The Homemade Heat Pump Manifesto

[bigsmile]

Originally Posted by AC_Hacker

Help is on the way...

-AC_Hacker

Got the file. Thanks a lot. With this, I may be finally able to put together a plan and get into action.

[AC_Hacker]

From what I have been able to gather, heat exchangers cost you about 10% in efficiency. I think this is one of the ideas behind HX systems... get rid of a HX and gain 10%.

I don't understand why no one is using DX floor heating, and cutting out that 10% loss also.

Seems to me that with copper tubing in the floor and all on the same level, the lubricant loss issue would not be so bad.

-AC_Hacker

[bigsmile]

If DX is used on the sink end, as well as the source end, this may be able to solve the problem of refrigerant imbalance between heating and cooling. During cooling, the refrigerant in the floor coil can be pumped to the ground coil to make up for the shortage, and an alternative air based evaporator coil will be used, which uses much less refrigerant.

[pachai]

Originally Posted by AC_Hacker

Great question...

What I need to do is over-charge the system, remove the electrical connecting parts, and submerge the whole frigging thing in a water tank and look for bubbles...that would do it.

Can you use soapy water from a spray bottle?
(That's how they find leaks on tires... :-)
Then you don't have to immerse a partly
electric device...

Seth

[strider3700]

copper in the floor isn't used very often anymore since it reacts with the cement and will develop leaks eventually. PEX is much more common but I don't know if it is ok with refrigerant or not.

[AC_Hacker]

Originally Posted by pachai

Can you use soapy water from a spray bottle?
(That's how they find leaks on tires... :-)
Then you don't have to immerse a partly
electric device...

Seth

Seth,

Thanks for your suggestion. I did try the soap-bubbles method previously and I couldn't find the leak.

...and you're right about the electronic part, but that is all easily detachable from the top of the compressor (one nut & three spade connectors), leaving nothing but hermetically-sealed mechanical parts.

The problem is that the leak is really small, about 2% a day, so I'm afraid I'll need to use extreme methods.

In the meantime, I'm running daily tests and carefully monitoring COP, to determine what the most efficient refrigerant charge level is. So far, the COP is still climbing.

Regards,

-AC_Hacker

[AC_Hacker]

Originally Posted by strider3700

copper in the floor isn't used very often anymore since it reacts with the cement and will develop leaks eventually. PEX is much more common but I don't know if it is ok with refrigerant or not.

Strider,

Yeah, I know about the caustic nature of concrete, and how it will eventually eat copper... it eats aluninum too.

But the potential for increased efficiency is really tantalizing. Another thing that I keep thinking about is the fact that the refrigerant lines would surely have a smaller diameter than PEX and a thinner, lighter floor might be possible.

This leads me to another problem...

I'm trying to think of a better way to do a warm floor, a way that would minimize thickness and especially weight but maximize conductivity.

Here's the problem as I understand it:

WET SYSTEMS

(PEX pipes cast, in place, inside of concrete)
So far, the most efficient floor construction for radiant floor is a concrete slab whose thickness is 3 to 4 inches thickness, well insulated from the bottom (3 to 4 inches, or more of high-density foam), with a good conductive layer on the top (no cover at all over the concrete, or thin linoleum or ceramic tile). The feed tubes are closely spaced (around 6 or closer), of larger diameter (5/8 or larger) and are near the bottom of the concrete slab. The thermal mass of such a slab is very large, it takes about two days, maybe longer to bring it up to temperature. But the temperature of such a slab is very even across the the top, so you end up with a huge, evenly-heated radiating area. All of this is done so that the feed temperature can be as low as possible often at 85 degrees, possibly as low as 80 degrees F, allowing the heat-pump to work at it's highest COP. The down sides are that the system doesn't respond to temperature changes quickly. it is very heavy and expensive. This system is reserved for new construction, on-grade work.

The next best is a thinner slab, maybe 1.5 inches thick of concrete or what is known as Gyp-crete, which uses Gypsum as a filler to produce a lighter floor. The conductivity of Gyp-crete is lower than cement and it is more expensive. Because if the reduced thickness, the surface temperature is less even and there are temperature peaks and valleys. To provide sufficient room heating, the feed temperature must be higher, somewhere between 90 and 100 degrees F. Here the heat pump COP will not be as high.

DRY SYSTEMS

(PEX or other pipe inside, or on, or under more common construction materials like dry-wall, plywood, wood or MDF etc. sometimes with aluminum fins to aid heat conduction.)
These systems have the lowest thermal mass, and can most easily adapt to temperature changes. They are least expensive to make, and can easily be used in suspended floors of conventional construction. Vlad has constructed such a system to heat a home he is building. He is using aluminum fins and also aluminum foil to direct heat up to the floor surface. I am very interested in how Vlad's floor will perform.
There is another way where foil-covered MDF "puzzle-pieces" with pre-formed grooves for the PEX, are nailed down to the top of an existing floor and the PEX is forced into the grooves and a covering material goes over the top.
Usually the the temperature peaks and valleys is higher for these floors than the 1.5 inch gypcrete floor and the feed temperature of these floors is higher, usually running around 95 to 105 degrees F.

So, with that as a history, I am trying to come up with a wet system floor that would use a material that would be lighter than concrete (density = 145 lbs/ft3), that has a much higher index of conductivity (10.0 Btu-in./hft2F)

It should also, if copper tubing is used, have a similar index of thermal expansion. PEX will flex, so expansion doesn't seem to be a problem.

I tried some experiments with PEX cast in concrete samples, but the experiments were poorly designed and didn't indicate anything conclusively.

I also tried concrete with aluminum bits as an aggregate, which was a disaster because alumimum reacts with concrete and causes it to foam up.

The material for the wet system doesn't need to be concrete...

I'm open to all ideas no matter how far-fetched.

Regards,

-AC_Hacker

[Jay-Cee]

AC,
I have to express my appreciation for this thread. I stumbled across it yesterday afternoon and have been like a kid in a candy store reading through the entire thing from page 1. Awesome job to you and Vlad and all others who went out and "did something". I have been entertaining the idea of GSHP for some time now and am thrilled to find a place where many of you have taken the dedication and initiative to go beyond the "entertaining the idea" stage and actually do something about it, whether right or wrong. Very impressive to all of you, I am excited to see how it turns out and am looking forward to taking some action myself.

I have noticed AC that you stress over and over in the post how "insulation, insulation, and insulation" should be one's first call to action so I will start there and also get the GSHP manual you talk about and begin educating myself further.

I have an incredible resource at my disposal in the form of a 15 acre lake positioned 65' from the back of my house and its 13' deep avg. I think this may allow for some very EASY and flexible testing scenarios this spring without all the sweat equity you gents have had to do. However, I don't know a whole lot about HVAC and especially the particulars regarding heat pumps and thermodynamics, but I have certainly learned a lot in the past 24 hours.

As to the radiant flooring, it would seem the most bang for the buck would be a "staple up" method underneath with the foil fins, covered on top with durock and tile flooring. The durock would be bonded to the original subfloor with screws AND thinset mortar, the tile would be bonded to the durock with thinset mortar creating a pseudo stone surface thickness of around 1". The question is, can you live with the inefficientcy of the not very heat conductive wooden subfloor standing between your heat source and stone floor?

Something that hasn't been mentioned previously, and I am surprised Vlad did not run into some of this, is the height factor for anything on top of the floor other than perhaps the MDF foil "jigsaw puzzle" you mentioned (I am familiar with). If a person were doing an entire floor, rather than just a room, and I am guesstimating here with Vlads floor. He put down 3/4" furring strips, then covered floor with 5/8" ply, so now he's at 1-3/8" up with a bare floor. What type finished flooring? hardwood, add 3/4". Tile with underlayment, add another 1-1/8". If he used tile (without cementitous underlayment) or hardwood, he has raised his floor by 2 to 2-1/4". Can he still open his exterior doors? Did he remove them and cut some of the header out and re-install? Will he have to remove/reinstall all of his interior doors as well (could undercut the interior door jambs and simply saw the bottom of the interior doors.)

It seems that if a person is not going to go some route with a poured concrete substance (sometimes the juice isn't worth the squeeze!) that doing a staple up scenario would yield as effective as what Vlad did. I say that because if we really look at what Vlad did.... he put radiant heating on the bottom of a new subfloor that sits on top of the old subfloor. Would his results be the same had he put it on the bottom of the old subfloor and sheeted the bottom of his unfinished (I am assuming they were unfinished) floor joists? I only bring this up as a consideration of cost/effort, not to undermine Vlads' work. He has done far more than I. Thanks again to all, I am loving this discussion/thread.

[Jay-Cee]

AC,

I was just sitting here thinking how can a person find a compromise between a full concrete floor vs a staple up scenario??? Perhaps this is a solution:

What if a person used metal studs as a concrete form to make "heat slabs" of a certain dimension? Galvanized metal studs used in commercial construction are thin, light, good transfers of heat, and cheap. They are shaped like a "U" and if layed flat, would yield a "tray" that was 1-1/2" tall x 4",6",8", whatever dimension you like depending on cost and availability.

Let's say you could find 2x6 metal studs on craigslist for a couple bucks each. You could run 1 tube down the middle, because they are easy to cut and work with, you can put them end to end to make a tray that you can run from wall to wall and then fill with the sakrete product of your liking. You can also do them 1 at a time making this a DIY dream. This would give you a preformed "slab" with a heat tube running right through the middle. Maybe you only place a run every 12" in the room, cutting your concrete weight down by x% but still giving you some mass. Fill the voids in between with foam and wood for subfloor support and something to fasten to. If you aren't worried about the raised floor height, this may be a great solution. If you used something cementitious as a substrate (i.e. durock, hardiboard, etc.) then you could further strengthen the thermal bond by applying mortar to the tops of your "heat slabs" as you go.

It seems this may allow each person to micromanage the weight/efficiency trade off .

[North_Pole_Guy]

AC Hacker,
I would like to start by saying great blog..
I live in the interior of Alaska and have been interested in GSHP for quite a while now, I worked for a commercial outfit about 14 years ago installing them in commercial and residential settings.
I am also very interested in re purposing other peoples "junk", I scrounged up a de-humidifier a few months back with the intention of hacking into it to make a small scale GSHP to test out. I have ground water at about eight feet and also heavy equipment to move dirt (40 acres) so burying field loops should be no problem just the material and diesel fuel cost to run the dozer.
I have been planning on building a new house for a few years now reading on super insulating and heat recovery ventilators and such.
what I am interested in at the moment (its cold outside) is more pictures/drawings of how exactly you tied the tubing from the compressor to the heat exchangers. I saw the pictures of where you cut the tubing apart and what you reused. I was wondering what controls the cold side temp above freezing your exchanger up, or is that a function of how much water is run through it?
I test ran the de-humidifier the other day after partial dis-assembly and the condenser got hot and the evap side got cold, I disabled the cooling fan by pulling a wire and ran it that way for a while.
thanks
Joe