The Homemade Heat Pump Manifesto

[Ko_deZ]

Bending pex
There was some talk about not being able to put pex close to each other due to the bends. Well, you don't need to make a perfect U. Strap the pex to the floor or whatever you need to connect it to, and make the end bends just like a C. That way the diameter of the bend is not very low (sharp bend), and you can get the tubes as close as you want. Although this is usually not the way one does this. Usually one would make a loop around the walls of the room (double back along outer walls, this to avoid the cold air falling from the walls to cool down the floor here more than elsewhere), and then loop in a labyrinth with twice the distance you want. Then loop back out again filling the gap you left looping in. This way you get 90 degree bends mostly, and you get the hot water looping in and back out, giving you the most evenly heated floor. If you go back and forth from one side only, the place where the water enters will be warmer than where it exits. Our feet are very sensitive and they do notice this.
As a small tip, if you do bend the pex flat, you can possibly save it by heating it with a hair dryer. Most of the time it will resume it's previous shape. You can also buy corner irons that the pex fits in, and that makes corners very nice. Not worth it imo. If a bend is a bit larger or smaller does not really matter that much.

Oxygen barrier
Pex for floor heating has an oxygen barrier. This is to avoid oxygen getting into the loops with the result that things starts growing in there. If you use oxygen barrier tubes you do not need to put any anti-freze or bacteria killers in the water. This greatly reduces wear on all couplings and pumps as the mixture you get often is some degree of corrosive.

Heated floor sounds
To avoid sounds in heated floors, which someone complained about, you need to make very sure that you do not have any air in the floor loops. Flush them with high speed water from the tap, one at a time. If you do not get this right, the thermal transfer will suffer, and so will the flow, noise and initial bacteria growth. Also when using heat spreader plates, get the snap in ones, and make sure that they are placed in their tracks with some fabric between. 90% of squeaking floors are because the aluminium heat spreaders are moving slightly against whatever they are placed in. Snug fit and fabring between fixes this permanently.

Parallel flow heat exchanger
Why not use paralell flow heat exchangers? Well, perfect opposite flow heat exchanger has the entering hot water exiting with the same temperature as the entering cold water, and the cold water exiting with the same temperature as the hot water enters with. That would be a no-loss (100%) heat exchanger. With a parallel flow one, a perfect sample will have cold an warm entering at the same point, and at the exit point they can only hope to have the same temperature. This makes a maximum efficiency of 50% heat exchange. It is logical when you think about it.

Hope this was slightly interesting. Also hope to get some new insight into this as there certainly are some things I do not understand fully, and that I am completely wrong in.

Now to explain what reading this thread has made me plan for. I want to make a huge accumulator tank. Firewood is almost CO2 neutral, so I will get a wood burning heater for the water tank. I am thinking something like 20000 litre ( 5000 gallon or so ), which to my calculations would last half a month if you draw the equivalent of 3kW continuously and estimate a drop from 80 degrees C to 30 degrees. In addition I will rebuild an old cold room cooler to be a GSHP that will start if the temperature drops below a limit (ca 30C), and electric heaters as an emergency solution if the temperature goes close to 10C. Solar panels will also be installed later, with the hot water heating the tank directly if the temperature is higher exiting the panels, else, it should be used as a heat source instead of the ground loop for the GSHP to increase COP on days with only a little solar input. An air exchanger should be used when air temperatures is the highest. This is a perfect job for the arduino to control. I am working on designing this whole system with temperature limits and connections in order to get everything just right. I will post drawings here when I get to it, but it will take some time. This huge tank will be placed in the basement of my yet to be built garage. Next summer I am guessing.

Thanks for all the good input guys. Homemade GSHP would never crossed my mind if it was not for you guys. Now my old cooler thingy will get some usefulness after all.

[peacmar]

Originally Posted by AC_Hacker

Sounds very interesting... please elaborate.

-AC_Hacker

And so I will....


The basis of this method is to provide a seemingly endless heat sink with wich to protect the adjacent areas to the heated joint. The act itselfis quite literally submerge all of the part except for the joint in a large quantity of water. I have worked on numerous commercial and home made tables but they all work with the same simple principle. The commercial units usually have an air bladder and a fixed height grate or deck to place the work on, the air bladder is filled and displaces water increasing the level till the part is mostly covered. My last diy copper core radiator I assembled by placing inside the cut off lower half of a 55 gallon drum, supported on cement "cinder" blocks, and used the garden hose to fill till just the end tank was exposed. Preventing from releasing the tube and fin core from itself. Now the design of such a thing is completley up to the operator and what ever may be easily available. When I made my hydronic space heater for our bathroom u used a metal brownie pan from good will. Cost $1. The only rull is to not fill it so full the torch splashes the water on the joint or it will weaken of course. I have even Hung a metal bucket over a pipe I was splicing a super heater into a dairy chilled to protect the dhw coil within. It's all a matter of creativity. But a very reasurring piece of mind also. And of course, the container must be metal so it doesn't melt down mid brasing.


-mark

[AC_Hacker]

Originally Posted by peacmar

The basis of this method is to provide a seemingly endless heat sink with wich to protect the adjacent areas to the heated joint.

Mark,

Great info. I can see how I could easily have used this technique when I was brazing copper tubing to my Brazed Plate heat exchangers.

Did you fabricate your hydronic space heater for your bathroom? Sure would like to see photos and learn more details about that one.

Also, earlier you mentioned about rinsing out (or off?) flux with muriatic acid... Is this an internal or external operation? Could you go into more detail there... I'm right at that step on a project I'm doing and I could use that technique right now. Do you need to neutralize the acid before gassing the system up?

And lastly, I don't know if you have a digital camera, but if you do, it would be extremely useful to have a few photos mixed in with your descriptions.

...if you need some help with how to get the photos into the blog, just let me know.

Thanks very much for your help.

Regards,

-AC_Hacker

[Piwoslaw]

Welcome to ER, Ko_deZ Thank you for the info.

Originally Posted by Ko_deZ

This is some really good reading (remove spaces, cannot post links yet):
http://www.ivt.ntnu.no/igb/forskning...v_bolighus.pdf

I edited the link to make it directly clickable. It does seem like a nice read.

Originally Posted by Ko_deZ

One thing that I don't see mentioned anywhere is one of the really huge advantages of having floor heating compared to radiators.

I believe it has been mentioned a few times that floor heating allows lower temperatures, which boost the efficiency when using low temp heat sources (heat pump, heat buffer, condensing boiler, etc.). There is also a thread dedicated to DIY hydronic floor heating which discusses the how-to side.

Originally Posted by Ko_deZ

Now to explain what reading this thread has made me plan for. I want to make a huge accumulator tank. Firewood is almost CO2 neutral, so I will get a wood burning heater for the water tank. I am thinking something like 20000 litre ( 5000 gallon or so ), which to my calculations would last half a month if you draw the equivalent of 3kW continuously and estimate a drop from 80 degrees C to 30 degrees.
[...]
This huge tank will be placed in the basement of my yet to be built garage.

A 20000 liter (5283 US gallon) tank will pretty much take up the whole garage:

But if you will be building around it, then this may work just as well as burying it underground. Osolemio mentioned that he has a large mass to store heat. I believe that the best use of a 20000 liter tank would be low temperature seasonal storage, and from there you can pump heat into a smaller (1000-3000 liter) superinsulated tank for home use. When you start building your system, please post about it in the heat accumulator thread.

[peacmar]

I will bring pictures to this discussion hopefully in short time. We lost both computers about 2 months ago during a thunderstorm and all of my archived photos are currently in limbo until they are replaced and the data retrieved. Only the power supply and mother board where damaged. The hard disks and my mass storage are still in tact, but we are waiting for the surge protectors "warranty" to replace what it didn't protect. Unfortunate situation but such is life.

The space heater is something of my own creation, I am building another for our front entry way, wich acts as an air lock between the kitchen and outside world. We bought the property 2 years ago, almost 3, and the first winter was terrible for heating.$3500 in lp and the furnace was new installed just before we purchased the house. Last fall I built a down draft gasifier wood stove which is outside in its own structure then piped into the house for heating needs. I was able to purchased a de-commissioned 500 gallon lp tank from the local co-op for scrap weight price, and that now serves as thermal mass storage so we have a total of 800 gallons of water storing heat. I now burn year around as this provides our hot water needs also until I can get solar up and functioning. But only burn once a week in the summer, and only consume about 2 plastic 5 gallon buckets worth if seasoned cord wood per week in the summer. I'm estimating about 95% efficiency in the burner. But ill wait to start a while new thread else where for that on another day....



As far as flushing with muriatic acid, that's for the internal of the system. When building from scratch, and there isn't an easy source of purge gas available. I flux both inside and out, then after all but the compressor are connected and brased. Careful planning and measuring of course. I carefully pour the acid through the system, using great care. Then flush with water into a bucket with baking soda to neutralize. If the acid has sat long enough all contaminants are removed. If the unit is small enough an old house hold oven set at 300 degrees can dry the unit out in an hour. Heat the oven, let it cook for half an hour. Then let it sit till it cools. This is generally only suitable for smaller assemblies. Then I pull a vacuum to remove any remaining moisture before the pump is uncapped and connected to the system. Now I have never actually charged any of the systems I have assembled, I usually work with local shops and help out with the fabrication part. They take care of that portion, but i am learning to become self sufficient in that respect.


-mark

[peacmar]

I also want to throw this out there while the discussion is fresh.

Muriatic acid is a very corrosive substance and great care should be taken while handling. Long cuff chemical resistant gloves and goggles are a must. Long sleeve clothing and pants of course should also be worn. Safety is ni joke. And neither is acid.

-mark

[AC_Hacker]

Ko_deZ,

Quite a thoughtful message you have posted here. I appreciate the fact that you took the time to communicate so well.

Originally Posted by Ko_deZ

This moves a whole lot of energy down into the ground and the groundwater that carries it. The speed of the groundwater is such that when winter comes, the heated water reaches the other end of the runway, where another set of wells are set up. These go to heatpumps for heating trough the winter. The warm water should give an exceptionally high COP.

This certainly shows that the thinking about geothermal is very advanced in Norway.

Originally Posted by Ko_deZ

Well spacing:
I have been checking a whole lot around this. I did at one point find documentation saying that unless you have moving groundwater, the temperature of the ground will be reduced quite a bit during the first 20 years or so.

This is in agreement with what I have observed. Even my small loopfield has shown significant change in temperature over the brief time (about a two-month period) that I did continuous-run testing. I know that I have ground water at the lower parts of my boreholes, but I have no information yet as to what flow-rate might be present, if any.

I have certainly moderated my thinking as to the amount of heat that may be available from a small loopfield. I still have more testing to do, but my current thinking is that unless my house insulation efforts are extremely effective, I will not be able to satisfy my entire heating requirements from my small loopfield.

The heating requirements for my geographical area are very, very moderate compared to Norway. An Air Source Heat Pump is able to function fairly efficiently through most of the winter here. There are however, periods when an ASHP must really labor at best or be temporarily replaced by another heat source, at worst. So, I am thinking about using my loopfield as a "fill-in" for thos periods when a ASHP is not operating very efficiently.

There is another line of investigation I am interested in and that is utilizing the loopfield for summer cooling, which will add to the available heat during the winter cycle.

Originally Posted by Ko_deZ

Well depth:
...The first hundred meters of ground have a temperature based on solar input... and the fact that very much of the ground is slightly radioactive... Some rock, like granite, is quite radioactive, and that gives a higher temperature...

It never before occurred to me that I might be living on top of a nuclear reactor.

Originally Posted by Ko_deZ

Water vs dirt in well
Around here (Norway), it is my understanding that we never use grout. The well is just left like it is (with a pipe protecting from collapse in the top soil part, we always hit rock after a short period here) allowing ground water to fill the well up.

Interesting... I had heard of Canadian wells that used small gravel around the well circulation pipe, instead of grout, to take advantage of ground water flow. I considered that but instead decided to replace the dirt that came out of the hole, so that it would be like it was (last out, first in).

Originally Posted by Ko_deZ

Series vs parallell, freezing and flow resistance
I see that AC hacker has his wells in series. That is not the ideal solution I am afraid.

My thinking was shaped primarily by work done by Oklahoma State University, which focused on ease of purging (serial is much easier than parallel) and by cumulative friction (parallel requires less powerful pumps). There was a 'rule of thumb' that if a serial branch exceeds a certain amount of friction (based on length, pipe diameter, % antifreeze additives, etc.) then it should be broken into shorter serial branches. By my calculations, I was close to the maximum, or not very much beyond. Now that I have pumped water through the entire system, I realize that I should have broken it at least once, possibly more, to reduce friction.

The considerations of heat flow when the working fluid approaches freezing were not known to me at the time. Thank you for the valuable explanation.

Originally Posted by Ko_deZ

The dude earlier talking about this, he was right. You should not have blown him off like that, he could very well have been a very useful and seemingly well educated asset here.

I highly regret that I discouraged a knowledgeable contributor to this blog. If you could be so kind as to tell me the number of the post (upper right corner) or supply the hyper-link to that post, I would like to review it.

We're all here to learn from our successes and our failures, including me.

Originally Posted by Ko_deZ

Oh, and by the way, the reduced friction and increased total waterflow of having them all in parallel does also increase the efficiency of the heat exchanger, so an added bonus there.

Quite so, as I have learned.

Originally Posted by Ko_deZ

Heat exchanger
Looking at the heat exchanger solution used with the brazed plate exchanger for the hot side of the heat pump, I cannot see the gain of this compared to having a coil or radiator in the accumulator tank. Given a long enough copper line and coil, that would remove the need for one water pump (and then also a COP ~= 1 device in the loop), and give less turbulence in the accumulator tank. Having the coil go from the top to the bottom of a tank would probably yield a higher top temperature and lower bottom temperature, making the undercooling better. Of course, getting the water from the bottom of the tank with a pump will do this too, but it will also stir the water, reducing the thermal gradient in the tank. In a very still tank you can get a significant thermal gradient, making the top maybe 50 degrees while the bottom is maybe 30. Having the condensed gas leave the tank at 30 rather than the average temperature of 40 would give a better COP.
A completely new system I saw just some months ago was claiming better COP with a tank with a very special design. They could achieve 80 degrees in the top and 30 in the bottom of a 1.80 meter tall tank. This tank had no water movement whatsoever (except convection), all energy input and draw was trough spiral copper tubes. If you wanted hot water, you went from bottom and to the top with the waterflow, giving you almost 80 degrees. For the floor heating your spiral started from bottom and stopped in the middle, giving about 40 degrees. They claimed a much higher efficiency than other accumulator tanks, and support for any heat source, they just placed the coils at the right positions to regulate the entry and exit temperatures. Ingenious and very, very expensive to buy. Seemed to be complex to make also, as convection was reduced to small areas around coils with perforated walls to avoid stirring up the water more than absolutely nessesary.

This is great info. One clear advantage of the brazed plate is that engineering data is available for them. I do not know yet how to calculate heat transfer for a copper tube in water, except to build and test.

I can clearly see the advantage of eliminating a pump (the compressor is the pump). So the hot side is kind of a DX in this configuration, right?

On the flip-side however, by using two brazed-plate HXs you do gain the advantage of modularity and portability, and the heat pump system is a hermetically-sealed whole, quite nice for experimenting.

Originally Posted by Ko_deZ

Floor heating is self adjusting
One thing that I don't see mentioned anywhere is one of the really huge advantages of having floor heating compared to radiators. Usually the air temperature to floor temperature is one a couple of degrees, while radiators are really hot, maybe 30 degrees warmer than the air. What this does is that if the outside temperature drops significantly, the indoor air will cool down. If we say that the floor-air difference was 2 degrees, and the air temperature drops by 2 more degrees, the difference is now 4 degrees. This will in essence double the convectional thermal energy dissipation from the floor, giving a huge change in supplied heat. Compare that to the radiator solution, the difference will only increase by a fraction of the original difference giving a very slight energy boost. For the radiators to increase the heating there must be a rise in temperature of the water and/or a higher flow rate. Compared to this, the floor heating is almost self adjusting. If the temperature drops, the energy given from the floor rises very quickly, avoiding the temperature fluctuations you often see with radiators when the weather change for the better/worse.

Like Piwoslaw said, and linked to, there is a discussion of floor heating. However, this self-regulating feature has never been addressed before. I would guess that you would need high-mass floors for this to work, right?

Originally Posted by Ko_deZ

Floor heating DX, not possible because of heat difference DX to floor heating has some problems as discussed here. One of the really big problems with it is that you really want to undercool the coolant as much as possible. This comes in direct confrontation with the wish for an even floor temperature. If the incoming superheated gas is 50C, and the exiting should be 30, you will need some really fancy heat protection for the first bit of pipe, reduced gradually to a very high dissipation solution with heat spreaders at the returning end. It is not a good solution.

Again, this info has never appeared here before. Thank you.

-AC_Hacker

[AC_Hacker]

Ko_deZ,

I would like to see what your thinking is regarding radiant floor construction, for low-temperature heating, such as GSHP and solar...

I have several questions...

Since this particular line of inquiry is better suited to the PEX floor thread, we should continue there...

[Ko_deZ]

Answered some questions regarding floor heating in the other thread. High mass floor not needed for the self regulating nature of heated floors.

Regarding brazed plate versus DX, yes, there are more options for how to connect different stuff to the HP with plates, but there are other ways of getting different temperatures from the HP that are way more efficient. The professional HPs usually has a shunt that swaps them between hot water and floor heating accumulators. That gives a very low COP when running on the hot water though, as the return from the condenser is very high. The most efficient way to do this would be to have a DX go trough three tanks in succession. The first accumulator tank will grab the superheat for heating tap water, the second one would be the accumulator for the floor heating. The last one is often called a undercooler, and reduces the temperature below the 30C or more in the floor heating tank. This can be used to heat outdoor stairs or stuff for freeing them from snow, heating a shed or something. This last one is in essence completely free energy. The compressor will have to do the same work as without it, but your ground loop must be a bit larger to accommodate the extra energy pulled from the hot side. This gives a better COP as well, and if used to heat rooms that should be cool, like bedrooms, it will give you a better COP as well. The logic that starts the HP would measure both tap water and floor heating accumulator and start if either of them drops below a certain point, resulting in heating both of them.
Obviously this could also be done using plates, but you would need three extra pumps for that to get the water to the accumulator tanks.
Oh, and also keep the heated floor accumulator somewhat higher in temperature than is needed, and put an electronic controlled mixer for the floor (some water going trough the accumulator and getting heated, while some just goes back in the loop). That would allow you to heat the tank for longer and making it warmer. This results in avoiding rapid starts and stops of the HP, as well as the water going to the floors having a more stable temperature. Should yield a higher COP as the HP needs a little time to really get going.

For the cold side it would be, if not very silly, so at least not efficient, to use a "cold accumulator tank". As the energy is pulled from the tank, the temperature drops in it. Since the incoming "hot" water from the ground loop mixes with the water in the tank, the water going to the exchanger will be colder than it would taking it directly from the ground loop. You will never be able to get the output gas from the evaporator very close to the ground loop temperatures. Using a plate exchanger you could get it a lot closer since you do not accumulate cold water. If you add three plate exchangers on the cold/low(?) side, you could have one from the ground loop first, one from an outside air unit, and lastly one from solar panels (in that order). Those sources needs their own pump anyway (but dropping the need for one from the cold accumulator -> higher COP), and if put in that order, the input to the compressor would be at least as warm as the ground loop. If outside ambient is higher than the ground loop, run the outside air loop pump too. And if sunny, run all of them. This should increase COP on days with warmer air than ground, and days with some sun. Maybe it is so sunny that air and ground loop pump can be stopped even, giving even higher COP. Not sure how hot one can actually make the gas going to the compressor before it becomes a problem. Maybe someone here knows?

[mgilbert754]

OK Im hooked! We have a 2600 sq ft house in the middle od SC. We have long hot summers and several months in the winter can get quite cold. Our house badly needs new window and wall insulation, but I'm really interested in DIY ground heat pumps. I"m lucky= we have nice sized pond in the back yard and I want to use that for a heat sink.
Can you give me more guidance in building the actual heat pump, including a desuperheater for hot water?