The Homemade Heat Pump Manifesto

[Hugh Jim Bissel]

So we've got holes in the ground with pipe sticking out of them at ground level. You can't have the pipes running on top of the soccer field, the poor soccer players have it bad enough running back and forth, they don't need a trip hazard as well. Besides, the pipe needs to be down 4' or so to insulate the cold water from the hot sun. So out comes the backhoe, and surprise, surprise, the utilities we marked (phone, data, and electric) were right where the locater said they should be. We went by the cad drawings when trying to figure out where the water line was though...silly us












The drawings said the water line should be under the street. Good thing the drawings were right on. Oh, wait...


It's just a flesh wound!


Now the trenches are dug, the horizontal piping is connected to the borehole loops...


And all the piping leads to the next post!

[Hugh Jim Bissel]

So we've got 230 boreholes. We divided them up into 17 different zones, and each zone has 13 or 14 boreholes in it. Those wells are plumbed in parallel reverse return (first in, last out) and each zone comes to the vault and has its own sensor ports and shutoff valves inside.


With all most all the pipes hooked up. 2" in and out to the wells, 6" in and out on the left end to the pump room and the buildings.


View from another angle. The original plan was to use the field for 3 different buildings.
1: The one we converted
2: one that was going to be built and was put on hold indefinitely before even a slab was poured (we'll probably end up converting another building on campus)
3: the warehouse building in the background of overview 2. (the piping has been run into the building, I'll be stubbing out piping in the next week or two when the parts come in, and it may be months or years before they hook up the water and use ground source units)

So in addition to the 2" well piping, the 6" is splitting up to feed those three buildings.


The only building using the field so far is the one story building in the background. the pipe runs under the parking lot and runs up the side of the building just to the left of the door and window.


And the pipe is somehow able to get past MS Windows Vista's Firewall! Found this just as it is online. Alas, the firewall has now gone the way of the bushes that used to surround the chillers (which cool the 3 story building). The lower left square of sidewalk was pulled up to run the pipes through, and the pipes go just to the left of the telephone pedestal on the far left.


Just inside the building is the pump room. The two biggest closer pumps run the water through the wells, and the two smaller pumps behind run the water through the building (only one pump of each pair runs at a time).



And thats about it for now. The building has been up and running on the new system for probably a month or so now (everyone had moved out to allow for some remodeling along with the HVAC conversion), and we just got grass seed down on the soccer field this last week. I need to get my hands on a camera and take some pictures of the pump room now that the pumps have pipe going to them, and take some of the inside piping and HVAC units.

[AC_Hacker]

HJB,

Great photos! And thanks for other information too.

Your info on grout:

50 lb grout mix + 250 lb silica sand

... though not identical, is similar to "MIX-111" which was hammered out (in open-source fashion) by a consortium of GSHP people when the state of New Jersey shut down that state's GSHP business until a grout could be developed to prevent cross-aquifer contamination and (only secondarily) to provide for good thermal transfer.

Here's a link to MIX-111:
Information Bridge: DOE Scientific and Technical Information - Sponsored by OSTI

Regarding where to put audio files, here's a link to the Internet Archive. You can upload any kind of file there:

Internet Archive

... go to the "upload" button at the upper right. It's free, also free of advertising, and will be there just about as long as the Internet exists.

I just checked and this blog doesn't seem to be part of the Internet Archive yet, but it should be in a few months. If it doesn't show up in a few months, we'll have to do something so as to insure that it does. This pool of information is too important to slip down the memory hole.
_ _ _

HJB, I really appreciate the knowledge you've gained working on such a large project. And thanks for stepping forward to share this knowledge and experience with us. You are in a position to have unique insights as to how this technology can be scaled down, so as to be accessible to average citizens who may not be wealthy, but who are smart and resourceful.

Best Regards,

-AC_Hacker

[Hugh Jim Bissel]

Originally Posted by AC_Hacker

Here's a link to MIX-111:
Information Bridge: DOE Scientific and Technical Information - Sponsored by OSTI

Regarding where to put audio files, here's a link to the Internet Archive. You can upload any kind of file there:

Internet Archive

... go to the "upload" button at the upper right. It's free, also free of advertising, and will be there just about as long as the Internet exists.

Their grout appears to be using cement as the main binder. Ours was just water, sand, and bentonite. (gotta be the sodium bentonite that expands to 10 times or so the volume in water) We used 5 to 1 sand & bentonite for max thermal transfer, but the ratio normally used is 4 to 1 because it pumps better.

I'll have to start working on cleaning up the audio and get them uploading. Thanks for that link!

In the class they were even saying if you took a (I think a specific brand or type, maybe variable speed compressor?) air to air heat pump and replaced the outside coils and fan with a refrigerant to water heat exchanger, and a water pump; you'd have a better ground source unit than what's on the market right now!

[Perry525]

This seems an excellent idea
I would mention two things.
One is that using a concrete slab as a heat source has its problems, in as much as, when the days start cold, it puts out heat, hopefully as a back ground heat but, once the sun comes out and the day warms up.... it is very slow to cool down. Whereas a light weight floor will react more quickly to warming up and more important will cool much more quickly.
Two, surely running the pipes say six foot down round the plot will pick up low level heat much the same and will be a lot easier to install. Keeping in mind that the heat exchanger will raise the output temperature.
Having thermal as a background heating system will extend the warm days of summer into the evenings and reduce the overall cost of winter heating. But, on cold winter days and nights it probably won't be able to keep the place warm.

[AC_Hacker]

HJB,

Originally Posted by Hugh Jim Bissel

In the class they were even saying if you took a (I think a specific brand or type, maybe variable speed compressor?) air to air heat pump and replaced the outside coils and fan with a refrigerant to water heat exchanger, and a water pump; you'd have a better ground source unit than what's on the market right now!

I've had the same thought myself.

Yes, I think that the variable speed compressor is the biggest part of the efficiency, but also the fan in the inside unit is continously variable too. I bought a little (9,000 BTU) mini-split to heat the core of my small house while I'm getting the GSHP up and running. This thing works really great, it's cheap to run and is unobtrusive. The fan on the inside unit runs all the time, but when the set-temperature is approached, the fan gradually reduces it's power and speed until it is just the faintest whisper.

And yes, I have considered modding my mini-split to use a ground source water loop... The R-410a would have to be evacuated & saved, then re-injected after the air-to-refrigeration exchanger was replaced with a water-to-refrigerant exchanger. but it's possible.

The only thing I can think of that could possibly be a problem is the controller on the mini-split... It monitors the temperature on the outdoor ait-to-refrigerant exchanger for temperatures that would suggest frost build-up. When this temperature is reached, it begins a de-frost cycle. With a ground source water-to-refrigerant exchanger, this shouldn't be a problem if the loop field is VERY large (aprox 2x the normal design size) or if antifreeze is used.

If you tried to mod the outside exchanger AND the inside exchanger for say, a radiant slab floor, the complexity would really increase. The temperature sensor and controller receiver is in the inside unit... also cooling with a chilled slab is not so successful because of condensing water vapor.
_ _ _

There are two efficiency ratings on these things, SEER and HSPF.

The SEER rating applies to the cooling function and the HSPF rating applies to the heating function. The current government-mandated minimum SEER rating is 13.

The other rating, HSPF is the heating efficiency rating. The current minimum is 7.7.

I wasn't so clear on this when I bought mine, and just went for a higher SEER, which I wrongly thought applied to both. Mine has just the minimum HSPF and I live in an area where heating predominates. It is still much more efficient than the heat source I was previously using (gas forced air), but I could have done even better.

The brand I got was a Sanyo. It had a HSPF (Heating Seasonal Performance Factor) of 7.7... they are now making a model with HSPF = 10.2.

Go to ebay and search for "heat pump inverter".

Where you live, you should give greater attention to SEER.

Best Regards,

-AC_Hacker

[Hugh Jim Bissel]

Originally Posted by AC_Hacker

The only thing I can think of that could possibly be a problem is the controller on the mini-split... It monitors the temperature on the outdoor ait-to-refrigerant exchanger for temperatures that would suggest frost build-up. When this temperature is reached, it begins a de-frost cycle. With a ground source water-to-refrigerant exchanger, this shouldn't be a problem if the loop field is VERY large (aprox 2x the normal design size) or if antifreeze is used.

If you tried to mod the outside exchanger AND the inside exchanger for say, a radiant slab floor, the complexity would really increase. The temperature sensor and controller receiver is in the inside unit... also cooling with a chilled slab is not so successful because of condensing water vapor.

It depends on what it's measuring the temperature of once it's converted. If it's necessary to keep that temp sensor, it should be measuring water temp since air temp is removed from the equation. I'm not sure I understand what you're meaning by "this shouldn't be a problem if the loop field is VERY large (aprox 2x the normal design size) or if antifreeze is used":

-Design size should take everything into consideration (heat added & removed from ground by heat pump, & avg ground temp) to result in a field which will keep the water temperature consistent over time (temp will vary month to month, but should be back to same temp in 1 year). I was given the impression if you size the system for your larger load (heat vs cooling dep on climate) your water temp will stabilize even though the heat pump is creating an imbalance (ie more heat added from cooling than heat removed to heat the building); it's only when the syst is improperly sized or installed that the temp will keep climbing or falling. I'm not sure how accurate this is though, since many installs around here have dry coolers* to pull heat out of the loops in the winter (though we do have a lot longer of a cooling season, and compressor heat is working against us instead of for us like in your primarily heating climate). The book talked about in cooler climates needing supplemental heat and not getting 100% of your heat from ground source. I'll need to look that up; I don't recall too much explanation why that is: perhaps not enough heat getting back into the ground to be sustainable. Someone in one of the ground source forum topics was talking about running the loop water through solar heaters. Doing that, or running a dry cooler (dry heater?!?) in the summer might be a solution to that issue.

-In a cooler climate (don't remember how cut-off is determined off the top of my head) you should use antifreeze no matter what. The defrost thermometer won't know it's looking at water or at what temp the water will freeze (if system contains antifreeze or not). Also, the whole idea of using water is that when it's 0F out and your air heat pump is freezing over, your water loop is 40F+ and doing just fine: the loop water in your climate shouldn't be approaching freezing, just like in mine it shouldn't be over 100F. The antifreeze, is because the water comes above ground it has potential to freeze: ie if power in the house was out and temp went below freezing inside. Poly pipe won't be affected by water freezing in it, but copper coils don't like it so much! We found that out the expen$ive way with the dry cooler on another of the ground source systems on campus!

*By dry cooler, I'm meaning coils the water travels through, while outside ambient air is blown through the coils (basically a giant radiator: no compressor involved). In our climate the water gets too hot, so we run the dry cooler when the air temp is significantly lower than the water temp (20F or greater difference is the number that's coming to mind).

------------------------

In regard to water to water, what I'd probably do would be to have one water to water unit dedicated to heating hot water in a storage tank before a regular or on demand water heater (or use a water to air unit that also has hot water preheat built in). Then run my in floor radiant as a pumped loop from/to that hot water storage tank. I'd also have a smaller water to air unit to cool in summer, and help warm the air up faster than the radiant could alone (cause of the mass of the slab/floor). That's just my thoughts, I don't know what a professional would do (though now that I've been through the install class I'm WAY too close to being a professional for my own comfort! )

------------------------

The other thoughts I've had about air to air heat pumps would be to have the outside unit enclosed so it was pulling air through tubes in the ground: Same idea as water to air: use the ground temp to raise the units input temp in winter and lower it in summer compared to the ambient air temp resulting in better efficiency. This is probably an in between in regard to efficiency: it would be better than a regular air to air especially during the extreme temperatures, but you'd need to bury a heck-of-a-lot of tubing to be able to even start to approach the efficiency of a water to air.

[AC_Hacker]

HJB,

On the whole I agree with your approach, but there are a few areas where our opinions diverge a bit...

Originally Posted by Hugh Jim Bissel

It depends on what it's measuring the temperature of once it's converted. If it's necessary to keep that temp sensor, it should be measuring water temp since air temp is removed from the equation.

I think that one approach here would be to identify what kind of component is used to monitor temperature. Most of them I've seen are pressed into the heat-exchanger fins. Most likely it's a thermistor (temperature-variable resistor). Then monitor the voltage across it at the moment when the unit goes into defrost mode. This would tell you what the set-point voltage would need to be. Then you could design a very simple thermistor circuit to measure outgoing water temperature on the ground loop side and design it so the voltage across it would be the same set-point voltage as before as the water temp approached freezing. This temp would be different depending on if you used antifreeze or not and if so, what kind and how much. You might get lucky, and just fastening the thermistor to the out-going water loop would work just fine. But it's something to look out for.

Originally Posted by Hugh Jim Bissel

I'm not sure I understand what you're meaning by "this shouldn't be a problem if the loop field is VERY large (aprox 2x the normal design size) or if antifreeze is used"

What I mean is that the heat pump can freeze water in the ground-loop heat exchanger if the loop field is relatively small. I know this to be true, because I have accidentally proven this, much to my embarrassment. If the loop field is sized large enough this will never happen.

If antifreeze is used, a smaller loop-field can be made before freezing of loop fluid happens. Economics of installation would favor the installation of a smaller loop field. However, the economics of long term efficiency might not favor a smaller loop field.

Adding antifreeze to the water also reduces the specific heat of water somewhat. The figure I have come across is that a 20% propylene glycol solution results in a 4% drop in specific heat, so this would require a slightly larger field.

See: Propylene Glycol based Heat-Transfer Fluids

There is also an increase in viscosity of the working fluid with antifreeze. This would call for greater power from the circulation pumps.

see: Calculating Friction Loss - Feature Articles - PM Engineer

The variety of issues are set out in the introductory paragraph of this paper:
http://www.google.com/url?sa=t&sourc...yA5OsjvRIZsgZg

The very next paragraph suggests that a computerized life-cycle cost analysis would be the best approach.

So, my thinking is that if you are building your own system, commercial price constraints are not exactly the same because you can do work yourself, working with your own tools, at your own speed. Working this way you can make choices that would be prohibitive if you were hiring a commercial installer, choices like making your loop field twice as large.

At a GSHP conference, I spoke with a contractor who knew of such a system in my area which did in fact use a 2X loop field running only water, and the efficiency was in fact enjoyably higher.

On the first page of this blog, I posted this graph:

...which was a test I did with my prototype heat pump on two 55 gallon barrels of water. The declining black line is the temperature of the barrel from which I am removing heat, not unlike the temperature of the water which is coming out of the loop field. Except the loop field temperature would look more like this:

On my firest graph, there is also a very jumpy light blue line which is the COP calculated for the last five minute interval (I multiplied the COP by 10 so it would show up properly in the graph. COP=60 really means COP=6). Notice that the trend line of the COP declines roughly parallel to the black line.

So all I'm really saying here is that since I'm building this thing myself, I'm trying to favor efficiency at every turn... so I prefer straight water for it's higher specific heat, I prefer water for its lower viscosity, I prefer a larger loop field for the lower seasonal decline in COP.

I really like your ideas and information about the "dry cooler".

Regarding hotter and cooler climates, you might be very interested in these maps:

Heating Degree Day Map:
http://lwf.ncdc.noaa.gov/oa/document...lheatingDD.pdf

Cooling Degree Day Map:
http://lwf.ncdc.noaa.gov/oa/document...lcoolingDD.pdf

You are quite right about exposed pipe, I'm planning to run pipe into my basement below the frost line.

Originally Posted by Hugh Jim Bissel

The other thoughts I've had about air to air heat pumps would be to have the outside unit enclosed so it was pulling air through tubes in the ground: Same idea as water to air: use the ground temp to raise the units input temp in winter and lower it in summer compared to the ambient air temp resulting in better efficiency. This is probably an in between in regard to efficiency: it would be better than a regular air to air especially during the extreme temperatures, but you'd need to bury a heck-of-a-lot of tubing to be able to even start to approach the efficiency of a water to air.

This is such a good idea!

Around here, where heating is the primary mode, some people are putting the 'outside unit' in the attic to benefit from the trapped heat (and leaked heat) there. I assume that the exhaust from the unit in the attic is completely directed out of the attic.

Thanks for all of your information...

Regards,

-AC_Hacker

[Hugh Jim Bissel]

Originally Posted by AC_Hacker

What I mean is that the heat pump can freeze water in the ground-loop heat exchanger if the loop field is relatively small. I know this to be true, because I have accidentally proven this, much to my embarrassment. If the loop field is sized large enough this will never happen.

You are quite right about exposed pipe, I'm planning to run pipe into my basement below the frost line.

Around here, where heating is the primary mode, some people are putting the 'outside unit' in the attic to benefit from the trapped heat (and leaked heat) there. I assume that the exhaust from the unit in the attic is completely directed out of the attic.

Thanks for all of your information...

I wonder what would be the most efficient way of raising your loop temp if it threatens to freeze. The choices I could think of were: use the units defrost cycle, shutting off the ground source and firing up your central heater, flushing/exchanging your loop water with cold tap water, or piping water through your hot water heater (ie flushing with hot water).

Once you've got your final, correctly sized system in place you shouldn't have to worry about freezing, but if you're loops are sized for your load and your water temp still drops year to year then you could think about a dry cooler or some other way to add heat back to the ground. You'd be in better shape than us with a dry cooler, cause you'd be only be running it during the summer, and could drain it before temps threaten to freeze.

Speaking of year to year temp drop, in the class the engineer said even a poorly designed/installed system will do great the first year. It's the 2rd to 4th year the bad ones begin to stand out (identified by water temp headed toward one extreme, rather than holding a cyclical constant)

Even in hot climates the pipe needs to be covered: First, if its not insulated where it comes close to the surface, even if there's no freeze potential, you're loosing/gaining heat to the outside air = less efficient. Second, the poly will begin to break down after a few years of exposure to UV (sunlight) so it has to at least be covered.

Putting the "external unit" in the attic in a heating climate is genius! It'll be less efficient for the little bit of AC you do (can counter that somewhat by having extra ventilation openings that are sealed off the rest of the year), but the rest of the time you're pulling the heat that rises out of the house right back inside!

No, thank you! I've enjoyed sharing, and brainstorming. I'm glad you're able to put it to good use, and I wish I was in a position to do the same. It's exciting to be able to help even a little. (cough-post-more-pictures-cough ) even if I can't mess around with a project of my own I can live vicariously through your project!

That Internet archive site, do you have first hand experience posting to the site? It looks like I can upload wave files and they'll convert to mp3 for me as opposed to me uploading an mp3? I gotta get cracking on that audio, I haven't even listened to any of it till just a few minutes of it just now. I have to get the voice recorder files into the computer, then I just want to edit out the extraneous parts, and also figure out which sounds better (recorded on voice recorder as well as direct into the computer, so I've got two different files of the same stuff). Hopefully I can get a segment or two uploaded by the end of the weekend! (as long as I don't get sidetracked modding my car for my upcoming trip)

[AC_Hacker]

I have been working on a Data-Logger to use for testing heat pump performance and just general data-gathering. I now have a working unit.

Here's the link to the detailed post:

http://ecorenovator.org/forum/projec....html#post3349

Regards,