Effect of Vent Size on Thermosyphon Collectors

This is a small test to determine the effect of changing the inlet and exit vent size on the output of a thermosyphon collector. 


Thermosyphon collectors operate on the chimney effect created inside the collector.  Solar radiation warms the absorber, which transfers heat to the surrounding air, warming it and making it lighter so that it rises out of the collector and pulls new air in.  This is a relatively weak driving force compared to fan forced collectors, so the collector must be designed for low resistance to get enough airflow to operate efficiently.

The collector used for the test is my shop heating collector ....


A lose rule of thumb for sizing the inlet and outlet vent area of a thermosyphon collector is to use one half the collector cross sectional area.  So, if the collector was 24 inches wide and 6 inches deep, the cross section would be 6*24 = 144 sq inches, and the rule of thumb vent area would be 50% of that, or 72 sq inches.


This 50% rule makes for some fairly large holes in the wall.

The question came up as to how much of a hit you would take if you reduce the vent area below the 50% rule.

This is a quick test to get an idea of the performance hit for reducing the area first by one half (25% of cross section), and them by 3/4 s (12.5% of cross section area).



Test of effect of vent size on thermosyphon collector output    
Time Vent  Velocity (fpm) Delta T (F)  Vent area (sqin) V*dT*A Performance (%) 
11:52 AM full open 130 29 72 271440 100.0%
12:04 PM half closed 158 34 36 193392 71.2%
12:14 3/4 closed 180 51 18 165240 60.9%



The vent area for the "full open" condition is 72 sq inches, and this is equal to 50% of the cross sectional area.

The "half closed" condition cuts this in half to 36 sq inches.

And, the "3/4 closed" condition (surprise) cuts this by 3/4's to 18 sq inches.

On the half closed and 3/4 closed, both the top and bottom vents were blocked by these amounts.


The velocity is measured in the middle of the top vent opening with a Kestrel wind meter in feet per minute.

The temperature at the inlet and outlet vents is measured with some hardware store alcohol thermometers.  The plastic around the bulbs was cut away so that they would not restrict airflow.


The V*dT* A column is the product of the (vent velocity)(Temperature rise through collector)(Area of vent)  -- this product is proportional to the collector heat output.  The actual heat output would be (V)(A)(density)(dT)(Cp)  Where V*A*density gives the weight flow, and Cp is the specific heat of air.


In each case, the data shown is an average of 3 to 5 readings taken at about 2 minute intervals.  There was very little variation from reading to reading for the same vent setup. 


The collector was given about 4 minutes to stabilize after each vent change.


The Performance column shows the comparison of the heat output of the baseline (full open) collector to the half blocked and 3/4 blocked conditions.


The significant fall in thermal output of the collector leads me to believe that you can't get away with vents that are too small.  It would have been nice to start with vents that are the same size as the cross section, and then go down to 75%, 50%, 25% ... to see what the effect was all the way down, but I did not feel like cutting larger holes in my wall to get to 100%.    The test clearly indicates that not nearly all of the air resistance is in the absorber -- the vent size and presumably the vent and cross section design have significant air resistance and changes to these to reduce air resistance might yield performance improvements.  Working on an optimal cross section and vent design would seem like a good project for someone -- maybe you?


Comments welcome...


upper collector vents -- this is the baseline, no blockage, 50% of cross section vent.



Measuring wind velocity and temperature.




Vents are 3/4 blocked with cardboard stapled over the vent openings.

Both the top and bottom vents were blocked the same amount.



Gary September 23, 2008, December 6, 2008