Flow Organizer Solar Collector Test  

This prototype of a window size solar collector with a “flow organizer” grew out of an internet discussion on how to prevent excessive heat buildup between an insulating window shade and the window glazing.  Nick Pine suggested that making the shade into some form of solar collector would keep the heat from building up, and provide some useful heat to the living space.  Version one of this combined window shade and solar collector is here:  http://www.builditsolar.com/Experimental/ColShade/colshade.htm. Nick further suggested that a version of the collector with a flow organizer might work better than version one, which is a straight through thermosyphon collector.  This is a first cut at a flow organizer version of the window collector. 

The results for a basic collector with flow organizer, and the results of applying three refinements to this configuration are show below in the “Test Results” section.  The Flow Organizer collector appears to do quite well. 

Solar Collector with Flow Organizer:

Most thermosyphon air collectors have an inlet at the bottom and an outlet at the top.  Cool room air enters the bottom, rises up through the collector, heats up as it passes through or along the absorber, and exits out the top vent. 

This collector has a “flow organizer” that allows the cold air inlet and hot air outlet to be positioned at the top of the collector.   The cold air passes between the aluminum sheet outlet vents and flows to the south side of the collector (see picture).  It then flows down the glazing, and through the screen absorber, as the absorber heats up the flow, it rises up the north side of the collector and exits out the aluminum outlet vents.  The flow organizer allows the cold flow to pass through the hot flow, and thus to flow down the glazing.  Having the cooler air flow down the glazing increases the efficiency of the collector, by reducing losses.  It also might be more effective in preventing backflow at night.

There is a good description of the flow organizer on David Delaney’s web site here:



Flow organizer collector prototype viewed from the south (sun side) without glazing.  The rectangular aluminum ducts are the outlet ducts.  The cool incoming air flows through the gaps between the ducts, and then down the glazing.  It’s hard to see, but the absorber is a layer of window screen suspended about midway back in the collector.  The ¾ by ¾ piece of wood visible near the bottom of the collector anchors the bottom of the absorber screen.


The flow organizer collector in position against an existing piece of glazing.
The picture is taken outside, and is looking north (this is what the sun sees).
The 3 small aluminum cylinders shield the temperature sensors from direct sunlight so that they can sense the air temperature (does anyone have a better way shield the temperature sensor?).


Inside height from bottom to flow organizer shelf = 40.5 inch
Inside width = 40 inches
Height above flow organizer shelf = 6 inches (without flow deflector)
Collector area = (40.5 + 7)(40)/144 = 13.2 ft^2

Inside depth = 7 inches, 3.5 inches front of absorber screen, and 3.5 inches behind screen

Inlet slot just north of glazing is 2 inches deep by 40 wide.  Area = 80 sqin

6 outlet ducts at 3 inches wide by 4 inches deep.  Area = 72 sq inches
The outlet ducts extend 6 inches above flow organizer shelf.

Temperature sensors:
There are 4 logged temperature sensors in collector.
All but the outlet temperature sensor are visible on the picture above. 

Outlet temperature – centered in middle outlet duct. 

Inlet temperature – in the inlet slot flow near middle of collector and just below the flow organizer shelf.  It is protected from the sun by a 2 inch diameter alum cylinder.

Lower collector temperature sensors

Two temperature sensors are located about ¼ of the way up from the bottom of collector.
One is just north of the glazing and the other is between the absorber screen and the back wall.
Both of these sensors are protected from direct sun by aluminum cylinders.

The absorber is 1 layer of black fiberglass insect screen that is suspended 4 inches from the back wall (3 inches from the glazing).   There is an open slot of about 2.5 inches at the bottom of collector.  Refinement 2 adds a 2nd layer of absorber screen. 

The back wall of collector is 1 inch Polyiso insulation board, and is painted black on the inside.

The other walls and bottom are Polyiso board without paint.  They are faced with aluminum foil.  

Test Results --A base configuration and 3 variations on it were tested. 

The table below gives a summary of the results.   

Base Collector: has one layer absorber screen, no flow deflector, square edge on inlet flow  

Variation 1: added a deflector that was aimed at deflecting the flow from the outlet ducts into the room.  The deflector did seem to increase the output noticeably (+15% ish).   I’m a bit at a loss to explain why this made so much difference. 

Variation 2: added a 2nd layer of absorber screen.  This actually appears to lower the output.  The flow rate is down, and the delta T from inlet to outlet is about the same. 

Variation 3: added rounding of the corner on the flow passage into the inlet slot.  That is, rounding the south edge of the flow divider shelf.  The area of the inlet slot was not changed, but instead of having a sharp corner where the air turns down into the inlet, it was rounded off.  This does not appear to make much difference.  It actually shows a slight reduction in output, but I think that is just the error margin in the test setup. 

The night measurements seem to indicate that not much air is flowing through the collector, which would be good.  I could not detect any flow velocity, but I can’t really read any less than about 60 fpm.  Smoke would be nice. 

Results Table





Add  Deflector (2)

Add 2nd Screen


Round Inlet Lip









2nd with



10:00 am




12 midnt

11:02 11:41a







42 30







70 64














91.0 87.1

Tglass side ¼ up







Tback side ¼ up













121.6 117.8
















90 fpm





105 100-


70 fpm (9)





85 80


18500 lum/ft^2





20500 21500














30.6 30.7


1897 BTU/hr





2150 2054

Qout adj






2203 2006







170 154

(1)   Base configuration as described in the geometry section above.  One absorber screen.
(2)   As above, with added alum sheet deflector that deflects the heated flow from the outlet ducts toward the room (see picture).
(3)   As above, with another layer of window screen added to the absorber – 2 layers total.
(4)   As above, with the south edge of the flow divider shelf rounded to improve inlet airflow.
(5)   Same as configuration 4, but run at night.

Base Configuration Energy out (Qout) was calculated as:

Qout  = (107F – 75.5F) (0.715 ft^2)(90 ft/min) (60 min/hr)(0.065 lb/ft^3) (0.24 BTU/lb-F)
        = 1897 BTU/hr

Adjusted Energy Out (Qout adj) is the energy out adjusted to the highest sun level of the test.   This is an attempt to compensate for the fact that sun level varied throughout the test.  Sun level was measure with an Onset light intensity logger, which gives light intensity in lumen per sqft.  It does not measure the full solar spectrum, but does give some idea how much the sun varied (see plot below).

Update 2/14/06:
This test incorporates the changes suggested by David. 

The changes are: 

-          Add a vertical flow divider that extends vertically down 2 inches from the edge of the horizontal flow organizer shelf.

-          Move the absorber toward the north (now 3 inches from back and 4 inches from front) 

-          Terminate the absorber 5/8 inch below the bottom of the flow organizer shelf

-          Suspend the absorber from a horizontal wire such that the top of the absorber does not have a “fat” support that impedes flow.

-          Also went back to one layer of absorber screen, since this seemed to work better in earlier test.

-          Dave’s drawing below:

Pictures with the changes:

Vertical flow divider added.

I am wondering if I should have used a light colored tape?  The tape gets hot the touch, and may be discouraging the downward airflow at the inlet by heating it to early?









Looking up at top of absorber, and into the outlet ducts.  The absorber has been moved north a bit, and the thick member at the top of absorber replaced by a suspension wire with tape over it.

Single absorber screen.













How did the Flow Organizer Collector compare with others:
A VERY rough comparison of the outputs of several air collectors: 

The Collector Shade concept (1)                    100 BTU/sqft-hr 

The Organized Flow collector                        190 BTU/sqft-hr 

The Garage Thermosyphon collector(2)          230 BTU/sqft-hr

 (1)   shade converted to thermosyphon collector: http://www.builditsolar.com/Experimental/ColShade/colshade.htm
(2)   My workshop thermosyphon heater: http://www.builditsolar.com/Projects/SpaceHeating/solar_barn_project.htm 
It may be unfair to compare the garage collector to the two window collectors, as it has the advantage of a 7 ft rise from inlet to outlet.

From outside with aluminum sheet flow deflector in place.  It deflects outlet duct flow into room.








With the inlet lip of the flow divider shelf rounded off to allow smoother flow into inlet slot.  It turns what was a square edge into a half cylinder.  It actually looks better than the picture.  Looking northwest.  Screen absorber starts just below the white south edge of flow divider shelf.







Odd stuff:
It seems odd that the collector inlet temperature was so much higher than room temperature.  Typically there was an about 20F difference between room temp and the sensor that measured the temperature right at the top of the inlet slot.  Is the inlet air getting warmed by the outlet ducts or air before it gets to the inlet slot?  Or, is the temperature sensor not getting a good reading?   The area of the inlet temp sensor is in sunshine.  I surrounded the temp sensor with an alum sheet cylinder to shield it from the sun – maybe the shield is not effective?   Any ideas on this?

I added the flow deflector to see if that would decrease this difference between room and inlet temperature, but it actually increased it, and also increased the output – odd. 

Looking at the temperature sensors that were located about ¼ of the way up from the bottom of the collector.  One is near the glazing, and the other half way between the absorber screen and the back (north) wall.  On most of the tests, the front sensor reads higher temps than the back sensor.  This seems like the wrong way round? 

Both of these sensors are in the sun, but are shielded by the same kind of cylindrical alum shields described above.  Any ideas on this?  Adding the 2nd layer of window screen absorber seems to increase this effect. 

I have to say this was a start, but not done to the level of precision I would like, and I think the numbers should be used with some caution.   It would be nice to do more of a side by side comparison under exactly the same sun conditions, and I am not sure I believe that the temperature sensors in the flow organizer collector are measuring the actual air temperature even with the shields.  I pass the results on because they seem interesting, and maybe someone will want to take this further, or make some suggestions on how it could be done better.

2/14/06 -- after playing around some more with the temperature sensors, and doing some additional measurements with a shaded thermocouple on a stick that I could move around from place to place, I feel better about the temperature sensors getting fairly accurate data for both of the tests.



Updated 2/14/06


Data Logger Plots: 

Configuration 2 -- Base configuration with outlet deflector added:

 Configuration 3 – outlet deflector + 2nd absorber screen: 

 Sun Intensity: 

Update 2/14/06 -- logger plot for the flow organizer with Dave's suggestions incorporated:

2nd Test with changes 2/10/06


Some Construction Pictures:




    Overall height    48 inches

    Height from bottom to flow organizer shelf 42.5 inches (40.5 inches inside)

    Overall width 42 inches (40 inches inside)

    Overall depth 8 inches (7 inches inside)


    Depth of cold air entry slot 2 inches


    Window screen divider is 4 inches for the back (north) face and 3 inches form the front (south) face.

    Area of downflow channel = 40*3 = 120 sqin

    Area of upflow channel = 40*4 = 160 sqin


    Upflow vents (hot air exit)  -- 6 total at 3 inches wide (E-W) by 4 inches deep (N-S) -- 6 inches tall.

    Upflow vent area = (3*4*6vents) = 72 sqin  -- about 50% of upflow channel XC


    Slots between the upflow vents (where cold air enters) are 3.66 inches wide.

     Area between upflow vents = 5* ( 2*1.83  + 5*3.66) = 110 sqin  -- about 90% of downflow channel XC


Sides, bottom, and back are made from 1 inch Polyiso board with alum foil faces.  South face of back wall is painted black using barbeque paint.    Glued together with "great stuff" PU foam.


The flow divider board is also 1 inch polyiso  -- it has 6 rectangular holes of 3 X 4 inches each to accept the alum upflow vents.   The slots between the upflow vents are 3.66 inches wide.


The upflow vents are made from light alum flashing bent into rectangular 3X4 inch ducts.  The ducts stick up 6 inches above the flow divider board after being pushed fully into the flow divider board.


The vertical screen is one layer of black fiberglass window screening.  The top end is fastened to the flow divider board at the S margin of the upflow ducts.  The bottom is weighted with a 3/4 by 3/4 wood piece, so that it hangs vertically.  The bottom of the screen is about 1.5 inches above the bottom of the enclosure.



My intent is to just push it against one of the acrylic panels on my outer garage doors to test it.


I will put some of the Onset temperature sensors in the up and downflow paths. 

If the velocity if high enough, I'll measure it with the Kestrel meter or the Testo stick. 

It would be nice to have some way (like smoke) to see the flow -- I have not had much luck with generating enough smoke to be useful -- maybe you have an idea on this.


I did not glue in the ducts or the flow divider board so that they could be changed easily.

Not sure what to do about the top lid for the box (near the top of the upflow ducts) -- if anything?


I'll give it a try on the next sunny day -- which may be tomorrow -- let me know if you have any ideas on things to change or measure.




Flow divider board






Cut out with razor knife -- very fast and clean




Upflow duct from alum flashing.                Upflow ducts taped to flow divider board.




Bottom of screen-- looking north.