Comparing the Performance of Two DIY Solar Water Heating Collectors -- CPVC vs Copper

Sometime back I had a brief go at a solar water heating collector made from CPVC with aluminum fins.  At the time it seemed interesting, but not as good as some of the other options for building collectors.

Scott Davis has taken a whole new look at the CPVC collector and come up with a design that places the risers closer together, and uses a flat aluminum sheet for the absorber plate.  This makes for a design that is inexpensive and easy to build.  Scott's initial testing of the collector indicated good performance.

  CPVC and copper solar collectors test
Side by side test of the two collectors.

I have to admit that I was a bit skeptical of the performance of the new design due to the poor thermal connection between the risers and the aluminum absorber sheet, but I've had a go at testing a collector with copper risers and six inch wide aluminum fins against Scott's CPVC design, and the CPVC does quite well -- it is only about 5% short of the performance of the copper/aluminum collector.

So, this page covers the performance test, a stagnation test, some IR pictures to try and see a bit more about what's going on, and a final comparison of the two collector designs on things like cost, build, durability, ... 

Comments and suggestions are most welcome.

Page Directory

The Test Collector

Test Setup


Analysis and Potential Improvements

Stagnation Temperature Test

Which Collector to Build?


More on Scott's CPVC collector:

Scott's YouTube video on his CPVC collector...

Scott is the founder of the Yahoo Simply Solar group, and this CPVC collector has been discussed quite a bit on this group...

Scott's CPVC Collector
Scott and his CPVC collector from his video...

The Test Collector

I built the test collector as close as I could to Scott's collector as described in his video.  About the only thing I changed was to tighten up the riser tube spacing to (hopefully) get a bit better performance.

The test collector absorber dimensions are 2 ft wide by 3 ft high.   This is the size I picked when building the Sun Simulator as the best compromise between not dimming the lights in the whole neighborhood when it comes on and having it larger enough to get good results.  For a real collector, you would, of course, want it to be much larger.

Here are a few construction pictures:

cutting collector tubes
Cutting the risers.
The shears they sell for cutting plastic pipe
work just as well.

Starting to put together the headers.
The short straight pieces connect the T's
to make each header.
gluing header CPVC collector
Solvent gluing the headers together.
assembling manifold cpvc collector
Gluing the risers into the top manifold.
finished top manifold cpvc collector
Finished top manifold with risers glued in.
bottom manifold cpvc collector
Starting the bottom manifold.
Glue the T's to risers first, then add the
short pieces of straight pipe between the T's.
finished cpvc absorber
Finished CPVC tube assembly.
cpvc collector alum absober fin
Cutting out the aluminum sheet for the
collector continuous fin.
This is 0.01 thick alum flashing material.
alum flashing in place on cpvc collector
First piece of flashing in place.
applying silicone to riser tubes
Applying the silicone to the riser tubes.
The silicone provides a thermal and
mechanical bond between the CPVC
and the aluminum absorber plate.
weighting alum until silicone sets
Weighting the alum to keep it in close contact
with the riser tubes until the silicone sets.
Finished cpvc test collector
The completed CPVC absorber in
its collector box.

In the pictures above, the CPVC is just as easily cut with the inexpensive shears they sell for that purpose.   Or, any kind of wood cutting saw you may already have will work.  The purple primer I used on the joints is not really necessary if you use the "Yellow" CPVC cement -- I just use the primer out of force of habit.  

I applied a fairly generous bead of silicone along the top of each riser before putting the alum flashing on.  I then flipped it over and added silicone as needed, and then went down both sides of each riser with my finger to form a fillet of caulk between the riser and aluminum flashing.  Then flipped it over, and put a flat piece of plywood over the alum with weights until the silicone cured.  Idea was to get as good a conductive path between fin and risers as possible.

The box that the absorber goes in is made from 3/4 plywood sides.  The back is 1 inch polyiso that is foamed in place with Great Stuff foam in a can.  The glazing for this test is a single sheet of about 3/32nd inch thick Acrylic sheet.   The center to center spacing of the risers is 2 3/8ths inches. 

Both the glazing and the back are setup to be easily removed so that IR pictures can be taken without the glazing (which is opaque to IR).

The collector goes together very easily.

The other collector is made with half inch copper riser pipes and aluminum fins.  The fins were pre-grooved to fin snuggly around the riser tubes.  Other than size, the construction is just like this collector...

Test Setup

I had planned to test the CPVC collector on the Sun Simulator, but I'm still having some trouble getting up to the full sun area light levels, and rather than waiting to get that sorted out, I decided to do a side by side outdoor test.   The side by side test method tests two identical size collectors outside in the sun at the same time.  Testing the two collectors side by side, its easier to measure which one is doing better without having to worry about variations in sun, temperature, wind, ...   This is similar to the solar air heating collector testing that Scott and I did last year, and these earlier side by side tests.


cpvc solar collector performance test
The two collectors each with its own reservoir.

Each of the two identical, insulated reservoirs was charged with 26.9 lbs of water at about 70F.  A collector's performance is proportional to the temperature rise it achieves in its reservoir.

The two collectors were pointed in the same direction at the same tilt -- the direction was set to provide nearly direct incidence.  An Apogee pyranometer was mounted on the collector glazing frame such that it measured the solar intensity in the plane of the glazing.

The water from the reservoir was pumped through the collector using  TopsFlo pumps.  The water from the pump entered the collector on the bottom right corner, and returned from the collector to the reservoir by a tube from the upper left corner of the collector.

Each collector is 2 ft wide by 3 ft high.  I attempted to make the area of the two absorbers exactly the same, but somehow the CPVC absorber ended up about a half inch wider, which should make very little difference.

Each collector is glazed with a single layer of 3/32 inch Acrylic plastic.  The glazing is mounted to a wood glazing frame, and this frame is set up for easy removal to take IR pictures without the glazing.

The temperature of the reservoir water was logged using Onset Computer loggers and thermal sensors.

The CPVC collector.
The clamps hold the glazing on and
allow for quick removal.

The copper/alum collector. 
Return tube from upper left corner with throttling valve.
Apogee pyranometer on the right side.
Topsflo pump
The supply line from reservoir with
the Topsflo pump.

The flow rate was adjusted so that is was approximately the same on both collectors.  The flow rate used was somewhat higher per sqft that is normally used on solar water heating collectors, but not a lot.  The higher flow rate should make both collectors slightly more efficient.

I would have preferred a cooler day, but that does not seem to be in the cards any time soon, so I went with the relatively warm day, but took the reservoir temperatures all the way up to 160F, so there was a good differential between ambient and collector absorber temperatures.

There are all sorts of things that happen in the course of these tests that are not accounted for: the reservoirs lose some heat through the insulation and pick a little heat from the sun, a little of the pump power gets into the water as heat, the wind goes up and down a bit, the sun varies a bit, and on and on.  But, the nice thing about these side by side tests is that all of these things happen to both collectors, so they don't effect the performance comparison between the two.


The plot just below shows the temperatures of the reservoirs attached to each collector.  The greater the temperature gain of the reservoir, the more heat the collector is producing during the test.


performance cpvc collector

Purple dash line                 CPVC Collector reservoir temperature (F)             

Green solid line                    Copper riser - alum fin collector reservoir temperature (F)

The jog in the two curves at about 12:50 pm was a stop to take of the glazing of each collector and get IR pictures of the absorber without glazing.  I then put the glazing back on and let the test go a while longer.  The two collectors were reoriented to face closer to the sun at the same time, which explains the greater rate of temperature increase in the reservoirs for both collectors from 1:00 pm to about 1:45 pm.  It looks like the reservoir temperatures would have made it up to about 170F had I continued the test longer.

The plot just below shows the solar radiation values with the pyranometer set to read the radiation on the plane of the glazing.  The other curve is the ambient temperature during the test as measured just behind the collector in the shade of the collector.  I think that the ambient temperature sensor location is not ideal, and the temperatures it registers are on the high side.  For example, at 12:12pm, the ambient temperature logger reads 86F, while the temperature in full shade on the north side of the barn 50 ft away reads 74F.  As you can see from the sun curve, it was a very clear day with good sun.  The wind was generally light.


Purple solid line                 Sun  (watts/sm)

Green line                            Ambient temperature (F)

This table shows start and end temperatures of the reservoirs and relative performance over the 4 hours of the test.

Collector Start Temp (F) End Temp (F) Temp Rise (F) Performance Relative to Base
Copper/Alum 70.9 156.3 85.4 base
CPVC/Alum 71.9 153.5 81.6 0.955

Again, I am surprised by how well the CPVC/alum collector performs relative to the copper/alum collector.  I thought that the line contact between the CPVC tube and the aluminum fin would take more of a toll, but apparently the closer spacing of the risers allows the CPVC collector to come within 5% of the copper/alum collector.   


I did notice a bit of film on the inside of the CPVC collector glazing.  At first I thought I had a small leak and that it was condensation, but when I took the glazing off, its more of an oily film.  This may just be associated with some outgassing of the new components.  It was easy to wipe off, and I'll see if it comes back again on future tests.

You can see the film in this picture to the left, and a place where I wiped across
it with my finger.

The CPVC collector tends to bow out a bit toward the collector glazing when the sun it on it.  I think this is due to the different expansion rates of the CPVC and the aluminum, but it may also be due to the fact that the absorber fits snuggly in the box and may be be restrained by the box.  In any case,  I would not restrict this bowing, as it probably helps to reduce the amount of thermally induced stress across the glue line.   It might be a good idea for tall collectors to cut the aluminum so that it is broken up into (say) 3 runs of 3 ft rather than one run of 9 ft.   It would also be good to hang the collector in the box from the top manifold, and then just support the bottom in such a way that its free to move vertically.  This is the same way I mounted my large copper/alum collector, and it appears to work fine and reduces thermal strain on the absorber plate.

Analysis and Potential Improvements

Glazing IR pictures

The IR pictures just below are taken around 10:45 am when the reservoir temperature is about 132F.  The copper/alum glazing averages about 6F cooler than the CPVC collector glazing, indicating that it is losing less heat through the glazing, which is likely caused by the average temperature of the cpr/al absorber running cooler than the CPVC absorber.

The Acrylic glazing is nearly opaque to the IR heat radiation that the camera senses, so the pictures show the temperature of the glazing, not the absorber.  The pictures are adjusted to show the same temperature range (70F to 130F) -- so, visual comparisons of the colors are valid.  The little tags (eg "Sp1 106.6") show the temperature in degrees F at that point.

cpr/al collector IR pic glazing
The copper/alum collector glazing temperatures.
CPVC/al collector
The CPVC/alum collector glazing temperatures.


Poly Glazing IR Pictures

cpr/al collector -- poly glazing
cpr/al collector through poly glazing
CPVC collector - poly glazing
CPVC collector through poly glazing.

The two pictures above were taken around 1:55 pm with reservoir temps around 160F.   The pictures were taken with polyethylene film glazing which is fairly transparent to the IR heat radiation the camera sees, so, to some degree, you see through the poly glazing and onto the absorber plate.  The fact that its showing temperatures that are lower than the reservoir temperature indicates (I think) the poly glazing is not perfectly transparent to the IR.

Looking at the cpr/al collector, it shows a temperature of 139F right on a riser tube, and 144F about half way out on the fin.  This is what I would expect -- in that the fin needs to have some temperature differential from the outer edge of the fin toward the riser to transfer heat to the riser.   I'd say that the the fact that its hard to even see where the risers are is a good sign that the thermal efficiency of the fin to riser joint is good.

The equivalent picture of the CPVC collector shows a temperature of 150F on the riser tube, and 143F on the fin about half way between the risers.  This is puzzling in that the temperature differential is in the opposite direction of what you would expect if the fin is transferring heat into the riser.  Not sure what is causing this -- it could possibly be a difference in the emissivities of the riser and fin, but they are both painted with the same flat black paint, so that seems unlikely.  Perhaps its an indication that the fins are not really doing much to transfer heat to the risers -- it makes we want to try a version in which the risers are as closely spaced as possible, and not fin at all is used.  Any ideas on this?

Overall, the average temperature of the CPVC absorber appears to be a few degrees hotter than the cpr/al absorber, as expected.


Fin IR Pictures -- No Glazing

These pictures were taken around 10:35 am with a reservoir temperature of about 132F.  The glazing was removed, and the IR pictures were quickly taken to try and get the temperature distribution across the fin and riser as it was with the glazing on.

cpr/al collector -- no glazing
cpr/al collector -- no glazing.   Upper left corner.
CPVC/al collector -- no glazing
CPVC/al collector -- no glazing.  Upper left corner.

On the copper/al collector picture, it looks like in some places the fins are doing a good job of getting the heat to the riser pipes without incurring a large temperature rise, but there are a few areas where the out edge of the fin is running hotter than one would like.  On the right hand riser its also pretty easy to pick out where the 3 inch wide backing strip of alum ends -- the temperature differential is definitely a bit better in the area where the backing strip is.

On the CPVC picture, there are some areas where the fins are running warmer than the risers as you would expect, but also a few mysterious cool areas on fins where they are running cooler than the adjacent riser -- puzzling.   In general the temperatures of both the risers and the fins are higher than the corresponding temps on the cpr/al collector -- probably indicating the not as good thermal conductivity of the CPVC.

Anyone see anything else in these pictures?


Potential Design Improvements

One variation that Scott has suggested and I would like to try is to reduce the spacing between the CPVC risers to as low as the Tee's will allow (about 2 inches).  I'd like to try this with and without the alum fin sheeting to try and see how much the alum is contributing (if anything).

Another possibility would be to keep the relatively closely spaced CPVC risers and work out a way to improve the thermal bond between the risers and the aluminum sheet -- perhaps a set of grooves in the aluminum that the CPVC risers could fit in?  Even shallow grooves might work pretty well given the close spacing of the risers?

Another possibility that would keep the easy build would be to use pre-grooved fins (like Tom's fins) and increase the riser spacing. 

Any other ideas? 

Stagnation Temperature Quick Test

Stagnation is the situation in which the collector is exposed to full sun with no flow of water through the collector to remove the heat.  The collector continues to heat up until it gets hot enough so that it can lose all the incoming heat via convection and radiation out through the glazing -- this can result in quite high collector temperatures.

Since everything was all set up, I decided to do a quick stagnation temperature test.

From the sun plot above, sun on the collector was close to 1000 watts/sm (full sun).   Ambient temperature was about 85F.

I removed the glazing and installed one of the logger sensors right next to one of the CPVC tubes.   It was just sitting loose next to the CPVC riser tube, and might have registered higher temps if it was held in good thermal contact to the tube. 

I turned off the flow and  re-installed the glazing, and watched the sensor temperature go up.  It  only took about 15 minutes to climb from about 140 F up to 232F (See plot below).  At that point I removed the glazing as I did not want to damage the CPVC. 

I think that this clearly shows that even for a single glazed collector, stagnation temperatures on collectors mounted at normal tilt angles are too high for CPVC.   The 232F where I cut the test off is already to high to expose the CPVC to and expect a good life.  Clearly the temperature would have kept going up and would likely have exceeded 250F -- too high.

The message here is that CPVC MUST be protected from high stagnation temperatures.  The steps that might work to do this that I know of are:

- Mount the collector at a high tilt angle -- I would say greater than 70 degrees.

- Do not use double glazing as this will increase stagnation temperatures.

- Some form of ventilation of the collector might work, but I would be sure to test it to be sure.

This is an area that could use some more work and testing.

Collectors mounted on vertical walls or at steep tilt angle will not experience high stagnation temperatures, and CPVC or PEX will work fine and provide  a good long life. 

Vertically  mounted collectors on walls work out well for space heating because the low winter sun comes in nearly perpendicular to the glazing on a vertical collector.  For solar water heating collectors, a somewhat oversized collector at a steep tilt angle also works well.  It does not overheat in the summer because of the high tilt angle, and the high tilt angle optimizes its performance during the winter when the sun is more scarse.  The end result is a high year round solar fraction.  Here is one example of a steeply tilted PEX collector with a 94% solar fraction...   There are a number of other examples of successful steeply tilted or vertical collectors on the solar space heating and solar water heating pages.

You could think about a system in which the collector is never allowed to stagnate, but I don't think its a good idea in that over time pumps fail, controllers fail, power fails, someone makes a change that causes a problem -- sooner or later your collector is likely to face stagnation. 

solar collector stagnation plot

Deciding Which Collector to Build

So, which is the best bet to build?  It probably depends on what your aims, skills, and budget are, but, I'll go out on a limb and offer the following comparisons in a few categories: 


Performance is a small plus for the copper/alum collector -- about 5%.  This is a relatively small gain.

Ease of Build

The CPVC collector is easier to build than the copper/aluminum collector.  I'd say it makes a good first project even if you have little to no DIY experience.

The copper/alum collector is also easy to build.  If you use fins with preformed grooves (like Tom's fins), then the only significant difference in the two builds is that the joints are glued on the CPVC and soldered on the copper/aluminum collector.  I know that soldering scares some people, but its really a very easy to learn skill, and you should not be put off by it.  With a little soldering practice on some scrap, I'd say that the copper/alum collector could also be a first project for someone with little DIY experience.


The table below is a cut at the cost of materials to do a 4 by 8 ft collector for CPVC and Copper/alum.

The cost difference is less than I thought it would be -- copper/alum  is about 10% more than the CPVC.  
Considering the performance difference, the cost per BTU would be about 5% more for the copper collector.
Did I miss anything?

  CPVC/Aluminum Collector         Copper/Aluminum Collector      
  Item Qty Unit Price Cost   Item Qty Unit Price Cost
Insulation 4 by 8, 1 inch polyiso 1 $20.00 $20.00   4 by 8, 1 inch polyiso 1 $20.00 $20.00
Risers 1/2 inch CPVC 20 $3.91 $78.20   1/2 inch copper type M 8 $11.22 $89.76
Manifolds  1/2 inch CPVC Tee 40 $0.20 $8.00   1/2 inch copper Tee 16 $0.65 $10.40
Glazing 2 by 8 ft sheet 2 $21.00 $42.00   2 by 8 ft sheet 2 $21.00 $42.00
Glacing closeout Foam closures 1 $7.62 $7.62   Foam closures 1 $7.62 $7.62
Alum fins alum flashing 0.012 32 $0.86 $27.52   Preformed fins 0.018  32 $1.50 $48.00
Box back 4 by 8 plywood 1 20 $20.00   4 by 8 plywood 1 $20.00 $20.00
Box frame 2 by lumber 24 0.5 $12.00   2 by lumber 24 $0.50 $12.00
Silicone caulk caulk gun tubes 3 4 $12.00   caulk gun tubes 1 $4.00 $4.00
Paint and misc paint, screws, 1 20 $20.00   paint, screws, 1 $20.00 $20.00
Grand Total       $247.34         $273.78

Larger collectors will cost less per sqft.
Prices will vary depending on where you live and ups and downs with time.
Collectors against a wall can probably drop the box back plywood.
Prices from Home Depot and

Keep in mind that the collector is normally only about a quarter of the cost of the full system.

I picked the 4 by 8 collector because a lot of people build in this size, but, in general, I think bigger is better if you have the room -- the price and labor per sqft go down as the collector gets larger, and you get proportionally more heat from the collector.

Stagnation Temperature Related Limitations

Stagnation temperatures must be considered when using a CPVC collector. The CPVC collector will work well when mounted vertically or at steep tilt angles (70+ degrees), but should not be used at more shallow tilt angles.   I would not use double glazing on CPVC (or PEX) collectors because the double glazing increases the stagnation temperatures.  It is a good idea to monitor collector temperatures the first season and make sure that your design is working as you planned. 

Another option might be to figure out a highly reliable (preferably passive) way to vent the collector during stagnation events.

The Cpr/Alum collector can be used at any tilt angle and with single or double glazing.  That said, its a good idea not to subject any solar collector at low to moderate angles to long periods of stagnation -- even commercial collector manufacturers recommend covering or otherwise protecting collectors that will be stagnated for long periods.

Durability and Life

I think that a well built copper/alum collector can be expected to last around 30 years.  This may require a glazing replacement at some point and a little painting maintenance on the box, but its a pretty durable collector with a good track record.

I think that the absorber of the CPVC collector is likely to have a shorter life, but its really hard to say at this point what it might be.  Maybe there is someone that has experience with CPVC in this kind of environment that could comment on this?

Other CPVC Collector Tests

SolarDan reported some testing of various CPVC collector designs on the SimplySolar forum... These tests were done pretty carefully with two identical size collectors each heating its own reservoir.

Dual Layer vs Single Layer CPVC:

Two of the test collector -- single layer on left and dual layer on right.

The dual layer of CPVC design of Dan's.

Dan did a unique double layer of CPVC shown in the right photo above and compared it to a single layer design with and without an aluminum fin plate.

When neither collecotor had the aluminum sheet bonded to the CPVC, the double layer outperformed the single layer by about 10% on heat output.

When aluminum flashing was added to the single layer collector (but not to the double layer), the results were close to a tie with a slight advantage to the single layer with flashing.

This pdf summarizes several of Dan's tests of single vs double layer with and without the aluminum sheet across the back...

Toward the end Dan experienced a loss of flow for unknown reasons and the resulting stagnation temperatures damaged the panel beyond further use -- so, be very careful of stagnation -- keep a high tilt angle.



What is your opinion?  Any other factors that should be considered?  (comments below)

If you build a CPVC collector, please let us know how it goes, and take lots of pictures.

Also, bear in mind that there are other choices besides these two -- look through this page to see more options...

July 27, 2012


blog comments powered by Disqus