The Homemade Heat Pump Manifesto - Page 57

dragvorl · 02-17-11, 07:26 PM

Hi guys, I joined the forum because of this thread, so I would like to thank AC hacker and all the others for the useful information I found here. I'm thinking about building a system like this, but I want to ask a few questions about heat exchangers. First, I've seen that water to water exchanger on your first heatpump Ac hacker, if I am not mistaking? Is that heat exchanger efficient enough for refrigerant gases as well, or it's a little less efficient in these heat pump applications? I'm thinking about a heat exchanger like the one on the picture, freon flowing in the copper tubes, and the copper tube would be put in a bigger tube which would be filled with water coming from the well? What do you think about it, would it be good enough for this application?

EDIT: I can't post links or images, so I left the link of the pic.

renewableenergyengineering.com/yahoo_site_admin/assets/images/Six-Heat-Exchanger-Coils.122132611_std.jpg

AC_Hacker · 02-17-11, 08:24 PM

Quote:

Originally Posted by dragvorl

...I've seen that water to water exchanger on your first heatpump Ac hacker, if I am not mistaking? Is that heat exchanger efficient enough for refrigerant gases as well, or it's a little less efficient in these heat pump applications?

The heat exchanger you are referring to is a brazed plate heat exchanger. It consists of alternating plates and the 'even pair' of plates carry one fluid and the 'odd pair' of plates carries a separate fluid. They are very high quality and can be used with many different fluids. I took a risk and bought a pair that were meant for chilling beer with cold water. But instead of beer in one circuit and water in the other, I ran refrigerant and water. It worked out pretty darn well. I have had problems with very slow leaks of refrigerant (and it's lubricant) but I now think I can fix that. I have since found that the same heat exchangers can be bought with 'sweat' fittings on the refrigerant side that can be brazed and offer superior sealing for refrigerants. The price from my source was the same.

Quote:

Originally Posted by dragvorl

I'm thinking about a heat exchanger like the one on the picture, freon flowing in the copper tubes, and the copper tube would be put in a bigger tube which would be filled with water coming from the well? What do you think about it, would it be good enough for this application?

If you are thinking about well water, the heat exchanger you described would be better than brazed plate, as it would not be likely to foul, compared to brazed plate, which certainly would foul in your application.

EDIT: I can't post links or images, so I left the link of the pic.[/QUOTE]

WELL THEN, HERE YOU GO...

-AC_Hacker

dragvorl · 02-17-11, 08:38 PM

Thanks for the swift answer. Do you have some idea of where could I find calculations for the kind of heat exchanger I've been talking about? I don't know if the same pipe lenghts as on an normal AC (refrigerant to air) would be good for the refrigerant to water heat exchanger.

AC_Hacker · 02-17-11, 09:04 PM

[QUOTE=AC_Hacker;11986]Hi morphector,

This is a good idea. It is very common to use water or concrete for short-term storage to keep the compressor from coming on too often. When the storage is used for short term like this it is referred to as a 'buffer'. When a 80 gallon (or so) tank of water is used this way it's referred to as a buffer tank. The heat is withdrawn from the tank perhaps several times per hour, to once every few hours, depending on heating requirements.

And if you really think about it, a GSHP system is using the earth for heat storage, but it is very long term heat storage, the heat is deposited maybe just by the sun during the warmer months, or if you are using a GSHP for cooling, by the sun AND you are storing heat that is extracted from your house during the summer, too.

But If I understand you, you are thinking about using a tank for medium-term heat storage. Water is a good choice because it is cheap and environmentally safe.

I'll simplify things a bit by assuming that the tank will not lose any heat and see if I decided to use something like this on my house, how it might work out. [You could do a similar analysis if you know what your design temp is, what your heat load is, how big your heat pump would have to be to do the job, etc, etc...]

1000 gallons of water would weigh 8350 pounds.

The hottest I might be able to heat the water up would be about 180F degrees. I'm choosing this number because most of the plastic piping and tank materials I'm familiar with will begin to soften above 180F. I also have to figure that my heat pump looses efficiency as it has to elevate temperature more, so I'll set my upper limit to 110F, and see what that looks like. [note that if you are considering central air, your heat pump will have to raise the water temps to 120F or higher]

The coldest I could go on the tank would be 32F because that's when ice forms, and liquid flow stops. I think a three degree margin of safety would be smart, so I'll say that my minimum temperature would be 35F.

If I were to consider the heat that could be stored in one pound of water over that heat range, I would multiply the heat range times the amount of water.

BTU = (T1 - T2) * W

BTU = (110 - 35) * 1

BTU = 75 BTU

But since I have 8350 pounds of water,

BTU = (T1 - T2) * W

BTU = (110 - 35) * 8350

BTU = 626,250 BTU

My house is small, at 700 sq ft, and it is well insulated, and the temperatures where I live are pretty moderate. Also, I have adopted the strategy that when temperatures plunge down to the low numbers, I start closing off parts of the house I don't absolutely need to heat. So I have learned that I can actually get by with 3/4 Ton (9,000 BTU/hr) of heating.

I have observed that on warmer spring and fall days, I don't need any extra heat at all when the outside temp is 55F or higher. So this would be the equilibrium temperature.

Around here where I live, the 'design temperature' for heat pump systems is 17F. In other words, my heat pump should be able to heat my house (for me, the part that I choose to heat) to 66F without any axillary power. I can expect that in any year there will be less than a week of days at this temperature. I do know that in a year we may get one or two days at temps LOWER than 17F, for instance, earlier this year it hit 12 for about 12 hours. I turned on the kitchen stove oven for this and used a resistance heater a bit. In thirty five years I have seen it go as low as 5F only once.

So, I'll use 66F as my balance point, and 17F as my design temp. The average between those temps is 41.5F which, interestingly, is pretty darn close to the average temperature here in the winter (statistical average Dec & Jan temp is 39.5F). Since I turn the heat off at night, I'll go with 41.5F as my average temperature.

The new heat pump that is now sitting in parts on my work bench, that am building should be able to give me 7,800 BTU/hour. Not so much by average American standards, but since I'm not an average American, it should do OK.

I'm going to make a simplifying assumption that if it takes 9,000 BTU/hr to keep my house warm on a 'design day' (17F) then it should take 4,500 BTU/hr to keep it warm on an average day. This is actually a bit of an over-estimate, as I have monitored my power consumption of my mini-split to be somewhat less than that, but I'll stick with the 4500 BTU/hr for simplicity...

So on an average day, my heat pump will produce heat for my house and still have a bit of margin of excess power that could be used to slowly bring the 1000 gallon tank up to temperature, but I will disregard that. At night when the power is cheaper and I don't need to heat my house, I could use use all or nearly all of the duty cycle to warm the tank.

To bring the tank from 43F (my typical unheated basement temperature in winter) to 110F would require 107 BTU per pound or 107 * 8350 = 893,450 total. If I assumed that night was 8 hours long and I was running a 100% duty cycle, I would produce 8 * 7,800 BTU/night or 62400 BTU/night.

So it would take me 893450 BTU divided by 62400 BTU/night = 14.3 nights or just over two weeks to bring it up to temperature.

So, now let's look at what that would do for me during the 'design days'...

During the design days, I would require 9,000 BTU/hr during the 16 hours I heat my house. This would be a daily heat load of 144000 BTUs. Since I could replenish 62400 BTU at night, my net daily heat load would be 144000 - 62400 = 81600 BTU/day.

So running only off of the heat storage tank during the day, and replenishing during the night, I could endure 626,250 BTU / 81600 BTU/day = 7.5 'design days'.

For my small house, that would work just great and I could get by quite nicely. But getting back to your question, when the oitside temps go lower than design temp, the heat in the tank would be used up more quickly.

Some improvements to this system might be the addition of a solar panel or two to help heat the tank on our rare sunny days. Where I live, the winter time is mostly overcast. But when it gets really cold, we lose our cloud cover and the sun usually breaks through, so that could really help.

BTW, here's a link to a technical illustration of a 1000 gallon tank.

It looks like 1000 gallon tanks cost about $600, shipping extra.

If you're considering glycol in the condenser:

The condenser is the heat exchanger that gets hot and is usually inside the house when the heat pump is used for heating.

But otherwise, glycol usually reduces efficiency by around 15% depending on the amount of glycol used, and it also has a higher viscosity than water... so if glycol is used, the loop field and heat exchanger and circulation pump in the glycol part of the circuit need to be sized larger to accommodate the changes.

Hope this helps...

Regards,

-AC_Hacker

AC_Hacker · 02-17-11, 09:16 PM

Quote:

Originally Posted by dragvorl

Thanks for the swift answer. Do you have some idea of where could I find calculations for the kind of heat exchanger I've been talking about? I don't know if the same pipe lenghts as on an normal AC (refrigerant to air) would be good for the refrigerant to water heat exchanger.

Great question.

You might try searching for "coaxial heat exchanger" or some similar variant, on some of the refrigeration or overclocking blogs like these:

If you find anything please report back, because other people really want to know this too.

Regards,

-AC_Hacker

dragvorl · 02-20-11, 09:11 AM

I think I actually found something that could give me a good idea for further thinking.

AC_Hacker · 02-20-11, 11:46 AM

Quote:

Originally Posted by dragvorl

I think I actually found something that could give me a good idea for further thinking.

Do you have any performance data on this unit? Were you able to locate any general coaxial performance data?

-AC_Hacker

dragvorl · 02-20-11, 12:00 PM

Nope, I did not, but I've found this interesting clip on YT.

Quote:

This is a heat pump water heater that I built, for heating potable hot water. It's based on parts from a 10,000BTU air conditioner, and various parts from the hardware store and refrigeration parts store. It provides about 3kW of heating capacity at 20C ambient temperature.

AC_Hacker · 02-20-11, 08:29 PM

Quote:

Originally Posted by dragvorl

nope, i did not, but i've found this interesting clip on yt. Youtube.com/watch?v=t5gf9dptcq0 the guy said it is an 8m long coaxial heat exchanger.

Quote:

this is a heat pump water heater that i built, for heating potable hot water. It's based on parts from a 10,000btu air conditioner, and various parts from the hardware store and refrigeration parts store. It provides about 3kw of heating capacity at 20c ambient temperature.

I saw the video, looks very interesting.

He's getting refrigeration parts here... which is good to know about.

They sell coaxial condensers of various sizes, might be possible to reverse engineer one with the help of their descriptions.

-AC_Hacker

* * *

Drake · 02-20-11, 09:09 PM

Photo of manufactured coaxial condenser Xgrs show smaller tube exiting larger tube perpendicular to coil. Homemade ones I've seen show small inner tube exiting inline and large tube from side. Is this difference of consequence?

02-17-11, 09:04 PM	#564
AC_Hacker Supreme EcoRenovator Join Date: Mar 2009 Location: Portland, OR Posts: 4,004 Thanks: 303 Thanked 723 Times in 534 Posts	[QUOTE=AC_Hacker;11986]Hi morphector, This is a good idea. It is very common to use water or concrete for short-term storage to keep the compressor from coming on too often. When the storage is used for short term like this it is referred to as a 'buffer'. When a 80 gallon (or so) tank of water is used this way it's referred to as a buffer tank. The heat is withdrawn from the tank perhaps several times per hour, to once every few hours, depending on heating requirements. And if you really think about it, a GSHP system is using the earth for heat storage, but it is very long term heat storage, the heat is deposited maybe just by the sun during the warmer months, or if you are using a GSHP for cooling, by the sun AND you are storing heat that is extracted from your house during the summer, too. But If I understand you, you are thinking about using a tank for medium-term heat storage. Water is a good choice because it is cheap and environmentally safe. I'll simplify things a bit by assuming that the tank will not lose any heat and see if I decided to use something like this on my house, how it might work out. [You could do a similar analysis if you know what your design temp is, what your heat load is, how big your heat pump would have to be to do the job, etc, etc...] 1000 gallons of water would weigh 8350 pounds. The hottest I might be able to heat the water up would be about 180F degrees. I'm choosing this number because most of the plastic piping and tank materials I'm familiar with will begin to soften above 180F. I also have to figure that my heat pump looses efficiency as it has to elevate temperature more, so I'll set my upper limit to 110F, and see what that looks like. [note that if you are considering central air, your heat pump will have to raise the water temps to 120F or higher] The coldest I could go on the tank would be 32F because that's when ice forms, and liquid flow stops. I think a three degree margin of safety would be smart, so I'll say that my minimum temperature would be 35F. If I were to consider the heat that could be stored in one pound of water over that heat range, I would multiply the heat range times the amount of water. BTU = (T1 - T2) * W BTU = (110 - 35) * 1 BTU = 75 BTU But since I have 8350 pounds of water, BTU = (T1 - T2) * W BTU = (110 - 35) * 8350 BTU = 626,250 BTU My house is small, at 700 sq ft, and it is well insulated, and the temperatures where I live are pretty moderate. Also, I have adopted the strategy that when temperatures plunge down to the low numbers, I start closing off parts of the house I don't absolutely need to heat. So I have learned that I can actually get by with 3/4 Ton (9,000 BTU/hr) of heating. I have observed that on warmer spring and fall days, I don't need any extra heat at all when the outside temp is 55F or higher. So this would be the equilibrium temperature. Around here where I live, the 'design temperature' for heat pump systems is 17F. In other words, my heat pump should be able to heat my house (for me, the part that I choose to heat) to 66F without any axillary power. I can expect that in any year there will be less than a week of days at this temperature. I do know that in a year we may get one or two days at temps LOWER than 17F, for instance, earlier this year it hit 12 for about 12 hours. I turned on the kitchen stove oven for this and used a resistance heater a bit. In thirty five years I have seen it go as low as 5F only once. So, I'll use 66F as my balance point, and 17F as my design temp. The average between those temps is 41.5F which, interestingly, is pretty darn close to the average temperature here in the winter (statistical average Dec & Jan temp is 39.5F). Since I turn the heat off at night, I'll go with 41.5F as my average temperature. The new heat pump that is now sitting in parts on my work bench, that am building should be able to give me 7,800 BTU/hour. Not so much by average American standards, but since I'm not an average American, it should do OK. I'm going to make a simplifying assumption that if it takes 9,000 BTU/hr to keep my house warm on a 'design day' (17F) then it should take 4,500 BTU/hr to keep it warm on an average day. This is actually a bit of an over-estimate, as I have monitored my power consumption of my mini-split to be somewhat less than that, but I'll stick with the 4500 BTU/hr for simplicity... So on an average day, my heat pump will produce heat for my house and still have a bit of margin of excess power that could be used to slowly bring the 1000 gallon tank up to temperature, but I will disregard that. At night when the power is cheaper and I don't need to heat my house, I could use use all or nearly all of the duty cycle to warm the tank. To bring the tank from 43F (my typical unheated basement temperature in winter) to 110F would require 107 BTU per pound or 107 * 8350 = 893,450 total. If I assumed that night was 8 hours long and I was running a 100% duty cycle, I would produce 8 * 7,800 BTU/night or 62400 BTU/night. So it would take me 893450 BTU divided by 62400 BTU/night = 14.3 nights or just over two weeks to bring it up to temperature. So, now let's look at what that would do for me during the 'design days'... During the design days, I would require 9,000 BTU/hr during the 16 hours I heat my house. This would be a daily heat load of 144000 BTUs. Since I could replenish 62400 BTU at night, my net daily heat load would be 144000 - 62400 = 81600 BTU/day. So running only off of the heat storage tank during the day, and replenishing during the night, I could endure 626,250 BTU / 81600 BTU/day = 7.5 'design days'. For my small house, that would work just great and I could get by quite nicely. But getting back to your question, when the oitside temps go lower than design temp, the heat in the tank would be used up more quickly. Some improvements to this system might be the addition of a solar panel or two to help heat the tank on our rare sunny days. Where I live, the winter time is mostly overcast. But when it gets really cold, we lose our cloud cover and the sun usually breaks through, so that could really help. BTW, here's a link to a technical illustration of a 1000 gallon tank. It looks like 1000 gallon tanks cost about $600, shipping extra. If you're considering glycol in the condenser: The condenser is the heat exchanger that gets hot and is usually inside the house when the heat pump is used for heating. But otherwise, glycol usually reduces efficiency by around 15% depending on the amount of glycol used, and it also has a higher viscosity than water... so if glycol is used, the loop field and heat exchanger and circulation pump in the glycol part of the circuit need to be sized larger to accommodate the changes. Hope this helps... Regards, -AC_Hacker __________________ I'm not an HVAC technician. In fact, I'm barely even a hacker...

02-20-11, 09:11 AM	#566
dragvorl Lurking Renovator Join Date: Feb 2011 Location: NS Posts: 7 Thanks: 0 Thanked 0 Times in 0 Posts	I think I actually found something that could give me a good idea for further thinking. Attached Thumbnails Last edited by dragvorl; 02-20-11 at 09:36 AM..

02-17-11, 07:26 PM	#561
dragvorl Lurking Renovator Join Date: Feb 2011 Location: NS Posts: 7 Thanks: 0 Thanked 0 Times in 0 Posts	Hi guys, I joined the forum because of this thread, so I would like to thank AC hacker and all the others for the useful information I found here. I'm thinking about building a system like this, but I want to ask a few questions about heat exchangers. First, I've seen that water to water exchanger on your first heatpump Ac hacker, if I am not mistaking? Is that heat exchanger efficient enough for refrigerant gases as well, or it's a little less efficient in these heat pump applications? I'm thinking about a heat exchanger like the one on the picture, freon flowing in the copper tubes, and the copper tube would be put in a bigger tube which would be filled with water coming from the well? What do you think about it, would it be good enough for this application? EDIT: I can't post links or images, so I left the link of the pic. renewableenergyengineering.com/yahoo_site_admin/assets/images/Six-Heat-Exchanger-Coils.122132611_std.jpg

02-17-11, 08:38 PM	#563
dragvorl Lurking Renovator Join Date: Feb 2011 Location: NS Posts: 7 Thanks: 0 Thanked 0 Times in 0 Posts	Thanks for the swift answer. Do you have some idea of where could I find calculations for the kind of heat exchanger I've been talking about? I don't know if the same pipe lenghts as on an normal AC (refrigerant to air) would be good for the refrigerant to water heat exchanger.

02-20-11, 09:09 PM	#570
Drake DIY Guy Join Date: Jan 2011 Location: Mpls,MN Posts: 315 Thanks: 2 Thanked 17 Times in 17 Posts	Photo of manufactured coaxial condenser Xgrs show smaller tube exiting larger tube perpendicular to coil. Homemade ones I've seen show small inner tube exiting inline and large tube from side. Is this difference of consequence?