Testing the Benefit of a Reflector on a Vertical Solar Space and Water Heating Collector

As previously reported, the heat output for my $2K solar space and water heating system during the spring has been a bit disappointing.  Spring is a time I would like it to be producing heat for both space heating and water heating, the collector is basically just keeping up with water heating needs. 

This section goes over one potential solution to increase energy production in the spring.  The solution is to add a reflector below the collector and extending out in front of the collector.  The reflector is angled such that it takes the rays from the very high in the sky sun and reflects them onto the collector.

This page goes over a test of a very rough prototype of this kind of reflector to see how effective it is in increasing energy input to the system.

The bottom line result is that the reflector is very effective in increasing the collector heat output.

Table of Contents

- Test Setup...

- Conducting the test...

- Results...

- Data Plots...

- Comparable day without reflector...

- Conclusions...

- Comparison to other reflector tests...

Test Setup

This is a test to verify that a reflector placed below and in front of the vertical collector for my $2K solar space and water heating system will actually provide a worthwhile gain in performance.  This is just a rough test setup and its ugly, but the idea is just to confirm that it works before spending the time to do a nicer version.  It is also larger than I would like to end up using, but its easier to get an idea of the benefit with a larger reflector.   Given that it does appear to be significantly improving performance, I will be going  ahead and doing a smaller "production" version.

test reflector for $1K system

The test reflector is made with a tacked together 2 by 4 frame.  The backing for the reflector is two sheets of  twinwall that were used for packing material on my last order (lots of holes in them).  The reflector material is some aluminized Mylar I had leftover from the Heliostat project.

The size of the reflector is 58 inches by 126 inches wide for 50.8 sf.

The up angle on the reflector is about 15 degrees as determined from this test on the simulator.   Since we are approaching the summer solstice the sun is near its highest elevation of the year.  This reflector angle results in an incidence angle for the reflected rays of about 45 degrees on the collector.  This seems about right -- if the tilt angle is increased, the effective area of the reflector is reduced, and all of the reflected rays are concentrated on a relatively small area at the bottom of the collector.  If the tilt angle is flattened out, the reflected rays cover more of the collector area, but the incidence angle of the reflected rays on the collector gets so shallow that I fear at lot of the reflected energy would just be reflected off the collector glazing. 

Conducting the Test

This reflector geometry produces a reflected light pattern on the collector around midday that looks like this:

reflected light pattern

You can see the top of the reflected rays pattern about half way up the collector.  The blotchiness of the patter is due to the aluminized Mylar not lying quite flat on the twinwall supports and on the twinwall having some wows in it.

Note that its not quite solar noon on the collector in that there is an area to the right side of the collector that is not yet getting reflected light, and there is an area to left side where the reflected light is hitting to the left of the collector on the wall.  This is why there is some benefit in having the reflector be wider than the collector -- see this test for more details on how the reflected ray pattern moves across the collector over the day...

Note also that the eave is shadowing about the top 1 ft of the collector -- this is even though the eave is only about 8 inches deep -- something to look out for on vertical collectors that you want to keep producing in the summer.

Its also (maybe) interesting to note that the underside of the eave (the soffit area) is lit up by reflected light.  I guess that may be light from the reflector that is then reflected off the surface of the twinwall collector glazing?  Its not all that bright, but certainly a non-zero loss.

To measure the solar radiation levels, I used an Apogee pyranometer on a rotating arm such that I could easily rotate it down to get a reading inside the reflector pattern and then up to get a reading above the reflected pattern on the top part of the collector.  In this way, I can get a reading at any given time how much stronger the radiation is in the reflected light pattern and use this to calculate the energy input gain for the reflector. 

To get around the blotchiness of the reflected light pattern I placed a flat glass mirror on the Mylar reflector such that the reflection from the flat mirror covered the pyranometer.  This flat mirror had to be moved as the test progressed to keep it centered on the pyranometer.

pyranometer placement

By knowing the size of the reflected pattern and the solar radiation levels inside and outside the reflected pattern, I can calculate how much energy the reflector is adding.

The plot of logged solar radiation readings for the day are at the end of this page.


This table gives the estimate for the benefit of the reflector for three times during the morning.    The clouds came in in the afternoon, so no data for the afternoon, but it should be symmetrical with the morning results. 

Time Collector Reflector Increase due to reflector
Radiation(w/sm) Power
Area reflected pattern
Reflector Rad
Energy Increment
Energy Increase
9:07 am 7.05 309 2178 2.97 1000 691 2052 94%
11:00 am 7.05 473 3335 3.47 900 427 1481 44%
12:11 pm 7.05 470 3313 4.53 966 496 2245 68%

So, this reflector that is only about half the size of the main collector certainly earns its keep with some pretty healthy increases in collector energy output.  A lot of this is due to the fact that the reflector is oriented nearly perpendicular to the incoming solar rays, so its area is nearly 100% effective.  On the other hand, with the sun elevation up at 65 degrees, the incidence angle on the vertical collector itself is also at 65 degrees and its effective area is only about 40 percent of its actual area.  That is, reflectors are very effective for vertical collectors in the summer.

I did not include the area of the collector that is shadowed by the eave in any of the above calculations, since this is basically lost area.

These pictures show the reflected patterns for some times during the morning:

reflected pattern
Reflected pattern at 9 am
reflected pattern 11 am
Reflected pattern 11 am
reflected pattern 12 noon
Reflected pattern 12 noon

Note how the early day reflected pattern is skewed off to the west -- its only around noon that the pattern aligns exactly with the collector.  The reflected pattern skews off to the east as the sun passes noon and heads west, but the clouds came out in the afternoon, so no pictures of that.  So, there is some benefit to making the reflector wider if that can be done easily.

This is a thermal image of the reflected pattern:

thermal image of reflected pattern
The ripples in the pattern are caused by the Mylar and twinwall not being flat.
Temperatures on the glazing range from about 148F down to 90F.

Recorded Data

solar intensity log

These are the solar intensity readings over the test.  The lower readings (around 470 w/sm) are taken with the pyranometer positioning arm rotated up so that the pyranometer is on the collector surface but above the reflected light pattern from the reflector -- so, these are the collector light intensities from the sunlight only.  The higher reading areas (around 950 watts/sm) are with the pyranometer arm rotated down to place the pyranometer in the reflected light area of the collector -- so these are reading sunlight plus reflected light.   A simple summary would be that the reflector about doubles the light levels and energy input within the reflected light pattern area.

solar tank temperature

This is a plot showing the tank temperatures for the day.  It shows a nice gain from about 105F up to 130F.  This is for about 170 gallons of water in the tank, and includes some use of hot water during the day -- you can see the dip one long shower at about noon. 

Note that the temperature inside the collector gets up to about 190F  -- this peak occurs during the time of day when the reflected pattern gets high enough on the collector that the collector temperature sensor is in the reflected light pattern.   While the collector is certainly capable of withstanding higher temperatures than this, it does give a bit of pause about what the no flow stagnation temperature might be, and is another reason for keeping the size of the reflector moderate.

Comparable Day Without Reflector

The plot just below is two days later than the plot above and without the reflector.  So, comparing the two similar days, the one above with reflector and the one below without reflector:

- Tank temperature increase with reflector was 25F, without reflector was 10F (so about 2.5 times as much heat added to tank)

The two days seem quite comparable to me:  ambient temps about the same, wind light on both days, sun levels were actually a bit higher on the day without the reflector.  The tank starting temperature on the no reflector day was a bit higher,  but on the reflector day, the tank climbed up through that temperature to end well above the no reflector day.  If anything, I think the day with no reflector was a better solar day.

sun intensity

sun intensity


When you look at the results for the sunny day with reflector and see that the reflector results in a doubling of radiation levels over a significant portion of the collector, and then look at the 2nd test on a very similar day and see the tank temperature rise for the day of 10F vs 25F for the reflector test day, it seems pretty conclusive that adding the reflector results in a very significant gain in performance. 

The actual tank temperature gain for the reflector was actually greater than the measured increase in radiation levels due to the reflector.  This may be due in part to the fact that the reflector increases the energy input into the collector without increasing the heat losses out of the collector glazing significantly -- this should result in an efficiency improvement.  Stating this another way, the efficiency of this kind of collector when the solar input is 1000 w/sm is about 52% while the efficiency with 500 w/sm is about 32%  (these are both with ambient temp of 60F, and absorber temp of 120F -- per this calculator...)   This is a key advantage of reflectors that I don't think gets enough attention or credit. 

In the end, I will try a "production" version of the collector that is smaller and (hopefully) looks a lot better, but I am optimistic that this will address the problem of marginal output in the spring. 

It should be remembered that for a vertical collector, the most effective season to use a reflector is in the summer when the sun is high.  In the winter, the same reflector with a different angle will have some benefit, but not as much as it does in the summer.  See the main reflector page for more on evaluating reflector gain by season.

Comparisons to Other Reflector Sizing Studies

It is interesting (maybe) to compare these results to the S. Baker table of benefits of a reflector in front of  a vertical south facing collector.  If you look at the first table where the reflector length is half of the collector height, for 44 degrees latitude, the benefit for June is 75% and this is for the case where the collector makes a 90 degree angle with the reflector.  This benefit of 75% is also what they show for collector where L/H = 1.0 and L/H = 1.5 -- so, they show no benefit in the summer for longer reflectors.  When the sun elevation is around 65 degrees, the reflection from parts of a reflector longer than half the height of the collector are reflected above the top of the collector.  I think that if they had shown angles less than 90 degrees in the table, that the longer reflectors (L/H>0.5) would have shown more than the 75% improvement.  Since I don't have access to the original paper, I don't know why they did not show angles where the reflector is tilted at less than 90 degrees.

Comparing the results of this test with the test on the earth-sum simulator of a square, south facing, vertical collector with a same size square reflector. 

The earth-sun simulator shows the radiation level on the collector with reflector to be 3.2 times the level without collector, whereas, the levels measure in the test above show the level on the collector with reflector to be 2.1 times greater than without reflector. 

So, this is a fair bit of difference.  Some of the reasons for the difference might be: 

- We are still 3 weeks short of the summer solstice, so the sun will be a bit higher at the solstice and that will drop radiation levels on the collector surface (according to the sun chart for our area, this is  about 5 degrees of solar elevation).

- The wall the collector is on faces 20 degrees east of south, so when the sun is shinning directly on the collector (12:11pm) it has not made its highest elevation for the day (the sun chart shows this to be about 6 degrees of elevation difference).

- The ground area in front of the collector is fairly light, so the whole collector area probably benefits a little in reflected radiation from this -- this decreases the difference between with and without reflector.

- Both tests are, of course, subject to some error.

Accounting for the first two items above would drop the 3.2 down to about 2.3 -- so, that accounts for quite a bit of the difference.


Gary June 4, 2011