The Homemade Heat Pump Manifesto

[AC_Hacker]

Originally Posted by tweeker

My RULE #1: If the A/C unit is working, NEVER BREAK THE UNITS REFRIGERANT SEAL! *Leave the pressurized part of the unit INTACT.

* * *

This is the time consuming part: Build a WATER TIGHT, INSULATED tank/container AROUND the EVAPORATOR & CONDENSOR. This is your FIRST fluid filled (COOL & HOT) HEAT EXCHANGER.

I like your approach. This isn't the way I've gone about it, but I like what you're doing. Certainly less HVAC tools needed, which is a barrier to many experimenters, and it avoids 'cutting into the system' which is probably the biggest discourager of all.

...So this is very good, I like it.

I came across this diagram of a Turkish active Heat Recovery Ventilator which uses a cross-flow heat exchanger in conjunction with vapor-compression heat recovery that might suit itself very well to your methodology.

This kind of device would provide fresh air, along with heat recovery, and some additional heating & cooling. If your house was really tight and very well insulated, it might provide all the heating/cooling you'd need.

* * *

Originally Posted by tweeker

Lets just say you have a 5,000 BTU A/C unit with an EER of "10", which means it is making 10 times the cool/heat per watt used, in this case 500 watts to make 5,000 BTU's = a "10" EER. This can only be achieved by HEAT TRANSFER - NOT watts TO PRODUCE HEAT in BTU's.

When you say, "10 times the cool/heat per watt used", that would imply a COP of 10 which is quite high.

I think we may have a different understanding here.

This morning I went to the Wikipedia page about EER and read it carefully, to make sure I wasn't missing something.

I like to look at COP, since it avoids using SI units and British units in the same statement, which can cause misunderstanding.

When COP is calculated for AC, the energy used by the compressor is subtracted in the equation when calculating the result, because the compressor's heat is waste heat in our process.

When COP is calculated for heating, the energy used by the compressor is added in the equation when calculating the result, because the compressor's heat is useful heat in our process.

So, from your example above, your 500 watt compressor generates 1706 BTU (500 * 3.421 = 1706 BTU) when it was run under test load.

The AC really does output 5000 BTU (1465 watts) of useful cooling power, and to do so, it had to get rid of the heat that was present under test conditions, and to achieve this, it also had to get rid of the heat generated by the compressor. In other words, the energy of the compressor has already been subtracted to get the 5000 BTU number.

So under test conditions, the AC consumed 1706 BTU of electricity (500 watts) and delivered 5000 BTU of cooling power (1465 watts).

The coefficient of performance is energy out / energy in, so...

In British Units...

COP (cooling) = 5000 BTU / 1706 BTU = 2.9

In figuring COP for heating, we include (add) the compressor power, since it is useful heat for us...

COP (heating) = (5000 + 1706) BTU / 1706 BTU

COP (heating) = 6706 BTU / 1706 BTU = 3.9

... if we wanted to do the same calcs using SI units...

COP (cooling) = 1465 watts / 500 watts = 2.9

COP (heating) = (1465 + 500) watts / 500 watts

COP (heating) = 1965 watts / 500 watts = 3.9

...same COP, either system.

So, while this is not "...10 times the cool/heat per watt used", it is very respectable and well worth pursuing. And I have found that in my experiments I could actually exceed the specs on the AC label and I have gotten COP of 4.5 under working conditions.

Best Regards,

-AC_Hacker

[tweeker]

Hello AC_Hacker,

I used the EER rating/Comparison since it is the nomenclature/verbiage used for Air Conditioning, hoping that people that only know how to read the EER of an A/C unit would be able to stay with me on my description of the ENERGY IN (watts) vs ENERGY OUT (BTU's).
Having children around me all the time, I sometimes get very elementary in my explanations.

Reference for the industry standard vocabulary:

A room air conditioner's efficiency is measured by the Energy Efficiency Ratio (EER). The EER is the ratio of the cooling capacity (in British thermal units [Btu] per hour) to the power input (in watts). The higher the EER rating, the more efficient the air conditioner. National appliance standards require room air conditioners to have an energy efficiency ratio (EER) ranging from 8.0–9.8 or greater, depending on the type and capacity, and ENERGY STAR qualified room air conditioners have even higher EER ratings.


I am familiar with C.O.P., and it is more "TRUTH TELLING" in the facts of the matter.

Tweeker

[smith]

Can you include some pics tweeker and maybe a short description of some of your currently running projects, When you have time of course

Thanks

[tweeker]

I am not very computer savvy. I tried to get my pictures linked into this forum, with no LUCK.

I can e-mail them to someone that has the ability to post the pictures?

Tweeker

[AC_Hacker]

Originally Posted by tweeker

I am not very computer savvy. I tried to get my pictures linked into this forum, with no LUCK.

I can e-mail them to someone that has the ability to post the pictures?

Tweeker

I'll help you...

I sent you an email with details.

-AC_Hacker

[randen]

Hello just doing some information gathering for obtaining some brased plate HXs. The GSHP I'm planning is utilizing two compressors rated 2T and 2.5T therefore a HX large enough for 4.5T (54,000) BTU. The two compressors in captivity are split air to air units complete with cap tubes. The 2.5 T has a new compressor the 2T is about 10 yrs with only a few hrs. I have gathered some stainless steel sheet stock (old cabinet doors from a comercial kitchen) for the enclosure. Although stainless can be nasty to work with the benefits no painting and no rust. 20yrs from now it will look the same as the day I had made it. The goal is to heat a 1600 sq ft shop in a Canadian climate with money left for the kids college fund. The shop is well insulated and the office portion has already in floor heating. The plan is to sequence the compressors such that say, after 8 hrs run time the second compressor should start until the thermostat cycles.
Questionable items: TX valves vs capillary tubes.The simplicity of brasing the cap. tubes in and recharging the system would be my personal favourite. Or is there an advantage in COP with the adjustability of the TX valve with the differing temps and associated pressures in GSHP application.?? The other item the reversing valves one for each compressor. The complexity would go up, switching the refrigerant, and switching from floor heating to fan centre (water to air) cooling. The shop always had airconditioning and some days is really needed.


Randen

[AC_Hacker]

Originally Posted by randen

Hello just doing some information gathering for obtaining some brased plate HXs. The GSHP I'm planning is utilizing two compressors rated 2T and 2.5T therefore a HX large enough for 4.5T (54,000) BTU. The two compressors in captivity are split air to air units complete with cap tubes. The 2.5 T has a new compressor the 2T is about 10 yrs with only a few hrs. I have gathered some stainless steel sheet stock (old cabinet doors from a comercial kitchen) for the enclosure. Although stainless can be nasty to work with the benefits no painting and no rust. 20yrs from now it will look the same as the day I had made it. The goal is to heat a 1600 sq ft shop in a Canadian climate with money left for the kids college fund. The shop is well insulated and the office portion has already in floor heating. The plan is to sequence the compressors such that say, after 8 hrs run time the second compressor should start until the thermostat cycles.
Questionable items: TX valves vs capillary tubes.The simplicity of brasing the cap. tubes in and recharging the system would be my personal favourite. Or is there an advantage in COP with the adjustability of the TX valve with the differing temps and associated pressures in GSHP application.?? The other item the reversing valves one for each compressor. The complexity would go up, switching the refrigerant, and switching from floor heating to fan centre (water to air) cooling. The shop always had airconditioning and some days is really needed.


Randen

This is a pretty advanced project, compared to what I have done.

First, look at the building with a vicious eye to reducing heat loss as much as possible. I posted a link to 'Mooney Wall' construction in the Conservation area, that should give you some ideas. The example they show results in a six or eight inch wall, but it doesn't mean you can't go thicker. And windows are almost always huge heat loosers. Daox had a very interesting post about DIY triple windows.

My inclination is to advise you to consider a separate HX for each compressor, and to run the floor-loop water in series through the two HXs. This wll give you some breathing room as far as HX prices, and also give you a 'soft fail' if one of your refrigeration units has problems.

As far as cap tube vs TXV, the recommendation seems to go with TXV over a Ton and cap tubes under a ton. There are cheap TXVs available on ebay. Make sure you get ones that are compatible with your refrigerant.

I just watched a 20 year old 3 Ton whole-house AC get uninstalled, and it used 4 cap tubes in parallel to share the duty.

...but if you already have the cap tube, use them.

If you are building your first heat pump from scratch, you are flinging yourself into the deep end of the lake.

Maybe go to HVAC Talk and solicit some seasoned advice, they have some experienced people there, if they will talk to you... well that depends upon the stars.

I'm happy to help you all I can, but this is pretty big and complex. What I have worked with is small and simple.

At any rate, stay in touch, this is just too interesting to miss...

-AC_Hacker

[Blauhung]

AC,

First off, I’m looking for polyisocyanurate sheeting in the Portland area and seems that most hardware stores don't seem to carry it. Wondering what you've been using in your walls and where you've been sourcing it from.

Also have done some digging through the local rules on digging and found this Oregon Administrative Rules 690-240. Section 690-240-0035 talks specifically about Geotechnical Holes which are what any closed loop borehole under 10' deep counts as. Only problem, once you go past 10' it looks like that's where the state of Oregon starts requiring permits as well. From talking with some people at the department of water resources, it sounds like I may be able to get a permit under the condition that I get myself certified as a professional engineer with the state which I'm currently working on doing (I'm a chemical engineer). In your case probably no one really cares as you haven't gone deep enough to cross into any aquifer (the regulations don't require any follow up until you get below 18' which then requires a well report being filled out and kept on file with the state).

Lastly found a better source for a really good thermodynamic diagram for the liquid/gas phase transfer region of propane
Propane P vs. S
This is what's known as a Mollier diagram which plots various constant property curves as functions of the pressure and Enthalpy of the system. Enthalpy is a thermodynamic property which basically refers to the total energy content of a given system. If people are interested I could probably be convinced to give a quick pictorial lesson on how to read this diagram for the purposes of sizing heatpumps.

[AC_Hacker]

Originally Posted by Blauhung

...what [have you] been using in your walls and where you've been sourcing it from.

I've been using EPS (AKA: Expanded Poly Styrene) foam (white) which I have been getting from any 'big box' store that has it cheapest. It has an R-value of about 4.5 per inch. The R-value table that is printed on the foam seems a bit ambiguous, but 4.5 is what I interpreted it to be. I know that there is a yellow foam with a higher R-value (also a higher price). The yellow foam seem to be an open-cell foam. After much searching, I came across information that the open cell foam loses R-value, due to gradual moisture entry into the foam over time, so this was the biggest decider.

This project has been going on for a long time, and when I started, the 6" (R-27) I was putting in the wall was so far over 'code' that I was sure I was super insulating. Now it just seems reasonable. My house is pretty small, with small rooms and increasing the inside wall thickness by 2" seemed like a lot, but I made the sacrifice.

At first I was cutting the foam exactly the width between the studs and pounding it in with a board and hammer, which meant that I was tapering some of the foam sheets, because not all the studs were parallel. It was terribly laborious, but the results are significant.

The method I finally settled on was cutting the foam about 1/4" to 1/2" too narrow and filling the gap with canned foam, trimming the excess when it set. This method gave much better seal, and was easier & faster. At first, I was using a very sharp butcher knife to cut foam, then my son came up with the idea of using a hot wire. I think the wire is the way to go, but our hot-wire setup needs some improvement to be really useful.

The one thing I learned about late was thermal bridging. I wish I had known about it sooner. I could have designed around it. I think it is a significant factor.

Originally Posted by Blauhung

Also have done some digging through the local rules on digging and found this Oregon Administrative Rules 690-240. Section 690-240-0035 talks specifically about Geotechnical Holes which are what any closed loop borehole under 10' deep counts as. Only problem, once you go past 10' it looks like that's where the state of Oregon starts requiring permits as well. From talking with some people at the department of water resources, it sounds like I may be able to get a permit under the condition that I get myself certified as a professional engineer with the state which I'm currently working on doing (I'm a chemical engineer). In your case probably no one really cares as you haven't gone deep enough to cross into any aquifer (the regulations don't require any follow up until you get below 18' which then requires a well report being filled out and kept on file with the state).

I really went to a lot of trouble to develop an electric hole boring machine, rather than a gasoline machine... I really didn't want to disturb the neighbors.

Originally Posted by Blauhung

Lastly found a better source for a really good thermodynamic diagram for the liquid/gas phase transfer region of propane
Propane P vs. S
This is what's known as a Mollier diagram which plots various constant property curves as functions of the pressure and Enthalpy of the system. Enthalpy is a thermodynamic property which basically refers to the total energy content of a given system. If people are interested I could probably be convinced to give a quick pictorial lesson on how to read this diagram for the purposes of sizing heatpumps.

This is a really good find. Please share anything you can about this diagram with us. I am very interested.

Thanks for the information.

Regards,

-AC_Hacker

[Blauhung]

Ok... let’s see what I have remembered from my thermo classes...

First off, some quick definitions of the terms (not everyone has taken thermodynamics classes)

• Enthalpy – the total amount of energy in a thermodynamic system. Think of it as how much work would have to be done to generate that system from absolute zero and zero pressure and in turn a measure of how much energy could be pulled out of a system when reversing the process. In the case of the diagram it is measured in kJ/kg where a Joule is a unit of work (Newton*meter), so any Enthalpy we pull off of here will then be multiplied by the mass flow rate of refrigerant to get to Watts.
• Entropy - a measure of the energy in a thermodynamic system not available to do useful work. At its simplest, the amount of chaos or disorder in a system that we can’t turn around and get something out of. On the diagram, these are shown as green lines of constant Entropy (kinda like a contour map).
• Adiabatic process – any process that occurs without the transmission of heat. For our purposes we will call both compression and expansion an adiabatic process, meaning that both the compressor or the end of the cap tube are perfectly insulated and we do not give the refrigerant any chance to exchange heat with the surroundings.
• Isobaric process – a process that occurs without a change in pressure
• Isothermic process – no change in temperature
• Isentropic process – no change in Entropy
• Reversible process – all work done to the system to change its state can be extracted back out as usefull work, pretty much another way to say no net change in entropy which again is the “useful work” content of a system

So this diagram shows values for Temperature, density, entropy and quality at given values for pressure and enthalpy

• Temperature = Red line in °Kelvin. Kelvin is a measure of temperature with the same value as °C, but its 0 point is absolute zero or -273.15°C (point at which all molecular motion stops) to get to K from °C add 273.15
• Density = blue lines. We don’t really care about this for our purposes
• Entropy = green lines. Measured in kJ/kg as with Enthalpy.
• Saturation line = curved black loop forming the phase change envelope. This line and anything outside of it can be considered 100% a gas/liquid or supercritical fluid (won’t delve into supercritical fluids in this writeup, but feel free to ask about or look it up)
• Quality = curved gray lines within the Saturation envelope. Quality of the gas is the mass fraction of refrigerant that is in its gaseous state in equilibrium at those specific values of pressure and enthalpy.

Here’s a quick more basic version of the graph related to refrigeration


At its heart, the refrigeration cycle has 4 elements that take place

1. Compression – points 1-2. Again, we are assuming that this is an adiabatic process and as such there is no heat transferred to or from the refrigerant. As we are performing work on the refrigerant there must be a net change in Enthalpy, but this can also be considered to be a reversible compression so there is no net change in Entropy and while compressing we follow a line of constant Entropy on the graph (the pic above has simplified this as a straight line) to our next point.
2. Condensation – points 2-5. For our purposes we will assume that there is no refrigerant pressure drop (isobaric process) across the condenser, which isn’t quite true, but it’s negligible. This means on the diagram the refrigerant crosses the phase change envelope at a constant pressure or a horizontal line. Also notice that for pure substances, phase change at constant pressure also takes place at constant temperature, this is what makes phase change systems so efficient at heat transfer as the deltaT across the heat exchanger stays much more constant.
3. Expansion – points 5-6. This is our systems TXV or cap tube. Again we can assume that we have this portion of the system perfectly insulated and no heat is transferred to/from the surroundings and is thus adiabatic. With compression we were adding reversible work to the system and thus followed a line of constant Entropy, but now there is no work being performed on or by the system, so we must follow a line of constant enthalpy or moving straight down. This means our expansion is not a reversible process and as such we must gain some entropy in the process thus loosing useful work from the system as gains in disorder which you can see as the lines of constant entropy are curved and from pint 5 to 6 we are now at higher entropy.
4. Evaporation – points 6-1. Same as with the condenser we can assume this to be an isobaric process and make a horizontal move to the right.

And now to play out a real world example...... (AC, I’ll use the output refrigeration temps and the pressures you listed way back in post 465 then show what an ideal refrigeration cycle would look like from that)
Tevapout = 64.2°F = 291K
Tcondout= 85.6°F = 302K
Pcond = 210psig = 1549kPa(absolute)
Pevap = 40psig = 377.1kPa(absolute)



AC, as you can see your temperature out of the compressor was actually lower then the one I have on the diagram following a line of adiabatic compression, this is due to the loss of heat to the compressor and thus reducing the amount of effective work done on the system.

That's enough for tonight, I'll try to go into how all this can play into sizing of heat pump components tomorrow if i have some time.