The most popular forms of solar heating seem uneconomical these days, with
high costs and low solar fractions (20% for some FSEC-certified water heaters),
so they seem to require customers who are concerned about climate change,
among other things. If solar house heating were to simply cost less than
other forms of house heating, more people might use it...
After hearing lots of doubt that low-cost solar house heating can work with
a high solar fraction outside of the Southwest from local architects and
reporters and others who read SBIC guidelines that say houses in Phila can only
be 60% solar-heated at best and seeing lots of "solar houses" that are only
30-50% solar-heated vs those "with no other form of heat," I think we should
dispatch Deployable Doubt Dispellers ("D-cubes") to regional Infestations of
Doubt, eg unboltable 8' cubes that stay 70 F for a few cold cloudy days, with
little windows so people can peer in to see big dial thermometers inside.
Lots of people (eg AIA and SBIC and SBSE) seem to have forgotten that things
like this can be engineered, notwithstanding high-school physics and houses by
PE Norman Saunders, who calculates needs for "purchased heat" with Gaussian
weather stats in the same way that other engineers calculate 100-year floods:
The ASHRAE 1954 Handbook gives the 1% and 2 1/2% temperatures for Boston as
0 and 8 F (-17.8 and -13.4C.) For -15C, the holding time for our houses is
three days, giving a standard deviation of 7.8 Kelvin degrees. In December,
the 1% temperature is 2.1 standard deviations down... and so can be expected
to occur about once every 4.5 years. This suggests the need to purchase heat
[or wear a sweater :-] in December once in 35 years.
NREL says 1000 Btu/ft^2 of sun falls on a south wall on an average 30 F day
with a 38 F max in Philadelphia. A direct-gain "mass and glass" D-cube (like
most solar houses) with an 8'x8' R4 south window with 50% solar transmission
could be 80 F on an average day if 0.5x8'x8'x1000 = 24h(80-30)G Btu, with
a thermal conductance G = 26.7 Btu/h-F max. Subtracting 16 Btu/h-F for the
window leaves 26.7-16 = 10.7 Btu/h-F for other surfaces, eg 4x64ft^2/R28.5
= 9 for the ceiling and non-south walls, with 2 2" polyiso foilboard "R13"
layers, counting R2.5 for the indoor foils, which makes G = 25. If the cube
temp falls to 70 F after 5 30 F cloudy days, 70 = 30+(80-30)e^(-120/RC) makes
time constant RC = -120/ln((70-30)/(80-30)) = 538 hours, so it needs thermal
mass C = RCxG = 538x25 = 13,450 Btu/F, eg 13450/4/8 = 420 nestable 9" square
x 13" tall 4-gallon ROPAK plastic water tubs stacked 7-high x 10-wide x 6-deep
in a 91" x 90" x 54" wall, over half the cube's volume.
An indirect-gain D-cube might have an 8'x8' isolated sunspace or air heater
with R1 glazing with 90% solar transmission (eg $1/ft^2 20-year corrugated
Dynaglas polycarbonate greenhouse roofing) with one-way plastic film dampers
and half the insulation, eg one 2" layer of foil-faced board for walls and
ceiling with G = 5x64/R15.5 = 21 Btu/h-F. If the air heate gains 0.9x8x8x1000
= 57600 Btu = 6h(T-30)64ft^2/R1 [the daytime sunspace loss] +18h(T-30)64/R16.5
[the night loss from the air heater] +24h(T-30)4x64/R15.5 [the 24-hour loss
for the rest of the cube], T = 98 F, If we keep it 80 F max with ventilation,
C = RCxG = 11,107 Btu/F, eg 347 water tubs stacked 7-high x 10-wide x 5-deep
in a 91" x 90" x 45" wall, about half the cube's volume.
An isolated-store D-cube might have less mass with a higher temp swing: 8'x8'
of Dynaglas solar siding might be T (F) inside for 6 hours on a Jan day if
57600 = 6h(T-30)8x8/R1+18h(T-30)8x8/R16.5+24h(70-30)4x64/R15.5, so T = 122 F.
If the cube needs (70-30)21 = 840 Btu/h and each stacked tub has 3.25 ft^2 of
surface and 5 Btu/h-F of slow-moving-air to water conductance, N tubs with
Tmin = 70+840/(5N) water can keep the cube 70 F. With an average water temp
Ta = (122+Tmin)/2, the tubs lose Et = 120h(Ta-30)64ft^2/R16.5 Btu over 5 days.
If the rest loses Ec = 120h(70-30)4x64/15.5 = 79,277 and 32N(122-Tmin) = Et+Ec,
N = 69 and Tmin = 72, so we might have a 7-high x 10-wide x 1-deep 91" x 90"
x 9" tub wall, about 10% of the cube volume, with a 1" foilboard wall between
the tubs and the living space, hinged at the top, with a passive Thermofor
vent arm to open it when the room cools to 70 F.
An active D-cube might have 32' of fin-tube pipe below a 2" foamboard ceiling
and walls and a 6'x6'x2'-deep unpressurized EPDM-lined heat storage tank on
the ground with a PV-powered 10 watt pump for heat collection. If 57,600 Btu
= 6h(Ts-30)64ft^2/R1 [daytime glazing loss] + 6h(70-30)4x64/15.5 [daytime loss
from the rest of the cube = 14,865 Btu] + 6h(Ts-Tw)160 [stored heat], then
the sunspace temp Ts = 48.5 + 0.714Tw, so Tw = 115 and Ts = 131. We might
distribute heat from the tank with more fin-tube under the floor or a damper
and Thermofor arm or a Honeywell 6161B1000 damper actuator (which only uses
2 watts when moving) to allow warm air to rise out of the floor... 84 ft^2
of tank surface with 5x84 = 420 Btu/h-F of slow-moving airfilm conductance
could supply 840 Btu/h at 70+ 840/420 = 72 F. Keeping the cube 70 F for 8
hours and 50 F for 16 hours on a cloudy day takes (8h(70-30)+16(50-30))21
= 13440 Btu, so the tank might store heat for 2x6x6x62.33(115-72)/13440 = 14
30 F cloudy days in a row. With a $60 1"x300' PE pipe coil in the tank, hot
water for outdoor showers could be an option for certain climates and seasons.
We might deploy D-cubes to Rifton and Port Jervis and Harlem, NY, Kempton and
Phila and Pottstown and Bryn Mawr, PA, a wintertime MREF or ASES conference
or DOE Solar Decathlon, the Chicago Museum of Science and Industry, and lots
of YMCAs and high schools.