Space Heating Performance for Sunspace/Greenhouse -- Not Optimized

This test using the Montana Experimental Solar Structure (MESS  -- got to have an acronym :)  looks at using the MESS to simulate a sunspace attached to a house for the purpose of providing space heating to the house.  While the MESS is not actually attached to my house, I do like the idea of using a low thermal mass sunspace to heat a house.

There is not much actual performance data for heating with sunspaces, and thought this test might be helpful for those trying to figure out how much heat they might get from a sunspace.

sunspace for space heating

This first test looks at the sunspace that has not been optimized for space heating.  The next test will see if some optimizations can improve the performance.

If you have a sunspace how about taking some measurements and send them in -- I can help you with what to measure if you are uncertain.

Configuration for Non-Optimized Test

Ideally, an attached sunspace or greenhouse that is used primarily to provide solar heat to the attached house would have these features:

  1. Large area of south facing glazing at a steep tilt angle for good solar heat collection.
  2. Low thermal mass so that the heat is sent to the house rather than warming sunspace thermal mass.
  3. Insulated east, west, and north walls and insulation on any non-glazed ceiling areas to reduce heat loss to the outdoors.
  4. An insulated floor so that the floor does not absorb solar energy.
  5. A heat distribution system (fan and ducting) that removes hot air from the sunspace as quickly as it is made and returns cool room air to the sunspace.

The idea of all these features is to harvest as much of the sunspace heat as possible for use in the house rather than having it go into heating the sunspace or greenhouse. 

This first stage in the testing measures how much heat the sunspace can produce when some of these rules are not followed.  The idea is to see what you lose by not observing all the rules.  The test green does follows rules 1 (large glazed area), rule 2 (low thermal mass), and rule 5 (a fan to remove heat as quickly as its made).  It does not follow rule 3 (no insulation on walls or ceiling), and does not follow rule 4 (the floor is just bare dirt with no insulation).

A latter test will look at the performance after adding wall and ceiling insulation and floor insulation to see how much difference it makes.

There are two weaknesses in using this freestanding greenhouse to model an attached sunspace:

- The inlet air for the sunspace is just outdoor ambient air instead of being air returned from the house being heated by the sunspace.  Since the sunspace is freestanding, it was not possible to pull the inlet air from the house.  I may try to work out a solar preheater for the inlets on the next test.   One result of the colder inlet air is that the collector operates a little more efficiently because the losses out the glazing are reduced.

- If this were an attached sunspace, its likely that only the south half would be used, and this would be placed directly against the side of the house, with the house wall forming the north wall of the sunspace.  This would eliminate the heat loss from the whole back of the greenhouse, which is significant (see below) for this uninsulated version.  One of the advantages of an attached sunspace is that the shared wall with the house helps both structures from a heat loss point of view.

Some pictures of the sunspace/greenhouse as it was for the test:


sunspace inside
Looking west inside the sunspace.
susnpace fan
sunspace exhaust fan near peak
sunspace inlet vent
Inlet vent is currently just two 1 ft square
openings to the outside
on the east and west walls.


sun conditions
Sun conditions for the test.
sunspace inside
pretty well sealed, but no insulation.

Looking east.





Space Heating Performance

This is the performance plot for November 28, 2012 with the greenhouse/sunspace in the configuration shown above (non-optimized).  Temperatures at the inlet, outlet and at several points in the structure were logged along with solar intensity. 

space heating performance of minamal sunspace

Red Solid  -- Space heating outlet temperature (F)   (exhaust fan inlet temperature)

Blue Dash -- Greenhouse inlet temperature (F)

Green Dash -- Ambient temperature (F)  (measured outside the north wall)

Aqua Dash-Dot  -- Temperature 1 ft above floor near west end (F)

Yellow Green Solid -- Solar radiation on glazing surface (watts/sm)

Purple Dash-Dot -- Temperature at peak near east end (F)

Purple Solid -- Temperature near center of space at 6 ft above floor (F)

The sun around noon was good at about 1100 w/sm -- this is measured in the plane of the glazing and includes snow reflection.  There was a faint but visible cloud layer in front of the sun much of the day.  By about 1pm the clouds had increased and the sun level dropped to about 750 w/sm.  The sun measurements were taken with an Apogee pyranometer mounted in the glazing plane.

Its interesting to the note that the fan inlet temperatures averaged 15 to 20F lower than the peak temperature and also lower than the temperature at 5 ft off the floor.  So, perhaps a better way is needed to pull the hot air out of the peak across the full width?  Not sure where the cooler air that is finding its way to the fan inlet is coming from?

During the highest sun periods, the sunspace heated the air from about 43F up to 83F, or about a 40F warmup -- probably about the right amount of heating.   If the inlet air had been 60F room air, then the exit air would have been closer to 100F, which seems about right.

Later in the day, when the sun dropped to about 700 w/sm, the air was warmed from about 46 F up to about 76F, or about 30F.   This is when a variable speed fan that could drop the flowrate a bit to allow for more temperature rise might be nice. 

Velocity and Flow

The exit vent velocity was measured several times during the test and ranged from 630fpm in the early part of the testing (noon) up to about 685 fpm at 2 pm.  The flow is measured in a 10 inch duct with an area of 0.55 sf.  This gives an exit flow rate of about  350 cfm to 375 cfm. 

The velocities at the inlets were also measured at around 200 fpm.  Each inlet is 1 sf, so the flow rate calculated using the center of inlet velocities would be 400 cfm -- more than the exit flow.  I attribute this to a fall off toward the edges of the square inlet openings. 

The flow rate is about 350 cfm and the volume of the space is about 1400 cf, so it takes about 4 minutes to introduce enough new air through the inlets to fill the space. 

The flow rate per sqft of glazing is about 350/200 = 1.8 cfm per sqft.  This is a bit lower than what is typically used for good performance in an active collector.

Floor and Wall Temperatures

I checked the dirt floor temperatures several times both in the sunlit areas and the shaded areas.  The results were a bit surprising.

The sunlit floor near the north wall started at about 97F just as the fan was turned on to start the test, and dropped to mid 70's as the test was going.   It reached about 60F by the end of the test.  So, it appears that there even though there was good sunshine directly on the dirt floor, there was enough airflow over the dirt to drop the temperature of the floor during the test.

The shaded floor just north of the low south wall was about 59F at the start of the test, and dropped down to the 36 to 40F range.  I believe that this was the influence of the cold air streams coming into the inlet vents located low in the east and west walls near the south wall.  That is, the cold vent air flows tracked low along the south wall cooling the floor.

All of the floor temperatures were just surface temperatures -- I did not see what was happening a ways down into the dirt.

The back wall temperature behaved similarly, stating at around 99F and dropping down to 60 to 75F during the test.   Again, the airflow appeared to cool the back wall once the fan was going. 


Living in Your Collector

Since one good reason to build a sunspace is to have a nice place to spend some time, I'll pass on some experiences from spending time in mine.  

Without the fan running, the space can be quite hot -- up to 120F and more.   Not a space you would like to spend much time.  But, a good space for doing laundry or installing a solar water heater.

With the fan removing heat the sunspace can be quite comfortable.  The temperatures drop down into the nice range, its brightly lighted, there is just a bit of air current, which feels good.   It would be a great place to have a cup of coffee and read the paper -- something you can't say about the inside of most solar collectors.

The fan I am using is a Dayton 10 inch 600 cfm fan from Grainger.   Its pretty quiet, but still too noisy to live with if its mounted right in the sunspace.   It might do better with some kind of isolated mount.  I think that choosing a quiet fan that produces the right amount of airflow is something that deserves a lot of attention in sunspace design.

Contrast to Active Collector

In this application, the sunspace is being used as an alternative to a conventional solar air heating collector to provide space heating.   Its interesting to note some of the contrasts. 

  1. The flow velocities are much lower in the sunspace -- in most places I can't measure them with a good air velocity meter and you really can't feel them, but in a conventional collector, air velocities are fairly high -- up toward 100 fpm. 
  2. In a conventional collector, there is usually a significant air velocity along the glazing, where as in the sunspace its too small to measure. 
  3. The surfaces that absorb the solar radiation are spaced further from the glazing, are larger, and run cooler.
  4. The flow pattern in a conventional collector can (it seems) be controlled more precisely with channels, baffles, screens, ... than for a sunspace. 
  5. The sunspace has more outer heat loss area and more potential leakage area.

Items 1 through 3 would seem to me to favor the sunspace as an efficient heat collector with low losses?

Items 4 and 5 would seem to favor the conventional collector?

IR Pictures

Thermal camera pictures of the glazing from outside and inside during the test at around noon.

IR glazing temps

At the time these images were taken, temperatures were: Peak of GH 95F, at 6 ft near center 95F, ambient 45F, Inlet air 45F.

Smoke Test

A smoke test was done to have a look at airflow patterns.  The 90 second smoke pellet was placed in the East inlet vent with the fan operating.

The inlet where the smoke pellet was ignited is low in the east wall near the south glazing (pictures above).  The smoke rushed into the inlet and proceeded west staying low and moving quickly.  As it approached the middle of the front wall of the GH, it encountered the stream of inlet air from the west vent and this caused it to spread out and slowly move upward -- it did not penetrate the air stream from the west inlet at all.   Over some time (3 minutes?) the smoke spread out horizontally and then upward in what seemed like a fairly slow and diffuse way.   It basically filled the whole GH with a diffuse smoke that gradually made its way up to the peak and then out the fan.

The flow rate is about 350 cfm and the volume of the space is about 1400 cf, so it takes about 4 minutes to introduce enough new air through the inlets to fill the space. 

Velocity Survey

I walked the GH with a hot wire anemometer trying to get an idea where the air currents were flowing.  Basically the air currents in nearly all of the GH were too low (less than 2 fpm) to measure with the anemometer.  The exceptions were within about 4 ft of the inlets and the exit fan.    So, any flow channels or paths that exist are too low in velocity to detect with the hot wire anemometer.

Next time I will see how the smoke pencil does on seeing where the air is flowing.



Some rough estimates of heat output and even rougher estimates of efficiency.

Looking at 12:30 pm

Solar input = (1080 w/sm)(200 sf)(1 sm/10.76sf) (3.412 BTU/h / watt) = 68500 BTU/hr

Heat output = (350 cf/min)(0.062lb/cf)(81.6F - 43.8F)(60 min/hr)(0.24 BTU/lb-F) = 11800 BTU/hr

An implied rough effic of = (11800 BTU/h)/(68500 BTU/h) = 17%


At 1:30 pm

Solar input = (770 w/sm)(200 sf)(1 sm/10.76sf) (3.412 BTU/h / watt) = 48800 BTU/hr

Heat output = (375 cf/min)(0.062lb/cf)(76.3F - 46.4F)(60 min/hr)(0.24 BTU/lb-F) = 10000 BTU/hr

An implied rough effic of = (10000 BTU/h)/(48800 BTU/h) = 20%


So, around 20% efficiency is not is not exactly stellar, but 12,000 BTU/hr from a modest sized sunspace is definitely a useful amount of space heating.

Just for comparison, its about equivalent to a 4000 watt PV array at the STC rating. 

I expect that once the sunspace is optimized for heating in the next test it will do better.

In the screen absorber collector testing that Scott and I did last year, the sunny winter day efficiencies were in the 40% ballpark.

Next Test

For the next test, I will insulate the floor and also insulate all wall and the north ceiling.   I'll repeat the same sort of test and see how much difference it makes.

I am inclined to think that the difference will be significant.   The sun shining directly on the dirt floor as it does now should result in a significant fraction being absorbed directly by the ground and not getting to the "house".  For the insulation, the total area to be insulated is about 300 sf.  If you assume that the average inside temperature is 70F and outside is 30F, and that the current plywood walls are R1, then the heat loss is (300sf)(70F - 30F)/(R1) = 12,000 BTU/hr.   The total solar incident on the 200 sf of glazing is about 60000 BTU/hr, and you would expect to only harvest about half of that in a good collector, so the 12K BTU/hr is a worthwhile fraction of that.

I guess we will see.

If you have any ideas or suggestions, please leave a comment on this page....


November 28, 2012