DIY Solar Hot Water Heat Exchanger
 

I just built a heat exchanger for my DIY solar hot water system.

I am planning on an INSTRUCTABLES on this, but I thought I would tease you all with some photos for now.

Basicially, it's a tube inside a tube to that anti-freeze that is heated by a solar panel can heat up otherwise cold water that then goes into an insulated storage tank (an old electric water heater) and then into my standard natural gas water heater.

This is sort of a little experiment, and I have never done copper soldering before, so, yes, I already know how sloppy it is!

Hot fluid from solar goes in and out the the skinny inner tube, and water to be warmed in the hot water tank thermo-siphons through the thicker outer tube through the T-connections.

See all photos at:
Solar Heat Exchanger construction - a set on Flickr

http://farm3.static.flickr.com/2652/...dcbac203f0.jpg
http://farm4.static.flickr.com/3527/...9cd94b654c.jpg
http://farm3.static.flickr.com/2528/...48c85a3a44.jpg
http://farm3.static.flickr.com/2566/...39dc2326c5.jpg
http://farm3.static.flickr.com/2629/...dde799beb5.jpg
http://farm3.static.flickr.com/2667/...4da7968be2.jpg
http://farm3.static.flickr.com/2428/...0d852bedec.jpg
http://farm3.static.flickr.com/2581/...006fc20e3f.jpg

Good stuff Ben! Gotta love the tusken raider impression, LOL! :D

As I said in my email, I really don't think that is enough surface area. You might need to run several of those lines in parallel.

Also, have you pressure tested it yet? I hear thats really the only way to test your soldered joints.

Good job! I was thinking about trying something like that after I saw the "side arm heat exchanger". It seems like a DIY friendly design. What do you figure the materials cost for one like your's? Were any of the fittings odd or hard to find?

Now you could wrap it in insulation to make it more efficient :thumbup:.

I did some more work on the heat exchanger today.
(Yes, I am planning on insulating it...)

The next big challenge is to get it to connect to the electric water heater through the heater element ports. Those are straight threads with a gasket. Regular pipe threads will not seal properly into it.

At the hardware store, I was able to find a pvc conduit connection that had straight threads and a flat surface where an O-ring could be added. I got a pair of those, along with a chunk of 1" pvc pipe, and two pipe to 1" pipe thread connectors. With all that, I could make a plastic adapter that would go from the water heater to the heat exchanger.

However, I still had to solder on the dielectric unions. These allow me to transition from copper to steel pipe, without getting lots of corrosion at the point the two metals meet. A union is also a point where you can unscrew the unit. It would be very difficult to attach to the water tank otherwise!

Here is the union soldered in place. Note that it uses a gasket inside against the flat surface. That's similar to how the water heater elements connect to the tank.
http://farm4.static.flickr.com/3509/...19fe374a43.jpg


Heres the adapter that I made to fit the heat exchanger to the tank.
http://farm3.static.flickr.com/2583/...8997c00321.jpg

A #34 O-Ring looked about right for it.
http://farm3.static.flickr.com/2628/...d559aaeb08.jpg

The O-ring slides right on and HOPEFULLY squishes in and seals to the tank.
http://farm3.static.flickr.com/2594/...fb553a810e.jpg

You can see that the arm of the heat exchanger has gotten rather long. Oh well, easier to shorten than lengthen.
http://farm3.static.flickr.com/2508/...1668c0b6e5.jpg

Not one but TWO sections of insulation are to go around the heat exchanger. I happened to have a piece to fit the 1" copper tube, and had another chunk that fit the first piece of insulation. I put the one part on, and then the second over the top of it.
http://farm3.static.flickr.com/2672/...96c14607f9.jpg
http://farm4.static.flickr.com/3505/...11a45436b3.jpg
When installed, the entire heat exchanger will be well insulated.

I still need to add some fittings to the end of the 1/2" (inside) pipe so that I can connect them to a pump and solar panel. I also need to attach a thermometer to the input of the heat exchanger and inside the hot water tank, so that I can see how much energy is coming from the solar panel, and how much of it actually gets into the tank.

So far, none of the parts have been overly expensive to hard to find, although some hardware stores have better selection than others. I stopped at a MegaStore today that carried 2" and 3" diameter copper fittings, but those get expensive real fast!

You will also notice that I am not taking photos in my driveway for once! Nope, I traveled over to a friend's farm where they are doing sort of an informal co-op on building wood gasifier technology. The guys there are supportive and have far more soldering experience than I.
I also managed to get a free 5 gallon bucket of apples and pears to run through my solar food dryer. :D

Just so you know, I'm pretty sure building code requires any heat exchanger that has antifreeze in it that connects to potable water to have a double wall with the space open to allow drainage should one part of the heat exchanger leak it will not contaminate the potable water... it's screwy.
one way that some heat exchangers are done is the cold water is in your center pipe then some soft copper tubing is wrapped around the outside and soldered on to keep it in place.

You would ALWAYS want to make sure to use a non-toxic anti-freeze with a single-wall heat exchanger.

I believe that most building codes specifying double-wall go back to when we only had very poisonous anti-freeze. But, building code comment is noted.

Right now, this is all just back yard experimenting.
The single tube in a tube is pretty easy to build and very efficient.

I just test fit the PVC tank adapter into the electric heater element port. It went right in very easily. The O-ring squished right up between the flat lip on the tank and the flat part of the conduit connector.

Also, I must have measured close enough, because I was able to connect BOTH ends of the heat exchanger to the tank at the same time. In the back of my mind, there was always a little bit of fear of the exchanger being just enough too long or too short for both ports to line up!

How is this working? I saved my old electric water heater when I put in the new one. I want to make a solar hot water system, but I don't want to freeze it and break something. It's hard enough to get my wife to sign off on something like this.

Well, it's not working, because it's not hooked up.

Last week, I took the "Intro To Solar Hot Water" workshop through the Midwest Renewable Energy Association.

The instructor told me that external tube-in-a-tube heat exchangers work well in that they are simple and don't clog with sediment the way some other heat exchangers can, but they are pretty in-efficient.

Also, after taking the workshop, I don't think I want any plastic involved ANYWHERE in my solar heating system. I used PVC connections to make a coupler to attach the heat exchanger to the hot water tank, but it was only because it was cheap and available. I couldn't find the right brass/bronze parts anywhere.

I have just been pretty busy lately, and we have hit our cloudy season about a month early around here.

I don't think I will have a "for real" system hooked up until the spring. Until then, I will keep learning all that I can about solar water heating, maybe take another workshop and learn more about solar sighting.

In the workshop I took, there was a photo of a tube-in-tube heat exchanger, similar to the one that I made, only it used 2" copper tube on the outside and FOUR 1/2" tubes running through the inside. That makes a heat exchanger with four times the surface area of the one that I built!

Building a heat exchanger of that style may be an upcoming project.

Could you use Ethyl Alcohol as antifreeze? Not sure what quantity you need, but it'd be fairly non-corrosive and very resistant to freezing. It would be non-toxic (well, depending on your definition - especially bad for a pregnant woman...) and easy to detect by its scent if a leak did occur. I don't drink beer or liquor, but I'll take Ethanol any day over ethylene glycol.

New tank adapters!
 

I now have metal tank adapters!

I was at the Milwaukee Electric Car Club EV Build Day today. A while back I gave my original electric water heater tank elements to two of the guys to see what they could come up with to re-use the heating elements to make an adapter for standard piping to go into the water heater tank.

http://farm3.static.flickr.com/2712/...ca9d046074.jpg

One member took the element, cut it out on the lathe, and tapped out the middle to 3/4ths" female pipe thread. I should be able to screw some standard pipe right into that.

The other member modified the element to sort of a scooped-out cone shape. That way we could weld a 1" pipe directly to it to be able to keep the larger diameter pipe, which may still allow for thermosiphoning from the tank, through the heater exchanger and back. I don't know how well this really would work or not, but I do know that bigger pipes are required for it to work.

It's pretty interesting how you can give two people the same problem and limitations, and they come up with completely different answers.

Either style connector will still use the original flat rubber washer to seal the adapter to the water heater tank, just like the original heater element was.

Materials used for tube-in-tube exchanger
 

I saw in a few of the photos you were using a PVC adapter to connect to the tank. As a plumber I would recommend using a CPVC adapter, the difference being that CPVC is suitable for hot water piping whereas PVC is for cold water use only.

One question I have is would the efficiency be greater is the solar heated liquid circulated through the outside tube instead of the inside tube? Im getting ready to build my own system and thought it would be better that way.

Hello Indyplumber, welcome to ER:)

No, it is more efficient when the warmer liquid is circulating inside the cooler liquid. That way the temperature differential through the outside wall is smaller, wasting less heat. In theory, if the heat exchanger was ideally insulated, then this wouldn't make any difference, but no insulation is perfect.

The PVC adapters were just because I couldn't find anything else that would fit the threads on the tank.

It was an off-the shelf part, at least good enough for testing.

The electric heating elements ports on the side of an electric water heater tank are straight "electric" threads. They are NOT compatible with steel or copper water pipe threads.

Later on, a friend of mine drilled out the old pair of water heater elements and tapped the inside to fit a standard pipe.

I don't know if it would make much of a difference or not if the hot fluid runs on the inside or outside layer of the heat exchanger. I kinda figured that INSIDE would be better, as the heat can ONLY go into the water. If the heated solar fluid is on the OUTSIDE, the heat can go inside to the water, OR to the outside of the heat exchanger. (Obviously, you would have insulation on the outside of the exchanger, but still, some heat would be lost.

This heat exchanger was more an experiment in soldering and design than anything.

The downside to this design is that IT JUST ISN'T THAT MUCH SURFACE AREA! A similar design, but with MULTIPLE tubes running through it would be better. For example, a 2" outside pipe with four pipes running inside it. Of course that would also be a little trickier to build, but also have 4 times the heat transfer!

Ben, what you need is an EGR cooler. The one I installed in my engine has some 30-35 small tubes inside:

With some luck you can find a truck-sized EGR cooler at a junkyard. Mine cost about 50€ ($70-$80) from a crashed Peugeot 307.

Sure looks like a heat exchanger to me!

rather then trying to get multiple tubes inside 1 larger tube why not go with multiple of the original all plumbed in parallel?

That would work too.

How many would be needed?

the math can be done.

What's the incoming temp and desired outgoing temp of the fluid you're heating, what is the incoming temp of the fluid doing the heating, what is the flow rate of both? Somewhere I found a heat exchanger calculator that will work out the necessary sqft of surface area necessary to achieve it and it will be able to tell you the outgoing temp of the heating fluid as well.

I recently was looking for a double wall flat plate that could take 50F to 110F using 140F water on the other side both flowing at 4 gallons/min The recommended model had 12.4 sqft of surface area and weighed 16.8 pounds. Not a small or cheap thing.

in this case it looks like the inside pipe is 1/2" so unless I screwed my math up you need about 92 feet worth of inside pipe to get the same surface area. You can improve things though by increasing the flow rate of the heating fluid. As well my calc was for a double wall with airspace. Single wall should be more efficient I still would be shocked if it's less then 50' though. 60F in 1 pass is not easy to do.

Yeah, math is sort of the boring part of all this, but it IS much easier to figure something out on paper before cutting and soldering copper.

The cutting and soldering part is much more FUN though!!!

Another take on all this is the fewer walls, the longer length, and the more heat the fluid can hold, the more efficient the system is.

A long coil of copper pipe inside a big tub of hot water is excellent for transferring heat!

At my house, I don't have a basement, and the solar panel will be ground-mounted, so it's not appropriate for a drainback system.

For a pressurized system, you really need some kind of heat exchanger. I have finally gotten my hands on a combined heat-exchanger and solar heated water tank. It's in my garage, and the weather is below freezing for a couple of months. Once we thaw out in the spring, I can test it out and start work of hooking up my actual - real-world - solar hot water system!!!

I can't believe that one can do better than this heat exchanger here:

$1000 solar water heating system -- storage tank

Like others, I would be concerned about cross contamination of anti-freeze with the drinking water at the solder points whereas, if one uses intact coil of pex tubing, it has plenty(!) of surface area for heat transfer, it holds a good bit of water in the tubing as a reserve, and I wouldn't have reservations about its integrity of the tubing and cross contamination (although they don't use anti-freeze, the water in the storage tank is not exactly potable having sat in the tank for a year or more).

I would have put a section of finned baseboard on it, thereby getting some free radiant heat on the vertical run. ( I have some lengths that I scrounged somewhere) FYI

I can't believe I missed this thread! Sorry to revive something long dead, but I didn't do it first...

Ben, how did this experiment turn out? Did it work? If so, did it do well? Did you retire the rig? I'm interested if you can't tell.

I know from experience that this type of supplemental heating can work really well. Not just for solar, but for other types of heat sources as well. From a design standpoint, maximum efficiency of the heat exchanger is not paramount. With a low-grade heat source, there are only a certain number of raw btu available to transfer each pass through, so even with a superconducting exchanger, the gains are not enormous like with steam or a flame. As long as the heat doesn't leak out of the loop, it will come back through on the next pass through.

With this "shotgun" heat exchanger, the best way to improve the efficiency is easy. The unit can be built just like Ben made his, then a length of refrigeration-grade tubing (or a quadruple helix or what have you) can be coiled around the straight length of hard pipe, in between the tees. The heat transfer fluid line can be teed off at the supply and return ends, so that the fluid flows both through the center of the water pipe as well as around the outside. This greatly increases the surface area of both loops, and allows for a higher flow rate through the heating loop. Both factors increase hx effectiveness. If you feel the need to maximize contact area on the outside wall, and don't want to solder the tubing to the pipe, toothpaste or zinc oxide ointment are nearly as good at conducting heat.

I also want to do like this something new for solar hot water system. But let me know that how it's working exactly and the materiel what should we collect how to start this? How many problems you face during this period?

"If you feel the need to maximize contact area on the outside wall, and don't want to solder the tubing to the pipe, toothpaste or zinc oxide ointment are nearly as good at conducting heat."

Jeff, with heat do these two materials get runny? I guess I could do the experiment and put some toothpaste in the microwave . . . .

This is a great suggestion as the standard heat conducting gells are expensive.

Thank you for this quick, easy and inexpensive hack.

Steve

The zinc oxide stuff is normally oil-based and it depends on what product you use. It is much more like heat sink grease than toothpaste. Once the joints heat up, surface tension sucks the compound into all the nooks and crannies. I usually use the chapstick stuff on close-fitting joints due to its tenacity. The fish oil and diaper rash stuff works well on joints that move against each other due to dissimilar expansion. Neither of these two really harden completely.

OTOH, toothpaste is a hardening product. It doesn't flow out when it heats up. Over time, the water evaporates out, and it ends up like stale polishing compound, cured drywall mud or hazy car wax. Being water-based, it is much easier to rework or clean up later.If the dried assembly is subject to rattling or dissimilar expansion, the toothpaste will eventually lose contact with at least one surface and crumble off where it can.