Une serre pour Sylvain / A greenhouse for Sylvain

Suppose Sylvain's Quebec greenhouse were a sort of quonset hut, a commercial
plastic film structure, with curved galvanized pipes,  that looks like this
>from  the top:

                 100'                                    .
 ---------------------------------------                 .
|                              |        |            .   .
|                An            | solar  |           .    .
|          terre noiratre      | closet?|          .     .
| . . . . . . . . . . . . . . . --------|   30'    . 13' .
|                              .        |          .     .
|                As            .  white?|           .    .
|                              .        |            .   .
 ---------------------------------------    East-->      .
.                                       .                .
.                                       .                .
.                                       .                .
.        shallow reflecting pool        .                .
.                                       .                .
.                                       .                .
.                                       .                .
. . . . . . . . . . . . . . . . . . . . .                .

         view from the top                       view from the east

On an average December day in Montreal, the outdoor temperature is about
21 F and about 540 Btu/ft^2/day of sun falls on a south-facing wall. This
is a not an easy climate for passive solar heating...

The reflectors above will augment the sun by about 50%, so the amount of sun
that falls on the south side of the greenhouse will be about 810 Btu/ft^2/day.
The reflecting pool also keeps vegetation from growing on the south side,
which would block the sun. It might be made from a single piece of EPDM
rubber, 20' wide x 100' long, costing 30 cents/ft^2. If the glazing were
glass, the pool would keep people with lawn mowers at some distance, but
better still to avoid mowing at all... And the reflecting pool can collect
and store rainwater for use in the greenhouse.

How much heat does this greenhouse need to stay warm on an average day in
December? Suppose it is a commercial greenhouse, made with curved galvanized
pipes on 4' centers and two thin layers of polyethylene film, inflated with
a very small blower. This would have a US R-value of 1.2. The curved surface
is approximately the shape of a cylinder sliced in half lengthwise, so it has
a total area of about PiDL/2 = Pi(30')100'/2 = 5,000 ft^2, divided into north
and south roof/walls with An = As = 2,500 ft^2. Let's ignore the east and west
endwalls and the floor for now. Suppose the temperature inside the greenhouse
is 68 F, 24 hours a day.

Then the amount of heat that is needed to keep it warm on an average
December day is Eout = 24 (68-21) 5,000 ft^2/R1.2 = 4.7 million Btu/day.
Equivalent to about 40 gallons of oil a day. That's one reason people
don't build real houses out of plastic film.

How much solar heat comes into this greenhouse on an average December day?
According to page 64 of the $35 NRAES-33 Greenhouse Engineering book (3rd
revision, August 1994, written by Robert A. Aldrich and John W. Bartok, Jr.,
published by the Northeast Regional Agricultural Engineering Service,
152 Riley-Robb Hall, Cooperative Extension, Ithaca, NY, 14853-5701,
(607) 255-7654), poly film has a solar transmission of .92, so two layers
should have a solar transmission of about .92 x .92 = .85, and the south-
facing wall area of the greenhouse is about 13' high x 100' long, so the
total solar energy that enters the greenhouse on an average December day
in Montreal will be Ein = .85 x 810 Btu/ft^2/day x 1300 ft^2 = 891,000 Btu,
about 20% of what is needed to keep this greenhouse warm...

Hmmm. How about putting some insulation in the north wall? Suppose we somehow
slip 3 1/2" of R13 foil-faced fiberglass insulation inside the double poly
film pillow of the north wall, with the foil side facing into the greenhouse.
That way, sun is reflected down from the north side, so the plants will grow
fairly straight rather than always leaning towards the sun ("heliotropism.")
We could paint the outside of the north plastic film white to make it last
a long time. Then the daily heat requirement of the greenhouse becomes

Eout = 24 (68-21) (2500/R1.2 + 2500/R13) = 1128 (2083 + 192)
     = 1128 (2275) = 2.6 million Btu/day,

equivalent to about 22 gallons of oil a day. Better, but still about 3 times
more than the available solar heat.

Hmmm. Let's lower the temperature of the greenhouse at night, to, say 48 F.
Then, if the December day is only 6 hours long,

Eout = (6 (68-21) + 18 (48-21)) 2275 = (282 + 486) 2275 = 1.75 million Btu.

Better, but still about twice as much heat as the sun will supply.

Well, how about a beadwall? (Or perhaps a soap bubble wall?)

If we screw on 1x2 wood strips to the inside and outside curves of the 1.66"
galvanized pipes, and cover the strips with a layer of thin flat polycarbonate
plastic on both sides ($5,000 worth, but it should last at least 10 years),
with some butyl tape and a 1/2" x 1/8" aluminum cap strip on the outside,
and fill the 3.17" cavity it with R3.5/inch polystyrene beads at night, that
will give the south wall an R-value of 3.17 x 3.5 = 11 at night, so

Eout = 6 (68-21) 2500/R1.2 = 588K Btu  south wall, daytime
    + 18 (48-21) 2500/R11  = 110K Btu  south wall, nightime
    +  6 (68-21) 2500/R13  =  54K Btu  north wall, daytime
    + 18 (48-21) 2500/R13  =  93K Btu  north wall, nightime,
     =                       845K Btu/day.

OK. This is slightly less than the average solar input of 891K, so we
can probably make this work. There is still a large heat loss through the
south wall during the day, but this IS a greenhouse... If we somehow made
it taller, we could collect more solar heat in the winter. Or perhaps we
should leave the beadwall closed and turn on some grow-lights when the sun
is dim, or before dawn or after dark in December, like Rudy Behrens does,
ie 10 watts per square foot of high pressure sodium lights. This would add
some backup heat. And in order to actually grow, vs just stay dormant in
winter, plants need a longer day. With this sort of backup heat, we could
probably keep this greenhouse at 68 F, 24 hours a day, which is desirable
for some plants, like poinsettias. Keeping them warmer at night makes them
round and bushy, vs thin and tall. Maybe this is the thing to do in Quebec,
where hydroelectric power is not too expensive...

BTW, I wonder if we could make this a bubble wall instead of a beadwall?
Steve Baer says he's tried this with solar collectors, and bubbles transmit
lots of light, in fact they even reduce the reflective losses from the glazing,
by serving as an index matching fluid :-) so perhaps the bubble wall should
stay in place during the day, too...

But Steve also says the walls of the bubbles themselves are so thin that they
don't block much radiation heat loss, so the bubbles wouldn't be great
insulators for a solar collector, with a high temperature inside. But it
seems to me that they might work well for a greenhouse, with a low temperature
inside... Steve said someone else built a bubble wall beadwall, and published
the results, and Steve was surprised at how good an insulator their bubbles
were. A couple of possible problems with this approach are that the bubbles
might collapse on themselves by gravity if the bubble column were more than
a few feet tall, or that we might have to pump so much air and bubble water
through the glazing that that mass flow would collect and transfer the heat
>from  the inside glazing to the outside glazing, making the bubble wall a poor
insulator, but Steve didn't think those would be serious problems.

Perhaps the bubble wall thickness and insulating value, especially for
radiation insulation, depends on the composition of the soapy water used to
make them. There may be a happy medium, some nice combination of soap,
detergent, glycerin, oils, dyes, etc., that will make for good shortwave
solar transmission and poor longwave IR heat transmission, a good "greenhouse
effect," or high-pass filtering effect, like glass. Perhaps the bubble liquid
should have one composition by day and another by night. Transparent bubbles
by day and opaque bubbles by night. That would not be hard to arrange. This
would be an interesting science fair project, that would not require a lot
of equipment to do.

Making a bubble wall with polyethylene film might be a problem as far as
the film life goes. Consider this quote from the CT Film application note
on page I-8 of the 95-96 Stuppy Greenhouse catalog ((800) 877-5025):

  Soap or detergent should not be left on film. If film is washed with
  soap or detergent, it is recommended that immediately thereafter the
  film be well but carefully rinsed with water. Do not use soap
  containing "Pine Oil" or other solvents.

Is this only a problem when the soapy liquid dries, or does the plastic
degrade if it is always wet? I don't know. CT Film's sales engineer
Warren Manning at (517) 423-675 may know, or he may know a chemist who
knows. It would be very exciting (to me :-) if someone could make a good
plastic film bubblewall. "Doing more with less," as Bucky Fuller used to
say. "Pathologically frugal," my friends might say. A 4 mil thick x
32' wide x 100' long roll of polyethylene greenhouse film with a 3 year
guarantee costs about $140, about 4 cents per square foot. It is recyclable,
and if one uses alumimum extrusion clamps to attach it, only at the edges
of the structure (this costs about 60 cents per linear foot), changing the
plastic film every three years is not much harder than changing a bedsheet,
on a calm day :-)

BTW, Stuppy also sells a product called "Varishade 2," which turns white
when it is dry and transparent when it is damp. This is a permanent product,
meant to be painted on plastic films. That's another way to control solar
intensity and heat losses. Poly film has almost no "greenhouse effect,"
unlike glass. It loses a lot of heat by longwave IR transmission (77%.)

Beadwall inventor Dave Harrison suggests NOT making a beadwall out of
poly film, because if the film rips or tears, the 660 ft^3 of beads that
get loose may be very unpopular with the neighbors, altho they are
biodegradable, as I recall. Perhaps one could make a good beadwall with
a layer of polycarbonate film on the outside and a layer of polyethylene
film on the inside...

Anyhow, the beads cost on the order of $1/ft^3, and they would fill up
78 55 gallon drums at 8.4 ft^3 per drum, or a row along the south side,
about 50 drums deep by 2 drums high, and one would need 50 vacuum cleaner
motors to make all this work, if one followed the well-proven Beadwall
design. They would want to be sequenced in operation, because altho they
would only operate a few minutes per day, the motors use 7 Amps each...
A more interesting but difficult to design alternative would have only
one or two vacuum cleaner motors and some holes between the drums to
make a common bead store.

So, what would this finally look like? The greenhouse would be 6' off
the ground, sitting on top of a row of drums stacked 2-high, on each of
the long sides. The drums along the south side would be welded together
and filled with beads, or perhaps a lot less bubble water, and the north
side of the greenhouse would have a solar closet holding up the north
benches, consisting of drums stacked 2-high and perhaps 1-deep, ie 100
drums full of water, which would have a layer of fiberglass insulation
on the south side, then an airspace, then a 4' layer of polycarbonate
outside of that. This might also be fed with a fan that brings down some
warm air from an air heater above on a sunny day.

The aisle down the middle of the greenhouse might be filled with an air duct
made from 55 gallon drums welded together end to tend, laid sideways, with
their tops and bottoms removed, with dirt and gravel on top to make a smooth
and elevated walking surface, 4' below the benches. During the day, warm air
>from  the peak of the greenhouse might be blown down and through this duct
with one or two fans, to store that heat during the day.

The north benches might as well be supported by 55 gallon drums as well, in a
peninsular layout with another 200 drums, inside the bead drums. They would
add some desirable thermal mass to the greenhouse, but they wouldn't store
much heat, since they would be at the average greenhouse temperature.
Surrounding them with a polyethylene skirt and blowing some warm air down
>from  the peak of the greenhouse during the day would make them into a lower
temperature solar closet.

What would the average drumwater temperature have to be to store heat for
5 cloudy days in a row? We have 300 drums, ie about 150,000 pounds of water
here (vs the 200 drums full of water that David Boyer and I put into his
2,000 ft^2 poinsettia greenhouse last fall, near Philadelphia), so the drums
will store 150,000 Btu per degree F above the greenhouse temperature.

With the beadwall closed, at 48 F, 24 hours a day, the greenhouse needs

         24 hours (48-21) (2500/R11 + 2500/R13) = 272K Btu/day

to stay warmish inside, so the average drumwater temperature has to be
at least 48 + 5 x 272K/150K = 57 F. This seems very doable.

The outer layer of drums on the north and south sides should be insulated, eg
with 3 1/2" of fiberglass insulation, covered with poly film. The north
insulation should be dark on the outside, with an airspace under the glazing.