I was jolted a bit when I read the amounts and requirements for commercial crew last week. Not that Boeing got more than SpaceX, but rather that $6.8B is budgeted for a developing a couple of commercial capsules to fly on existing launchers. Boeing’s $4.2B doesn’t bother me as it should be expected of traditional contractors, SpaceX’s $2.6B does.
I am not going to suggest that human transport vehicles to LEO should be trivial or cheap. I do suggest that SpaceX requiring $2.6B to finish development of  a capsule that is supposedly almost ready to go, even including the maximum 7 flights, is well out of step with the reported development costs of that company. It has been said quite often that SpaceX developed the Falcon 1, Falcon 9, and Dragon 1 for under a billion dollars.
My question is, why should capsule development cost on the order of triple that of an entire Falcon 9 launch vehicle? This is not to trivialize the work to be done, or the paperwork involved, but seriously asking what is so complicated and expensive inside that vehicle that drives the cost far beyond that of entire launch vehicle families?
This post might be considered snark. That is not the intention. It is a serious puzzle that affects development of the solar system. A puzzle that I can’t yet understand. I am not real interested in the standard NASA bashing response of how the government drives up costs that we have all rehashed a hundred times. I am interested in any solid information on the hardware/software cost drivers of this bus without a main engine.
johnhare
Latest posts by johnhare (see all)
- Regenerative cooled turbopump - March 16, 2023
- Second Guessing Starship - September 19, 2022
- Projectile Fusion - May 19, 2022
There’s enough satellites out there, that even if you’re doing something relatively interesting and unique (like an ISS cargo capsule) you can buy many essential components off-the-shelf or near enough to it. GPS, star trackers, flight computers, that sort of thing.
It’s a lot harder to buy space-rated ECLSS systems, controls and displays, and so forth off the shelf. Not as much of a market.
Even if the “safety factor” is not considered, that alone will balloon your costs significantly.
Either the Dragon (manned) capsule is not as far along as it appears in public or the cost to meet the NASA requirements under the fixed price contract are much, much higher than under a Space Act. I can actually believe its a combination of both.
Elon has simply figured out the game. The paperwork will say $2.6 billion. Actually he will spend $600 million or so including the cost of six flights. $2 billion will go directly into MCT and he never has to go public. Game over baby.
Could some of that cost be for man rating the F-9/Atlas? That is one reason I thought SNC lost as they don’t have a launcher and had no control over that.
Five answers that have legitimacy.
1. The hardware is not available off the shelf, or even nearly so, which balloons costs when many systems must be designed and developed from scratch.
2. The Dragon that was unveiled was not nearly as far along as Elon and company suggested, which means that there is a tremendous amount of work yet to be done.
3. NASA’s arbitrary requirements drive costs beyond reasonable expectations.
4. SpaceX is charging what the market will bear regardless of actual financial requirements.
5. The launch vehicle (that I said was already available) must be man rated to some NASA requirement that really jack up the costs.
Emotionally I lean toward #4. Logically though I believe a mix of all the above was involved with no single culprit to be dealt with.
I’d venture that a big hunk of that money is in the abort system. The crew is going to ride inside what has become a rocket itself, along with a large quantity of highly toxic hypergolics.
“It has been said quite often that SpaceX developed the Falcon 1, Falcon 9, and Dragon 1 for under a billion dollars.”
Um, I don’t think so…
USASpending.gov shows :
(http://www.usaspending.gov/explore?tab=By+Prime+Awardee&fiscal_year=all&contractorid=120406462&contractorname=SPACE+EXPLORATION+TECHNOLOGIES+CORP.)
Almost $1.7B in Nasa grants and contracts alone.
Maybe $1B in invested ‘seed’ money from private investors ???
(This actually seems in error, I though I found $2.2B last time I looked)
-G.
Snyder,
Thanks. Often the most valuable answers are to questions I didn’t think to ask. The sub $1B has been mentioned enough times that I didn’t bother checking.
#6. becomes that my basis for comparison was off which seriously affected my reasoning.
Try looking at the award dates (in particular starting at the bottom of page four.) SpaceX began in 2002.
Anyway, summarized I show only about $11M before 2012.
Ken Anthony,
“Elon has simply figured out the game. The paperwork will say $2.6 billion. Actually he will spend $600 million or so […] $2 billion will go directly into MCT”
FAR rules makes that difficult. You have to demonstrate your costs, and the government can make you justify them at any time (which is why you make sure you “charge” everything, including every hour you spend justifying costs.) It’s the reason why major contractors tend to actively avoid making changes which will save money. Unlike normal commercial contracts, they can’t charge the same price for the same service if their actual costs drop.
“Actually he will spend $600 million or so including the cost of six flights.”
SpaceX is currently charging NASA over $130m for CRS flights. There’s no way that (non-reusable) crew flights will be less than that.
born01930,
“[man rating Atlas] That is one reason I thought SNC lost as they don’t have a launcher and had no control over that.”
Boeing has no “control” over Atlas. They may co-own ULA, but their CCtCap bid wasn’t via ULA. ULA would be listed on their bid as a subcontractor, which in turn will sub-contract out to LM, which in turn…
There’s nothing to prevent SNC from doing the same thing, and no reason NASA wouldn’t have accepted that. It’s the normal way of doing NASA contracts. SpaceX’s lack of endless chains of subcontractors is the one that’s unusual.
Snyder/John,
From the original SpaceX website: “The total company expenditures since being founded in 2002 through the 2010 fiscal year were less than $800 million, which includes all the development costs for the Falcon 1, Falcon 9 and Dragon. Included in this $800 million are the costs of building launch sites at Vandenberg, Cape Canaveral and Kwajalein, as well as the corporate manufacturing facility that can support up to 12 Falcon 9 and Dragon missions per year. This total also includes the cost of five flights of Falcon 1, two flights of Falcon 9, and one up and back flight of Dragon.
“The Falcon 9 launch vehicle was developed from a blank sheet to first launch in four and half years for just over $300 million. […] The Dragon spacecraft was developed from a blank sheet to the first demonstration flight in just over four years for about $300 million.”
The figure for F9 v1.0 is backed up by the NASA OIG report which estimated the “firm fixed price” for Falcon 9 at $300m (exactly $299.9m {laughs} plus $146.3m for two test launches, which they contrasted with $1,382.7m prediction for a simple cost-plus “acquisition” and $3,799m for a more traditional NASA program.) Some slides, apparently from the report: http://www.nasa.gov/pdf/586023main_8-3-11_NAFCOM.pdf
So, I guess that makes it:
F1 – $200m
F9 – $300m
D1 – $300m
F9v1.1/CRS – $1,700m (Snyder’s figure)
D2 – $2,600m (CCtCap bid)
Hmmm. Inflation?
Ken Anthony,
“summarized I show only about $11M before 2012.”
If you switch the USAspending.gov list from “List of Transactions” to “Summary” and scroll down, then click “List View” under the year-by-year graph. You could have saved yourself some work.
2003: $0.5m
2004: $3.4m
2005: $0.03m
2006/2007: nil
2008: $4m
2009: $25.7m
2010: $115.3m
2011: $194.6m (Subtotal: $343.53m)
2012: $256.3m
2013: $594.2m
2014: $497.4m (TTD: $1691.43m)
Thanks Paul, I’ve revised my blog post…
$33.9 million by 2009 with $25.7m coming in 2009.
So by their 7th year about a tenth of their funding came from NASA.
“Actually he will spend $600 million or so including the cost of six flights.â€
SpaceX is currently charging NASA over $130m for CRS flights. There’s no way that (non-reusable) crew flights will be less than that.
The price to NASA is $130m. The cost to SpaceX is less. Price includes profit.
Also, Elon understands how to get manufacturing cost down (Tesla experience.) I’d wager the crew Dragon 2 cost less to produce than the cargo Dragon 1. I believe v2 uses fewer Dracos and the 8 SuperDracos are 3D printed for a drastic reduction in cost and production time.
FAR rules makes that difficult. You have to demonstrate your costs
I know this is sacrilege to say, but mom and pop businesses have been dealing with rules like this for ages. Auto body repair shops, for instance, will often give bogus quotes to customers to give to their insurance companies so each will get a better deal.
SpaceX’s lack of endless chains of subcontractors is the one that’s unusual.
Does this make it easier or harder to pad the numbers? When I worked at one of the regional FAA HQ I was astonished that forms of padding were institutionalized in everything they did (over ten years ago… I’m sure… Oh absolutely… this doesn’t go on anymore.) Back then, for instance, we could only by computers off a negotiated government contract for $20k a piece which otherwise could have been purchased for $2000. I usually bought about ten at a time which was similar to dozens of departments where I worked.
Demonstration of costs sounds like cost plus accounting.
— john hare says:
September 23, 2014 at 5:28 pm
……
4. SpaceX is charging what the market will bear regardless of actual financial requirements.
5. The launch vehicle (that I said was already available) must be man rated to some NASA requirement that really jack up the costs.
Emotionally I lean toward #4. Logically though I believe a mix of all the above was involved with no single culprit to be dealt with.–
I think it is mainly 4.
Basically SpaceX has a valuable [as it’s proven] rocket.
And there is no real competition as no other US company can be expected to provide NASA with crewed vehicle, in a timely and dependable manner. And NASA is hassle. Or SpaceX could make more money without dealing with NASA- as it has proven rocket which can get satellite customers.
One also say SpaceX is providing an argument- if you do things my way rather only NASA way, then it could be cheaper, but since NASA has decided [and may be adamantly required by congress in terms safetly concerns of crew trump everything] then NASA will pay for this costs.
It seems to me that Musk strongly wants to do NASA crew. The result of it will proven manned spacecraft- which money in bank if he wants to do Manned Mars.
He will be the guy with spacecraft that that’s got track record of working. He will be the go to guy for everyone who wants to go to Mars and has money to do it.
Boeing *could* do the same thing, but Boeing not interested. Or it’s familiar to what Bigelow is trying doing with space habitats- though it seems Musk would be a much better position.
So Musk wants the crewed contract as it relates to future plans, but it would have lost opportunity in terms of distraction from focusing more on satellite launches- or nearer term it’s a cost [and investment in long term].
Or as CEO who had convince shareholders, making lower priced bid, would be a hard sell.
But couple things wonder about, SpaceX has already got paid for crewed milestones, are we sure this contract is not including all of it- so it will be the total amount SpaceX will be paid related crewed launches.
Another thing is how falcon 9 are going to launched testing dragon, before crew are actually send on “the maximum 7 flights”.
Another aspect is maybe the reusable falcon 9 is not working out as well as expected.
Table revised thanks to Paul
The more money Musk can make off of NASA, the more money he has to develop his next generation of vehicles.
SpaceX likes lots of hardware testing, while NASA wants lots of documentation to make sure things are thought through.
Set aside at least $1billion for 6 ‘certified missions’ to ISS. And then $600million-$800million for documentation/reviews because NASA wants lots of documentation and SpaceX is trying to innovate a lot.
Leaving $800million to $1billion for actual development/tests. DragonFly program involves many drops and hops, with toxic propellants and parachutes. Test flight to ISS. Probably an uncrewed launch to orbit and back. Develop and test their own custom NDS docking adapter, since the NASA provided ones are incompatible with Dragon2. NASA is concerned with both Dragon2’s lack of radiation hardened electronics and SpaceX developing their own spacesuit, leading to lots of testing. Probably lots more too, since ‘NASA’ is paying.
Elon said at the Dragon2 reveal they might be able to finish for $500million. Big difference between getting the hardware good enough to try out versus demostrating and documentating the risks down to NASA acceptable levels.