Message-ID: <kZzo1W200VcJA5iE5S@andrew.cmu.edu>
Reply-To: space+@Andrew.CMU.EDU
From: space-request+@Andrew.CMU.EDU
To: space+@Andrew.CMU.EDU
Date: Thu, 15 Mar 90 02:12:34 -0500 (EST)
Subject: SPACE Digest V11 #151

SPACE Digest                                     Volume 11 : Issue 151

Today's Topics:
       SPS cost-effectiveness (was Re: Large vs. Small scale.)
----------------------------------------------------------------------

Date: 15 Mar 90 03:28:31 GMT
From: cs.utexas.edu!mailrus!b-tech!kitenet!russ@tut.cis.ohio-state.edu  (Russ Cage)
Subject: SPS cost-effectiveness (was Re: Large vs. Small scale.)

[Please follow-up to whichever group is appropriate.]
In article <1990Mar11.152345.11614@cs.rochester.edu> dietz@cs.rochester.edu (Paul Dietz) writes:
>This isn't obvious to me, Russ.  Yes, you won't need big vacuum
>chambers (although one would still need a gas-tight vessel for some
>processes, like deposition of a-Si:H from a plasma), 

I doubt that such processing would be used.  It would be much simpler
to vaporize elemental silicon with a solar furnace and deposit it on
a substrate.  It could be doped either while being deposited or
afterward, by ion implantation.  No pressure chambers required.

>but other costs
>will be much higher than for ground-based manufacturing.

Some of them will be.  Some will be exactly the opposite.  For instance,
process heat will be INCREDIBLY cheaper.  A square km of mirror, massing
a few tons, intercepts better than a gigawatt of sunlight.  A plant
separating ferrous metals using the carbonyl process could process its
own mass every 8 hours [hearsay], with nothing more than solar heat required.

Incredible amounts of energy are not needed.  Example: The Apollo 12
lunar samples ran about 20% FeO by weight.  Assuming 1 MT/yr of regolith
into orbit, I calculated how much power it would take to break down 200
kT of FeO into Fe and O each year.  I got a figure of about 24 megawatts.
If the objective is to produce multi-GW powersats, making 24 MW of power
locally is a pretty simple first part of bootstrapping the operation.
(This is assuming that the refining cannot be done thermally; mirrors
are cheaper, simpler and lighter than PV panels by a long shot.)

>The cost of
>keeping people in orbit is currently on the rough order of $100K *per
>hour*.  Even if you drop that by a factor of 100, space-based labor
>will still be extremely expensive.

I think I know where you got that number:  

($100K/crewmember-hr)*(24 hrs/day)*(11 days)*(7 crewmembers) = ~$185,000,000

or on the rough order of the price of a Shuttle launch, with the
capacity of a Shuttle crew and the length of a Shuttle orbital stay.

Guess what?  You're comparing apples and oranges.  Having people work
a 3-year hitch, with NO OTHER CHANGES, drops the cost by two orders
of magnitude right there.  They will go up on cheaper vehicles or in
bigger groups; a 50-seat Spacelab-style "Shuttle Bus" module could
certainly be built, and would cut personnel launch costs by nearly
another order of magnitude.  We're already down to investment banker
wage costs, without yet replacing Shuttle as the launcher.

Compare the cost of saturation diving to divers operating from the
surface to deep depths, and you'll find that the technology and
logistics have enormous effects on the cost per worker-hour even
here on earth.  You can't make meaningful comparisons without
taking these things into account.

>Similarly, the cost of
>transporting the equipment and spare parts to the factory will be much
>higher, per pound, for space manufacturing.

Very true!  On the other hand, if the proper technologies are chosen
to use the properties of the space environment to good effect (cheap
heat, high vacuum, etc.) each pound may be able to do many times the
work it could down here.  For instance, a square foot of solar panel
receives many times the sunlight in space that it could on earth (at
least 4x in summer).  A gadget to sublime or sputter silicon onto a
substrate must be much bigger and heavier here.  Et cetera.

Without looking at all the angles and opportunities for economies,
you can't make realistic projections of the cost of SPS.  I lean
toward the L5/SSI/LLNL-type thinkers on this one, they are more
often right about the technical details and possibilities.
-- 
  I am paid to write all of RSI's opinions.  Want me to write some for you?
(313) 662-4147        Forewarned is half an octopus.
Russ Cage, Robust Software Inc.            russ@m-net.ann-arbor.mi.us

------------------------------

End of SPACE Digest V11 #151
*******************