DIYbanter - whither the ice house?

Al Bundy wrote:

Procuring the ice is labor intensive...
There must be a better and cheaper way.

Using NREL's TMY2 ("Typical Meteorological Year") hourly weather data file
for Philadelphia, one of 239 US cities listed at

http://rredc.nrel.gov/solar/old_data/nsrdb/tmy2/

10 OPEN "tmycool" FOR INPUT AS #1
30 FOR H=1 TO 8760'read ambient temperature every hour for a year
40 LINE INPUT#1, S$'input a line from the TMY2 file
50 TEMP=VAL(MID$(S$,27,4))'ambient temp (F)
60 ATEMPT=ATEMPT+TEMP'accumulate average yearly temperature
70 IF TEMP 32 THEN FDHT=FDHT+32-TEMP'accumulate freezing degree-hours
80 IF TEMP 32 THEN MDHT=MDHT+TEMP-32'accumulate melting degree-hours
90 NEXT H
100 PRINT "average yearly temperatu";ATEMPT/8760
110 PRINT "freezing-degree days:";FDHT/24
120 PRINT "melting-degree days:";MDHT/24
130 CLOSE 1

this fine BASIC program yields:

average yearly temperatu 53.62591 (F),
freezing-degree days: 339.4584, and
melting-degree days: 8232.917,

like heating and cooling degree days (F) for 32 F houses. The number of
"freezing degree days" (FDD) is the total equivalent number of days in
which the outdoor temperature is less than freezing during a typical year,
times the difference in degrees between 32 F and the outdoor temperature,
on each of those days.

The FDD for a particular climate measures how easy it is to make ice. We
can freeze a pound of water with 144 Btu at 32 F, and a water surface has
a thermal conductance of about 1.5 Btu/h-F-ft^2 in still air, and a cubic
foot of water weighs about 64 pounds, ie 5 pounds per inch of depth, so
1 FDD can freeze an ice layer 24hx1.5/(144x5.3) = 0.05" thick, on top of
32 F water. In Philadelphia, we might freeze a layer of ice about 339x0.05
= 16" thick on some sort of specially-designed icemaking pond...

Melting degree-days determine how big an icehouse must be to stay cold
for a year. Bigger icehouses work better. An L' icecube surrounded by R40
insulation contains about 64L^3 pounds of water, ignoring the insulation
thickness, and initially stores about 9200L^3 Btu of coolth with a yearly
heat gain of about 24hx8200x6L^2/R40 Btu in Phila, for a minimum L = 3'.
A 5x5x5 = 125ft^3 icecube inside a 10' strawbale cube might stay partially
frozen all year, if it were first frozen solid and it only had to keep
itself and some vegetables cool, vs say, air-conditioning a nearby building.

Freezing a 5' cube takes about 5^3x64x144 = 1.15 million Btu. An A ft^2
roofpond can make 24x339x1.5A = 1.15 million, so A = 94 ft^2, eg a 10'x10'
salt water pond under IR-transparent polyethylene film in freezing weather.
Winter winds and night sky radiation would cool the pond more, as would
evaporation, without the poly film.

A low-power pump might flood the pond at night and let it flow back into
a tray above an 8'x8'x4' thick 256 ft^3 water wall inside a 10' cubical
outdoor fridge. The wall might be welded-wire mesh and 2x4 shelves under
plastic film 55 gallon drum liners, as Anna Edey used to store solar heat
in her Cape Cod Solviva greenhouse.

An air-antifreeze heat exchanger and fan might replace the roof pond...
Something like a $35 used 1984 Dodge Omni auto radiator with its 12V fan
or Magicaire's $150 2'x2' SHW 2347 heat exchanger, which can transfer
45K Btu/hour between 125 F water and 68 F air at about 800 Btu/h-F at
1400 cfm (like a window fan) with a low airflow resistance and an air
pressure drop equivalent to a 0.1" column of water. A 100 watt fan might
run 2.3million/(800x(32-25)) = 410 hours at 25 F to freeze the wall,
consuming 41 kWh/year, ie about $4 at 10 cents/kWh, while providing
256 ft^3 of year-round refrigerated space inside the 512 ft^3 box, like
17 home fridges. Is it time for neighborhood icehouses yet? Fill them
with beer to facilitate apres-lawn-mowing-male-bonding? And what will
we do about lawns, after the oil runs out?

Nick