Terry,

If you read my "Science and Society" article, you'll note SPS
launched from Earth may compete economically with base power plants
(nuclear, coal - capital costs about $1-4/watt) with current launch
costs of $10,000/kg (i.e. we don't necessarily need lower launch costs)
on three conditions:

* transmission efficiency is high enough (> 40%?)
* solar panel costs come down to about the $1/peak watt level
(terrestrial solar panels aren't far; solar panels designed for space
use are 100's of times higher right now)
* (this is the key one) - the mass per kW in space can be kept below 0.1
kg/kW.

If the first two aren't met, it really doesn't matter how low launch
costs are, solar panels in space won't be competitive with ground-based
energy under any conditions.

But that third constraint is the toughest. Panels have been deployed in
space with about 20 kg/kW; commercial modules seem to be available at
about 5 kg/kW. It's not clear how light we can go. The constraints in
space are very different than on the ground, where rigid strength is
essential in the presence of wind and rain and other assaults. You write
that an SPS has to "resist [...] maneuvering and solar pressure induced
torques" - a very light one can rather use such things to move itself
around. If we have to launch from earth, all we need do is get to low
orbit, and then the light SPS components can move themselves under their
own electric power, using solar electric propulsion technology, with
very minor propellant requirements. Once in place, say at GEO, they can
use solar sail techniques for station-keeping - there are plenty of
mathematical analyses around that'll help you do that very precisely
with whatever the thrust you happen to have - which is what you seemed
to suggest in your notes as well...

For reference, the Planetary Society's solar sail uses thin mylar of
0.02 kg/m^2, the equivalent of less than 0.1 kg/kW if they included thin
film solar cells of 20% efficiency. So that number may not be
technically impossible - with nanotechnology, there's still a lot of
"room at the bottom".

But if 0.1 kg/kW isn't possible, 1 kg/kW means we need launch costs of
$1000/kg, ten times below today's. If we can't actually do better than 5
kg/kW that seems to be commercially available now, then we'd need launch
costs of $200/kg.

In other words, it's not a factor of 2400, but a factor of 50 that may
be necessary. At least by those numbers.

There are of course mass and cost constraints on the power transmission
components and other pieces of the design not included above, and the
cost of ground equipment - these may in fact be the real limitations.
Higher ground-side power density limits ground-side costs but probably
raises space costs - I believe the Japanese are looking at some laser
designs though that would have quite small ground footprints.

It's not clear we have sufficient energy from other sources to meet the
needs of Earth-based industry through the end of this century. With
fusion we would, if a practical design for that were ever economically
competitive. Fission will require new designs that may be very expensive
also, if it's to expand the 50-fold or so factor needed to power the
planet. Wind, wave, etc. probably can't supply enough power. If those
are accurate statements, that may leave just terrestrial solar vs.
space-based solar, which was the main subject of my article.

And in partial defense of Mike Combs - he was mainly talking about
O'Neill colonies; the SPS issue was just an angle on it. The design
there of course assumes construction from lunar materials and launch
from there rather than from Earth. That would certainly be much more
energy efficient even without "mass drivers" - would it be more cost
effective? David Criswell's lunar solar power proposal is another
alternative - building the thing on the moon itself, without bothering
with the launch infrastructure. At very large scale, if the robotic
manufacturing techniques he describes are actually possible, it really
would be very cost-effective. Like your Mars colony, it would grow
itself out on the Moon with little need for further input from Earth.
But I consider that pretty speculative, and unlikely to much precede a
Mars colony in fact - mid-century perhaps. Earth-launched SPS could be
done much sooner - if the mass/launch and other (hopefully easier)
constraints were overcome.

Arthur