Dxgshpwh (ground source heat pump water heater)

[jeff5may]

I have been off the forums for awhile, life to do and such... sounds like you are all-in on this project! Post lots of pics of what you do.

The heat exchanger in the water tank will do its best at the bottom. When cold water enters, it naturally sinks, so that's your best spot for heat transfer efficiency. Many DIY people have left the heating elements alone, by either inserting the hx loop through an existing fitting (inlet, outlet, or drain), or by bringing the water to the hx and back with pipe and pipe fittings. The big purpose is to have "less custom" fittings that don't take machining tools to fabricate. You can then buy a prefab, pretested hx that will be a proven winner. Less guesswork for the wary.

With a hermetic compressor and just a single refrigerant path, oil control is a non-issue. The smallish compressors have a copious sump built into the shell. The built-in muffler also acts as a suction line accumulator. With the evaporator dropping way down into the ground, I would size the tubing small to maximize velocity to keep the oil from pooling in the bottom of the hole. A curly-q every ten feet of height would trap the stuff stuck to the walls. Might not need it, but it's cheap insurance.

[MEMPHIS91]

Jeff,
I've been all in on a long list of crazy home improvements that I have not posted to a forum about. So I am REALLY all in on this one. Thank you for your awesome advice.

I will look at more post on the site to see what people have done. I believe I have only looked at two so far. (Got suck on the bore hole problems) Prefab sounds like an excellent option.
Bottom of the tank it is then.

---"With a hermetic compressor and just a single refrigerant path, oil control is a non-issue. The smallish compressors have a copious sump built into the shell. The built-in muffler also acts as a suction line accumulator. With the evaporator dropping way down into the ground, I would size the tubing small to maximize velocity to keep the oil from pooling in the bottom of the hole. A curly-q every ten feet of height would trap the stuff stuck to the walls. Might not need it, but it's cheap insurance."---

In your opinion would two lines of 1/4 inch 30-35 feet deep be ok?
And can you explain your curly-q idea a little better? I think I get the basic idea, you are giving the lubricant a small place to -get a foot hold(kinda like rock climbing)-. Is that right?

Awesome info and food for more thought. Thank you

[jeff5may]

For the tubing, check Xringer's thread on the airtap A7 HPWH he installed. He successfully integrated a solar/ashp/oil furnace powered water heating rig into his house for hot water no matter what. Last I heard, his solar was doing most all of the water heating, with the Airtap kicking in during high demand. Since the tank was rigged in, he has burned very little fuel oil at all. Way down from thousands of dollars a year.

http://ecorenovator.org/forum/geothe...er-heater.html

Randen also turned a 5000 btu window aircon unit into an air-source water heat pump, with much custom work done throughout. He is just a mechanical master genius. Everything he does comes out looking awesome, even the stuff that doesn't work! I hate to see a good-looking piece of machinery destroy itself, but it looks ALMOST as good dead... kind of like taxidermy.

http://ecorenovator.org/forum/geothe...heat-pump.html

If you look at a larger split-unit install manual, the manufacturers have pictures and guidelines for where and when to put the oil traps in the lines. They work just like a p-trap in a drain. When the flow stops, the extra oil gets stuck in the trap instead of finding the bottom of the plumbing. When flow starts, the oil is carried back to the compressor much quicker.

Rather than jog the line laterally for the trap, you can loop it once for the same effect.

With natural refrigerants (propane, propylene, butane, etc.), the oil is much more soluble and miscible in the refrigerant, so oil flows much easier and better through the system than with "modern"(r134a, r410a, r407c, etc.) refrigerants. As a result, it is usually ok just to size the tubing small enough to get decent velocity, so the foamy froth can make it a decent way back up the tube before the gas and oil separate completely.

[jeff5may]

Originally Posted by AC_Hacker

I don't have any experience with DX, so I'm trying to imagine what the behavior of the vaporized refrigerant would be, based on what I have witnessed so far...

I initially imagined that there might be a condensation problem, but then I realize that condensation will not happen in the low pressure side. Another aspect would be that the temperature of the liquid-flashed-to-gas that is exiting from a high-pressure tube into a low pressure area will be the very lowest in it's trip back to the the compressor, so the delta-T will be greatest and the heat migration into your vaporized refrigerant will be maximum at that point. The delta-T will decline as the vaporized refrigerant gathers more heat. So there could be an advantage to having all that take place at the deepest, most wet, most temperature-stable part of the bore hole.

The opposite scenario would be that if the liquid-to-gas transition happened at the top of the hole, the vapor would have a longer distance, and longer time to gather heat. The construction and maintenance would be much easier. Also, if you went for a TXV, it would need to be at the top of the hole. But I think that a TXV for such a tiny compressor would be just too much techno for what should be a simple, elegant project.


I would lean on the second scenario, with the cap tube before the loop. While the subcooling of the high pressure liquid refrigerant could help superheat the exiting vapor to above that of the ground temperature, the heat gained would not be stellar. Most of the extra heat would be radiated into the ground anyway. Giving the liquid/vapor mix more distance to gain ground heat would be much more beneficial. Remember, liquid moves slowly, gas moves fast through the system.

In the IGSHPA manual (CLGS Installation Guide (#21020)), in the section on the loop fields, it looks at various ways of doing loop fields, both trenches and boreholes. There was even a small discussion on putting two loops in the same hole, and I believe that the performance increase was about 3% to 5%.

I'm afraid I don't understand how two loops would remedy the lubrication issue?

-AC

P.S.: This manual is a tad expensive ($45.00), but it is so amazing in the breadth and depth of information regarding Ground Source heating and cooling, that you should seriously consider getting a copy. It can sometimes be found used, sometimes in libraries. But considering the information, it is a miracle. As I recall, the manual does not directly address DX, but most of the principles are the same.

Also consider if there is water in the bottom of the hole. Migrating or not, a massive body of water in direct contact with your copper pipe is going to transfer orders of magnitude more heat than dry or moist earth per square inch. 10 Meters of 1/4 inch pipe may be good for up to 15000 BTU/H if the bottom half of it is sitting in an aquifer.

[MEMPHIS91]

I have read each of the other threads completely now. I'm a plumber by trade so p-traps I can understand. Do I need loops going down and coming back up or just on the way back up?

I'm think maybe just one run of 1/4 down the bore hole then, with the cap tube at the top. I'm sure this time of year the water table is at maybe 6 feet down. Should I not fill the bore hole? If water is best then filling with sand will only make it worse? I'm sure some dirt will fall in over time though.

I made a sketch up of what I am thinking. Any thing I'm totally forgetting?

http://ecorenovator.org/forum/attach...r-img_0584-jpg

-- Next project, running my 4 ton heat pump in a loop in my pond. --

[AC_Hacker]

Originally Posted by MEMPHIS91

...I'm sure this time of year the water table is at maybe 6 feet down. Should I not fill the bore hole? If water is best then filling with sand will only make it worse? I'm sure some dirt will fall in over time though...

We looked at this image previously in the post:

In the lower right corner you will find instructions regarding the desired way to backfill a bore hole that passes through an aquifer. I saw this graphic quite some time ago, and it was from a site that may not exist anymore, but the site was in Canada, and the distances for fill that they describe were specifically suited to that part of Canada where they were operating.

If you scale their advice to your knowledge of your location, you'll be good to go.

The gravel fill principle is to put the larger gravel (pea gravel) at the bottom, up to the level of the top of your aquifer. Then graduate to finer granularities by following with a small (6" or so) layer of finer gravel, then a similar layer of some coarse sand above that, and then finish off with whatever you want... dirt, bentonite, etc.

-AC

[jeff5may]

The lines in your sketch are backwards. As a general rule, the liquid line should be smaller than the suction line. In your sketch, the borehole loop is part of the suction line. It would be better to use 5/16" from the cap tube to the compressor muffler and 1/4" everywhere else, if you even need it. The only tubing I have used over 1/4" in my beasts has been used to adapt up in size to fit compressors and txv fittings. The situation changes rapidly with long distances and larger capacity systems.

[MEMPHIS91]

AC, Somehow I didn't read the diagram I just looked at the pictures..... Pea gravel it is then.

Jeff, Really? 1/4 inch all the way? In the water heater and bore bore? That would make things SO much easier.

[jeff5may]

The newer minisplit units are using long parallel runs of small diameter, thin wall, ribbed tubes in their heat exchangers. The superconducting supercomputers have led them in this direction. Apparently, some pressure drop is a good thing. More surface area is always better, as is turbulent flow.

There are charts and formulas all over the web (and this site) that can help you predict how your design will act before you actually install it. I used coolpack and the SWEP literature a lot to make things jive. The literature ac_hacker cited is also awesome.

For a 4 ton rig, you will need to do your homework to ensure the rig will be reliable. For this half ton build, it will probably work well without lots of modeling. However, it will be good practice to get your hands dirty and for measuring performance. I perfected my brazing skills on small systems and built up my tool set.

[MEMPHIS91]

Originally Posted by jeff5may

The newer minisplit units are using long parallel runs of small diameter, thin wall, ribbed tubes in their heat exchangers. The superconducting supercomputers have led them in this direction. Apparently, some pressure drop is a good thing. More surface area is always better, as is turbulent flow.

There are charts and formulas all over the web (and this site) that can help you predict how your design will act before you actually install it. I used coolpack and the SWEP literature a lot to make things jive. The literature ac_hacker cited is also awesome.

For a 4 ton rig, you will need to do your homework to ensure the rig will be reliable. For this half ton build, it will probably work well without lots of modeling. However, it will be good practice to get your hands dirty and measuring performance. I perfected my brazing skills on small systems and built up my tool set.

Really now? That sounds VERY interesting.
Coolpack looks like quite the learning curve for me, but I will already downloaded it and starting to play with it. Basically I just need to see how long my hx needs to be based on all the other data I have.

Yeah the old 4 ton can wait. I got a water heater to build. Thanks for the info.