re: increase solar flux and pv lifetime
9 jun 1997
>nick pine wrote:
>> >a normal pv module with full sun on it, 25c ambient air, and 1 m/s
>> >wind will attain a cell temperature of 60-63c. increase the irradiance
>> >to 2.5 suns and the temperature will reach 100-112c.
>> not if there's some 60c water trickling over its face...
norman saunders suggested this, btw, saying pv panels are weatherproof anyway.
>the glass covering the panel has an antireflective coating designed for
>an interface with air.
the glass has an expensive weatherproof 1/4 wave matching coating? is it etched
as described in duffie and beckman to reduce the air-glass fresnel loss from
((1.5-1)/(1.5+1))^2 = 4% to 1%? what's the overall air to cell transmittance?
>by running water over the panel constantly you'll lose photons to
>(a) reflection at the air/water interface,
water has a lower index than glass (1.33) and it may not have to cover the
whole front surface, since glass (1w/mc) and epoxy are fairly good sideways
thermal conductors. a 2nd cover plate would reduce evaporative loss, if we
wanted to use the "waste heat" (85% of the input :-) polycarbonate plastic has
an index of 1.6. the back of the panel might be insulated, or it might live
in a sunspace, with its "waste heat" warming air that flows into a house.
>(b) absorption by the water,
water films reduce glass-air-glass reflections, because of the lower index
of water. steve baer says you can see this by noticing that the edges of a
stack of glass panes become less visible when wet. do they reduce air-glass
reflections too? duffie and beckman say
the reflectance of a smooth water surface for solar radiation is of the
order of 5% [why more than glass, with a smaller index?] at near normal
incidence. as radiation is transmitted through water, the longer wavelengths
are absorbed near the sufrace. measurements... indicate that little radiation
at lambda > 0.7 um is transmitted through a meter of clean water, but over
95% of the radiation in the wavelength range 0.3 to 0.6 um is transmitted
through this much water.
>and (c) reflection at the water/glass interface.
about ((1.33-1.53)/(1.33+1.53))^2 = 0.5%, with ordinary glass. seems like
no big deal, especially if the water doesn't cover the whole surface.
>i don't know how noticeable these effects would be, but you're also faced
>with the problem of cooling the water (or sourcing cool water)
you might consider this a water heating opportunity, not a cooling problem :-)
as does dr. richard komp, firstname.lastname@example.org, author of practical photovoltaics,
who sells hybrid panels with 150w of electrical output and 1600w of water
heating output, using 1:1 mirrors with half the usual number of ordinary pvs.
the maine solar energy primer describes how to build these panels, as does his
book, as i recall. losing heat is a lot easier than gaining heat...
>and pumping it over the face of the panel in an evenly
>distributed, fairly laminar fashion.
non-uniform random trickling might work... two suns with 2 gpm/m^2 would make
the water temp rise about 4 c at most. if the moving water film has a large
thermal conductance, and a strip of glass 1 meter long x 1/8" (0.3 cm) thick
x l cm wide has a sideways conductance of 0.3cm/100x1mx1w/mc/(l/100) = 0.3/l
w/c and collects 2kl/100 w/m^2, this leads to a very small distance between
trickles, l = 0.25 cm for another 4 c, if i did that right. maybe the water
does have to cover the whole face, or it needs a good thermal connection to
the back of the cells, or the panels need fins on the back...