I've decided that the best way to fix this is to take it from the
horse's mouth. He goes in W/kg and so does OECD and CRC books, so
that's what I'll use.

Entering Space PVA est.: 49 W/kg (make it 50 W/kg so easier to figure stuff.)
After Columbia PVA est.: 200 W/kg (actual SOA)
After Columbia TDG est.: 500 W/kg (a guess...heavier per unit area,
but a lot more juice per unit area.)

SOA = State Of the Art (used in CRC texts), PVA = PhotoVoltaic Array;
TDG = ThermoDynamic Generator (used in Space Nuclear Power, Orbit
Series and was it...Living In Space by Frank Caprara or Island In The
Sky by Piers Bizony, mebbe it was just Marcus Lindroo stuff...space
station expert at www.abo.fi/~mlindroos ...it is in English, BTW.)

Transmission efficiency, 50%, I think we can agree on that.

Fraction of SPS that is actual power generating element:
Entering Space: 25%
After Columbia, crewless SPS estimate: 40%
INSEA, crewless SPS estimate: 25%

(INSEA uses higher safety factors and deployed load factors so things
don't get to expensive during on-orbit integration and tug
operations.)

Entering Space GEO current: $40,000/kg
After Columbia GEO current: $75,000/kg
After Columbia normalized for 25% payload est: $46,875/kg
INSEA current equivalent: $2500/kg

The difference is mostly due to inflation

Ascent Costs of Only:
Entering Space $/W: $3,200/W (book says $3,300/W)
After Columbia current PVA: $938/W
After Columbia current TDG: $375/W
INSEA PVA: $50/W
INSEA TDG: $20/W
After Columbia ground plant costs: $0.30/W

The last is based on informal quotes I got from the Global Petroleum
Show 2006 for large scale gas turbine units. GE Jenbackers came in at
$0.24/W. This is of course for locally delivered shaft power, not
electrical power (which would be only slightly higher.)

Based on After Columbia's TDG estimate differences to Entering Space,
a factor of 16 is gained in favor of SPS. The resulting total cost
requirement is $272kg/GEO, which I think is impossible from Earth.

I don't have the numbers with me so I'm going to guestimize the GTO
ascent costs of Bluestar/Yellowtug (leaving out the expendable
Yellowfly) and using ion propulsion; GEO delivery is about $950/kg
(GTO is about $885/kg)

One thing that would have to be developed for my TDG estimate to be
realized is a completely closed turbine fluid cycle, which as near as
I can tell, has never been done (closest thing: nuclear BWR; close
enough to suggest that it is possible in the near term.) One of
Zubrin's assumptions was that you'd stick your ground based solar
plants in desert areas with few clouds. One of mine is that you'd
also intersperse them with wind turbines in such a way as to avoid
disrupting generation (which is easier for a TDG mirror concentrator
than it is for a PVA based system, where shadowed cells would actually
drain power.) The closed turbine fluid loop would also be needed
there, because water is expensive in the middle of a desert...it
wouldn't have to be _quite_ as closed as a space TDG.

There are other advances that would increase the specific power of an
SPS that would also benefit ground based solar powerplants. So, by a
factor of 3.5 ascent costs only, and probably several times that with
the hardware, the SPS is dead. In a nutshell, the SPS will not happen
until we are a multi-planet society, period.

From the moon, with the assistance of mass drivers and advanced
propellant tugs, low enough GEO ascent costs (i.e. in the $40/kg
range)

For sort of a baseline idea of my INSEA crewless SPS estimate, an SPS
beams power to somewhere it is a major pain in the backside to
generate power...downside is I can't think of any. Only thing I can
think of is beaming it to areas that have lost their normal source of
power to an emergency. (Coastal areas use tidal and ocean water solar
thermal cycling in deep water (these enormous oil-rig like constructs
electrolyse water into hydrogen and oxygen; some of it is shipped to
launch sites as hydrogen, most of it is Sabatier'd into methane, the
heat generated recycled back into electrolysis, and this replaces
natural gas); desert areas use solar/wind interspersed plants, and
most areas I can think of that are inland enough that tidal power is
unavailable have conventional hydro or geothermal sources available.)

Anyway, back to the SPS. Say the minimum practical size for a
crewless SPS is 5MWe delivered, which is a bit small, but should be
enough to power about 5000 homes at 1000 W/ea average (a bit on the
slim side, I think.) So...at 50% system efficiency, this asks for a
10MWe raw SPS (which probably works out to 12MWt, but who gives a
hoot?) At 500W/kg, that translates into 20,000kg of payload, or
50,000kg of satellite, which in turn translates into a 52,750kg GTO
delivery (so 2750kg of ion engine fuel...typically xenon gas...is
consumed, I wonder how much that would cost?) This is actually quite
reasonable for spacecraft size.

Delivery to GTO at $885/kg would cost $46.7mln. With hardware costs,
that would come to somewhere on the order of $150mln. That's about
half as much as a comsat.

Hmm...what if it was a comsat? All the payload gobbledegook on board
would probably reduce its RF usable power to 5MWe, or about a thousand
times as much as a real comsat with a similar total cost (i.e. Orbital
Sciences Star-2 based.) One of these bastards on 19.2degE could
conceivably replace the entire SES Astra fleet! Sound like a cool
idea?

Terry