re: solar hot water heater for home: give me a push...
25 oct 2002
anthony matonak wrote:
>even in the dead of winter most places (below the arctic circle) get
>an average of a few hours of sunlight every day...
nrel's direct beam sun on a 2-axis tracker averages show phila gets
2.4 kwh/m^2-day in december. burlington gets 1.3. binghamton gets 1.2.
portland gets 0.9. seattle gets 0.7... 0.7 hours per day or 1.4 every
other day or 4.9 hours 1 day a week or 1 sunny day every 2 weeks or
2 sunny days at the end of the month?
>say you were to use a heliostat to focus the light from a 4 by 4 meter
>square mirror array onto your hot water heater through a window.
aim your 13.1'x13.1' mirror into a 13.1'x13.1' north bubblewall window
onto your 13.1'x13.1' water heater target? :-) duane says a single
heliostat mirror can't concentrate. he's made 8'x8' (6 m^2) heliostats.
still large for a window. we might use 12 4'x4' heliostats...
>if you get, say, the equivalent of three hours of full sunlight at
>your location on average then that's 48 kwh of energy. with about
>3414 btu per kwh that makes it about 163,800 btu per day on average.
a typical house might need 400 btu/h-f of heat, eg 24h(65-30)400 = 336k
btu/day in december in phila. a 4'x4' heliostat with a 90% reflector and
a window with 80% solar transmission might provide 0.8x4x4/10.76x2.4x3414
= 7800 btu/day, or 94k btu for 12 of them, ie 28% of the house heating.
maybe heliostats are more useful for water heating or lighting or cloudy
day heat stores than average-day house heating in december. we might say
they aren't very useful for lighting, if they only work 2 days per month.
>...mirrors and trackers aren't free but you can get the electronics
>from duane at http://www.redrok.com/main.htm pretty cheap and he has
>lots of information for building your own heliostats from commonly
the ones that look at a reflected sun image have problems. his latest
version that looks directly at the sun seems promising... he says he's
sending me one "soon" :-)
>you might get a week or two without sunlight sometimes but a large
>enough tank that is well insulated will help with that...
we need 5x336k = 1176k btu for 5 cloudy days, eg 1176k/(180-80) = 11760
pounds or 1470 gallons or 184 ft^3 of water cooling from 180 to 80 f.
a 256 ft^3 4'x8'x8' plywood box might do, with 2 layers of r1 glazing
over an 8'x8' side for a target.
if it's indoors, it doesn't need as much insulation, and its waste heat
warms the house on an average day. r30 insulation (r19 fiberglass plus
1" foil-faced foamboard) on 5 sides would make its thermal conductance
64ft^2/r2 for the target plus 192/30 for the rest, ie 38.4 btu/h-f, so
it would lose 24h(180-65)38.4 = 106k btu of heat on an average day. hmmm.
maybe we need insulation on all sides, and a pump, and a ceiling target,
an easier aiming geometry that could provide more useful light. then the
box loses 24h(180-65)8.5 = 23.5k btu/day, about 3 heliostat's worth...
on an average day, we might heat the house with a 300 ft^2 south bubblewall
window shining onto a hot vented slab, or trickle low-temp water between
a dark metal roof and a polycarbonate cover, and heat the house with
a hydronic slab.
we need to store about 18h/24h(336k) = 252k btu of overnight heat in
december. with a 10 f day/night temp swing, the hydronic slab needs
25.2k btu/f of heat capacitance. a cubic foot of concrete stores about
25 btu/f, so a 1000 ft^2 slab would have to be about 1' thick, or 4"
thick, if we store 200 gallons of low-temp water...