|
|
re: composite solar system question
26 apr 2001
anthony matonak wrote:
>...experiments with mirrors and average panels proved disastrous enough
>in the past that they specifically state that all warranties are void
>the moment you use mirrors or other concentration.
there are many ways to do something badly :-)
>...i could well imagine some kind of lightweight aluminum frame supporting
>mylar film attached around it's edges only and stretched fairly flat...
you might imagine a commercial plastic film greenhouse sliced in half
lengthwise, with parabolic bows on 4' centers and mylar greased to the
underside of a large piece of 5 cent/ft^2 4-year uv polyethylene film
with the aluminized side towards the film to protect it from moisture.
poly film is stretched over greenhouse bows with aluminum extrusions
costing about $1/ft, along the ground and at the endwalls... 1x3s with
deck screws are frugaler. the poly might last 10 or 15 years if it's
facing north and whitewashed or covered with shadecloth on the outside.
>> >...would there be ordinary conditions which would cause overheating?
>> >say 100% relative humidity on a hot august day with 100 f air temp,
>> i wouldn't want to live in a place like that...
>i wouldn't want to live in a place like that either but i do know such
>places exist...
the worst climate for evap cooling in the nrel book looks like key west,
with an average 84.3 f in august with w = 0.0185, ie 90.4% humidity and
pa = 29.921/(0.62198/w-1) = 0.92 "hg. a square foot of 140 f water would
lose about 100(6.2-0.92) = 528 btu/h there. not much different from phila.
we might make the sump a cooling lagoon with more surface, or float a
layer of poly film on top to reduce evaporation while still cooling all
night at about (140-84.3)1.5 = 84 btu/h-ft^2, if water consumption were
an issue.
>at what temperatures do panels start to lose lifespan significantly?
it seems safe to say that they will last a long time at 3 suns if we don't
exceed the noct temp (60 c, ie 140 f?) at 70 f and 250 btu/h-ft^2, 1 ft^2
of 15% panel needs to lose 213 btu/h. still air with no heat sink and both
sides cooling makes 213 = (t-70)2ft^2x1.5btu/h-f-ft^2, so t = 141 f.
>...a square 120 ft^2 panel measures almost 11 feet on a side and if it's
>pointing to either horizon it'll be mostly up and down. a pump which can
>only pump water a maximum of 9.5 feet isn't going to cut it all by itself
>even if it can water 7440 ft^2 of panels that only stand 3.3 feet tall...
so maybe we need a couple of sumps and pumps at different heights. i like
the idea of laying the panels flat on the ground in a pool with an overhead
linear parabolic reflector. nrel says phila gets 729 btu/ft^2 of beam sun
in january, when the average outdoor temp is 30.4 f. multiply 4x, remove
20% for the mylar and water surface, and 2333 btu/ft^2 (684 wh/ft^2) per
day falls on the panel surface, making 103 wh/day-ft^2 at 15%. a 4'x32'
long panel pool under a 32'x16' reflector might make 13 kwh/day in january.
a draindown pool with a cover might also collect about 254k btu of heat
and lose 3h(140-30)128x1.5 = 63k btu, for a net gain of 191k btu/day,
enough for lots of hot showers and some space heating.
>i imagine that you could set up a stationary array, with oversized
>mirrors, such that it'll get fairly good concentration through most
>of the day...
one version being a winston concentrator...
>...i don't think this will give the same kind of performance as a tracker
>that tracks the sun east/west...
agreed, altho there isn't much difference in wintertime.
>but it might give you three times the energy of a fixed
>panel without mirrors and it wouldn't require any curves.
collectors with winston curves (2 simple parabolas) have been well-proven
to collect about the same amount of sun as a fixed flat plates with equal
area and solar exposure. pvs can use 1/3 of the usual cell area, with cells
costing about 40x more than mirrors. concentrators also allow collecting
solar heat at a higher temperature with less thermal loss from a smaller
collecting surface, and less of that costly surface, vs a flat plate.
>discarding the tracker changes the economics significantly. 40 sq feet of
>pv panels costing a total of $1920, add another 320 sq feet of mirrors
a 3:1 winston would use about 1900 ft^2 of mirror. a half-greenhouse
(a 2'x20' pool under a 20'x8' tall reflector made with 6 $4 bent 12'
2x4s on 4' centers) would only use 240 ft^2 of mirror film...
>costing $320...
or maybe $200, as a half-greenhouse, with no pumps.
>pump(s), hoses, mounting, etc, maybe another $260 or so for a total of
>some $2500. on a good day it'll produce 1440 watts, sans 20 for the pump(s),
or maybe 40x103 = 4120 wh on an average january day.
>for ($2500/1420=) $1.76 a watt...
that works for small values of 1440. but are we talking power or energy?
on an average phila jan day, 1000 btu/ft^2 of energy falls on a south wall,
so a 15% pv panel with no concentration would collect about 0.15x1000/3.41
= 44 wh/ft^2, vs 102 with concentration. crudely speaking, we've decreased
the cell cost to $4x44/103 = $1.71 per peak watt. same ball park.
>too bad all the solar buy down programs insist that you don't void the
>warrantees on any of the equipment they pay you for or it might actually
>be possible to build a system for less than what the buy downs are.
one might retrofit a reflector :-)
nick
|
|
