Space Forum / Shuttle / October 2005

Correction to my first question...$500 million per flight?

Dave Mitsky

$750 million per shuttle mission looks like it might
be a fair price compared to other, unmanned missions.
The three-constellation Inmarsat 4 system, for example,
will cost $1.5 billion all told, of which the actual
cost of launch is roughly the same as the claimed
shuttle marginal launch cost above.  Inmarsat costs
$500 million per mission, and it doesn't have to keep
astronauts alive going up *and* down.  

- Ed Kyle

He said marginal cost, not average cost. There's a difference.

Take a closer look at that table. Notice that the annual cost of the
space shuttle program is practically independent of the number of
flights. One time, for laughs, I computed the correlation coefficient
between the annual shuttle budget and the number of flights. It came out
close to zero, in fact slightly negative if I recall. That tells you that
the bulk of the shuttle budget is overhead, and that the cost of adding
or deleting a single flight is very low. That's the marginal cost that
Andrew is referring to. It's the actual cost of the ET, SRB refueling,
and other costs that are directly attributable to that flight, without
including any of the overhead. If it assumed that the shuttle will
continue flying ISS flights anyway, then the cost of adding an HST flight
is close to the marginal cost. It's actually more than the cost of adding
another ISS flight due to the special CAIB recommendations (and NASA
raising-the-bar requirements) that kick in for non-ISS flights.

The corroborating evidence of this is that there are multiple instances
where a shuttle flight was added in a fiscal year (STS-94, STS-103) where
NASA covered the cost of the flight without an additional appropriation,
and similar instances where flights were deleted without significant cost
savings. And of course there's the post-accident standdowns where the
program continued to consume its annual budget without flying at all.

Mind you, the concept of "marginal cost" has become somewhat obsolete now
that the orbiter fleet has a firm retirement date and a fixed (though not
yet publicized) number of flights to go until that retirement. At this
point, adding a shuttle flight means extending the retirement date of the
shuttle fleet (and extending the annual overhead spending) such that the
new flight now costs the average cost instead of the marginal cost, or it
means accepting the opportunity cost of deleting one of the ISS flights
that the shuttle would otherwise have flown.

Of course, now that NASA has settled on a shuttle-derived architecture
for the VSE era, some unknown portion of that overhead spending will
continue after the orbiters are retired. I suspect that portion is larger
than many people hope it will be.

I'm sorry, but no.

Marginal cost to fly up a Shuttle refurbishment mission:

$100 million launch costs + new instrumentation

Marginal cost to fly up a new HST:

launch costs + new instrumentation + new telescope

Let's assume the instrumentation cost of a new HST would have been the
same as the instruments built to re-equip the old one, which seems fair.

So, what are the marginal costs of building a new telescope *and
launching it*?

I find it fairly implausible you'd get an expendable launch in that
weight class (and to a moderately high altitude) for significantly under
$100 million. And that's before you even take into account the costs of
building a new telescope; even with several produced, those don't come
cheap.

Given that the current cost *overrun* on the JWST is $1.08 billion, I
find this very hard to credit.

mikharakiri_nospaum@yahoo.com wrote in news:1128544832.171097.86730
@g14g2000cwa.googlegroups.com:

Depending on what you think the cost of a shuttle repair mission is, that's
true. Johns Hopkins thinks that such a Hubble sibling can be built for as
little as $0.8-1.0B.

<http://www.pha.jhu.edu/hop/>

You're right - it's a *lot* higher. $4.5B and still climbing:

<http://www.newscientist.com/article.ns?id=dn7908>