The Homemade Heat Pump Manifesto

[AC_Hacker]

Originally Posted by Thepprof

Actually I'm just trying to understand more regarding HVAC...

I hope you read the signature block of my posts... it's no joke, I really am not an HVAC tech.

However, your reply gives me a better idea of what you are trying to do.

I think you are correct to be concerned with liability and the use of R290 in a rental, etc.

That being said, and since this is for educational and experimental purposes...

I would say that charging by weight is the best way to go if you are using the original refrigerant on a piece of unmodified equipment and have a good scale. There are some loose estimates of refrigerant weight correction factors when changing to a different refrigerant. But if you do change refrigerants, all of the careful design considerations that went into your unit, all go out the window, and you really are flying by the seat of your pants.

Method #1:

My limited experience is with R22 and R290, and in this case, I found that for R290, the pressures, both high pressure and low pressure, were about half of what was specified for R22. I did some adjusting of the pressure levels until I got a visual indicator (frost) that everything was working properly. I also put some measuring instruments on the machine, and logged data, and put the data into a spreadsheet that had formulas that would calculate efficiency over a 5 minute time period. I used the efficiency numbers to 'tweak' pressure until I reached maximum efficiency.

Method #2

randen used a much smarter way than my method... he measured temperature output and also amps. He had a heat pump, and looked for the coldest output from the output of his evaporator HX (as I recall). He attempted to reach the coldest temperature possible, and also, the minimum current used. So there was a bit of a balancing act there.

Method #3

This would go for the optimum super/sub heat values. Be aware that these will not be the same values as for R22 or R410a. There was a good discussion previously on the Manifesto thread that addressed the optimum super/sub heat values when using R290. There are also 'smart' field pieces that have those values built in, and show super/sub heat... with everything calculated automatically. I suspect my life would be much improved if I had one of these things.

Method #4

This would be your method, wherein you use an IR thermometer and look for a full and even cooling (or was it heating) of your coil area. Personally, I thought that this was pretty darn smart, if you have an air cooled or heated coil. It might not work so well with a brazed plate HX, or a tube-in-tube HX.

So, to my knowledge, these are the approaches to use, in order to get an optimum charge for your machine.

Since most of us here are hacking existing systems, and changing them in ways that are not in keeping with how they were originally designed... and also that we are doing things that would make any properly schooled, god-fearing HVAC tech have uncontrollable conniptions, we have to develope and use experimental methods. And the method you have devised is now part of our possible approaches, our way of working. Thank you for your input.

I would like to hear more from you about what you are learning here... and I would also like to hear from other experimenters as to their ideas and ways of approaching the problem you have brought up.

Best,

-AC_Hacker

[hikerjohnson]

Hi AC, folks,

I am a new-ish lurker, and have devoured the entire thread, and a large chunk of the rest of ecorenovator over the last few weeks, in search of projects and priorities for upgrading my own home. Looking into ground source heat pumps, I found a very good (it seems) sizing manual from Bard HVAC.

Since I cannot post a link yet, just google the following string (without the quotes): "bard hvac pdf 2100-099(N)"

http://www.bardhvac.com/digcat/volum...100-099(N).pdf

It will be the first PDF you see on the results.

Separate question: does anyone know of a ready source for ground temperatures in the southern Maine/NH area?

[AC_Hacker]

Welcome aboard!

Originally Posted by hikerjohnson

Hi AC, folks,

I found a very good (it seems) sizing manual from Bard HVAC.

...following string (without the quotes): "bard hvac pdf 2100-508"

http://www.bardhvac.com/digcat/volum...e/2100-508.pdf

It will be the first PDF you see on the results.

What I am finding looks like installation instructions for air conditioners. Is this what you have in mind?

Originally Posted by hikerjohnson

Separate question: does anyone know of a ready source for ground temperatures in the southern Maine/NH area?

This should answer your question:

Best,

-AC

[hikerjohnson]

Sorry, that is a screw-up on my end. Not sure how I did that.

The document to look for is Bard 2100-099(N)

http://www.bardhvac.com/digcat/volum...100-099(N).pdf

I have been reading it and while it does have a lot of good information on practical aspects of installation, it still tip-toes around the subject of science-based field sizing. I'm trying to run through the heating field size formula, to see if it yields reasonable-sounding numbers...

[hikerjohnson]

Well, I would like someone to vet my conclusions, but I think that the Bard formula for ground loop sizing might be a little padded...

See the formula on page 25 of the manual,

Lh = 12000BTu/h/Ton(COP-1/COP){Rp+(Rs)(Fh)}

(t1-Tmin)

where:

Lh = heating length of groudloop in feet per ton of heat pump capacity (feet)
COP = Coefficient of Performance of heat pump (unitless)
Rp = pipe thermal resistance (Btu/h*ft*deg. F)
Rs = soil thermal resistance (Btu/h*ft*deg. F)
Fh = heating run fraction (unitless)
T1 = low soil temperature (deg. F)
Tmin = design minimum etering fluid temperature (deg. F)

I inputted the following (guesstimates, all)

COP = 2.5 (on the lower end, and very sane)
Rp = 0 (the pipe is so small compared to the soil as a fraction of the system)
Rs = 1/0.7 ( 0.7 pulled from the ASHRAE handbook as a median value for soil thermal conductivity for sandy damp soil, and inverted to give thermal resistance)
Fh = 1 (full-time running on a design day)
T1 = 47F (from the soil map in previous posts)
Tmin = 35F (pull the fluid to just above freezing, is this sensible?)

The value I obtain is 3003 feet of pipe needed per ton of heating.

Wow! Can anyone see where I am completely off the reservation?

[AC_Hacker]

I looked at the formula.

Right now, I have some stuff that needs doing, but later on today I can try running the numbers for my locale, in which I know what size typical loops tend to run.

I do know that the local (Portland, OR) industry estimated requirement for one Ton is 225 feet for a borehole (450 feet of pipe), and you can see from your ground temp chart what the ground temp is here. For you the loop requirements will be greater, but 367% seems a stretch. More pipe will mean lower temperature drop over the heating season, and greater efficiency.

For a slinky trench, here (Portland, OR) it is something like 80 feet of trench, 4' loops, 700 feet of pipe at a 6 foot (minimum) depth, to achieve one Ton.

There are many factors to consider, and the IGSHPA (Oklahoma state) manual pretty much covers all the angles.

Also, call some experienced local installers and ask for "seat of the pants estimates". Well diggers in your area probably also know.

* * *

Did you already do a heat load analysis on your house? The lower the heating requirements, the smaller the loop field, the less the expenditure.

Best,

-AC

[hikerjohnson]

Hi AC,

I did a heat load analysis for my house in it's current state (pretty good for 1981, not so much for 2011) using HVAC-Calc (yup, I paid Don 50 bucks for it, and to be honest, it was worth it given how long it'd take me to do by hand)

On a design day I am looking at a load of 30,719 BTU/hour. That include my basement, which I don't ordinarily heat, dropping the heating load to closer to 22K Btu/hour. This house has electric resistance heat which I do my best to shun, a tiny coal stove in the basement, and a big honkin' woodstove on the first floor, which does a remarkably good job of heating the house most of the winter. I have made up the difference in the bedrooms with small oil-filled electric heaters. So far, so good (2/3 of a winter under my belt at this place)

So, 31KBtu/hour is approximately 2.5 Tons, and I would like to size my field for 3 to 3.5 tons, provided that the math works out for field sizing. This will probably give me an entire loop that I could cut out in the event of some catastrophe, like a puncture. This 31kBtu/hour value makes some sense, based on the amount of wood I was burning, but that really is a very rough guesstimate.

I plan on calling a couple of local geothermal outfits, but I want to have my own research more or less done, so I can figure out what they are telling me.

I intend to duplicate your soil conductivity test sometime in the fall, when I have all the bits and pieces to do it accumulated. Soil here is deep sand (near a river, but not near enough to do a pump and dump)

In the interest of full disclosure for all reading this (and so you know I'm not trying to make the thread veer off-topic), I am planning on DIY-ing my heat pump, and field, but it will be on an available-cash basis, and probably 3 years from completion at this point, if I had to hazard a guess. First step is sizing and calculating, second step scrounging, third and final step is building.

[hikerjohnson]

Hi guys, AC,

I found this paper that, although lacking a few details, seems extremely relevant to our working toward homebuilt heat pumps. Summarized, it is the senoir design project of some mechanical engineering students who were charged by Waterfurnace to design, build, and test a 3-ton heat pump capable of delivering output temperatures above 140 degrees F. Also valuable is that it is designed to be hooked to a ground loop circulating at 32degrees F.

They succeeded, mostly. Of value in the paper are component selections for the compressor, heat exchangers, and expansion valves, approximate charge values for R-134a, and the plotted outputs at various system temperatures.

Seems to give a very good starting point for someone piece-mealing their heat pump from new or scavenged parts.

Most noteworthy (and logical) bit of information in the paper is that they slightly undersized their exchangers, and could likely have gotten output temps above 140F with larger/more plates.

http://www.ipfw.edu/dotAsset/239446.pdf

[AC_Hacker]

Originally Posted by hikerjohnson

...I found this paper that, although lacking a few details, seems extremely relevant to our working toward homebuilt heat pumps.

Thanks, both of these papers look to be very useful... I have added them to my permanent collection. The heat pump building document is very good to come across.

I did look into running some numbers on the formula that you were working with in the above post.

Here is how I interpreted the formula:

Lh = ((12000 BTU-h/Ton)*(COP-1/COP)*((Rp+(Rs)(Fh)))/(t1-Tmin)

I also found a source for the Rp value HERE.

I also found a handy tool for converting various metric values into what we call the English system HERE.

The other values will be local to you.

Let me know if what you come up with in your calcs.

I think that a three year project time line is pretty reasonable for a DIY GSHP.

The loop field will be the largest part of the project. If you hire it out, it will be the largest cost... if you DIY, it will require the greatest effort (and cost).

The heat pump part can be complex, but it's pretty straight forward and not too expensive... success there can help propel you through the rest of the project.

Again, I suggest that as part of your project, you undertake a full and fearless analysis of the heat loss of your house. Your heat loss tool is a good one, I have tried it. It can come pretty close to determining your heat load. But what it doesn't do is to suggest ways to improve your insulation, or to reduce your infiltration losses.

Investigate an energy audit, or find out exactly what that it consists of and DIY. Get or do the blower door test, since infiltration is the #1 cause of heat loss. Consider re-insulating your house. Probably the least painful way to do this is to "outsulate" your house. There is an outsulate thread in progress right now, and the author of the thread (Mikesolar) is not only doing a great job, he is taking very useful photos of the project.

What have you found out about your soil conditions? Have you asked your state of local geologist? I did and got some very useful info.

You might try the basic water drill like I did in THIS post? It's cheap and easy and can give you an idea what lies in store. I was able to get down about 14 feet in less than 10 minutes. It can teach you a lot.

Thanks again for your document contributions, they are a very useful addition.

Best,

-AC

[hikerjohnson]

Thanks for the kind words and advice, AC.

I have not spoken to a local geologist in the recent past, but I have spoken to my neighbors on both sides, and our hyper-local geology is topsoil for 10-14 inches, followed by seeming bottomless sand, and a water table at no more than 10 feet of depth, sometimes 8 feet if you dig in a low spot. (Did I mention I'm only a few hundred feet from a lazy little river?) When I said my well was about 18 feet down, I forgot that the well head is at a high point on my property, skewing the impression of where the water table lies.

I also recently spoke to a local well driller, nice old fart named Tom, who confirmed that drilling is easy on my road, since it's all sand for some tens of feet before he hits ledge (what we call living bedrock here).

I'm not sure how much more there is to be gained by consulting a local geologist, as New England is famous for being solid rock in one spot, and 30 feet of sand ten feet over. Never mind the glacial erratics thrown in for variety...

The general schematic for what I'm planning follows below:



I want to do at least three tons of ground loop, and preferably four. Two reasons for this: I plan, someday, to add on a large garage with a mother-in-law apartment on top, and so I would like to have the capacity to enlarge the sytem. Second, I like the idea of redundancy (I fix submarines for a living, redundant is a Good Thing.).

Beyond the thinking above, I have friends with excavators, so what's the difference between three trenches and four? It's a marginal material cost, to be sure.

Work is taking me away from home for a couple of weeks, but when I return, I should have everything I need to duplicate your soil conductivity experiment, and then I will have some hard numbers to help evaluate the loop field design process.

Alright, dinner is ready, more to follow.