Small and crude GSHP for a Michigan garage

[pick1e]

*This thread began as some questions on AC_Hacker's "The Homemade Heatpump Manifesto" thread but was moved here by Daox. AC_Hacker has a lot of excellent information in his thread, but there is also some bad information. He doesn't want to open his methods for discussion, so I will address it here I guess. This will also be a repository for logging my own progress/successes/failures/methods/results.

I too am headed down the DIY GSHP path. I think my project should be easy to accomplish. I have a 1 car detached garage that I would like to heat a little bit. I'm not living in it, it's just a workshop, so I'd just like to keep it from freezing in winter, maybe 45 degrees or so would be good, and raise it up to 60 or so when I'm working out there. I hope to use a window air conditioner unit and a water ground loop.

I have some questions though. What's the deal with straight down and up loops in PEX tubing? What kind of flowrate are you using? I would rather dig less holes and use some copper tubing, but I guess I'm curious if that's worth it, maybe the thermal transfer rate in the soil is the limiting factor? I mean suppose you buried the same size loop of copper instead of PEX in the same hole, would you still be limited to the same amount of heat gain?

I'm not a HVAC guy but I do have an engineering background so I know the thermo basics. I borrowed a heat transfer book from a friend (mine is long gone) and am trying to figure how much tubing I need... But I don't know the temps of the in and out water until I rig up a test I guess. Any insight here? What is a reasonable temperature of a cold evaporator? I could use that as a theoretical for max heat gain I suppose. I didn't know if I could use the ground temp as the theoretical exit temperature though. Do you guys with working systems see the exiting water reach ground temp or still slightly below?

I guess a big and most immediate question I have for you HVAC gurus (the answer to which could through a big monkey wrench into my plans) is can I extend the tubes leading to the evaporator coil? My plan was to literally add a few feet so that I could just hang the coil in a tub of water kept warm by the ground loop.

Thanks for any insight

[Daox]

pick1e, I moved your post to get its own attention and to keep AC Hacker's thread on topic about his own project. I thought I posted saying this but apparently didnt. Sorry for the confusion.

[pick1e]

Since posting my initial questions, I have found some answers which I will address here.

Originally Posted by pick1e

What's the deal with straight down and up loops in PEX tubing?

At first glance it looked like they were using PEX tubing which I'm familiar with, but it is of course HDPE.

What I meant to ask was why is there a single U shaped loop instead of some other formation. The answer is that it is simple and effective. Multiple loops in the same borehole ARE more efficient, but only slightly. You could have something more like a W in the borehole but the pipes end up sharing most of the same heat, thereby being less efficient than two separate boreholes with to U pipes.

Originally Posted by pick1e

What kind of flowrate are you using?

I haven't seen that addressed here, but it should have some effects. There should be a threshold of water velocity where water flowing below that velocity will not move heat fast enough, and anything over that threshold is more than enough. I suppose that is why it's not talked about, it's easy enough to get it past that design threshold so most people aren't worried about it. But, if you designed your system well, you would know the necessary velocity of water in the tubes, and use a suitable pump that would move water slightly faster than that threshold. Anything more powerful would be overkill and therefore a loss of efficiency.

Originally Posted by pick1e

I would rather dig less holes and use some copper tubing, but I guess I'm curious if that's worth it, maybe the thermal transfer rate in the soil is the limiting factor? I mean suppose you buried the same size loop of copper instead of PEX in the same hole, would you still be limited to the same amount of heat gain?

I will be experimenting with copper tubing. By my calculations, it should be more efficient, able to gain more heat with shorter lengths of tubing. The reasons for using the HDPE sound like:

a) it's cheaper
b) it lasts forever (relatively) underground while copper may corrode and leak over time

But the fact is, while the thermal conductivity of soil is generally between 1 and 2 W/(m*K), the thermal conductivity of HDPE is about 0.5. So effectively it is insulating the ground loop water from the ground. Copper has a thermal conductivity of 401 W/(m*K), far above that of any soil (on this planet) so your limiting factor in the thermal transfer network of a borehole would be the soil and not the tubing.

I will be testing with copper because I am not installing a system that I intend to last forever, only several years. And I am very curious to see how much a ground loop can be reduced by using a more thermally conductive material. My theory is that it should be much more efficient in the short term.

[pick1e]

I have acquired my refrigeration unit. But...

Is it an Air Conditioner?


Or is it a Heat Pump?


Well, it's going to be a heat pump anyway.

At first it wasn't working. The fans blew nicely but I wasn't getting any refrigeration. I figured it was just too cold in my garage already (about 40 degrees) for the thermostat to kick it on. I first tried to warm up the thermostat bulb in my hand but it wasn't enough. I finally ended up just removing the thermostat.



I finally realized that since I have an analog system, the thermostat was electrically just a simple circuit disconnect. So I simply bypassed it by running the lead from the rotary switch directly to the compressor wiring harness. Hooray! Refrigeration is go.

Hopefully I can carefully remove some pieces of the unit in order to build or fit some sort of container for a heat exchange bath around the evaporator coils without removing them. I hope I can completely avoid disconnecting any lines and the whole mess of capturing refrigerant, brazing, and all that. Luckily on this unit the incoming and exiting refrigerant lines are on top, so hopefully I can carefully remove the unnecessary parts and get a container in there.

Here you can see the lines entering and exiting at the top of the evaporator. I've drawn a blue box showing where I intend to build the water tub. I may be able to run the ground loop directly from this box but we'll see.


Here is a top view with the top plate removed, I figure if I can remove the styrofoam and squirrel cage fan I should have enough room for a container for a heat exchange bath.


I said crude right?

[pick1e]

Tonight I intend to test the ac heat pump with a garden hose as a stand-in for a heat pump. It's about 30°F outside and about 40°F in the garage and falling so the heat output should be obvious.

Here are some models of the situation:



The first is the overall model of the AC unit. It requires 1050W of juice from the wall (manufacturer's spec) and is supposed to put out 10,000 BTU/h which is equal to 2931W. The 1050W input doesn't contribute to the heat output enough to be considered. Most of it is turned into mechanical energy to turn the fans and compressor. The refrigeration loop is what we are concerned with. It pretty much has the evaporator, condenser, and compressor. There is a negligible amount of heat added from the compressor and other things like friction.

Below that is the control volume model for my water tank. I will pump water in from the garden hose, and out a drain. The evaporator should pick up energy from the passing water and cool it before it exits. When I tested my water yesterday it was coming out at 57.4°F so I've estimated 55°F at the inlet. The exit temperature will be related to the flow rate, but for the sake of calculation I used 40°F as a starting point in order to find a viable flow rate. I'd want to get this value as close to 32°F as possible to maximize efficiency but I figured 40°F was safe. This may differ though because I'm using a much larger tub than I need so there will be more error.

I'm assuming that the water will be enough to effectively transfer heat to the evaporator at any reasonable flowrate. Since the evaporator coil heat exchanger is designed for air, it should more than necessary to get the same amount of heat from water.

So we have the following calculation scribbled down:



So for a temp drop of 15°F I should need a flow rate of about 5 L/min. This would be 80gal/h in water pump specs. I also figured 1.33 gal/min so that I can measure it by adjusting the flow rate to fill 1 and 1/3 gallon jugs in one minute.

We'll see how it goes.

[mrd]

2931 W is the spec under expected conditions, but submerging the evaporator in water could certainly change that number. If you know your test flow rate accurately, then you could plug in your actual dT and solve for Q as the variable..

[Patrick]

Originally Posted by pick1e

I will be experimenting with copper tubing. By my calculations, it should be more efficient, able to gain more heat with shorter lengths of tubing. The reasons for using the HDPE sound like:

a) it's cheaper
b) it lasts forever (relatively) underground while copper may corrode and leak over time

But the fact is, while the thermal conductivity of soil is generally between 1 and 2 W/(m*K), the thermal conductivity of HDPE is about 0.5. So effectively it is insulating the ground loop water from the ground. Copper has a thermal conductivity of 401 W/(m*K), far above that of any soil (on this planet) so your limiting factor in the thermal transfer network of a borehole would be the soil and not the tubing.

I will be testing with copper because I am not installing a system that I intend to last forever, only several years. And I am very curious to see how much a ground loop can be reduced by using a more thermally conductive material. My theory is that it should be much more efficient in the short term.

I think this choice might come down to the cost/benefit ratio. With copper, you have 400 times the conductivity of the soil, which is overkill. With the HDPE, half the conductivity of the soil. If you put in twice as much HDPE as you would have copper, you will get the same heat flow (limited by the soil). So does copper cost more than twice as much as HDPE? If so, it would be cheaper to use the plastic. IMO, copper is a total PITA to work with.

[pick1e]

Originally Posted by mrd

2931 W is the spec under expected conditions, but submerging the evaporator in water could certainly change that number. If you know your test flow rate accurately, then you could plug in your actual dT and solve for Q as the variable..

Exactly right. Especially because I wanted do a quick and dirty test- I used a (relatively) large tub instead of making a small box simply because it was quick and easy, so heat transfer was quite inefficient. Plenty of water was flowing down the drain before giving up its heat.



This was taken this morning because I forgot to take one last night before firing it up. The water is extra dirty only because I used a dirty bucket from outside to scoop out the top bit of water so I could remove the drain hose without mess

It started to freeze up after about 15 minutes:



While I let it thaw I turned off the water supply and fashioned a cardboard shroud around the coil to keep the warm water closer to the coil. This gave me an actual dT of about 5.6°C and a dQ of 1955 W.

With this value I come up with a theoretical rise in temperature assuming no other heat loss from the garage (false) of 102m^3 at 5°C of 54°C per hour or 13.5°C per 15 minutes, about how long I ran the system at a time. Considering there is no insulation on the building and it's full of stuff, I'd say that's a fair calculation.



I still had no forced convection in the tub aside from the trickle of water which the evaporator was certainly not designed for.

But, it did work for a while which was cool for about 1 hour of effort putting it together and about an hour of testing. In 15 minutes it was able to raise the air temperature in the garage by 4ºF which was fair considering the garage is completely uninsulated at the moment, it was below freezing outside, dark, and the door was cracked to let the hoses in.

So I'm just about done! All I need to do now is drill some holes in the ground, build some ground coils, build some solar panels, reconfigure the AC unit with proper heat exchangers, develop a control system and insulate the garage!

[pick1e]

Originally Posted by Patrick

I think this choice might come down to the cost/benefit ratio. With copper, you have 400 times the conductivity of the soil, which is overkill. With the HDPE, half the conductivity of the soil. If you put in twice as much HDPE as you would have copper, you will get the same heat flow (limited by the soil). So does copper cost more than twice as much as HDPE? If so, it would be cheaper to use the plastic. IMO, copper is a total PITA to work with.

I definitely agree with your logic. I do want to try copper though for a couple reasons:

1) I don't need it to last forever.
2) It's a small project.
3) Nobody uses copper and I'm always up for a challenge.

Hopefully I can get away with shallow boreholes (maybe 20 feet?). My friend who installs fences for a living suggests that in this case I SHOULD be able to pull them out of the ground and recycle them in the future.

If I were heating my house and I planned to live here a long time it would certainly be different but I just want to heat the small garage with as little disruption to the yard as possible.

[pick1e]

oops, double post