Posted by Gary on August 1, 2004, 4:57 pm
I have an attached garage that I use as a shop (no cars allowed). The
garage door is 8ft by 18ft and faces South --it constitutes about half
of my total available south exposure.
I have been thinking about ways to make this garage door opening work
to collect solar energy for heating the house. I have worked out one
scheme that seems promising to me.
Here's how it works:
Add a fixed layer of glazing (e.g. polycarbonate panels) about 4 to 6
inches South of the existing garage door. Mount the glazing panels
inside of the existing garage door framing. The new glazing panels
would be removable (perhaps 4ft wide by 8ft high). They would remain
in place for the heating season. The glazed panels would be attached
to the garage door framing in such a way as to allow little
infiltration (unlike my current garage door :-).
Just North of the new glazing add a layer of black metal window
screening. The screening could be supported off the back surface of
the glazing panels, but would be spaced off the glazing by about 2 to
3 inches. When the garage door is down, the screen would be about
midway between the new glazing panels and the door.
Paint the South face of the garage door a darker (but not too ugly)
color to increase its solar absorption.
There is an attempt at an ascii sketch at the bottom of this message.
Note: The garage door consists of 5 horizontal panels, each 19 inches
tall and 18 ft wide -- they roll up on a set of tracks. The panels
are metal with about R5ish insulation embedded. There is weather
striping all the way around, but its hard to seal something so big and
flexible very well.
Operation as a solar collector:
When the sun is shinning, raise the garage door about 10 inches. This
makes a 10 inch gap at the bottom of the door, but also
opens a horizontal slot at the top of the door as the top panel
starts to get pulled North along the horizontal part of the door
track. The gap is about 3.5 inches wide. With the door in this
position cool garage air near the floor enters the gap below the door,
makes it way up through the space between the new glazing and the
garage door, and picks up heat from the door and from the screening.
This is a bit like the free convection solar heater that I use on my
barn (described at http://users.montanadsl.net/~reysa/ )
Based on experience with the barn heater, I think this would do a
good job of heating (probably overheating) the garage space,
but I would really like the heat to go to the house, not the garage.
So, add the following components to transfer most of the collected
heat to the house:
Add a 12 inch tall vertical divider that runs East-West across the
garage ceiling about 18 inches North of the plane of the garage door.
The top of the divider is in continuous contact with the ceiling.
This makes a cavity that traps the heated air (which is buoyant)in the
space just above the 3.5 inch top gap. The top panel of the door
also helps to guide air into this cavity area. I have a similar
(unintended) cavity on the barn collector that is formed by the
ceiling joists that tends to confirm that this approach might work).
Add a duct with an inline electric fan that connects from the cavity
to the house space. The fan would have a simple thermostatic control
that would turn the fan on when the air in cavity reaches (say) 80F.
A manual override switch would allow the fan to be turned off when heat
is not wanted. Or, perhaps, a PV panel driven fan? A simple return
grill in the garage/house wall would allow return air to enter the garage.
I suppose that the opening and closing of the garage door could be
automated based on a sun sensor, but this seems like overkill in our
case where the garage door opener is very handy to the house.
When the sun is not shinning, just shut the garage door. In this
night time position, There is less heat loss then there was because
the added glazing decreases infiltration and increases R value.
If heat is desired in the garage instead of in the house, just open
the garage all the way. If things are getting too warm, just close
the garage door all the way. I suppose some thermal storage in the
form of water barrels could be added just North of the garage door if
some carryover through the night is desired. If I leave the glazing
up in the spring and fall, I can open the garage door, and have a shop
with a view :-)
Here is an attempt at a Nick-ulation of the performance for Billings, MT.:
Net Collector area = 90% of 144 ft^2 = 130 ft^2
Transparency of PC glazing = 0.95
Absorption = 0.95
NREL says that for the month of Nov (kind of an average performance
winter month), Billings sees 1080 BTU/ft^2 per day of radiation on a
South facing vertical surface. Average temperatures are 35.1F
average, 44.5F maximum, and 25.6F minimum. So, average day
temperature = 39.8 and average night temperature is 30.4F.
For the about 6 hours of collection time,
Collector gain = (Acol)(Trans)(Absorb)(Daily Rad)
=(130 ft^2)(0.95)(0.95)(1080 BTU/day)
= 128K BTU/collection period
Collector loss = (Acol)(Ucol)(Tcol - Tamb)(Hours door is open)
= (130 ft^2) (1.1 BTU/ft^2-F-hr)(120F - 39.8F)(6hr) =
= 69K BTU/collection period
Tcol is a guess at the average collector temp
Ucol is loss coef for single glazed col, including rad
Net gain = 128K - 69K = 59K,
or 59000/(90000 * 0.85) = 0.8 gals of propane at 85% efic
or, (30 days)(0.8gals/day) = 24 gal/month
or, (24 gal/mo)($.30 /gal) = $1 per month
If you do all the heating months from Oct -> March in the same way,
the totals come to 149 gallons of propane per year. The savings are
largest in the shoulder months, and smaller in Dec and Jan (due to
less sun and colder temperature). I did assume that I could use all
the heat for these months, which depends on the house heat loss, but
this is (unfortunately) not a problem for my house :-)
These seem a bit conservative to me in that no credit is taken for the
decrease in nighttime losses from the garage due to the new glazing,
and some heating occurs in other months.
Here is my first cut at pros and cons.
The system should generate heat equivalent to about burning about 150
gallons of propane a year at 85% efficiency for a saving of $00 in
our area (and going up each year).
The cost of the system is low -- I estimate about $40 for materials.
The payback on materials would be less than two years. Or, one
could look on it as an inflation protected, tax free 57% return on
I like the way the system makes multiple new uses of the existing
garage door to: 1) provide movable insulation, 2) provide the
collector vents, 3) provide part of the absorber, and 4) provide a big
window/collector when opened all the way.
The system should have a long life with little maintenance.
The garage should be more comfortable in the winter. Some of the
collector heat that is intended for the house will "leak" into the
garage. Also, the heat losses from the garage should be reduced,
which should keep it warmer.
The heat loss from the house should be reduced a bit, because the
garage shares two walls and a ceiling with the house.
I like the simple way in which heat can be directed to either the
garage or the house.
It seems like a nice simple system that anyone can understand and use.
You lose easy access to the garage through the large door (e.g. for
cars). This does not matter to me, but would be a problem if you
insist on putting your car in the garage :-). One potential way
around this is to integrate the glazing panels into a 2nd set of
doors that open outward on vertical hinge lines, but this does make it
a more complex carpentry project.
I am a bit concerned about making this look OK from outside. Any ideas?
I need to work out some way to be able to rip very long boards on my
table saw -- now, I just open the garage doors to do this.
A final thought -- it seems like there ought to be some way to
integrate this whole concept into a new garage door design (one that
preserves the ability to open the door as well as collect solar
energy) -- if widely applied, this might be worth:
(10^2 heating days/yr)(10^5 BTU saved/day)(10^6 houses?)
= 10^13 BTU per year ??
Any comments would welcome -- Gary
<- Duct toward house <-- | | | <-- South wall
| | |
3.5in gap ----> \ +----+
\ . |
\ . |<-- new glazing
+ . |
| . |
Garage door -> | . |
+ . |
| .<------ Screen absorber
| . |
+ . |
Garage | . |
| . |
+ . |
Gap --> . |
--------Floor ---------------------------- Driveway --------------
Posted by nicksanspam on August 2, 2004, 11:11 am
Nice. I have an 8'x9' south garage door...
I saw an arrangement like this at an auto repair place in Brooklyn,
but the panels stopped 2' short of the ceiling. Polycarbonate expands
It's simpler and more efficient to open the door all the way. I'm reluctant
to do that automatically because of burglars. A house alarm system might
prevent the door opening if nobody's home, but then we don't get the heat.
Last winter, I insulated my garage walls and ceiling and put a layer
of greenhouse poly film over the outside of the door. I was pleased with
the light and heat when the door was open, but it was more difficult to
work on large objects, since everything had to go through the man door on
the opposite wall. Poly film is very transparent to IR, but it's cheap
and it comes in large pieces. The view is cloudy...
Would that interfere with opening the door all the way?
How many cfm?
Or a $5 line voltage thermostat.
I suppose you will want to prevent reverse thermosyphoning at night.
I would definitely want to do that. Perhaps Duane Johnson can design
and sell a kit that provides a pulse (like a finger on a push button)
in the morning and evening, with a door position sensor to make sure
the door is open during the day, vs vice-versa.
As I recall, black aluminum window screen is about 0.5, so 25% of the sun
would find its way back out the glazing, with a reflective door... 0.95
would require a door reflectance rho, where 0.25rho = 0.05, ie rho = 0.2.
Green is 0.5. Flat black is 0.04...
That depends on the upper door air gap and the fan cfm.
About 10K Btu/h. A 1000 cfm fan would have a 10 F temperature rise.
I like Grainger's 4TM66 $3 5850 cfm 90 W reversible 16" window fan,
with 3 speeds and a thermostat.
Another pro: I really enjoyed the light with the door open.
Much better than working under the single bulb in the garage.
It may not raise the cost much...
A 4'x8' flat polycarb panel probably needs wood on 2' centers to avoid
too much oilcanning. Can the wood act as a spring to prevent it entirely?
Hinges or easily removed panels.
Posted by Gary on August 2, 2004, 9:49 pm
Hi Nick -- thanks for the comments -- a couple questions below if you
have time. -- Gary
I agree that if I only want to heat the garage, it would be better to
just open the door all the way, but if I want to get the heat to the
house via the duct/fan arrangement described below it seems like I
have to do something like the partial door opening scheme above to
collect the heat for the fan intake -- or is there a better way?
I think that if I just opened the garage door all the way, and had
(say) a fan that would blow air into the house that the garage would
overheat, and also that I would not get as much heat to the house?
A house alarm system might
I guess one advantage of rural MT is that there aren't many burglars :-)
I am leaning toward adding a column in the middle of the 18 ft span.
This makes two 9 ft wide openings. I think I could then make the
glazing panels look something like garage doors, and have one (ir
nire) of them actually hinged on a vertical hinge line to allow easier
The 12 inches is as far down as the divider could go without
interfering with the door. When the door is raised 10 inches, the top
of garage door matches up pretty well with the bottom of the divider
-- I think this will help to channel hot air into the cavity formed by
the divider, and then into the duct.
If I were to use the barn collector as a guide it has (10
vents)(0.5ft^2) = 5 ft^2, and a typical sunny day vent velocity is
140 fpm -- this would give
(5 ft^2) (140 ft/min) = 700 cfm
or, if scaled down by the collector area (144/160) (700) = 630 cfm ??
Yes -- I was thinking about a motorized damper that is activated with
the fan. Or maybe just the poly film "check valve".
Yes, 0.9 -- I thought that SunTuf quoted 0.95, but when I went back
and looked it is 0.9.
I like the way that you can see through one layer of window screen --
it seems to me it makes it look better from outside, and also gives
you a view from inside.
If I am understanding what you are saying, half of the incoming light
goes through the screen, and part of that is reflected back by the
door, and half of the reflected part gets back through the screen --
for the green this would be: (0.5)(0.5)(0.5) = 12%lost?? I think
something just a bit better than green at 0.5 would be OK with me.
I did some hour by hour simulations for Billings using the TMY2 data,
and the assumption of 120F for the collector temperature hurts the
performance quite a bit when outside temperatures are very low (high
losses), or solar gain is not high (eg partly cloudy) -- ie the model
calcs losses with a 120F collector temperature that would probably
never be reached on a partly cloudy day.
I guess I need some way to vary the collector temperature according to
the actual conditions?
I suppose a more complex control system and a variable speed fan would
The hourly simulation actually shows some peaks above 20K BTU/h
A 1000 cfm fan would have a 10 F temperature rise.
?? not sure what this means ??
I'll take a look at that.
I may have been a bit hasty assuming all of the heat that is generated
could be used -- the hourly simulation shows times during mid day on
warm days when the supply exceeds the demand by quite a bit --
especially in the warmer months.
Yes, I open the big door sometimes in the summer now, but its not
something you want to do when it 10F outside :-)
I have rebuilt my barn collector so that the corrugated PC panels are
supported every 2.5 ft perpendicular to the corrugations -- this makes
an amazing difference -- there is almost no buckling or deflection
now. This is true even thou the wood supports are only 3/4 by 3/4 wood.
Posted by nicksanspam on August 3, 2004, 11:32 am
Warm air rises, but it seems to me you still need the screen to avoid
storing sun in the concrete floor, and maybe to reduce the amount of
light and radiant heat when you are working in the garage.
Maybe not. The tradeoff is keeping the temperature and daytime loss
through the glazing low. In full sun, 0.9x250x130ft^2 = 29.3K Btu/h
passes through the glazing. If the air near the glazing has temp T (F),
(T-40)130 is lost through the glazing. The 3.5"x18' slot has 5.25 ft^2
of area. With 70 F air in the bottom slot and a 7' height diff between
top and bottom slots, Q = 16.6x5.25xsqrt(7(T-70)) cfm will flow into the
garage, carrying I = (T-70)Q Btu/h of heat. So we have something like
this, viewed in a fixed font:
--- | 1/130
29.3K Btu/h | I -->
29.3K Btu/h = (T-40)130 + 231(T-70)^1.5, or T = 70 + ((265-T)/1.78)^0.666
Plugging T = 100 F into the right side makes T = 182 on the left.
Plugging T = 182 in on the right makes T = 82.9 on the left.
Repeating makes T = 91.8, then 91.1... Not too hot.
You might collect 22.4K Btu/h of useful heat
with a 76% efficiency.
You may not need it, if the garage ceiling is insulated. You might add
thermal mass to the ceiling.
You may not need a duct, if the fan is near the ceiling.
If the house temp varies much, you might also use a differential thermostat.
We might make this from standard parts, eg a thermostat in a glazed box.
If rho = 0.2 = 0.5g + 0.04(1-g), you might mix 35% green with 65% black.
You might redo the calcs as above.
Maybe. You could put that into the simulation.
Moving 10K Btu/h from the garage to the house with
a 1000 cfm fan requires a 10K/1K = 10 F temp rise.
More thermal mass in the house or the garage ceiling might help.
Should be fine, with the glazing and screen in place.
You can see through Dynaglas, with distortions... Flat polycarbonate
(eg Lexan rolls from GE) is more windowlike, but subject to scratching.
Posted by Gary on August 5, 2004, 4:20 am
Your idea is that the screen would absorb some of the light, heat up,
and cause air around to rise to the ceiling where it could be fanned
into the house?
I wonder if the screen could roll up like a window shade? Then the
screen might be able to basically control whether heat went mostly
into the garage (screen up), or mostly into the house (screen down)?
I did this, and it looks pretty good -- Tcol varys in sensible ways
with the sun input, and ambient temperature.
Your equation: Q = 16.6x5.25xsqrt(7(T-70)) cfm
Is this for a chimney? Is the 16.6 empirical or?
I guess that on the actual system, I could vary the 5.25 ft^2 gap
area, and/or introduce flow restrictions if I wanted somewhat higher
Tcol at the cost of lower efficiency?
On cabinet doors that have a central panel surrounded by a frame, the
central panel expands across the grain, while the frame expands with
the grain. There is much more expansion across the grain, and to keep
the panel from buckling or cracking, it floats in a dado in the frame.
I wonder if the same technique could be used for these garage door
glazed panels -- i.e. the PC glazing panel would float in a dado in
the frame, perhaps with light caulking to prevent infiltration?
I am thinking of the frame breaking the glazing up into panels that
are probably no more than 2 ft by 4ft ish.
Not sure what you meant by using the wood to act as a spring?
Anyway, it seems like I could go ahead and build the glazed
panels/doors, and then experiment with gaps, screens, dividers, ducts
I ordered one of the Testo 405's that mentioned in another post --
looks like it should make velocity measurements a lot easier.