On 2005-10-24, wrote:

Staple black aluminum window screen into the south wall cavities to
compress 3.5" R11 batts to 1.5" R6.2 batts and screw clear
corrugated polycarbonate Dynaglas or Sun Tuf "solar siding" over the
wall to make a thermosyphoning solar air heater which gains 0.9x1050
= 945 Btu and loses 6h(70-48.7)1ft^2/R1 = 128 Btu by day and
18h(70-48.7)1ft^2/R6.2 = 62 by night, for a net gain of 945-128-62 =
755 Btu/ft^2-day, vs the original wall which loses
24h(70-48.7)1ft^2/R11 = 46 Btu/ft^2 on an average January day in
Berkeley.

Well, this is a nice idea, but I'm afraid that aesthetic concerns
preclude it. Unless there is some way to make this work with cedar
shingles on top? :-)

As to the computations, let's see. 1050 Btu is the daily insolation
per ft^2 in Berkeley on an average January day, and 0.9 is a
correction factor for vertical walls given the angle of the sun in
January, yes? My interior design temperature is 68F, not 70F, but
that's a small difference. Why is the heat loss resistivity during
the "day" R1 instead of R6.2? 48.7F must be the average January
temperature (average low temperature?), seems like it would be more
accurate to use one temperature during the 6 hour "day" and another
during the 18 hour "night". Not that it really matters, as the heat
loss is an order of magnitude less than the solar heat gain.

Did I get all that right? Thanks for the education. Oh, and why does
the thermosiphon require a 2" air space? I assume the 1050 Btus/day
is an average of sunny and cloudy days, how much insolation can I
expect when it is raining? It does that alot in Berkeley in January,
and so I imagine this heat source would be very uneven day to day.

Cheers, Wayne