The Homemade Heat Pump Manifesto

[Pegasus]

Originally Posted by Xringer

Please correct me, if I'm wrong about this.
I've never used 220 appliances while overseas.

Well, the rated input of all (to my knowing) electric appliances in Europe is 220-240VAC, 50-60Hz (1PH), my donor A/C included. 220V AC is the Greek Power Standard thats why i wrote it.

[Pegasus]

I am really tempted to make a copper tube HX for the ground loop side, "drown" it in to a barrel and use a submerged pond circulation pump. For testing it i found on ebay a cheap China Made one with 1400Lt/H (370 USG/H) flow and 2m Head, consuming 28W.

As i wrote on my previous post i have 6 ground loops in parallel at a 2m depth from where the HX will be placed.

Does any one (AC hacker, randen, ...) have any suggestions for the sizing of the HX (Copper diameter, Length, Spiral spacing - diameter ect) for my 13000 BTU compressor?

[Xringer]

Originally Posted by Pegasus

Well, the rated input of all (to my knowing) electric appliances in Europe is 220-240VAC, 50-60Hz (1PH), my donor A/C included. 220V AC is the Greek Power Standard thats why i wrote it.

I guessed right! So, that only leaves the possibility of systems
with 4-way valves that require DC voltage.?. Maybe 24 volts?

I see one 24V coil here:
Search Results - PexSupply.com
But it might be a 24VAC coil, since that seems to be a common
thermostat control voltage in the USA..

When I checked the 4-way valve control relay on the main board in my Sanyo,
it was connected to the 240vac input line. No transformer or diodes in that circuit.

[BradC]

Originally Posted by Pegasus

Does any one (AC hacker, randen, ...) have any suggestions for the sizing of the HX (Copper diameter, Length, Spiral spacing - diameter ect) for my 13000 BTU compressor?

I spent an inordinate amount of time developing spreadsheets for this stuff when I was going to use 7/8" tube coils in a 44 Gallon drum as a condenser. Need some more info.

If you can give me details like

- Temperature of water in drum
- Temperature of ground
- Evaporating or Condensing temperature

13000 BTU is approximately 3.8KW/h

If you are circulating 1400 Lt/h (as you stated earlier) you will see a temperature fall (water in/water out) of about 2.5 Degrees C when evaporating. Condensing you will see more as you need to dissipate the heat generated by the compressor also.

You need to take into account your loop friction which will reduce your circulation rate.

I note you want to use this for both evaporating and condensing (heating and cooling). For condensing you need about 5.5 times the tube surface area you need for evaporating as you are basically using gas to copper which is far less conductive than liquid to copper.

A rough back of a napkin coil to condense 4.5kw comes to a gas-side surface area of just under a square meter (0.987 to be accurate).

That's 16M of 7/8", 29M of 1/2" or 41M of 3/8".

Let's go with 7/8 as you can buy 18M coils of that. A 16M 7/8" evaporator will hold 5L of refrigerant or about 6KG of R22!

When you are condensing, you need somewhere to store that extra liquid!

You also need to keep the water in the drum moving quite fast. Slow moving water builds up boundary layers around the tubes and reduces heat transfer. This is why tube-in-tube or PHX are so good, as they keep the flow turbulent.

I did a proof of concept with 6M of 1/2" tube bent in a coil and submerged in a laundry trough to condense an old 2HP window unit I have here. The water really had to be moving to keep the heat transfer going.

The other thing about a flooded drum like that is your condensation temperature is going to be far higher than it would be with a proper heat exchanger due to the very small temperature difference across the copper coil. You really need to keep the cold incoming and the warm outgoing separated in a counterflow arrangement.

Remember, the higher the condensing temperature, the lower the COP, and thus the higher your power bill!

To build a coil that comes even close to my PHX condenser I'd need over 124M of 7/8" Copper tube. At the best price I could find locally that costed out to over $1400, and the result (if I could even build the thing! - 7/8 is a BEAR to work with) would be less than half as efficient as a good PHX.

A 1M2 PHX should be available on fleabay for a couple of hundred $$. 16M of 7/8 will set you back a couple of hundred also.

<Edit> all the calcs are based on a ground loop temp of ~20C and a condensation temperature of 26C. If you condense at 40C you can halve the size of your exchanger at the expense of a large power consumption increase.

[randen]

-AC_Hacker...just saying...

The Heat exchangers are indeed larger. If anything maybe a little oversized. The HXs are removing essentially all the heat. The return (approach) temps are 1 Deg C. differential. Water in refrig. out. The next cold snap I will do a little experimenting. Yesterday the outside air temp was a balmy 17 Deg C.

Randen

[AC_Hacker]

Originally Posted by randen

The Heat exchangers are indeed larger. If anything maybe a little oversized. The HXs are removing essentially all the heat.

...then how about the ground loop flow rate through your HXs?

This would call for more pumping power. To double the flow rate in your loop, you will need 4x the pump power...

-AC_Hacker

[Pegasus]

Originally Posted by BradC

A rough back of a napkin coil to condense 4.5kw comes to a gas-side surface area of just under a square meter (0.987 to be accurate).

That's 16M of 7/8", 29M of 1/2" or 41M of 3/8".

Let's go with 7/8 as you can buy 18M coils of that. A 16M 7/8" evaporator will hold 5L of refrigerant or about 6KG of R22!

That's huge, therefore out of the question!!! I'll stick to the plated HX.

But i am still puzzled for what is the best solution :

- Reverse the refrigerant flow with the existing Reversing Valve
OR
- Reverse the Water flow (as Randen did) and get rid of the rev. valve

???

By the way now i can post some pictures:

The compressor :



The Reversing Valve:



The cap Tubes (insulated probably against vibration and friction):

[BradC]

Originally Posted by Pegasus

But i am still puzzled for what is the best solution :

- Reverse the refrigerant flow with the existing Reversing Valve
OR
- Reverse the Water flow (as Randen did) and get rid of the rev. valve

Oh that's easy. Reverse the water flow. Remember, condenser needs to be about 5.5 times the size of the evap. By keeping the refrigerant loop constant and reversing the water you only need properly sized HX units and one TXV. Much simpler design and implementation for the refrigerant side.

[Pegasus]

Originally Posted by BradC

Oh that's easy. Reverse the water flow. Remember, condenser needs to be about 5.5 times the size of the evap. .

Why is then, that most of the commercial HP' s use a reversing valve.

[BradC]

Originally Posted by Pegasus

Why is then, that most of the commercial HP' s use a reversing valve.

Convenience rather than efficiency. Nobody condenses at temperatures low enough to really maximise efficiency.

Look at it this way. If your condenser needs to be about 5.5x the size of your evaporator to get great efficiency, how good is the efficiency of the units that have the same size either end? My ASHP has a condenser about twice the size of the evaporator. As far as efficiency goes it sucks, but it does the job and is affordable. Would people buy a unit that was three times bigger and vastly more expensive if it was more efficient? Mostly not. It's a design compromise between cost and efficiency. A tradeoff we DIY's don't have to really make. We can afford the extra time, effort and outlay to make the most efficient system possible.

I could use a 1M2 condenser (which is about the surface area of my evaporator) for my GSHP unit, but my condensing temperature would be about 50 degrees C. No better than my current ASHP unit.

*OR* I could have halved the size of my HX at the expense of doubling my water flow. I did the sums and the larger HX will pay for itself in under 24 months when you take into account the extra electricity required to run a bigger pump to double the water flow.

In my "Brads Gas and Gear" thread, I posted an Openoffice spreadsheet that will allow you to calculate the surface area required to achieve certain HX goals.

I use that in combination with "Coolpack" to define my requirements.

Use Coolpack with your cycle specifications and it will give you your discharge temperature for a target condensing temperature. Feed those into my spreadsheet along with your water in temp. I adjust the water out temp to achieve the flow available from my bore. It's probably *** backwards, but it made sense when I did it.

Coolpack will give you your system heat load (the heat being moved + your compressor heat). Use that in the spreadsheet and it will tell you your required surface area.

In my case I iterate the values until I reach the surface area of my HX (8.6M2).
I know my bore water is 20C. I know I flow ~30LPM. I know I'm moving 18KW of heat. Coolpack tells me I have a system heat load of ~21.01KW if I evaporate at 2C and condense at 26C. My spreadsheet tells me to dissipate 21.01KW at 30LPM My outlet temperature is going to be ~30 Degrees C.

I estimate I'm going to condense at ~25C, so I feed that into Coolpack and get a discharge temp of 43.6. I put that into my spreadsheet and it tells me I need a HX of 7.07M2. That's a bit low.

I'll go down to 24C then. Coolpack tells me I discharge at 41.3C. Put those figures into the spreadsheet and I get 8.65M2. Close enough!

So with my bore and pump, I should be condensing at 24C. With an evaporating temperature of 2C and a 5K Superheat. This gives me a COP of 7.2.

As the heat load in the house drops and I can raise the evaporating temp, my cop creeps up towards 10.

Now, as you lower the condensing temp, you lower the total system heat load. You could continue to iterate until you converged on an answer, but my best guess is probably close enough.