I love SpaceX. Such a wealth of interesting speculation to be had out
of the privately-held SiliValley info-is-currency-model sparse and fluid
facts!
I tend to agree with you about the initial price decrease, Bill - it
came from Elon & Co recognizing and realizing the large efficiencies
available from bringing modern commercial materials and methods (and
some very smart tradeoff choices, EG cheap gas-gen engines and a
somewhat larger overall rocket) into the rather old fashioned ELV
industry. There actually is data that SpaceX's basic ELV development
and production costs were significantly lower than the previous industry
norm - some NASA paper comparing the basic F9 development cost to first
flight to NASA costing models, IIRC.
As for subsequent price increases being related to the presumed
additional expenses of developing booster reusability, correlation is
not causation. I could argue, for instance, that price progression was
far more likely just a standard commercial pattern of lowballing to gain
initial market share, then increasing prices to what the traffic will
bear at the desired market volume. I couldn't prove it either, but I
could probably find considerably more precedent to back the position in
such a non-commoditized craft-built price-opaque industry as space launch.
That aside, I would certainly expect an F9R stage to cost more /to
build/ than an early expendable F9, from the higher parts count alone.
How its total per-mission cost differs from that expendable over
whatever number of missions it ends up flying, well, I'm sure we'd both
love to peek into SpaceX's books. Absent that, though, about all that
can be safely said is that the recovery/reflight processing would have
to be quite surprisingly high for them not to come out ahead on the
operation as it settles down into high-rate ops.
best
Henry
On 4/4/2020 9:34 PM, William Claybaugh wrote:
George:
But facts matter: the price decrease came from pulling the cost out of expendable rockets.
The subsequent conversion of those then very low priced ELV’s into somewhat higher priced RLV’s (in price per unit mass to orbit) does not reveal any new information about RLV’s; nor can it serve as an existence proof that those RLV’s are lower cost than the lowest cost ELV’s. Indeed, the data slightly support the conclusion that the ELV version of F9 was cheaper.
Bill
On Sat, Apr 4, 2020 at 6:37 PM George Herbert <george.herbert@xxxxxxxxx <mailto:george.herbert@xxxxxxxxx>> wrote:
There’s a certain amount of pointlessness to the theoretical
argument; SpaceX semi-reusable Falcon 9 and Heavy are far cheaper
to buy than any competition, and have been for some time.
Whether more fully reusable, bigger, methane vs RP-1 next step
works or not there today are only companies and agencies and
countries who understand the current Market and who don’t.
IMHO
Boemart seem unclear. ULA gets it but is hampered by parents.
NASA is in parts aware and in parts unclear. ESA and Arianespace
are politically blinded. China is aware but continuing national
programs of record while it experiments with innovation and gave
up on foreign commercial launches. Russia... hard to tell. But
politically hampered.
-George
Sent from my iPhone
On Apr 4, 2020, at 2:08 PM, William Claybaugh
<wclaybaugh2@xxxxxxxxx <mailto:wclaybaugh2@xxxxxxxxx>> wrote:
Henry:
When RLV’s will be more economic than ELV’s remains unclear to me
and many others. I doubt I have ever suggested that they would
never be lower cost.
Our dispute has always been about when, not if.
Bill
On Sat, Apr 4, 2020 at 3:03 PM Henry Vanderbilt
<hvanderbilt@xxxxxxxxxxxxxx <mailto:hvanderbilt@xxxxxxxxxxxxxx>>
wrote:
Bill,
And you've never disputed the matter with me or anyone else
here since? Okay...
best
Henry
On 4/4/2020 11:46 AM, William Claybaugh wrote:
Henry:
You really need to revisit your assumptions about me.
Griffin and I proved that in 1994.
Bill
On Sat, Apr 4, 2020 at 12:44 PM Henry Vanderbilt
<hvanderbilt@xxxxxxxxxxxxxx
<mailto:hvanderbilt@xxxxxxxxxxxxxx>> wrote:
Ah! We agree that reuse is a benefit!
This is progress...
best
Henry
On 4/4/2020 11:21 AM, William Claybaugh wrote:
Henry:
Have it your way if you wish.
I am certain that Elon will tell you that the first
benefit of reuse is in spreading depreciation; and,
that spreading amortization is a very second order effect.
Bill
On Sat, Apr 4, 2020 at 12:10 PM Henry Vanderbilt
<hvanderbilt@xxxxxxxxxxxxxx
<mailto:hvanderbilt@xxxxxxxxxxxxxx>> wrote:
Bill,
Terminological quibbling aside, what I'm talking
about is something alien to the cost-plus trad
space industry: Commercial businesses ferociously
controlling their costs, both upfront and ongoing.
Up-front costs are like poison, while ongoing
operating costs are merely like heroin. Both are
worth considerable effort and ingenuity to
minimize. And SpaceX, in successfully going for
reusability, has avoided both a big initial chunk
of poison and a fair-sized heroin habit, both
implicit in the trad cost-plus approach to
eventually flying circa sixty booster cores a year.
Given we once again seem to be talking past each
other - it's good to be back! - perhaps best we
simply continue to disagree about this being a
significant part of why SpaceX is cleaning the trad
industry's clock.
cheers
Henry
On 3/24/2020 8:19 AM, William Claybaugh wrote:
Henry:
Terms matter: what you are talking about is
depreciation, not production savings.
I’m will to be educated but I would be shocked if
making 1/5 as many vehicles resulted in a
production system 1/5 the previous size: that is
simply not how production works.
There are high fixed costs in any production line
as well as minimum costs.
Bill
On Tue, Mar 24, 2020 at 8:49 AM Henry Vanderbilt
<hvanderbilt@xxxxxxxxxxxxxx
<mailto:hvanderbilt@xxxxxxxxxxxxxx>> wrote:
Bill,
The long-term production difference in
question, by definition, is a factor of five
times. Not 2:1 either way around a base of
12/year.
SpaceX knew this going in. Being sensible
people /not/ locked into the established way
of doing things, they likely would have set up
a production establishment for sixty expended
cores a year very differently than they did
the plant for ~12 5X reused cores. Twelve a
year, as you say, is pretty much craft
production - modest production tooling and a
lot of very skilled hand labor, low plant
investment but relatively high ongoing
personnel cost. 60X a year is still not
exactly Willow Run, but sensible people
planning that would very likely invest
considerably more in plant and tooling so as
to not require5X the skilled personnel plus
2nd and 3rd shift differentials, working in
~2X the modest 12/year plant (assuming it was
originally run one-shift).
Yes, I oversimplified by saying "1/5th the
size of production establishment". Thought I'd
allowed for that sufficiently with "(to a
first approximation)", oh well. And yes,
"size" was not quite the mot juste; "cost"
might have been closer to what I was driving at.
My basic point: SpaceX gambled on 5X
reusability to greatly reduce their up-front
investment in, and ongoing cost of, F9 booster
production. And they seem to have won. By a
quick count, 92 F9 booster core flights so
far, and already over half of those (51) have
been used boosters. The used proportion will
only rise from here. And they did this on the
up-front investment for a dozen a year.
In other words, one of the reasons they're so
far ahead of the game now is they gambled and
won bigtime on a major-capital saving shortcut
at the start. I hope that's clearer.
Henry
On 3/23/2020 2:00 PM, William Claybaugh wrote:
Henry:
It isn’t clear to me that there is all that
much difference between making 12 per year
and making 6 or 24.
One saves the material costs and the marginal
labor cost but the infrastructure doesn’t (or
at least shouldn’t) change much over that
range of production.
That said, if you optimize your system for
four units per year you will find making 24
more costly than a line optimized for
twenty-four.
But rates of a few dozen per year—or a few
hundred—all fall into “craft production” and
are not going to show economically
significant variation on production costs.
The benefit of even a few reuses is in the
depreciation of the hardware cost over
multiple launches.
Bill
On Mon, Mar 23, 2020 at 2:24 PM Henry
Vanderbilt <hvanderbilt@xxxxxxxxxxxxxx
<mailto:hvanderbilt@xxxxxxxxxxxxxx>> wrote:
Another way of looking at this that I
think is relevant: 5-reuse boosters
allows SpaceX to support a given flight
rate with (to a first approximation)
1/5th the size of production
establishment they'd need for fully
expendable operations.
Henry
On 3/23/2020 8:12 AM, William Claybaugh
wrote:
Robert:
There is too little data to make any
assertion about stage longevity at this
point.
However, ignoring propellant and launch
operations costs, five flights per
booster would indicate a cost per
booster at 20% of the manufactured cost,
not including refurbishment between
flights. The former is around $30-35
million, so $6-7 Million per flight,
again, not including refurbishment. If
an overhaul costs more than about $6
million, it would make more sense to
simply build a new five launch lifetime
stage.
We may note that compared to a $50
million price, these depreciated stage
cost estimates suggest either a good
deal of profit or that other costs
(launch operations, refurbishment) are high.
Bill
On Mon, Mar 23, 2020 at 8:51 AM Robert
Steinke <robert.steinke@xxxxxxxxx
<mailto:robert.steinke@xxxxxxxxx>> wrote:
From hobbyspace.com
<http://hobbyspace.com> about the
latest Falcon 9 launch:
" A first stage engine shut down
prematurely (just before staging)
but had no effect on the mission as
the other 8 engines made up the
difference. The booster also failed
to make a successful landing on a
sea platform. This was the fifth
flight of this booster."
That was after a previous launch
attempt aborted due to slightly high
power.
Wonderful demonstration of
engine-out fault tolerance, but it
does look like the rocket is showing
some wear and tear after 5 flights.
What does this do to their economics
if stages need an overhaul/have an
increased chance of loss of vehicle
after only 5 flights?