The Homemade Heat Pump Manifesto

[AC_Hacker]

Have you thought of planting trees around your garage? That's what they used to do.

-AC_Hacker

[AC_Hacker]

I started the continuous heat pump run on 1/12/2011 and it us now 1/16/2011, so I've had five days to collect data and impressions.

• One of the first is that the ground loop I so laboriously constructed is a real asset, and the possibilities of how I can use it are coming to me almost as fast as I can write them down.
  
  
• The same can be said for the little heat pump. I can see ways that having a DX side instead of a water HX would be handy. For instance it would be more efficient to have a DX heat-gathering side in the ground. Less hole to dig. Also, having a DX side on the condenser side could be handy, because it is already part of an air conditioner, and would make the building of a GSHP shop warmer really easy.
  
  
• But as it is, with water-in-water-out, the refrigeration circuit is really small, so lubrication transport is a non-issue. Additionally, and probably the most compelling reason, is that changing the configuration is really a snap. All I need is some tubing and hose clamps and I can make it from a heater to a cooler to a de-humidifier to anything I want.
  
  
• There is a large and growing puddle in the cellar where the heat pump is. At first I thought it was rain seeping in where my loop field pipes enter the basement. But everything is dry there. Now I realize that there is continuous condensation happening on the outside of the pipes and tubing connecting the cold side of the heat pump to the ground loop. My initial fix is to get some pipe insulation and snug it up around the exposed pipe to keep the basement air away. I'm sure that this will work but I now recognize that the wet pipes also present an opportunity, and that is that they are de-humidifying the basement. I just need to capture the water. So... free de-humidifying... great!
  
  
• Another opportunity I see is that the incoming temp, around 40 F is also the temperature of my refrigerator. So, if I built an insulated box with water loops in the walls, I'd not only be getting cheap basement heat, but also essentially free refrigeration!
  
  
• Another opportunity is if I built a waste-water heat recycler and ran the loop water around that, the delta-T would be greater, improving the efficiency of waste water reclamation, a plus!
  
  
  
  
  
• Also, it has been raining a-plenty (water condensing from the sky) the last few days. I just went outside and measured it with an accurate thermometer and it read 56 degrees F... that's warmer than the temp of the ground, So another opportunity is to have an additional heat loop in a rain barrel (they fill up pretty fast) that could be switched in when ever the barrel temp exceeded the ground loop temp. A good application for a differential thermostat.

This has certainly been a very interesting and useful endeavor.

-AC_Hacker

[Vern2]

This post was off topic.

[AC_Hacker]

My vote is for more insulation in the shop...

...ant trees.

-AC_Hacker

[AC_Hacker]

I'm in the process of gathering the pieces for Heat Pump #2.

I've done loads of testing using my first, 'Proof of Concept' heat pump.

With the first one, I pretty much threw it together just to see if the mystery of refrigeration was within my grasp. I didn't know how to size it or anything, I just guessed... and it worked.

So with this one, I'm trying to apply what I have learned from my first attempt, and also the great advice I have gotten, and actually design a Heat Pump.

I'd like to thank the following people who have volunteered valuable suggestions, both here at EcoRenovator and other relevant blogs. Their contributions are being applied both directly and indirectly to this project:

• Fred_Fredowski
• bigsmile
• Gray Mole
• DetroitAC
• sheepdog
• Vlad
• OldChap
• wdrzal
• mytekcontrols
• BradC
• strider3700
• Jim-L.
• Ward Cunningham

...big thanks, guys!


THE DESIGN LOAD

My loop field was tested and built to yield about 12,000 BTU/hr and so far it's looking like I'm close to that. I may be on the low side, but I haven't done a high enough level of heat extraction to know what the limit really is.

So I decided to build this one to be larger than the first heat pump, which seems able to extract about 4,000 BTU/hr, and smaller than the designed maximum of my loop field, 12,000 BTU/hr...


COMPRESSOR SELECTION

So then I looked over my compressor list:

And the third compressor from the top, which has a capacity of 7800 BTU/hr seems like a happy alternative.


HEAT EXCHANGER SELECTION

On my spreadsheet...

• Column A is the compressor's capacity.
• Column H is what the evaporator's target heat load would be
• Column I is the compressor's capacity + the heat produced by the compressor.

So columns H & I give me some important information for sizing my heat exchangers (AKA: HX).

Evaporator HX => 7800 BTU/hr
Condenser HX => 10335 BTU/hr

I decided to go with brazed plate exchangers because they're small, available on ebay, and fairly cheap compared to other configurations.

Then I started pouring over the Braze Plate heat exchanger engineering specs from various companies. The search terms, "brazed plate performance curves" was where my search started.

The first thing I noticed was that the heat exchangers that were on ebay at an acceptable price, usually didn't have proper engineering specs available.

So I made the the following assumption:

In spite of what the manufacturers of brazed plate heat exchangers want me to think, the brand of the HX is not as important as the physical configuration. So if I could find industrial literature that specified a particular configuration, particularly if other manufacturers were in general agreement, I would not be limited to a particular manufacturer, but could select the configuration I wanted, and at the price I wanted.

Through all the searching, I realized that water-to-water exchange characteristics wouldn't do me any good, because refrigerants have very different thermal characteristic than water, so I had to disregard all the water-to-water thermal data. Sadly, the most abundant data seems to be of that type.

Refrigerant data for brazed plate exchangers could be found, by several different manufacturers, each selling several differnet sizes of brazed plate HXs. Here is a partial list of manufacturers who make stainless steel brazed plate heat exchangers for use where heat will be exchanged between refrigerants and water:

Focusing on these companies when searching for HX performance data should greatly improve the task.

Something else I noticed was that a heat exchanger has different heat transfer characteristics, depending on if it was being used as an evaporator HX or a condenser HX. In every case, the condenser HX transferred more heat than the evaporator HX. The amount of difference varied but the average was very close to 16%. I assume that this is due to the presence of liquid refrigerant in the condenser HX, compared to all vapor in the evaporator HX.

This is all a happy coincidence, because the condenser HX will need to transfer the extra heat that the compressor generates, and it is able transfer more heat due to the presence of liquid at that point in the refrigerant circuit.

I did find models with relevant data that were very close in size to heat exchangers that were readily available on ebay, so I selected first for proper heat transfer, then for proper fittings (barbed or pipe fittings on the water side, sweat fittings on the refrigerant side), then for price.

Here's what I chose, 7.5" x 2.9" x 30 plate, 3/8" sweat & 3/4MPT, $80 ea.:

My requirements, according to my compressor specs

• evaporator = 7800 BTU/hr
• condenser = 10335 BTU/hr

According to the selection tables, this HX heat transfer is

• evaporator = 9792 BTU/hr
• condenser = 11771 BTU/hr

...so I will have a modest excess of HX capacity for each HX, all for a pretty good price of $160 + S&H.


THE METERING DEVICE

To make things easy, I can simply re-use the cap tube that is on the donor de-humidifier. This approach should work out fine, because I am not exceeding the general performance specifications of the compressor / HX combination.

However, I am intrigued by the idea of a screw-adjustable metering device for the following reasons

• Ease of adjustment
• Ability to easily experiment with changes in heat-load, charge-level, etc.

The down side is that very small capacity screw-adjustable metering devices are not common. So, I still need to investigate whether a fractional-Ton (about 5/8 Ton) TXV can be found.


CONTROL ELECTRONICS

I have started working with a local guy, Jim-L and also others at a local technology development group 'DorkbotPDX', who are helping me develop control electronics for this unit.

The current electronics are based on a small Arduino clone called Teensy that is designed by a local Dorkbot member. Thanks to Ward Cunningham (inventor of wiki) who pointed me in the direction of 1-wire sensors for monitoring, and strider3700 who spelled out how exactly it is done.

Last night I met another guy at Dorkbot who has developed an Arduino-variant datalogger with real-time clock and SD card data storage. This unit looks to be the ultimate solution.

Good HVAC control capability seems to be the missing link at this point.

-AC_Hacker

[Piwoslaw]

Originally Posted by AC_Hacker

I started the continuous heat pump run on 1/12/2011 and it us now 1/16/2011, so I've had five days to collect data and impressions.

• There is a large and growing puddle in the cellar where the heat pump is. At first I thought it was rain seeping in where my loop field pipes enter the basement. But everything is dry there. Now I realize that there is continuous condensation happening on the outside of the pipes and tubing connecting the cold side of the heat pump to the ground loop. My initial fix is to get some pipe insulation and snug it up around the exposed pipe to keep the basement air away. I'm sure that this will work but I now recognize that the wet pipes also present an opportunity, and that is that they are de-humidifying the basement. I just need to capture the water. So... free de-humidifying... great!

One word: Insulation. And not the normal foam stuff used for insulating hot water pipes, there is rubber-based insulation for cold pipes which does not soak up water. And no spray foam either.

Or you can catch it in a bucket and have distilled water for whatever purpose.

[AC_Hacker]

Originally Posted by Piwoslaw

One word: Insulation. And not the normal foam stuff used for insulating hot water pipes, there is rubber-based insulation for cold pipes which does not soak up water. And no spray foam either. Or you can catch it in a bucket and have distilled water for whatever purpose.

Good thinking.

I think you are correct that the rubber-based insulation is better, but for a different reason that one would think.

I have completely stopped the condensation problem and the fix is to keep the moisture-carrying air away from the cold pipe and heat exchanger surfaces.

The rubber-based insulation is more pliant than the foamed-plastic stuff, and it works better to make an effective seal.

But if the geometry of the condensing surface is simple, the foamed-plastic will work too.

On a different note, I did learn from attending a solar panel installation workshop, that temperature-cycling such as found in domestic hot water flat-plate collectors will break down the plastic-foam insulation fairly quickly. The rubber-based stuff is not as seriously affected.

Regards,

-AC_Hacker

[philb]

A newby question just popped into my mind. If you are producing cool on one side of the unit and heat on the other side, could you use both? That is, pump the hot into your hot water tank with the heat exchanger before further cooling it in the ground? So, as long as you are using electricity to power the unit you might as well get all the benefit you can.

BTW, I barely understand what the HVAC stands for.
This thread has improved my definition of the word drastically!

[Drake]

Is your heat pump design reversible to "cool" radiant floor in summer?

[electrostunt]

Cooling a room with running cold water in a radiant floor is not very efficient. You get a really cold floor in a hot room... (heat rises... )

You can use a fan convector as the common choice.

There is also some test being done by putting the floor pipes and heat spreaders in the ceiling and running cold water in it. This should not be used for heating, as that would be inefficient.