The Homemade Heat Pump Manifesto

[BradC]

Originally Posted by AC_Hacker

I really appreciate your idea about brazing. It had already occured to me to do that, but the thought of how much heat it was going to take to braze the threaded pipe connections stopped me in my tracks... I was afraid I might un-braze the HX. Now that I know that it is possible to braze this up correctly, I'll give it a try.

Steel transmits heat a *lot* slower than copper or Brass. Wrap the HX in a wet (not damp, Dripping wet) cloth and heat the brass/copper mostly. The process of heating the copper/brass will transmit heat through to the steel. If you try and heat the steel to temp first you'll probably melt the brass in the process. It's a delicate balancing act. I've not brazed a PHX, but I've brazed brass fittings to steel drums.

Originally Posted by AC_Hacker

Your GSHP cooler project sounds wonderful.

I'm not doing anything new by any stretch of the imagination, but thus far I've spent about $300 in parts so I'm *way* ahead of the curve.

• Originally Posted by AC_Hacker

  What are the daily high temperatures you have there?

  The days we need the A/C are between 32 and 40 Deg C
  
  

  Originally Posted by AC_Hacker

  And what is the temperature of the water you are bringing up from your aquifer?

  Generally between 18 and 20 Deg C
  
  

  Originally Posted by AC_Hacker

  Are you also using Propane?

  Yep, the Nazis that regulate refrigerant here have the whole show so far closed it's almost impossible to get accredited to handle refrigerant without tossing in your career and doing a 4 year apprenticeship. Not a good move for a 30-something with a mortgage.
  
  

  Originally Posted by AC_Hacker

  What is the capacity of your compressor?

  I have 4 I'm playing with. The main two are as follows :
  One is a single phase ~7KW cooling capacity recip, and the other is a 3 phase ~18kw scroll. I've got a VSD for the Scroll, but I've not hooked it up yet, so most of my experiments thus far have been with the 7KW recip, or a couple of small rotaries (~14cc displacement) from window A/C's
  
  

  Originally Posted by AC_Hacker

  What type of metering device are you using?

  Right now I'm using the TEV that came with the first fan coil unit.
  
  I've just acquired 4 EEV's (three little ones ~4KW each and one ~15KW). I've just ordered some stepper driver chips and I'll build my own EEV controller to play with.
  
  

  Originally Posted by AC_Hacker

  What are the details on your heat exchangers?

  Well, somewhat agricultural to start with, I've tried ~18M of 7/8" coil in a 208L (We call them 44 Gallon, but I think they are ~55US Gallons) drum. Worked ok-ish, but after seeing what you are up to I'm looking for a PHX.
  
  

  Originally Posted by AC_Hacker

  How about your water pumps?

  Only 1 water pump as I'm using DX head units. I'm using a three phase bore pump I usually use for reticulation.
  
  

  Originally Posted by AC_Hacker

  Are you using one well for a source and the other as a sink?

  Not yet. I have one bore and I'm pulling from that at just below surface level and re-injecting back into it a couple of meters deeper. This is working for my experiments but I'm not convinced of the long term viability of this. I need to get things stable enough to do some longer term testing to watch the temps coming out of the ground.
  
  

  Originally Posted by AC_Hacker

  Are you using this system to cool the air?

  Yes. Currently I'm using the DX fan coil I got with the condensing unit, but I've just picked up a second one that I'm re-plumbing to run multi-head. Ideally I'd like to be running a chilled water loop rather than DX, but I just had a quote for over $2,000 for a couple of chilled water fan coils, whereas the last DX unit I got for $20 from the recycling station.
  
  

  Originally Posted by AC_Hacker

  How big is the air HX?

  Physically or in capacity?
  One is about 60cm x 45cm x 50cm and ~7KW total cooling, the other is about 120cm x 60cm x 45cm and *should* be good for about 14kw total cooling (far bigger than I need thus the need for the EEV)
  
  

  Originally Posted by AC_Hacker

  How many CFM are your fans?

  Both FCU's are good for about ~1000CFM.
  

Now, I "cheated" to start with. In exchange for a donation to a nominated charity, I managed to score a 7KW split system heat pump. I started by installing this and converting it to R290. I've then been plumbing various heat exchange devices in series with the air cooled condenser to experiment with lower head pressures. Out of the box the unit has a COP ~2.5, and thus far I've been able to double it by reducing the condensing temperature down to about 28 degrees.

My aim is to get the EEV's running, plumb in the second head and get the control algorithms stable. Then play with some PHX units in place of the existing condenser to try and get the condensing temperature down further. Also with the EEV's, I *should* be able to raise the evaporating temperature in response to load drops. Of course the closer the evaporating temperature to the condensing temperature, the better the COP. The other thing I need to look at is a dedicated submersible pump, preferably with speed control.

Because I inject the return bore water below the surface, the pump is effectively pumping against zero head, so it's power consumption is less than it would normally be.

Once I get that stable I'll try replacing the compressor with the scroll and see if I can improve it further by slowing the compressor as the load decreases.

Ultimately all I'm doing is taking a conventional split system heat pump and making it more efficient. I'm lucky in that I have an practically endless supply of ground water at stable temperatures.

What I did was find a scrap metal dealer than gets occasional loads of large A/C gear, and I go up and scrap it (break it down into clean as possible constituent metals) in exchange for any bits I might need. In other words, I get to spend the odd Saturday cutting A/C units apart with a 4" grinder and I can take home the bits I want.

[BradC]

Oh, while I remember. If you don't go the brazing route for the PHX, there is a gooey refrigerant compatible sealant affectionately known as "snot". It's product name is Nylog. Removing the teflon and using some of that instead might solve your problem. It remains elastic and is pretty tenacious stuff. I know it's available in the US, and it's not very expensive.

Don't overdo it though as if it gets into the refrigerant loop it might block your capillary tube.

[AC_Hacker]

BradC,

I gotta say that I am completely in awe of your scrap supply. I need to snoop around here to see if I can find a cornucopia like you've found.

Ebay doesn't look so good to me any more.

Originally Posted by BradC

Steel transmits heat a *lot* slower than copper or Brass. Wrap the HX in a wet (not damp, Dripping wet) cloth and heat the brass/copper mostly. The process of heating the copper/brass will transmit heat through to the steel.

Good advice on the brazing. Regarding heating the copper first, I also read some similar advice on a HVAC blog, describing brazing up a Danfos TXV. Thanks for that.

Originally Posted by BradC

I'm not doing anything new by any stretch of the imagination, but thus far I've spent about $300 in parts so I'm *way* ahead of the curve.

Well, you may not be pioneering any new technology, but in my opinion, being able to convert scrap into functioning, efficient HVAC equipment really is different. Not only because you are able to make what you need, that you may not be able to afford otherwise, but also because you have avoided an additional burden on the old ecosphere.

I'd say that entitles you to a magnificent trophy, so here it is:

This trophy presented to BradC
for world-class scrounging.

I think you can probably print it out and put it on the mantle, or hang it on the wall.

Originally Posted by BradC

Yep, the Nazis that regulate refrigerant here have the whole show so far closed it's almost impossible to get accredited to handle refrigerant...

Yeah, I heard that the ozone layer destruction was much more serious in Australia than in other places. I heard stories about sheep going blind, and school children being required to wear hats when playing out side at school. Is this true?

I personally take the ecology warnings pretty seriously. It's a very long way to the next habitable planet, and I'm probably on the 'Do Not Fly' list.

Originally Posted by BradC

One is a single phase ~7KW ...the other is a 3 phase ~18kw scroll...a couple of small rotaries (~14cc displacement) from window A/C's.

You're working with some pretty good sized stuff there. You must have noticed that my stock is quite small in comparison. My philosophy is to make insulation my first priority, and to make up the difference using the smallest, most efficient HVAC technology I can build myself.

Originally Posted by BradC

Right now I'm using the TEV that came with the first fan coil unit.

I'm right now in the process of designing my next unit, and I'd really like to use a TXV in it. I have a Danfos TXV that uses inserts to give it a wide range. I selected an insert for about 2.6KW. I know that cap tubes are used in units up to about 5KW, but I really want to be able to change the metering with a screwdriver, rather than the cap tube twiddling method. There's a lot of experimenting I just haven't done because of the hassle of changing cap tubes.

Originally Posted by BradC

I've just ordered some stepper driver chips and I'll build my own EEV controller to play with.My aim is to get the EEV's running, plumb in the second head and get the control algorithms stable. Then play with some PHX units in place of the existing condenser to try and get the condensing temperature down further. Also with the EEV's, I *should* be able to raise the evaporating temperature in response to load drops. Of course the closer the evaporating temperature to the condensing temperature, the better the COP. The other thing I need to look at is a dedicated submersible pump, preferably with speed control.

I think I was perhaps too quick to bestow the trophy... This sounds really interesting! I was thinking about something like this just a few days ago, how cool it would be to have automatic, programmable feedback in the metering device.

I don't know if you know about this one, but I came across a paper that was about computer modeling of the action of cap tubes. The action of a cap tube is way more complex than I would have thought, and as I recall they are able, within a certain range, to adapt to changing conditions. But programming them is pretty crude, what with cutting and brazing and all...

Originally Posted by BradC

Well, somewhat agricultural to start with, I've tried ~18M of 7/8" coil in a 208L (We call them 44 Gallon, but I think they are ~55US Gallons) drum. Worked ok-ish, but after seeing what you are up to I'm looking for a PHX.

I really like them. They can be expensive, though, especially in larger sizes. I did try to make my own heat exchanger... I bought some copper coils to make a tube-in-tube exchanger, and I realized that I had spent almost as much on the copper as a new Brazed Plate would cost, and I still wasn't sure what the capacity of the homemade unit would be or its efficiency.

By the way, earlier in this blog I published a formula for figuring the heat transfer capacity of a Brazed Plate HX. I now realize that I was overly optimistic. I have changed the formula (the final multiplyer was changed from 5000 to 1950) and it works pretty good for brazed plate HX in the 7.5 inch x 2.8 inch x (any number of plates). For other dimensions, it’s way off.

I am working on a revised approach that should simplify selection. Until then, I would advise using a program called FlatPlateSelect, and double-checking that with data from brazed plate selection charts, from more than one manufacturer.

I have gotten selection advice from seasoned HVAC technicians, completely different results from the computer program, and still different results from the charts... however, I think I’m starting to zero in on this one.

Originally Posted by BradC

Only 1 water pump as I'm using DX head units.

OK, you are using refrigerant-to-air heat exchangers, right?

Originally Posted by BradC

Ideally I'd like to be running a chilled water loop rather than DX, but I just had a quote for over $2,000 for a couple of chilled water fan coils, whereas the last DX unit I got for $20 from the recycling station.

Wow, “over $2,000 for a couple of chilled water fan coils”, that’s way too steep.

How about this... I came across this paper called, “Low Temperature Heating and High Temperature Cooling”, and researching that idea a bit led me to information that the Germans are experimenting with radiant cooling in the ceiling. The idea is that to achieve the same level of comfort (which is not the same as ‘the same temperature’), the temperature of cooling water in ceiling panels can be warmer that water in air-cooling heat exchangers, thus the power required is less.

I don’t know how things are where you live, but used copper solar heating panels go a-begging here. And the panels, which have water conductive tubing in them, could stripped out of the collectors and used like the Germans are using them.

Might be worth a try in a small room...

I don’t know if you are a digital camera kind of person, but it would be great to see some photos of what you’re doing.

Best Regards,

-AC_Hacker

P.S.: I'm hearing terrible stories about walls of water raging aross parts of Australia.

Sounds just horrible, any of that happening near you?

* * *

[Vern2]

This post was off topic.

[Xringer]

Originally Posted by Vern2

I found this out of the box thing of interest because it does not use a HX, just some pipe stuck into a tank. I suspect replacing the condenser, don't know. It seems to be a very popular hot water tank add on. This unit is there A7 12,000 btu unit. Next month the 24,000 btu unit will be available, Feb 2011.

This link is their install demo. http://www.airgenerate.com/installation_demo.php It seems like an efficient way of heating water. Do you think this design may take the place of the HX and still use a ground loop for heating water for floor.

This unit has been around for a while.
I think they got hit by a hurricane a while back (Houston maybe?) and went off the grid for a while.
The next time I saw them, their price had jumped up about 40%..

Humm, I should make DIY unit for summer only..
I wonder how hot the out-put coil gets on my new dehumidifier?
Have to check that out this spring.. Or maybe in July..
(Man, do I ever hate the dead of winter)!


If my basement wasn't like a meat locker, I could use one of these..

[AC_Hacker]

Originally Posted by Vern2

I found this out of the box thing of interest because it does not use a HX, just some pipe stuck into a tank.

If I'm not mistaken, it does use an HX, in fact, two of them. One is an HX (air-to-refrigerant) to draw heat out of the air and the other is a HX (refrigerant-to-water) to put heat into the water tank. The "pipe" that sticks into the water tank is the second HX.

Also, in many states the HX that is in contact with potable water must be of 'double wall' construction.

Originally Posted by Vern2

It seems like an efficient way of heating water.

For sure... Not as efficient as solar, otherwise it's about as good as it gets.

Originally Posted by Vern2

Do you think this design may take the place of the HX and still use a ground loop for heating water for floor.

I'm not quite following you here... Please elaborate.

-AC_Hacker

[Vern2]

This post was off topic.

[AC_Hacker]

Originally Posted by Vern2

Was just thinking storage on peak demand would be better. But I guess your thinking is to have a continuous source of 90 degree water without storage, to floor loops.

OK, now I understand... I agree with you, I think the water storage tank is a really good idea. In fact, I have my eyes open, trying to find a small used water heater, 30 or 40 gallons or so, to use for testing. I don't know how low the temp switch on an electric water heater goes, but if it's around 90 degrees, I could just wire my heat pump to that. Wouldn't that be easy! What would be ideal would be to have two temp switches, and have the other one set a few degrees lower for the electric heating element, so that if the heat pump failed, or if the outside temperature was just too cold for the heat pump to supply enough heat, the water heater could kick in.

Another bonus of the tank idea is that it would prevent 'short cycling'. I've done a heap of tests with my little heat pump and I see a repeating pattern that it takes about 5 minutes to get to the beginning of the efficient range. So running my particular heat pump for shorter durations is inefficient, and I think running for maybe 20 to 30 minutes would be optimum. The length of these periods will be different for every heat pump and heating situation, but I think the basic pattern will be pretty much the same.

By the way, I've been running my little heat pump (about a third of a Ton) continuously for about three days now and it's actually picking up a goodly share of my house heating requirements.

In all honesty, it has been unseasonably warm the last few days, so I'm not ready to start bragging... not just yet.

-AC_Hacker

[Vern2]

This post was off topic.

[S-F]

Originally Posted by Vern2

AC Hack,

And yes R30 in roof and R11 in walls. Plus insulated garage doors.

You might first want to think about R60 in the walls and something more in the roof. R11 really isn't that much. I'm about to start construction and the friend who's helping me build recently finished his father's house. R30 in the walls and R60 in the ceiling. He only needs to run his central AC for 20 minutes a day to keep it cool in the SC summer.