a bubblewall house SOLAR.KnowledgePublications.com

a bubblewall house
6 jun 1997
consider a 12' cube with a south bubblewall sandwich containing some drums
full of water, as in the report "liquid foam--greenhouse insulation..." by
john f. groh & t. l. thompson, u arizona, ca 1974, the subject of us patent
no. 3,672,184, now expired, i guess, based on an earlier swedish technique
for insulating shop windows at night. groh, thompson and otho wells at u new
hampshire insulated commercial plastic film greenhouses by filling up the
curved walls with tiny cold soap bubbles, which have r-values comparable to
fiberglass, eg r11 for 2" of foam with 1/16" diameter bubbles at a mean
temperature of 50 f, by measurement. it worked, although they had practical
problems with foam generation, leaks, overflows, voids, and freezing bubbles.

smaller and colder bubbles are better. this approximate formula fits their
data: r = (0.372/d+5.05)(-0.318t+33.3)/17.4 = 0.71/d -0.0068t/d -0.092t +9.7,
where r is the us r-value, d is the bubble diameter in inches, and t is the
mean foam temperature in degrees f, so it looks like a 6" r20 wall with 30 f
air on one side and 130 f on the other requires d = 0.166/(r-2.34) = 0.0094",
ie 106 bubbles per inch, or 240 micron bubbles. an r40 wall might have 112
micron bubbles. i wonder what the r-value is of enviro-bubble floatation's
(superfroth@aol.com) 5 micron bubbles. when does this trend end? a solid
water wall (d=0) would have a much lower thermal resistance... 

the cube might have two vertical greenhouse polyethelyene film pillows on
either side of 24 dark plastic drums full of water, stacked 6 wide by 4 high,
with a thermal capacitance c = 24x55x8 = 10,560 btu/f. the outer pillow might
be filled with bubbles at night and inflated with air during the day in the
winter, and vice-versa in summertime. the inner one might be deflated when
the rest of the cube needs heat, and filled with bubbles at night, or when 
the rest of cube is warm enough.

let's try to find the average drum temperature t on an average january day in
philadelphia, with 1,000 btu/ft^2 of sun falling on a south wall and 30 f
outdoor temperature. the energy that flows into the cube would be about
1,000x0.8x12x12 = 115k btu (or about 20% more, with a frozen pond to the
south.) with r20 non-south walls and an r20 outer bubblewall, the energy that
flows out of the cube would be something like

    6h(t-30)12x12/r1.2    = 720t   - 21.6k   outer bubblewall, daytime
  +18h(t-30)12x12/r20     = 129.6t -  3.9k   outer bubblewall, at night
  +24h(70-30)12x12x5/r20  =        + 34.6k   rest of the cube, all day

  = 849.6t + 9.1k btu.

if the energy that flows into the cube is equal to the energy that flows out,
on an average day, 115k = 849.6t - 9.1k, so t = 146f. that seems too high,
without a selective surface, and it makes the mean foam temperature higher 
than 80 f. let's assume the average drum temperature is only 130 f. 

the 5 non-south sides of the cube lose about 24(70-30)144x5/r20 = 34k btu/day.
in cloudy weather, with the outer bubblewall filled all the time, the drums
might have an average temperature of 100 f, so the south wall might lose about
24(100-30)144/r20 = 12k btu/day, for a total of 46k btu/day, cooling by 46k/c
= 4.4 f per day, so the cube might stay warm for about (130-80)/4.4 = 11 days
with no sun at all. bigger cubes could stay warmer longer.

nick