That. You need to re-qualify the engine after each launch. The cost of labour far exceeds the cost of materials, you need to strip and re-build the thing, and you ask yourself just how much money you can save through re-using the first stage. I don't think it's much.
But the Merlin engines have been designed from the start to be reusable and dependably reignitable, whether at sea level or upper atmosphere. They're pretty amazing-- essentially the pinnacle of +50 years of engine design. Tom Mueller is a huge badass. :) http://en.wikipedia.org/wiki/Merlin_(rocket_engine)
And check out that bad boy roar: http://www.youtube.com/watch?v=wkdReoxGHG8
I don't have a link on hand, but Elon has said multiple times that the majority cost of the Falcon-9 is the raw material. One of the reasons they've been able to drop the price significantly is by using more efficient manufacturing techniques, requiring less material and creating less waste. (e.g. stir-friction welding)
It's also worth noting that Falcon-9 can still complete a mission with one (and perhaps two?) engine failures, which gives quite a more comfortable margin of error. There's a big difference between "perfect" and "near perfect" when it comes to engineering these things. ;)
SpaceX disagrees. They expect to cut the cost of launch by 1 order of magnitude
https://en.wikipedia.org/wiki/SpaceX_reusable_launch_system_...
That's how politically-designed boondoggles like the shuttle work. It's not inevitable, if you actually spend the engineering effort on reliability and repeatability.
If NASA operated an airline, they would probably be tearing down and rebuilding every jet engine after every flight, and a ticket would cost $100,000.
Sure - for human payloads I'd want to be damn sure my process is good, like, hundreds to thousands of missions deep before I trusted it.
But that's totally unnecessary for an unmanned payload! If the cost of launch drops enough, you can fully justify launching 2x the payloads if you expect maybe 1 in 10 failures due to the mode of launch.
I suspect it's possible to do a damn site better then that, but for NASA its never been an option. If it's reusable, they can't let a mission fail because they'll only get punished and funded according to the failures, even if they specced everything expecting 1 mission to possibly not go off right.
Space-X is great for the sub-set of things that don't require insurance, and don't therefore need to use corner-case mil-spec stuff with the corner case pricing.
For others, the cost of losing a $1-2B bird on top of a $100m/cheap rocket is shitty math. Nobody is going to insure the top of it, so ultimately the "waste" is akin to a form of insurance.
Everybody knows this already, so I'm not sure how sympathetic a hearing it is going to get. It will be great PR though to hopefully spurn <designs> that fit the new framework...and thus expand the market for space-x and hopefully limit the superflous use of corner-case technology for mundane/run-of the mill applications (at the tax payer's expense).
Now the Falcon 9 is an all-new design, and their track record isn't all that shiny - there is a history of missing the target orbit by a fairly significant margin. It's going to be downvoted, but it had to be said.
I wouldn't be surprised if this flight went without insurance.
Or having their failsafe system activated after one of their engines blew up on a single mission, causing them to be in position to release the small secondary payload (which the customer did not particularly need in free flight apparently) slightly late, in a zone in which NASA had veto rights on due to proximity to the ISS. Rather than adjusting their orbit to deposit the secondary payload in the proper orbit, as is technologically feasible, NASA exercised their contractual veto and told them to drop it early.
That calculus changes if your rocket launches start to get cheaper. It changes by a lot if your rocket launches have reasonable but predictable failure modes - which is something you get from volume.
As it is, rocket launches are relatively infrequent and expensive - which means its impossible to figure out the amortization of costs, and its not worth building a 10m satellite if your launch costs 100m (since if you can raise the latter, you can almost certainly get more for a better satellite too).
And the two expenses buoy each other: if launches were inexpensive, you could spend a lot less on fault-tolerance in the payloads, because you could just launch a lot more payloads or even support on-orbit repair infrastructure.
You should send them this post of yours so they stop wasting all that money! I bet they never thought of what you said.