Index Home About Blog
From: Dani Eder <>
Subject: Re: The solar power satellite.
Date: Tue, 11 Nov 1997 23:04:29 GMT

Sunil Gupta wrote:

> I was under the impression that current PV's work most efficiently
> with near UV rather than near IR, due to the fact that UV has more
> energy than IR. I also seem to remember that the earth's atmosphere
> absorbs IR more than UV proportionally.

Your impression is incorrect.  Photovoltaic cells require a certain
energy photon to separate an electron-hole pair.  For silicon the energy
required is equivalent to near-IR.  Any photons with less energy than
that required are wasted by not being absorbed.  Photons with more energy
than that required waste the excess portion as heat.  This is why the
theoretical conversion efficiency of a single type of cell in sunlight is
limited to about 25%.

There are PV arrays using layers of cells with different required
energies, so as to capture more of the solar spectrum efficiently.  They
can reach in theory about 40% efficiency.

Dani Eder

Subject: Mirrors on Solar Arrays (was: Re: Energy by Airwaves?)
Date: Mon, 15 Mar 1999 15:53:32 GMT

In article <7cicl8$cft$>, (Frank Palmer) wrote:
>  Wi> (The sunlight-to-electricity-in-orbit conversion is still as
>  Wi> inefficient as ever,

"As inefficient as ever?"  A decade ago, the commercially available space-
qualified solar cells were 15-16 percent efficient.  Today, you can buy 25%
cells off the shelf, and higher efficiencies are available in the lab.

--I think that the Artemis site has a review of solar power technology I
wrote a few years ago.	It's missing the figures and tables, but it'll give
you overview.  Check:

>  Wi>but the theory is that that can be compensated for
>  Wi> by adding some more cheap mirror surface.)
> Which brings up the question as to why NASA doesn't use
> mirrors now.
> I see all those solar panels on pictures of satelites in
> orbit, they seem to be in direct light, with no mirrors aiding
> them.

Mirrors are not nearly as good an idea as you'd think.	The first problem is
that the increased solar flux on the cells increases the temperature.  It
turns out that increased temperature decreases solar cell efficiency.  To a
quick approximation, if you double the solar flux on a cell, if you do
everything else right, you increase the actual power out by about 50%.	And
you can do worse than that if you're not careful.  Solar arrays are very
sensitive to illumination nonuniformity; if your mirrors have ripple in them,
you can actually have worse performance from the illumination enhanced
mirrors than from the original cells.  Now, you can make a perfectly flat
mirror, but that is heavier and more expensive, so you're up against the
trade-off: why not just add more solar array area?

(The 50% figure, by the way, is only an approximation good near 1x solar
illumination.  At higher intensities, the power out gets worse and worse.
It's very technology dependent, but somewhere around 10x illumination, you
get a * decrease* in power with additional illumination.)

The increased solar intensity can also cause degradation if you're not
careful. The Carisa Plain installation found greatly increased degradation
rates (yellowing of the polymeric encapsulant) due to higher solar intensity
on their mirror-enhanced panels.

There are a couple of other drawbacks as well, which I won't go into.

Now, despite these caveats, some new solar arrays *do* have integral mirrors,
typically trough concentrators.  This is partly because newer solar cell
designs have lower temperature coefficients, partly because new solar cells
designs are more expensive (and hence the mirror area looks more attractive
than solar array area) and partly because new communication satellites have
very high power demands.  There are two trough designs commercially available
at the moment, the "Astro-Edge" array, from Astro Aerospace (which I believe
will be flying on some next-generation communications satellites), and also a
concentrator design from Composite Optics, which is flying if I recall
correctly on a technology test satellite.

--there are also the fresnel lens concentrators, such as the SCARLETT array
flying on the Deep Space 1 Mission.

You should be able to find some more info if you rummage around on the NASA
Glenn Photovoltaics and Space Environmental Effects home page,

(cross posted to

Geoffrey A. Landis
physicist and part-time science fiction writer
(and professional space solar array scientist: will design high-efficiency
photovoltaic devices for food)

Subject: Re: Mirrors on Solar Arrays (was: Re: Energy by Airwaves?)
Date: Tue, 16 Mar 1999 19:05:56 GMT

In article <> Doug Jones, writes:

>So here's some food for thought: have you considered transparent
>diffractive optics (maybe diffractive plus refractive) which would
>concentrate just the cell's optimum wavelength?  If a fresnel +
>diffractive concentrator can be made cheap & light enough, it might
>allow high concentration without cooking the cells.  This would be
>"high grading" of sunlight, using only the portion that can be easily
>used and wasting the rest.

Yes, many variants of such ideas have been considered.  The "cheap enough and
light enough" part is the tough part-- the question is, can you beat the
solution of just increasing the solar array size?  The solar cells themselves
can incorporate a reflective filter in the coverglass to divert the portion of
the spectrum that's not useful-- this is known as "blue-red reflector" covers
(actually "UV-infrared" reflectors.)  But it's a complication and an added
expense that only adds a very small amount of performance.

>A more broadband solution would use a diffractive trough concentrator
>to produce an intense rainbow at the target, with several types of PV
>cell matched to the spectrum,

Yep-- this is the "Rainbow concentrator" technology.  It was an approach
looked at by Varian in the early 1980's, which is being revived by a group
operating out at JPL.  The theoretical numbers look good-- the question is
whether it will work well in practical applications.

>including perhaps fluorescent converters for the blue end of the spectrum.  IR
>downconverters perhaps, to put the peak into the 1100 nm region?

I looked at spectral shifters a long time ago, and found that they were way
too inefficient to help.  The technology may have changed enough to make it
worth looking at again, and there are some possible tricks you can play, but
it's harder to do than you would think, and downconverters are in general not
very efficient.  (the exception is coherent downconversion, but unfortunately
sunlight is not coherent.)

 Geoffrey A. Landis
 physicist and part-time science fiction writer
 (and professional space solar array scientist)

From: (Henry Spencer)
Subject: Re: Mirrors on Solar Arrays (was: Re: Energy by Airwaves?)
Date: Sat, 20 Mar 1999 19:01:08 GMT

>> (I was under the impression that solar-thermal-driving-a-turbine had
>> higher conversion efficiencies than PV cells. In particular, no quantum
>> efficiency problems. Is my memory wrong? ...

Solar-dynamic systems have been perennial also-rans for a long time.  In
particular, the space station *almost* had solar-dynamic power; in fact, I
think that happened more than once.  For that application in particular,
solar-dynamic has a couple of big advantages:  the higher efficiency
permits smaller surfaces which have less air drag, and with suitable use
of phase-change materials, you can store energy for the night half of the
orbit as heat and dispense with the big battery bank.  (For a permanent
space station, station mass is much less important than resupply mass
flow, and battery replacement and rocket fuel for air-drag compensation
are big resupply items.)

>Turbogenerator SPS's never got as much attention as PV.  Main reason is:
>if the assumption is Earth-launch of all components, that drives you
>stongly toward the lightest designs.  But if ET resources are used,
>turbo might be more favorable...

Unfortunately, while solar-dynamic power systems can use mirrors made from
extraterrestrial resources, their turbogenerators are relatively heavy
pieces of precision machinery... while the simpler sorts of solar cells
are not prohibitively hard to make from lunar (etc.) materials.

Almost certainly the best way to make an extraterrestrial-materials
powersat is to use thin-film amorphous-silicon solar cells, which are not
as efficient as the more conventional (for space purposes) crystalline
cells, but are easier to make and can be made *extremely* thin and light.
The good old days                   |  Henry Spencer
weren't.                            |      (aka

Date: Wed, 11 Aug 1999 00:35:41 -0700
From: Doug Jones <>
Subject: Re: SPS != Solar Panels RE:Feasibility of SPS...

Folks, you're all losing track of the real cause of solar cell
degradation in space.  It's not micrometeorite impacts- those are
dealt with by a thin cover glass- but crystal defects caused by
ionizing radiation.  Darkening of the cover glass is a secondary
effect, but solar protons are the main concern. Annealing at 100-200
C will reverse most of these aging effects, so one need only scan a
large magnifying glass over the array to extend the system life.
One can envision Sojourner-sized "spiders", crawling slowly about
the arrays carrying fresnel lenses...

At there is a good
briefing on space solar cell history and technology.

Benjamin Diedrich wrote:

> wrote:
> >
> > Ian St. John <> wrote:
> > > Your vulnerability is the radiator. You need a "cold side", and with
> > > the equivalent of a vacuum bottle all around the only means is
> > > radiation, which take a lot of area.
> >
> > Yes, but would building/designing a redundant radiator be cheaper in the
> > long run? Would it be worth it to avoid solar panels? A solar dynamic
> > setup could be done as a grid of "nodes", each with a turbine/bubble/radiator.
> > This sort of design could be "plug and power", when a node is damaged, a
> > new one is popped onto the grid, inflated, and starts generating.
> I doubt it. Solar arrays already are made as a grid of photovoltaic
> cells. If an area is damaged, it reduces the total power of the system,
> but it is localized to that one area. Once an area accumulates enough
> damage, it can replaced by pulling the area out and plugging in new
> solar panels. Early SPS may never be visited by maintenence crews, so a
> photovoltaic system would be ideal. It would gradually degrade in
> performance, rather than having discrete failures of critical
> components, like any part of the heat engine in a dynamic system.
> > J05H
> Ben

Doug Jones, Freelance Rocket Plumber

Index Home About Blog