[XorNot]

I said this in another thread, but part of the issue is failure intolerance.

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.

[001sky]

This is true for a sub-set of unmanned payloads.

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).

[XorNot]

Right, but part of the reason we launch hugely expensive satellites is because launching a rocket is hugely expensive.

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).

[001sky]

^^^ Yes, this is exactly the 'white space' for Space-X.

[ef4]

The $2B satellite is equally absurd, and only costs that much for the same reasons that the launch costs $400M. Everything is bespoke and produced in the maximal number of congressional districts, with no meaningful competitive pressure to drive prices downward.

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.

[rjurney]

You really think SpaceX loads aren't insured? This seems impossible.

[HarryHirsch]

If you look at previous launches of new designs you will see that the first few flights are invariably not commercial, they carry research spacecraft. These are the qualification flights; until the vehicle has a good track record the satellite operator won't find anyone to insure their bird.

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.

[mapt]

Missing the target orbit by a significant margin on all their 9 flights?

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.

[lotsofmangos]

SpaceX appear to have gone with failure tolerance. Their first stage is practically a raid array.

[sitkack]

Teardown and rebuild doesn't increase your safety margins, they reduce it. If I rebuilt my Toyota every 10k miles I'd have a lot more breakdowns.