Solar Hot Water Collectors

[Ric Murphy]

Recently got back to working on a project that I started a few years ago (one of many!!). I'm building 2 prototype solar hot water collectors of a slightly different design than most. The box consists of an inner and outer shell with 1" insulation between. The outside dimensions are 28"x74". These dimensions were selected in order to use double pane patio door glass which is readily available for free. I've been saving 4 sets from a house retrofit about 20 years ago.The boxes are made of aluminum to minimize weight. The insulation consists of 3/4" foil backed polyisocyanurate (R5.0) and 1/4" foil faced bubble wrap (R4.3). On the inside of the box is a single absorber plate also of aluminum. The absorber plate has a 1" gap on all sides and is elevated 1" off the box floor allowing hot air the circulate around all surfaces. The absorber plate is mounted on 1" sq tubing riveted to the box floor. The coil is made from a single 50' piece of 3/4" flexible copper tubing so contains no internal joints. The original design called for 3-24" coils (25' in each for a total of 75') but I found I could not form the inner most loops without the tubing going flat. The current layout was much easier to form. The coil will be attached to the absorber plate with copper clips and rivets thru the absorber plate into the 1" sq tubing. I'm looking for a compound to use between the copper coils and the aluminum absorber plate. The compound needs to act as an adhesive as well as assist with heat transfer. Not sure what to use at this point so if anyone has any suggestions I'm all ears. Once I find that item it shouldn't take long to finish them up. I plan to collect and store the hot water during the day for use during the night in the greenhouse. To start I'll simply circulate the water thru a used truck radiator with a shroud and A/C fan. Some pictures below...
Ric
384

385

386

387

[Paul Bailey]

Hi Ric : I see you have put alot of work into this..and $$$. 3/4 " tube is not cheap!!. Still unsure of a bonding agent that will give you both ( thermal and mechanical coupling. . Alot of the collectors nowadays are using much smaller tube (3/8 ") that is compression crimped to plates the full length of the tube (Google sunstrip I think is an example) > You may have to switch to a copper backplate and solder the coil to it .Paul:)

[Paul Bailey]

Yep ! Another reply> As you are aware this bonding will make or break its performance. Here is a nice picture /example of radiant plates(bonded ) Vs. those who choose none and then later complain of poor performance and Efficiency.

[Ric Murphy]

Paul,
Thanks again for your input. Cost hasn't been too bad (I don't think anyway). Cost me about $450 for all the aluminum (material and labor). Since the copper tubing was purcashed 2-1/2 years ago it was a fair deal cheaper than today. After my problems forming the small coils with the 3/4" I was going to go back to 5/8" (which was what I was originally going to use until someone talked me out of it!!!). The price today of 5/8" is the same as I paid for the 3/4" 2-1/2 years ago!!! Thats when I decided to revise the coil layout to what you see in the pictures. I realize now that I should have used copper for the absorber plate but will finish these 2 with the aluminum. I recognize the importance of the bonding agent......your 2 graphics convey that very well also. So for the moment I'm at a standstill. I found the link below from the BuilditSolar website. This guy mentions he used a Dow Corning product between aluminum and copper but doesn't mention it's name (about half way down the page on the right). His absorber plates were the type you described. My main reason for going with a coil layout was simply to eliminate any joints inside the box...zero failure points. Another problem with my design is the panels won't self drain. Fluid will always stay trapped in the lower half of the coils. Either have to use air or syphon the drain completely if ever needed.
Ric
http://users.telenet.be/hagim/zonne_energie/building%20solar%20collector.htm

[Ric Murphy]

Well I see I haven't worked on this project since last Feb....where did the summer go???
As of today both copper coils are formed and mounted on the absorber plate. The absorber plate/coil assemblies are mounted into the inner pan. Have not found a suitable compound to use between the copper coils and aluminum absorber plate so will likely complete the 2 prototypes without and take the efficiency loss. Next step is to paint the interior with black high heat paint and mount the glass. I'll likely build some type of temporary mounting fixture to test. Hopefully can start plumbing the tanks, pump and radiator next week. Some pictures of todays progress below.
Ric
570

571

572

[Joe Blake]

Sweet looking piece of work. Deserves to work on that basis alone.

Joe

[Ric Murphy]

Thanks Joe,
I wish good looks counted for something......I've got a lot of good looking projects kicking around that never panned out!!! How come the butt ugly ones seem to work best??
Ric

[Doug Arndt]

Ric, I presume you've considered silicone and rejected it, would you (or others), for my education, explain why ? I know it's not a heat transfer product, but it's better to not use it ?

Thanks, Doug

[Rob Beckers]

Hi Ric,

To echo the prevailing sentiment: Beautifully constructed! What are the pans made off? Is this stainless steel, or aluminum?

To echo another of the comments: Unless you get a proper thermal connection between the tubing and the collector surface I don't think this is going to work very well. If the only connection between the tubing and the plate is the very small surface area of the tube that is clamped onto it, then you will have a very large resistance (and corresponding temperature drop). It takes a lot of heat to warm up water, and a corresponding large heat flow (current). The resistance to heat has to be low for that to work.

How about thermal grease, as used for electronics? A 1 pound can goes for $15 at Digikey (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=251G-ND). Maybe E-Bay has even better deals?

Another thought would be to use a thin sheet of copper, attach it to the backing sheet (pop rivets or something), attach the copper tubing, and then solder the tubing to the copper sheet.

-RoB-

[Ric Murphy]

Rob/Doug,
The pans are aluminum. I had them fabricated at a sheet metal place in Simcoe.
Silicone was one of many compounds I considered. Silicone should withstand the heat and usually has good adhesion. But it's not paintable so would have to be applied after painting. I don't perceive it to have any heat transfer properties so it would serve no purpose that I can see. Actually I think of silicone as more of an insulator and in this case feel it might impede heat transfer. Someone else had suggested thermal grease but whenever I think of grease in a high heat environment I just envision a mess. One other compound I considered was what we used to call "potting" compound. We used it in an electrical connector that was on a wiring harness that went inside an Allision automatic transmission. The harness was exposed to extreme high heat, huge temperature swings and tranny fluid. This compound was a 2 part epoxy and took about 24 hours to cure at room temp. We accelerated the cure by baking the harnesses in an oven at around 150F for about 1-1/2 hours. The compound cured to a very hard, glass like state. Had great adhesion too but don't think it had very good heat transfer properties. Plus the consistancy was very runny until it set up so using it to form a bridge between the coil and absorber plate would likely be messy.
So at this point I can either finish assembling and test or wait until I find something suitable. I can always go back later and open up the boxes to add. That would give me some data for comparison as well.
A few years ago I built a heat exchanger for the greenhouse. I store rainwater for irrigation in an underground tank. In the winter the water temp is around 45-50F which I thought might be too much of a shock for the plants. So in an attempt to pre heat the water, I formed a double coil of 5/8" copper (50' of tubing). The coil went into the hot air duct of the furnace. If I remember correctly I was able to raise the water temp from 50 to 70F in one pass. However I used a very, very low flow rate to achieve this increase. I never did use it as the water temp out of the tank didn't seem to bother the plants. I'll dig it out today and post a picture of it later.
So back to the solar panels...
When I try to understand the "magic" that takes place inside one of these here's my train of thought....
The final "transfer" takes place inside the coil between the hot metal and the cooler liquid. As the metal gives up it's heat to the liquid it must be replaced. The surface area of the coil must be insufficient on it own to do this efficiently? So the absorber plate is used to collect additional heat from the sun. The size of the absorber plate will control how much heat is collected. The amount of contact area between the plate and coil will determine the volume and speed of heat transfer. Sounds simple so far. So I have to assume that air itself is not a good medium for heat transfer. But what if the air inside the panel can be heated to a very high temp and be maintained? By insulating the box and hopefully achieving a good seal between the glass, how hot do I think I can get the air inside? No idea!!
As far as efficiency is concerned, how important is it? Depends I guess. In this case I think of efficiency as the ability to heat the max volume of water to the highest temperature in the shortest amount of time for the least amount of $$$. But what if you don't have a use for all the hot water you can produce? Initially I'd be happy to be able to warm up my 8'x20' power shack in order to improve battery performance in the winter. In the summer I'd use it to heat my water in the house. How well will these 2 panels work in those applications? Again I don't know (I don't know much do I !!!) but am close to finding out.
While pondering this issue last week I came up with another design that I want to test. This one uses no coils at all. Inside an insulated box (same as the ones I'm using now) there would only be an absorber plate. I'm going to use a piece of metal roofing with the ridges. The absorber plate would be the same size as the one on my current design and elevated 1" off the floor of the panel. At the top of the plate will be a copper header with holes drilled in it every 1/2" or so and the ends will be capped. Liquid will enter the header in the center, exit thru the small holes and run down the absorber plate (like a small waterfall). At the bottom of the plate will be a funnel that will collect the heated water and return it to the holding tank with the help of gravity. Obviously for this to work the panel must be elevated above the tank height. With this design 100% of the absorber plate surface area will be in direct contact with the water. Sounds simple....should be efficient. If I can just finish these 2 panels then maybe I can get to it!!!!
Thanks for all the feedback so far. Will post some pictures later today.
Ric

[Vasile Brindus]

Hi, Ric!


Actually I think of silicone as more of an insulator and in this case feel it might impede heat transfer.


The silicone caulking it's silicone rubber, I think. It's a thermal insulator, so it's not good for your application.


Someone else had suggested thermal grease but whenever I think of grease in a high heat environment I just envision a mess.


Yes, it will be messy, especially if dust gets in. I don't thing that will melt and flow at temperatures below 100C, though. But it's easy to check :)


One other compound I considered was what we used to call "potting" compound. We used it in an electrical connector that was on a wiring harness that went inside an Allision automatic transmission. The harness was exposed to extreme high heat, huge temperature swings and tranny fluid. This compound was a 2 part epoxy and took about 24 hours to cure at room temp. We accelerated the cure by baking the harnesses in an oven at around 150F for about 1-1/2 hours. The compound cured to a very hard, glass like state. Had great adhesion too but don't think it had very good heat transfer properties. Plus the consistancy was very runny until it set up so using it to form a bridge between the coil and absorber plate would likely be messy.


If this is the compound used in automotive applications that I know, it's main purpose is to keep electronics from dismantling/failing because of the vibrations and shocks.


The final "transfer" takes place inside the coil between the hot metal and the cooler liquid. As the metal gives up it's heat to the liquid it must be replaced. The surface area of the coil must be insufficient on it own to do this efficiently? So the absorber plate is used to collect additional heat from the sun. The size of the absorber plate will control how much heat is collected. The amount of contact area between the plate and coil will determine the volume and speed of heat transfer. Sounds simple so far. So I have to assume that air itself is not a good medium for heat transfer. But what if the air inside the panel can be heated to a very high temp and be maintained? By insulating the box and hopefully achieving a good seal between the glass, how hot do I think I can get the air inside? No idea!!



The air inside is hot, but far from the temperature of the black plate (think of your car dashboard when the car was left in full sun in a parking lot, compared to car's interior air temperature). The air is NOT a good thermal conductor, and it's not good in storing the heat, either. Efficiently transferring heat between air and the pipes requires a larger contact surface and the air to be in constant movement (think your car's radiator, or the coils of your air conditioner). While I have no practical experience in building solar panels, simple physics tells me that the solar panel's efficiency is highly dependent of the low thermal resistance in transferring heat. That's why all the manufacturers do something about it. Put it in another way, if you don't have a good heat transfer, you can just get rid of the plate, since the heat collected by it doesn't reach the pipes and the water :)


While pondering this issue last week I came up with another design that I want to test. This one uses no coils at all. Inside an insulated box (same as the ones I'm using now) there would only be an absorber plate. I'm going to use a piece of metal roofing with the ridges. The absorber plate would be the same size as the one on my current design and elevated 1" off the floor of the panel. At the top of the plate will be a copper header with holes drilled in it every 1/2" or so and the ends will be capped. Liquid will enter the header in the center, exit thru the small holes and run down the absorber plate (like a small waterfall). At the bottom of the plate will be a funnel that will collect the heated water and return it to the holding tank with the help of gravity. Obviously for this to work the panel must be elevated above the tank height. With this design 100% of the absorber plate surface area will be in direct contact with the water. Sounds simple....should be efficient.

It will certainly work, but what about the water evaporating and then condensing on the front glazing? It will reduce the sun energy reaching the absorbing plate. It will then drip and you must have a way to let it get out of the panel. Also, a wooden box will not perform well in a humid environment, wet insulation will lose its efficiency, and metals will corrode in time.

Regards,
Vasile

[Ric Murphy]

Todays progress.
Panels have 2 layers of insulation between the inner and outer pans. First layer is 3/4" polyisocyanurate. Second layer is a bubble wrap type insulation. It actually has a higher R value than the polyiso
573
574
Inner and outer pans assembled
575
Heat exchanger for my greenhouse furnace
576

Ric

[Ric Murphy]

I may have found a suitable compound to use between the coil and absorber plate to improve heat transfer. Seems to meet all the critieria for the application....with stands sufficient heat/cold, good adhesion, paintable, cures hard, easily applied, cost effective, locally available thru Fastenal and has some heat transfer qualities. I'm hoping someone more knowledgable than me can comment on the specs.
The compound is Lab Metal from Alvin Products. It's available in a variety of different size containers including a tube for use in caulking guns. The spec sheet (http://www.alvinproducts.com/Products/Content.asp?id=5) states thermal conductivity as "1-2 BTU per sq ft per hour per degree F". I found a chart for Thermal Conductivity of Metals here
http://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html
It shows aluminum as 118 - 144 and pure copper as 204 - 223 Btu/(hr oF ft)
First question - are these all the same unit of measure?
If so then the lab metal seems quite low compared to aluminum and copper.
Any comments or opinions?
Thanks,
Ric

[Vasile Brindus]

First question - are these all the same unit of measure?


I think there's a mistake in their datasheet, it should probably be 1 Btu/(hr oF ft). See http://www.answers.com/topic/thermal-conductivity


If so then the lab metal seems quite low compared to aluminum and copper.


If we consider 1.5 BTU/(ft h oF)=2.58 W/(mK), looking at the table (http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html)we see it conducts heat similar to marble (2.7) and a little better than brick (1.31), which doesn't look good at all.

OTOH, the silver thermal grease used on CPUs is in the same range (http://www.answers.com/topic/thermal-grease) (2-3). So the material has basically the same as the regular thermal grease. Even a thermal potting compond such as Loctite 3860 (http://henkelconsumerinfo.com/products/henkel.datasheets.Search.pdf?BUSAREA=0006&DOCTYPE=MSDS&LANG=EN&COUNTRY=US&MATNR=715477&VKORG=3450) stands only at 1.25.

The key here is keeping the distance between the two metal parts as small as possible and the contact surface as big as possible. It's easy to do between a CPU and its heatsink, but unfortunately both of these conditions are hard to meet when fastening a long round pipe to a metal sheet plate :(

So basically this is almost the best you can get, unless you do fundamental changes to your design (e.g. flat pipe, or some way to mechanically embed the pipe into the plate). Or, as suggested above, use some copper sheets mechanically and thermally attached to the big plate and solder the pipes to those (tin has a conductivity of about 36). Of course, you have to decide/test how much efficiency is enough for your application.

[Ric Murphy]

Thanks Vasile. You confirmed what I thought.
Ric

[Doug Arndt]

I'm wanting to build a flat plate collector using pex tubing, and your heat transfer question has me wondering too. I did find this link http://www.epoxies.com/therm.htm
Couldn't find any pricing info however.

[Rob Beckers]

Very interesting info regarding the thermal epoxies Doug! There is an online form to ask for pricing (and it looks like they do mail orders). If you find out, please let use know.

Here is a link to tests with several solar collectors (http://www.builditsolar.com/Experimental/PEXCollector/SmallPanelTests.htm) that use copper/alu/PEX as their tubing (and are otherwise identical). The conclusion there is that PEX incurs a loss of around 10% vs. aluminum. Not so bad at all. By the way, they use regular silicon caulk to fill voids, to improve heat transfer between plate and tubing.

-RoB-

[Doug Arndt]

Thanks Rob, it was that very link that got me to considering the pex collector. I figured the 15% loss could be made up by just making it bigger. My brother is planning on converting apx 900 sq ft of his existing metal panel pole barn into a shop and his south wall is open and clear, perfect for a solar heating, radiant floor (just gravel now).
With the cost of copper the pex option may be the answer, but we're just at the talk stage. (A good enough reason to get a subscription to Home Power, and join this forum :) ) Been fun lurking and learning.
Anyway, the question of silicon and this thread got me to wondering if that (silicone) is a good solution, or just better than nothing, or if something else would be better .... Then again, if we're only talking 15%, which i can oversize for, maybe no big deal.

[Ric Murphy]

Hi Doug,
Thanks for the epoxy link. I originally considered using PEX tubing before going with copper. There was some reason (that for the life of me I can't recall now!!!) that I decided against it. PEX would have been much cheaper and easier to form than the copper was. Might want to check both the max temp rating and UV resistance on the PEX before you get too far. I think one of those might have been an issue.
Ric

[Doug Arndt]

Hi Ric, Yes, good points. I have NO experience, this will be the first !
Going only by the comments and tests at the web link Rob provided.
One item was using Pex-AL-Pex, believe that gives greater strength and better heat transfer, and higher temp rating I think. In the collector that author built, the collector plate was made up of several AL panels, that were formed to wrap aound the Pex tube, (up and over), then added another panel underneath, this all screwed tight to a plywood backing. Seems like that should take care of the UV issue, as the tubing is enclosed. Stagnant temps could still be an issue however. I'm planning on a winter radiant floor (concrete) heating, and mounting the panels vertical. This should cut down on summer overheating as the roof and wall will place them in a shadow then.
Much to learn.

[Rob Beckers]

Take a look at this design guide for PEX (http://www.toolbase.org/PDF/DesignGuides/pex_designguide.pdf).

The long and short of it is that PEX is not UV resistant at all. So, it either needs to be wrapped completely, or (maybe easier), used on the shaded side of the collector. Either way it would seem a very good idea to wrap the tube with aluminum to use as much surface area for heat transfer as available.

Temperature wise PEX seems a good alternative; the ASTM standard that rules PEX tubing requires it to stand up to 99C (210F) at 150 psi for 48 hours (that's just about boiling water at atmospheric pressure), or indefinitely at 200F and 80 psi. I have not looked up the temperatures for glycol, if your collector gets this hot for long durations it probably means you'll be replacing glycol frequently due to breakdown of the stuff. So, with some way to avoid complete stagnation it looks to be a good alternative.

Now who's going to build one! :cool:

-RoB-

[Doug Arndt]

Thanks Rob, great link, I stored that pdf for future reference.

It's interesting that one of the "benefits" listed (p12) is
"Energy Efficiency - Pex piping minimizes heat transmission through the pipe wall"

Not exactly what one would want for a collector, yet it seems to work good enough.
And it's used for radiant heating.

In addition to the UV issue, another is avoidance of contact with petroleum products.
Is silicone a petroleum based product ?

[John Canivan]

Interesting idea with a spital flow tube, Ric, I've been using 3/8" serpentine copper flow tubes in the past. now I'm doing it without copper, keep those good ideas rolling along

YouTube - Our Big Sun

[John Canivan]

Ric
I don't think you'll have a problem with flow restriction, but I think you could make better use of the copper and do a better job of bonding it to the absorber plate.

YouTube - How to build a Serpentine Collector

[John Canivan]

Photovoltaic applications have become very popular lately.
Just plug it in and watch the meters spin backwards...
Solar heating systems require more planning and require space for collectinf heat and also space for storing heat... All solar applications require a lot of space to be practical, but I believe that solar thermal applications for home oners is under rated. Although large solar thermal applications have a smaller initial cost and a shorter pay back perios than photovoltaic applications. I'm not saying that PV is unimportant. I am saying they are both important.
Energy independence is possible but the planing of a solar home should begin before a house is built. Solar heated factories may also be very practical.
Heat may be used for more than keeping toes warm. It may be used to generate income.
Here is an example:

YouTube- Solar Workshop

[Sarah Asher]

Here's a link on how to build your own solar collectors:

http://www.jc-solarhomes.com/how_to.htm