|Most collectors that heat water have a supply manifold along the
bottom and a return manifold along the top --these two manifolds are
connected by a bunch of vertical risers with heat absorbing fins
attached to them. Water flows in the supply manifold, up the
risers where the solar heat warms the water flowing in the riser, and
then into the return manifold to go back to the storage tank.
It has been suggested by Alan Rushforth that for some of the very wide collectors that DIYers tend to build for space heating that an alternative in which the "risers" run horizontal (hisers?) and the manifolds run vertically on the right and left might have some advantages. The hiser collectors would use less material because the manifolds are shorter, and it would have far fewer riser to manifold joints -- these joints are expensive and time consuming, so cutting them back is good. The horizontal risers might also result in a more uniform flow distribution for the collector. Uneven flow distribution can be a problem on large collectors, so a more even flow distribution would be a plus.
This is a small test to take this horizontal riser collector idea a bit further.
In an earlier test, I took a cut at measuring the evenness of the flow distribution in my prototype copper pipe aluminum fin collector by removing the glazing, and measuring the temperature rise along each riser using an IR temperature gun. For this test, I turned the same collector on its side so that the risers run horizontally and did a number of temperature and flow measurements to see how the horizontal "riser" arrangement works out -- thanks to Bob Allan for suggesting this idea.
In a nutshell, the hiser collector looks promising to me, and (I think) deserves to be tried on a real, large collector.
I did these tests
Fill up test...
Flow Distribution with Normal Flow
Drain back test...
Fill up test with low flow...
Flow Distribution with Low Flow...
The next few pictures show the test setup.
The collector is set up with the riser nearly horizontal -- they have an about 1/4 inch per foot slope for drain back.
This is a drain back system. The drain back tank is the galvanized stock tank. Water is pumped from the tank with a submersibe pump to the lower left corner of the collector. Water returns to the tank from a return line connected to the upper right corner of the collector. A valve is included in the supply line to adjust the flow rate.
These pictures show the details of the collector absorber grid construction.
Lower left corner of collector.
The black hose is the supply line
from the pump.
Upper left corner of collector.
The elbow should really be a T.
Upper right corner of collector.
The return flow exits out the back
of the collector here and to the
Orange return line (see other pics).
Lower right corner of collector.
The elbow here should really be a T.
Note that the 1) the manifolds are only half inch pipe (same as risers), and 2) that elbows are used instead of T's at the two corners -- these two features are not usual for collectors -- the effects are discussed at the end.
The glazing was removed for the test so that the IR temperature gun could be used to measure the temperature of the risers.
The collector was installed such the the hisers had a slope of about 2 inches total toward the supply end of the collector for drain back.
This was just observing how long it took for the collector to fill when the pump was turned on, and how the fillup progressed.
The flow rate for this test was 1.25 gpm. This is about 0.04 gpm per sqft of collector, which I would call a good, normal flow rate. See below for low flow rate test.
As you might imagine, the collector filled from the bottom up, hiser by hiser. I could see the progression by just measuring the temperature of the risers with the IR temperature gun. So, first the lowest hiser temperature went down as water started flowing into it, then the next one up, and so on until flow got to the top hiser. There was no flow out the return line until the top hiser was getting water. It took about 1 minute for the collector to fill up and start flowing water out the return line.
The drain back test was done by turning off the pump, and disconnecting the supply line from the pump so I could see the drain back flow. This probably results in a little faster drain back since there is no pump resistance, but it makes it easier to see how its progressing.
The drain back was without drama, and took 1 minute 15 seconds. A nice uniform flow.
After all flow stopped, I blew into the return line to force out any remaining water. The result was about half a cup of additional water -- this seems OK to me.
It would be nice to compare it to the same collector in the vertical position, and I'll do this if I get a chance, but the drainback time seems about normal to me.
The table below shows the temperatures and temperature rise for each hiser.
Measuring hiser temperatures.
This is an average of 4 runs done during very clear skies.
|Left side of hiser temp||Hiser temp rise||Right end of hiser temp||
|85 F||7 F||92 F|
So, for example, the 2nd hiser up from the bottom has a left end temperature of 81.5F, a right end temperature of 96.3, and a temperature rise of 14.8F.
The temperature rises are reasonably uniform indicating that all hisers are getting flow.
There is certainly no bottom to top pattern of lower flow at the bottom risers or lower flow at the top risers as one might anticipate could be a problem with the hiser arrangement. The variations seem of the same order as I've seen on other flow distribution tests.
The lower temp rise and higher flow rate in the top riser is probably mostly due to the elbow at its right end rather than a T.
I adjusted the flow rate down to 0.68 gpm, or 0.02 gpm/sqft. This is lower that I think should be used on collectors, but I was trying to see if it would result in problems for the hisers.
The fill up went exactly as it did for normal flow, but just took quite a bit longer. The hisers filled starting with the bottom hiser, then next up, etc. until the top on filled. At that point water started flowing out the return line.
This all took about 2.5 minutes.
Note that is was possible to adjust the flow so low that the return line would not stay full -- in this condition, the flow rate went down to a very low level because the without the return line full, there was no pressure recovery to help the pump produce more flow. So, this is something to look for, but its a problem not related to the hiser configuration.
I did a couple runs of measuring hiser temperatures with the low flow.
The results were very similar to the normal flow with somewhat higher temperature rises as one would expect with the lower flow.
There was no indication that the lower flow resulted in flow distribution problems for the hiser arrangement.
More work could certainly be done on this, but the test appears to support:
- The horizontal riser arrangement results in a reasonably uniform flow distribution between the hisers (about the same as when the collector is used in its normal vertical position).
- The drain back with the hisers sloped at about 0.25 inch per foot appears to work well and result in good drain backs in reasonable times.
- The fillup is pretty much the top to bottom process one would expect with no surprises.
The test collector is a pretty small collector, and it would be nice to do this kind of test on a larger, wider collector.
If I had any reservations about the hiser arrangement it might be that as the hisers get longer, the drain back might get less certain. I suspect that it would be OK, but it would be nice to test how well a (say) 20 ft long hiser sloped at 0.25 inch per foot drains.
Its important to note that this prototype collector has a couple features that are not normal and they may have some effect on the results.
- The manifolds and risers are both half inch copper pipe. Normally the manifolds would be larger diameter (say 3/4 or 1 inch). The small diameter manifolds were used to see if on a relatively small collector like this you could get even flow distribution without the larger manifolds. The gain is that building the absorber is very easy for a DIYer.
I'm inclined to think that the small manifolds make life a little harder for the hiser collector, so if normal (larger) manifolds were used, I think the results would be the same or better.
- I plumbed the first and the last riser to manifold joint with and elbow rather than a T -- this seemed a sensible thing to do at the time, but the earlier flow distribution test shows that the risers with the elbows get a bit more flow because of the reduced flow resistance. So, use T's not elbows for all the riser to manifold joints.
Gary Nov 6, 2010