and perfectly smooth. Also, does the compression fitting procedure call for using some kind of assembly lubricant on the threads and perhaps face of the metals in the compression area? Lastly, exactly what is the specified torque that is called out for your exact fitting? Too much torque can weaken the union or even ruptured the copper... too little torque can come loose. A properly executed compression fitting, due to the forces, can actually 'weld' the mating metal faces, when analyzed on a microscopic level.

Originally Posted by wsexton1

Additionally, I've estimated the water flow through the ground loop to need to be around 1.8m3/hr (7.9Gpm), for turbulent flow. Requiring around a 160W circulating pump. Is this similar to your loops pumps?

This sound about right. As you probably know there are some very sophisticated pumps available now that can sense and self-adjust to changing requirements.

Originally Posted by wsexton1

Only I happen to have a central heating circulating pump banging around, and I'm more tempted to reuse that than go ahead and buy a great new pump!

Use what you have! Pumps are easy to change out.

As a request from me to you, it would be very useful to me and all the other readers, if we had a better understanding of the local temperature characteristics you are subject to, so if you could go to Degree Days dot Net and calculate your degree Days in both Fahrenheit and Celsius it would be very useful. In fact, if you could tell us what your local temperature monitoring station location is it would be interesting to see what the recorded historical temperature patterns actually are.

Good luck with your project, William (also my son's name)...

Best,

-AC_Hacker