Update added April 23...
This is a test of a thermosyphon solar water heating collector. This collector uses the design in which horizontal manifolds at the top and bottom of the collector are connected by multiple vertical riser tubes. Each riser tube has an aluminum heat transfer fin painted black to absorb the solar radiation and transfer the heat into the tubes. The collector used for this test is the first prototype I made of the copper tube, aluminum fin collector made back in 2008 -- its described in full here...
For thermosyphon systems, the storage tank that is heated by the collector is placed above the collector. The collector inlet (on the bottom of the collector) is connected to the bottom of the storage tank. The collector outlet (on top of the collector) is feed into the tank hear the top.
As sun heats the water in the collector, it becomes less dense and rises out the top of the collector and into the tank. This water is replaced by cold water from the bottom of the tank flowing into the bottom of the collector. As long as the sun continues to heat the water in the collector, a circulation loop is setup in which water flows continuously out the bottom of the tank, through the collector, and then back into the top of the tank.
When the sun goes off the collector, the water in the collector cools, and circulation no longer occurs since the water in the collector is more dense than the tank water and naturally stays in the collector.
So, in this system, the circulation is all by natural forces -- no pumps needed. And, the control is automatic -- no controllers needed.
These systems are very simple and are widely used in some parts of the world.
I did the test on this copper/aluminum collector to see how it compared to a similar test done on a collector that uses PEX tubing and aluminum heat transfer fins -- described here...
All in all, I would say that this thermosyphon water heating system performs well. While the system shown here is just a very crude prototype, I believe that an actual system could be built that would: 1) cost about $300, 2) perform well, 3) last a long time, and 4) look nice.
I am puzzled by some aspects of the performance (see below) -- maybe there is a thermosyphon knowledgeable person out there who can explain this?
Compared to the popular batch solar water heating systems, the thermosyphon offers a system that is nearly as simple and cheap to build, but does not have the larger temperature drop over night that batch systems experience.
More on thermosyphon systems here...
This is a very crude thermosyphon solar water heating system. It was built just to do this one test, so the construction was for speed -- not for life or looks -- I know its ugly, but it can be made to look just fine for real applications.
The collector used is described in detail here... This is the original prototype of the copper/aluminum collector built in 2008. It has been sitting in the yard, pointed south and stagnated since the end of 2008. The wind has dumped it over on its glazing twice -- careless of me to let this happen. But, the collector survived all this very well -- its in great shape inside and outside.
The collector is glazed with a single layer of SunTuf corrugated polycarbonate glazing. I pulled a corner of this back to install the heat sensor on the absorber, but otherwise the collector is unchanged from the original prototype.
The collector area is about 32 sqft.
The storage tank is a galvanized stock tank -- 45 by 21 by 12 inches tall with rounded ends. The tank is vented to the atmosphere -- ie not pressurized. For this test, the tank holds 40 gallons.
The tank is insulated all the way around with 2 inch thick (R13) polyiso insulation.
The bottom of the tank is about 6 inches above the top of the collector.
The outlet from the tank to the collector inlet makes use of the tank drain fitting, and is near the bottom of the tank on the east side.
The inlet to the tank from the top of the collector uses a bulkhead fitting to enter the tank just below the water line near the top of the tank.
I was not able to completely avoid bubble traps in the collector plumbing because on this collector the inlet and outlet come out the back of the collector.
It may be that the these small bubble traps at the inlet and outlet cause the mysterious "no flow" episodes discussed below?
I filled the tank with a garden hose by running water into the tank outlet and letting the
fill water run through the collector and come into the tank via the tank inlet.
Idea was to get all the air out of the collector, and get the thermosyphon circuit full of water.
Top black dashed line -- Collector temperature (F)
Red dashed line -- Collector outlet temperature (F)
Purple line -- Tank temperature at mid tank (F) -- not that tank was highly stratified, so this is NOT an average tank temperature
Blue line -- Collector inlet temperature and bottom of tank temperature (F)
Sickly green line -- Solar intensity (watts/sqm)
A mostly sunny day with some very thin high clouds, and an occasional lower altitude cloud drifting in.
Light winds most of the day going to strong winds in the late afternoon.
Ambient high temperature for the day was 70F
So, the plot above shows the log of temperatures and sun for the full day. The collector was allowed to just run for the full day. I did not shutoff the flow to test for stagnation temperatures as was done on the PEX collector.
Basically, the collector does a nice job of heating up the water in the tank over the course of the day. The tank temperature became highly stratified as the day progressed. The water on the top of the tank being 100F or more, and the water at the bottom of the tank remaining around 60F through most of the day. You could put your hand slowly down into the tank, and actually feel at some depth a very rapid transition from hot to cold. This transition zone moved downward in the tank as the day progressed.
The tank insulating cover was off for half hour periods a couple times, so the actual performance would have been somewhat better.
The puzzling thing (to me) is the periodic episodes in which the collector outlet temperature drops suddenly and the collector temperature increases at the same time. This occurs about 10 times. It appears to me that for some reason flow stops during these periods. This causes the collector outlet temp (as measured at the tank inlet) to drop (since there is no hot water flow), and the collector temperature to rise (since there is no water flow removing heat). I would guess that these episodes are caused by an accumulation of bubbles at some point in the system? Then (I guess) when pressure builds enough the bubbles get pushed out, and normal flow resumes. All this does not appear to significantly effect the performance of the collector, but it would be nice to understand and eliminate it -- do you know the answer? -- please let me know...
The temperature rise through the collector is about 80F, so the thermosyphon flow is not strong enough to keep this as low as one might like. The hotter collector that results from the large temperature rise no doubt causes increased heat losses and somewhat lower efficiency, but it still does the job, and the system is so simple that it seems that the gain in simplicity is worth the drop in efficiency. Maybe systems designed from the ground up for thermosyphoning can do better than this?
The Collector Efficiency Calculator says the drop in efficiency due to this 80F temperature rise compared to a 10F rise is about 10%.
The high temperature rise does allow the top of the storage tank to heat up to much higher temperatures more quickly, and this highly stratified tank can be helpful in recovering part of the hot water supply to full showering temperatures more quickly.
The difference between collector temperature and collector water outlet temperature is fairly small indicating that the copper/aluminum construction is doing a better job than the PEX did in transferring sun heat into the tubes.
I redid the plumbing at the collector outlet to eliminate any possibility of a a bubble trap. The new arrangement is shown just below, and the old arrangement is shown in the pictures above "Collector outlet and tank inlet".
New plumbing at
collector outlet to eliminate the possibility
of a bubble trap.
The idea was to see if this stopped the incidents of no flow that show up in the performance plot above for the 22nd. See the plot below for the results.
It appears to me that this did the trick. I don't see any of the episodes where the collector temperature rises and the tank inlet water temperature falls that can be seen on the plot above, and indicate to me that no flow is occurring. On today's plot there are a couple drops in tank inlet temperature, but they are not accompanied by rises in collector temperature, and, for the most part, they seem to correlate with drops in sun light intensity from clouds passing.
So, I'm about ready to believe that you just have to be very very careful about the plumbing for these thermosyphon collectors. No dips at all that could catch bubbles.
Note that on both days, I flushed the collector completely of air before starting by basically filling the tank through the collector with a garden hose to flush out any air.
So, it does not matter how good a job of getting the air out you do before you start, the collector makes its own bubbles -- presumably from water vapor.
Interesting also how you can see the temperature stratification band reaching the mid tank temperature sensor at about 2:20pm and causing a rapid rise in temperature at that depth as the band passes through that level.
The temperature rise through the collector is still a pretty strong 55F.
But, all in all, I think this configuration performs pretty well, and given its extreme simplicity makes a pretty attractive solar water heater if your climate (no freezes) and plumbing system are compatible with it.
I've received a few suggestions that I will probably try to incorporate over the next week or so. If you see any plans for a well proven thermosyphon water heating collector, please point me toward them.
The Weather Gods signaling end of test.
Gary April 22, 2010