Friday, July 31, 2009

July Performance for the $1K Solar Water Heating System

The $1000 Solar Water Heating System has been in operation for nearly a year now.  I've been logging the tank, collector, and ambient temperatures and posting the plots of these temperatures at the end of each month.

So, here is the July one:

As you might expect for the middle of summer, the solar fraction was 100%.

This month was a test of the strategy to keep the stagnation temperatures within the limits of the PEX tube used in the collector, and it appeared to do OK with a maximum collector temperature of 195F.

More details on the $1K system...
Performance plots for the last 10 months...



  1. Looks great Gary. I have just started looking for a logger to keep track of my new system. I have a temp HOBO logger that I will use initially, then hopefully I will get a logger set-up a little more complex, similar to yours.
    I just got the tank finished. My system is based on your $1000 DHW system. So far so good. My tank (150 gallon+) has been online for two days now for two days, and I am up to 118 degrees. I have a heat exchanger built from 60+ feet of 1/2" copper tube. I need to measure the incoming and outgoing once I get the tank up to operating temps and see how well I will do. Hopefully I can turn my electric hot water heater off soon!!!

  2. Hi Doug,
    Two days to get 150 gallons up to 118F sounds good. If you started around 60F, that's about 36,000 BTU per day -- not counting any hot water you may have used.
    I think that you may find the 60 ft of copper is not enough, but I'm not sure. Please let us know how it works out. Matt has sent in a description of his system, which also uses a copper heat exchanger. He used 4 (I think) coils of 3/8 inch copper hooked up in parallel so that 1/4 of the flow goes through each. I'll put his pictures up sometime this week. He is going to measure the temperature drop through his heat exchanger, so between your data and his we should get some idea how much copper is needed to keep the temperature loss down.
    I use the Hobo U12 loggers on my system -- I just set them for a 5 minute interval, and they will do a whole month. I use the 4 channel one.
    I've used them for years and like them a lot.
    Also -- congrats on being the very first person to leave a comment on the new blog! I guess I don't have to worry about being overwhelmed with comments :)

  3. I doubt 60' will be enough to completely bring the water up to temperature fully. I plan on adding more sections as I get them built. I will add another 30' this weekend, bringing the total to just shy of 100'. I will keep records and let you know what I get.
    Had a failure on one of my thermistors and had to change it out. The one I replaced it with was not in the same specs, and is not allowing the differential controller to function properly. I have another thermistor (matched) to install, but haven't had time to do it yet. Hopefully tonight when the roof cools a bit. Consequently, I have been running the pump manually, which doesn't work near as well.
    I plan on doing a write-up when I get the system completed. I will let you know when I do!
    Cheers, Doug

  4. Just did a quick and dirty check to see how my heat exchanger is working. Again, I have approximately 64' of 1/2" copper tube in the tank. The tank has approximately 150 gallons of water currently.
    Tank Temperature Start 119.9 degrees F
    Tank Temperature Finish 118.3 degrees F
    Incoming water temperature is 70.1 degrees F
    Outgoing water temperature is 99.5 degrees F
    This test was under full flow of a sink nearby my storage tank, with the hot water running for approximately 2 minutes full on.
    Not too bad for a rough test. I will likely push to double the size of the heat exchanger I currently have. That does reduce the temperature that the hot water tank has to heat up though. I have it set to 120 degrees F.

  5. Hi Doug,
    Good stuff -- thanks.
    You might want to do a test at about 2.5 gpm, since that's a typical shower flow rate.
    Please let us know how it works out as you add more.
    Are you adding to the length of the coil, or adding new coils in parallel?

  6. Which leads to my question, is one long coil better than many short coils in parallel...assuming the same total length?

  7. I will actually be adding to the overall length. I'm not an engineer, but I would imagine the overall surface area to volume ratio involved is what makes the difference, and not the configuration?
    I am actually not adding coils of copper tube, but rather a soldered configuration. I will have to send a picture. Just a bunch of compact bends with street elbos and a regular elbo, about 30" long..... Fits the depth of the tank well, and is compact.
    I will have to measure the rate of my sink.

  8. Here is a picture of my heat exchanger.

  9. Impressive! Looks like lot of work. I figured surface area was the main consideration in a heat exchanger, but every time I think of the price of copper my nose starts to bleed! I wonder about alternate ideas? Would aluminum fins added to copper risers increase the 'surface area' in an exchanger, thereby reducing the cost .... somewhat? Would the wet environment cause too much interaction between the two metals? How about some Twinwall plates with manifolds top and bottom? I imagine polycarbonate has its own issues in a DHW system like BPA? Of course the work to construct something like that might outweigh just buying copper.....there goes my nose! ;-)

  10. If you have the right type of flux and a flat bench to set it up on, it is not too much work. I make "U's" first, then tie them all together on the bench at the end. I make one flat section at a time, then connect them in sections. It works out pretty easy. The first one I did, ended up having leaks, since I had poor flux. After I got the right flux, it went very easy.
    I have toyed with the idea of adding fins of copper. That would eliminate the corrosion issues. I do have copper spacers which add surface area. Maybe my next step will be to try adding copper to the existing setup and see if it improves efficiency. I think I can get copper plate from a local supplier. I'll give that a try and report back.....

  11. Doug:
    I built a similar HX for my 100 gal tank.
    It is 2 parallel runs of 3/4" copper, 50' each.
    These are in a rectangular pattern, 17"x19" about 9 rectangles each.
    I got results similar to Gary's HDPE / PEX test.
    The outlet was nearly equal to the average temp at 1.25 GPM. 2.5 GPM had raised the water from 55 to 97 with a tank temp of 105.
    There is a noticeable amount of pressure loss with than many bends. This is the reason for the 2 parallel runs.
    If I did it again, I would use 1" copper. The cost / sq ft area is about the same and it has alot less cutting / soldering.

  12. Hi Rowland,
    There is also the large coil of PEX pipe concept that I used in my system -- described here:
    See the part on heat exchanger. The idea is the coil itself holds so much water (12 gallons) that the water already in the coil will supply almost all hot water demands. Since the water in the coil is up to full tank temperature, the HX is 100% efficient for this part. After that, the PEX coil has so much surface area, that even though its not very conductive, it does well.
    I will also try to put up the pictures of Chad's heat exchanger and Matt's as well. I guess you can't have too many ideas :)
    Kevin's system uses a hybrid copper/PEX heat exchanger, and this also appears to work well:
    Chad's system is here:
    But, I still need to had the pictures of his heat exchanger.

  13. I have not noticed any drop in water flow in the current setup. Our hot water tank is plumbed (poorly IMHO) with 1/2" flexible copper tube feeding. Thats not to say I wouldn't notice the drop if I double the size. I think that does present a good reason to run the sections in parallel though.
    I am very interested to test the current exchanger with some copper plates soldered on to increase the surface area. Any thoughts on this guys?

  14. Hi Doug,
    The pressure loss in half inch copper at 3 gpm is about 0.078 psi/ft of length. So, 100 ft would add about 8 psi of loss -- a fair bit. I think this is a pretty good argument for parallel coils.
    The 8 psi would go up to 20 psi at 5 gpm.
    For the copper coil in a tank situation, the heat transfer bottleneck is the transfer of heat from the mostly still water in the tank into the pipe wall. The transfer from the pipe to the moving water inside the pipe is more efficient because the water is moving, and you don't get a layer of static water right next to the wall that heats up and then stays there and reduces heat transfer. So, I think that the fins might be pretty effective in that they connect the pipe to more of the tank water. I'd certainly like to hear how it turns out.
    I am wondering a bit how the solder will hold up when its immersed in hot water all the time. I think it will probably be fine, but I did have to replace a galvanized metal strap I had across the top of my tank after a year because the galvanize coating was corroded. This is a worse case because it was exposed to both water vapor and air. But, you might inspect the solder joints on your exchanger after a couple months.
    Another idea for a heat exchanger that Alan uses is to immerse a small tank in the large tank, and plumb the incoming cold water through the small tank. This has the same advantage as the big coil of PEX in that the small tank stores enough preheated water to satisfy a demand or two. This is a picture of his setup:
    His small tanks are pretty large because the system if for an apartment building. For a home the tank could be (say) 10 gallons. I've just not thought of a place to get a good small tank to use.

  15. Yes, Gary, I think PEX is the way to go. But here in Nova Scotia, PEX is almost as dear as copper. The best price I can find is $300 Can for 500 ft of 3/4" PEX ULT (whatever ULT stands for?) but I'll keep checking around. The site you link to from your page offers a nice price but you have to take the dollar conversion into consideration and shipping would be a killer!
    I had forgotten about Alan's tank in tank exchanger. Might be able to pick up a used 10 Gal water heater for next to nothing. With a few plugs and maybe a coating of epoxy, it should do the trick. Could even hybrid in a short coil of copper to assist with some 'on-demand' heating when the inner tank is depleted.
    So many ideas! I'll have to pick one and run with it when I'm ready to build!

  16. hi Rowland,
    Yes -- I've heard the same thing from other Canadians -- pipe prices are higher.
    Be sure that for whatever you get it has the approval for potable water. Down here that means its stamped "NSF" -- not sure what the Canada equivalent is. Ult sounds like it might be what they call utility down here, and its not approved for potable water.
    You might check with Kevin, Chad, and Cory as they all live in CA and built systems -- they are listed in this section:


Note: Only a member of this blog may post a comment.

/* Start Analytics ---------------- */ /* End Analytics ---------------- */