The Homemade Heat Pump Manifesto

[pick1e]

Originally Posted by AC_Hacker

No, this is not correct. I used the 'LS press' reading to refer to a PT chart for R-290 to see what the temperature should have been at that particular pressure (actually I did a curve fit on the PT chart for R-290 and derived the formula for the entire PT chart in order to save myself from interpolating every value). Then I subtracted the actual temperature from the calculated temperature to determine Sub Cooling.

Ditto Super Heating. I derived values for Sub Cooling and Super Heating, to be used in system tuning. Since the only fine adjustment I could make was refrigerant pressure. I used an excessively high Sub Cooling value to discover that I needed to lower my refrigerant pressure. I did adjust that, and the efficiency increased accordingly.

That is interesting. I don't remember you mentioning that in the thread, sorry if you did. Hopefully I will avoid that necessity by not disassembling my AC unit. I would like to use it as is, sacrificing some efficiency for simplicity.

Originally Posted by AC_Hacker

I used a power meter to determine power into the compressor for each time interval. I also knew the volume, and thus the weight of the water for each time period.

Okey dokey, thanks.

[Vlad]

Originally Posted by AC_Hacker

So, yes I have been planning to do a radiant floor all along.

It's not going to be easy, because I live in a very old house and a two-inch thick concrete suspended floor is just not in the cards. But I am looking into ways of getting a high efficiency floor that doesn't weigh as much as two inches of concrete.

Right now, I'm favoring a built-up floor made of layers of Wonderboard brand tile-backer board. I have done some preliminary tests on various materials and Wonderboard looks better than Durock and Hardybacker board.

I have also done some experiments with unusually high thermal transfer aggregate for concrete, but so far, no love. Nevertheless I still think it has promise.

AC_Hacker

Try design I used in my house. It is light, strong, easy, cheap,....
The heat you potentially loose (I mean go down and not up) depends on density of material and area of contact(mostly). When you use cement board and fill gaps you loose more. In my case I used 1x2 strips of wood. The tube has good contact with upper plywood through aluminum heat plates and has no contact with sub-floor. The only contact upper layer has with sub-floor are wood strips(low density compare to cement and fraction of total area). Also I used aluminum foil for reflection of heat up toward top layer. If you want to be more sophisticated you can use special bubble foil insulation. I used foil because I needed 3000 sq-ft and it would cost me a fortune.

Unfortunately I can not tell you about performance because I started from house to outside and you started outside first. But I can not see no problems.

If you need more details or pics I can post them.

[AC_Hacker]

Originally Posted by Vlad

Try design I used in my house. It is light, strong, easy, cheap,....
Unfortunately I can not tell you about performance because I started from house to outside and you started outside first. But I can not see no problems.

Vlad, I have been wondering how the floor part of your project is coming along.

Originally Posted by Vlad

If you need more details or pics I can post them.

Yes, this would be a very good idea. Since it is still fresh in your mind, and all the details have not gotten lost.

I'm sure your floor will work well, I sure would like to see some performance figures, when you get the water flowing.

I started a thread which begins here for DIY radiant floors. Since that is exactly what you are doing, it would be be a likely place for your photos and info.

And since it relates to the homemade heat pump thread, maybe you could post a link to it here.

I'm looking forward to seeing it.

Best regards,

-AC_Hacker

[AC_Hacker]

Yesterday I did a test of my homemade heat pump. This is the first test with the heat pump finally hooked up to the loop field.

Here's a photo of the setup.

"a" is the homemade heat pump that I built about a year and a half ago. I've learned a lot since then but I think I can learn more from this unit, before I build the next one.

Here is an older photo of the heat pump, showing better detail...

"b" is the white "source" barrel (so named because the ground loop is where the source of low grade heat is coming from). This is the barrel that has the water coming from the ground loop. I have a quarter-horse sump pump pushing water through aproximately 720 feet of high density polyethylene pipe buried in the back yard. The water then returns from the loop and enters the brazed plate heat exchanger on the left side (evaporator side) of the heat pump. After exiting the heat exchanger, the water flows back into the 'source' barrel, and the cycle begins again.

The photo below is a detail of the source barrel and the sump pump. I measured the sump pump and it was drawing around 250 watts of power, ok for initial testing, but way too much for normal operation.

"c" is the sink bucket (blue 5 gallon bucket)

The photo below is a detail of the 'sink' bucket (so named because this is where the high grade heat is going to). Here the water is pumped from the bucket into the brazed plate heat exchanger which is on the right side (evaporator side) of my heat pump. After exiting the heat exchanger, the water flows back into the bucket.

I measured the pump, (actually a heavy duty aquarium pump) and it was drawing around 25 watts of power.

The photo below, is a closeup of the point where the capillary tube enters the brazed plate, evaporator side (left side) heat exchanger. The frost gives me an indication that things are working. I should have positioned the cap tube at the top of the exchanger, so that gravity would be working for me. I will incorporate this improvement into the next unit.

The chart below shows data gathered during the test. I ran the test for an hour, taking readings every five minutes.

Even though I wasn’t monitoring compressor power, I can tell that the COP is way down, and experiments are in order to determine optimum refrigerant charge levels .

Also, for a heat pump of this size, the 250 watts being consumed by the sump pump is extreme and I need to find other pumping solutions that will reduce this.

Regards,

-AC_Hacker

[Daox]

So, it looks like you pumped ~2,800 btu in 1 hour. Does this sound right? If it is, that seems a bit low? Is that due to compressor size or something else?

[AC_Hacker]

Originally Posted by Daox

So, it looks like you pumped ~2,800 btu in 1 hour. Does this sound right? If it is, that seems a bit low? Is that due to compressor size or something else?

"...a bit low...", you are being too kind.

Actually, it's awful.

I was expecting something in the neighborhood of = or > 5118 BTU per hr.

As I recall, in the beginnng, even before I made my first post on EcoRenovator, my very first test yielded really dissappointing results. In fact right there, in the beginning, I was ready to give up. But when I tweaked the charge level a bit, things really started to happen.

So, I'll be experimenting with charge level, water flow rates and also possibly heat exchanger sizes.

Regards,

-AC_Hacker

[pick1e]

I would call it pretty good for a quick and dirty test actually.

2800 Btu in an hour but if you look at the first half hour I get about 4300 Btu/hr. That's pretty close to the AC unit rating of 5000 Btu/hr... I don't know how they test them but I'm sure they give the max efficiency rating (probably 1st 10 minutes), not steady state rating. Pretty good I'd say for using the equipment for something other than its intended purpose and design conditions.

Not to mention even if you consider the lower value of 2800 Btu/hr that's about 820W, and you're only burning 575W (300W compressor, 25W aquarium pump, 250W sump pump) for a COP of 1.4. So you're not losing ground.

Also the small bucket of water causes the dT to rise relatively rapidly. I bet you would get a better reading with a larger volume of water (big drum?) with insulation. You are probably losing a fair amount of heat to the air at 100F with convection. But mostly reducing your dT/dt I think would reduce the measurement error.

Most importantly it looks like you're reaching steady state with the ground loop so that is bonus.

[AC_Hacker]

Originally Posted by pick1e

I would call it pretty good for a quick and dirty test actually.

Thanks.

I did another test today, and the compressor has a slow leak (always did), so by default I will be comparing progressively lower charge levels.

Today I was monitoring the power used by the compressor, so I was able to calculate the COP over each five minute period.

The power used by the pumps is not reflected in this chart.

The COP line is multiplied by 100...

... I'm also getting my data-logger set up to use in the testing, but it's not ready for prime time yet.

I think you're right about heat loss in the heat-sink bucket, most importantly is the fact that the bucket is sitting on the bare concrete floor!

Next test will have the bucket sitting on 2 inches of foam.

Part of the discontinuety in the data is due to the fact that the analog clock I'm using has floppy hands. I have a digital clock for further readings.

Regards,

-AC_Hacker

[bigsmile]

May I ask why you need to drill for closed loop water exchange setting? All the source I learned seemed to use trenches instead of wells. I'm really confused because usually they drill wells (shallow ones) for direct exchange setting.

Sorry it may seem to be a lazy question, since I may be able to find the answer by reading through this thread. But it's simply too much reading, with over 40 pages spanning almost two years. I very likely will take my time and read through it anyway because there is so much useful information. But for the moment, I'm really burning to find the answer for this really simple question.

And congratulations on the success and thanks for sharing this.

[AC_Hacker]

Originally Posted by bigsmile

May I ask why you need to drill for closed loop water exchange setting? All the source I learned seemed to use trenches instead of wells. I'm really confused because usually they drill wells (shallow ones) for direct exchange setting.

bigsmile,

Welcome to the conversation.

FYI, it is possible to search particular threads for certain key words, to find out what you want, but I will try to answer your question...

My whole reason for doing this is that I am aware that GSHP is the most efficient way to heat a house. I'm also aware that the entry cost of a GSHP is so high that the only people who can afford a GSHP are the people who don't actually need the operating cost savings that a GSHP can provide.

So I have taken upon myself to see if it is possible for a normal person, of normal means, to build their own homemade ground source heat pump from cast-off parts.

In other words I am trying to democratize the practice of Ground Source Heat Pump construction.

I live on a city lot and the loop field would have to go in my back yard.

I learned that foot for foot, full-depth boreholes are more efficient than trenches and slinkies.

I didn't go with trenches, because I couldn't see myself digging trenches by hand, and I didn't think I could build a power trenching machine.

I did think I could build a powered drilling machine.

Initially, never having drilled a hole any deeper than a hole for a fence post, I had no idea just what I would encounter if I drilled holes, and I planned to do some holes with a depth of 50 to 100 feet each. I thought if I drilled slowly, in time I would achieve my goal.

My next step was to hand-drill a 12 foot deep test hole, and perform a thermal test on it to see how much hole I would need to satisfy my estimated heat load. My test indicated that I would need about 200 feet of hole.

Then, I tried to drill deeper holes, but my equipment and technique caused too many problems to proceed.

(At this point, you may want to take notice of the work that Vlad has done because he has built a very serious drilling machine that looks like it will be able to do what my equipment could not)

In the course of my drilling, I did learn a bit about the soil conditions in my yard. I learned that there is a 'hardpan' layer about 17 feet down, that is holding up a water table which manifests as wet, coarse sand. This high water table and wet coarse sand is very favorable for heat transfer.

Based on this, I decided to go with 16 holes each 17 feet deep, with connecting pipes at a depth of 2 feet. The frost line here is 9 inches deep.

So the 16 holes with an effective length of 15 feet each would give me 240 feet of borehole, more than enough, I reasoned, to do the job.

But getting back to the borehole vs trenching approach, if you have a large enough area, and if you can get and run a backhoe, trenches just might be the way to go. However, the deeper you can go, the more stable is the ground temperature...there is that trade-off. I think the most often quoted figure is 'deeper than 25 feet', there is not much significant temperature swing. So I realize that my loop-field will have temperature swings, especially the more shallow parts. The same for trenches & slinkies, so the heat pump needs to work harder, later in the winter.

If I could have gotten a backhoe into my backyard, I might have tried trenches, I did consider it.

If you have more than a passing interest in this endeavor, you might want to read the whole thread, because there's a lot of info from my experience, and the experience of others, and also many very informative information resources that you can link to from the thread.

Best of all, there are now some other brave souls who have taken up the challenge and are trying their own projects, borrowing some methods, inventing some others, all driven by their own unique skills, gumption, and understanding.

In more ways than one, we are breaking new ground!

I invite you to join us.

Regards,

AC_Hacker