A high-performance, cost-effective, solar-heated greenhouse

Commercial greenhouses that operate in the winter are usually made of two
layers of polyethylene film, statically-inflated with a small blower, after
the film is stretched over curved steel pipes on 4-6' centers. The steel pipes
typically form a quonset hut or gothic arched structure, about 13' tall at the
ridgeline, for a 30' wide house. These greenhouses are fairly inexpensive:
A 30' x 96' greenhouse costs less than $3,000, including the poly film cover
and ground stakes that serve as the foundation. Three people can put one up
in less than one day.

But, they also use a lot of energy in the wintertime. Roger Williams mentions
that a standard commercial greenhouse like his 30' x 208' tomato greenhouse
in New Bruswick, Canada, would typically require about 12,000 gallons of oil
to heat, in that 9,000 degree-day climate. He was able to reduce the oil
consumption of his greenhouse to less than 4,000 gallons a year, by various
means, including the use of better insulation and solar heat :-)

What can be done to make commercial greenhouses more solar-heated?

Here is a drawing of a typical $3,000 commercial greenhouse...

                    Greenhouse  96' long x 30' wide
          :                                                 :
          :                                                 :
        < A                                                 : 
        < P                                                 A <
        < R                                                 : 
          :                                                 :

The house usually has a large fan at one end, about 23,000 cfm, mounted
chest-high on one endwall, and a large motorized damper at the other end. 
There is usually a large gas or oil heater hanging from the ridge pipe.
When the greenhouse begins to heat up from the sun in the wintertime,
the fan turns on and the motorized damper opens, to cool the house.
When the sun goes down, the heater comes on...

The greenhouse may have 4 benches, as shown, each one being about 4' wide,
with 2' aisles between the benches. The benches are typically supported on
a 2 x 4, cement block or steel framework, at a height of about 33". Since
the greenhouse walls are curved, the area near the walls has low headroom,
typically about 4' 5" a foot away from the wall, 5' 5" 2' away, and 6' 7"
3' away from the wall. This headroom can be increased 1-2' by using longer
ground posts.

The first thing that one might do to help this picture, energy-wise, might be
to put some dark-colored 55 gallon drums full of water under the south bench,
to support it. 55 gallon drums are free for the hauling in many places, and
the benches need some sort of support anyway. This would make the benches a
little taller, since 55 gallon drums are about 23" in diameter and 35" tall.
So along with this, one might pile up 2" of gravel or dirt in the walkways.

                    Greenhouse  96' long x 30' wide
          :                                                 :
          :                                                 :
        < A                                                 : 
        < P                                                 A <
        < R                                                 : 
          :               white poly film                   :


The drums would be standing on end, and dark colored, so the low-angle
winter sun would shine on the sides of the drums. If the south bench
were close to the south wall of the greenhouse, with no usable aisle
on the south side of the bench, because the curved greenhouse wall has
low headroom close to the wall, one might as well put some white poly
film on the ground to the south of the south bench, to reflect about
50% more sun onto the drums. There might be a double row of drums under
the 4' bench, about 96 drums altogether... It would make sense to use
an 8' width of white poly, extending it under the drums to keep them
from rusting, if they are steel drums in contact with the ground.

Ordinary steel drums last about 20 years, when filled with water, when
kept dry on the outside, I think. Adding a quart of oil to each drum,
along with the 450 pounds of water, increases the lifetime, I've read.
Food-grade drums, that contain things like molasses, usually have a plastic
coating on the inside of the steel, which should further increase the drum
lifetime, when the drum is filled with water. An empty drum weighs about
15 or 20 pounds. Plastic drums should last a long time, but they are harder
to come by than steel drums. One might have to pay $5 for a plastic drum...

An average day in the month of January, in the Philadelphia area, has
an outdoor temperature of about 30 degrees F, and on an average January day,
about 1000 Btu/ft^2 of solar energy falls on a south-facing vertical surface.
Each pair of drums under the south bench has a south-facing area of about
2' wide x 3' tall, ie 6 ft^2, so, ignoring the sun that is absorbed by the
plastic glazing, each pair of drums receives a daily solar input of about

Ein = 6 ft^2 x 1000 Btu = 6000 Btu.

Each drum has a surface area of about 25 ft^2, so if the average
day/night greenhouse temperature is, say 60F, each pair of drums
that contains water at a temperature of Td, higher than 60F, loses
energy to the greenhouse over 24 hours, as the water cools off. The
energy lost over 24 hours would be about

Eout = 24 hours x (Td - 60) x 50 ft^2/R1 = 1200 (Td - 60).

If Ein = Eout, over 24 hours,

Td = 60 + 6000/1200 = 65F. Not very warm...

But suppose we put a 12' poly film skirt around the drums, over the top
of the drums (and under the bench itself) and down the sides to the ground.
This would increase the R-value and decrease the surface area of the "solar
closet" under the bench...

In that case, the amount of heat that would leave the drums over a day
would be about

Eout = 24 (Td - 60) x (2 x 3 + 2 x 4 + 2 x 3) ft^2/R2 = 480 (Td-60),

so if Ein = Eout, again, Td = 60 + 6000/480 = 72.5F. Better...

Suppose we put some insulation on top of the drums, under the poly film.
This will help keep the plant roots from cooking, as they sit in peat
moss or pots on top of the benches, and make the drum water warmer. If
the tops of the drums are covered with 2" of beadboard, or the benches
are filled with mostly dry vermiculite or peat moss, most of the heat
loss of the drums will be through the poly film side walls of the area
under the bench, so Eout will be approximately

Eout = 24 (Td - 60) x (2 x 3 + 0 + 2 x 3) ft^2/R2 = 288 (Td - 60), 

so if Ein = Eout, again, Td = 60 + 6000/288 = 80.8F. Better...

We are trying to increase the drumwater temperature here, in order to
better provide overnight heat, at least, and ideally, enough heat for
a few cloudy days in a row.

According to Roger's estimate, a 30 x 96' greenhouse, with a double poly
inflated cover, would need about 3600 Btu/hour/degree F to stay warm
inside in the winter, with no sun, or about 86.4K Btu/degree day. So
in the Philadelphia area, a 96' x 30' greenhouse would need a daily
heat input of about

Eday = 86.4K x (50F - 30F) = 1.72 million Btu/day

to stay at 50F inside on an average day, with no sun. Call it 2 million
Btu/day, for a round number. This is the heat equivalent of about 20 gallons
of oil, as burned in an old oil burner... If the 80.8 degree water in
the 55 gallon drums above could be cooled quickly enough to heat the
greenhouse to 50 degrees F for a while, the useful stored heat in the
drum water might be about  

Estored = (80.8F - 60F) x 96 x 450 lb/drum = 900K Btu, 

which is enough heat to keep the greenhouse warm for about a half-day
without sun, eg overnight...

By the way, in some cases, commercial greenhouses use very efficient,
internally-vented gas heaters, called CO2 generators, as well as externally
vented heaters, to raise the CO2 level inside the greenhouse, which plants
like a lot :-) Here is some text from the Stuppy Greenhouse catalog:

   Plants must absorb carbon dioxide in combination with water, soil
   nutrients and sunlight, to produce the sugars vital for growth. A
   shortage of any of these requirements will retard the growing process.
   Normally there are approximately 300 parts per million of CO2 in the
   atmosphere; when this level is increased to over 1,000 ppm, it results
   in higher production and better plant quality. The Johnson Generator 
   [18" diameter x 20" high, 25 pounds, 20-60K Btu/hr, 8 lb CO2/hr, $465]
   provides up to 1,500 ppm per unit in an average 24' x 200' greenhouse or
   an equivalent 50,000 cu. ft. volume, **based on one air change per hour**.
   [Emphasis mine.]

   Nighttime levels in a greenhouse range generally from 400 to 500 ppm 
   due to plant respiration. Shortly after sunrise this level will drop
   to normal atmosphere (300 ppm) due to the plant using the early light
   to start photosynthesis. After 3-4 hours of early morning sunlight,
   the CO2 level can drop to around 100-150 ppm. Then growth is practically
   stopped. Supplemental CO2 added during this period can substantially
   increase your plant and flower production. By adding CO2, especially
   during the winter months when greenhouse ventilators are closed and
   when low CO2 concentration becomes a limiting factor in growth, users
   are obtaining yield and bloom quality comparable to that which is normally
   associated with spring and summer conditions...

Reducing the use of ventilation to the outside to cool the greenhouse
on sunny winter days also helps increase the retention of CO2, if a
CO2 generator is used, as noted above. (I think a CO2 generator would
make a fine backup heater in a solar greenhouse.) Reducing sunny-day
winter ventilation also helps in maintaining a high humidity level in
the greenhouse. It takes about 1000 Btu to evaporate a pound of water,
so the less we vent humidity to the outside, in the winter, the less
water and energy are needed to keep the greenhouse humidity level high.

Preheating water is also sometimes a good thing, in a greenhouse.
Plants don't like being watered with very cold water. Roger's tomato
greenhouse consumed about 400 gallons of water a day, and he preheated
this water by running it through a long 1/2" poly pipe along one wall.
A poly pipe for water heating might also run under the south bench... 

What's next? Well, we could put 55 gallon drums full of water under
all the benches, since 55 gallon drums are free, and the benches need
some support anyway. This would help keep the greenhouse from freezing,
at least, but it seems like just doing this would not add a lot to the
useful stored heat, because the new drums would have a drumwater temperature
of 60F (or less, since they would be near the floor), so they wouldn't
do much to help the greenhouse stay warm at 50F, which is not much
cooler than 60F. With a 10F difference between the water temp in the new
drums and the greenhouse temp, there will not be much heatflow from the
drums to the greenhouse air, at night. Still, the idea of that much free
thermal mass is appealing: 300 more drums, each containing about 500 pounds
of water, with a temperature fluctuation of 10 degrees F, will store about
10F x 300 x 500 = 1,500,000 Btu of heat, about 15 gallons of oil's worth,
almost one cloudy day's worth for the greenhouse...

About now, we might ask, if this is so easy to do, why don't more commercial
growers do it? I think the answer is that they haven't thought much about it.
They are more interested in plants and beautiful flowers than Btus. A lot of
them don't even have engineering degrees :-) Some of them come from families
who have grown plants in greenhouses for five generations, and they know what
works, and they are not too inclined to experiment with anything new, which
might cost them an entire crop of dead plants, or poinsettias that are
practically worthless, because they bloom the week AFTER Christmas...

So they just fire up the old heaters and pass on the costs to their customers,
making their lives simpler, and we pay high prices if we want to eat fresh,
red tomatoes or oranges in December. (One of the reasons that Roger Williams
got interested in building a tomato greenhouse, in northern Canada, was that a
couple of children in the small town where he lived got scurvy in wintertime.)

So, where were we? Oh yes, now we have lots of thermal mass, but it's
not very warm yet, and the heat transfer rate is too low. Considering only
the drums under the south bench, with 600 ft^2 of drum closet surface area,
losing heat at 1 Btu/hr/F, if the drums were heating the greenhouse entirely,
at 60F inside and 30F outside, the greenhouse will need 108K Btu/hr, so the
drums would have to be 108K/600/R2 = 360 degrees warmer than the greenhouse.

Hmmm. Something is wrong here... The heat transfer rate is too low.

If there were a motorized damper in the closet, to let some air flow past
the drum surfaces, the heat transfer area would increase to 2400 ft^2,
with an R-value of about 1, so the drums would need to be about

108K/2400 = 45 degrees

warmer than the greenhouse. This is the right direction, but it is still
too large a temperature difference... If the air were moving at 10 mph, the
R-value of the air film at the drum surface would decrease to about

1/(1+ 10/2) = 1/6,

so the temperature drop would decrease to about 7 degrees, which is OK.
I guess the bench needs a fan underneath, now, at one end, with a gravity
damper at the other end now, not a motorized one. In this case, one might
stagger the drums a bit, to allow air to flow more freely between them,
down under the length of the bench. This might reduce the number of drums
in a row from 96 to 90... That's OK.

How big a fan? 108K Btu/hr at, say, another 10F drop makes the fan want to be
about 10K cfm. 10 mph is 880 fpm, so the effective cross sectional area
for airflow, under the bench, wants to be about 10K/880 = 10 ft^2.

This could be done by placing the drums about 3' apart, but that would
make the bench 7' wide, so it is probably better to stagger the drums,
and make them about 6" apart, at their closest points, with a 3" air gap
between the outside walls of the drums and the poly sidewalls. A small
horizontal air gap, or some air leaks along the base of the poly skirt,
near the ground, would also make air flow more easily. In this case, the
airspeed might be about 3000 fpm at 10,000 cfm, but the air would actually
flow more slowly at a lower cfm, which would depend on the characteristics
of the fan. 

E. C. Geiger at (800) 4GEIGER or (215) 256-6511 sells a 1/3 HP, 431 rpm,
36" diameter, two-speed, $700 fan, rated at 10,300 cfm free air delivery, 
and 8800 cfm at 0.10" of water. The stock number is DC36F, on page 118
of their 1995 catalog. 

At this point, one wonders if the fan should be near the top of the house,
to push warm air from the peak down a diagonal poly duct along one endwall,
thru the length of the "solar closet" under the bed...

The sun that comes into the greenhouse in January, at a low angle,
comes thru "the south wall," an effective area of the length of the
greenhouse x the height. In the Philadelphia area, I would figure this as

1000 Btu/day x 96' x 13' = 1.25 M Btu/day,

for a 30' Geiger "Northerner" greenhouse, with a peak height of about 13',
on an average January day. At 86.4K Btu/dd, the greenhouse would lose about

(60-30)x86.4K = 2.6 M Btu/day, on an average Jan day...

So if the greenhouse were entirely solar heated, the average (day and night)
indoor temp would only be about 30 +1.25/2.92x30 = 44F. But if there were
snow or a white piece of poly or perhaps a shallow reflecting pool on the
ground, in front of the south wall, outside the greenhouse, that would roughly
double the solar input, making it close to 100% solar-heated in January...

Here is one possible "final" version of a commercial solar greenhouse:

                    Greenhouse  96' long x 30' wide
          :                                                 :
          :                                                 :
        < A                                                 : 
        < P                       A                         A <
        < R                                                 : 
          :               white poly film                   :

                          reflecting pool


Looking at this from the south side, not to scale:

          :                                                 :
        < DAMPER Poly film ductP<==FAN                    FAN <
          :  P                 P                          P :
          :  up               down                       up : 

The central fan would be the main one used on a sunny winter day. The east
fan would be used for extra cooling capacity in the summer, if needed.

The central fan would push warm air from the peak of the greenhouse into
a 30" poly film duct near the ridge that went to a motorized damper at the
west end of the greenhouse. In external cooling mode, this air would pass
through the poly duct and out the motorized damper, which would be open.

Another duct would run north and south from the central fan, to connect to
4 down ducts that went into the tops of the benches at the center of the
greenhouse. This air would emerge from the ends of the benches into up-ducts,
which would return the air to the peak of the greenhouse, allowing the
peak of the greenhouse to be warmer than the air in the bench area. One
of the down ducts might contain a CO2 generator...

The space under the benches would look like this, from the east:

     |     4'     |
   _ :___bench____:
        polyfilm            In winter, the drums would be kept at, say,
       pfoamboardp          130F, weatehr permitting. This would store
   3'  odrum drumo          about 200K lbs x (130F-80F) = 10 million BTU,
       ldrum druml          or 5 cloudy day's worth of useful heat.
       ydrum drumy
white poly film...---------

And the space under the bench might look like this, from the top:

     :           :          In the summer, the central fan could run up to 
     : drum      :          24 hours a day, partially storing summer heat
     :           :          in the drums during the day, and cooling the
     :      drum :          400 drums and the greenhouse at night. In the
     :           :          Philadelphia area, with 50% shadecloth, the 
     : drum      :          summer sun input in June would be about 1000 BTU
                            per square foot per day, or 
			    3000 Btu x 1000 Btu/ft^2 = 3 million Btu/day.

If all of this daily summer sun's heat (100%, with NO external ventilation
during the day) were stored in the 200K pounds of drumwater, this would raise
the water temperature by 15 degrees F.

If the drumwater were at 130F in the winter, and the house were at 60F,
the plant roots, in peat moss, would be at a temperature Tr, where
Tr is about 60 + (130-60)/R11 = 66F. Water evaporating from the plants
would actually keep the roots a bit cooler than this...

Roger Williams suggests that in a new greenhouse, solar heating can be done
by storing energy in the earth below, eg by burying 10" plastic pipes, one
on each side of each bench, and blowing air through them, without using all
of these drums. The success of this technique for heating and cooling a
greenhouse would depend on the height of the water table and the nature of
water flow underground...

I've heard that solar greenhouses are used extensively in Japan, where
energy is more expensive. Does anyone know anything more about this?