back
5 comments
1,600M / 18 means $88.8M per launch.

Supposedly they've been selling launches for ~$74M per launch.

https://nextspaceflight.com/rockets/3/

If its profitable at $74M then I'd imagine its more profitable at $88.8M. There could be some extra costs for dealing with government customers, but I imagine a lot of those $74M launches were also government customers. Lots of estimates of their cost per launch is somewhere <$20M.

Even more significant is the timing, it's 18 launches by the end of 2027. Even before New Glenn and Vulcan broke they couldn't have scheduled a single launch before the end of 2027. Their manifests were full because they were years late.

SpaceX could easily charge more for timeliness. Its way more important than price per kg for a military customer.

The standard catalog price for a dedicated falcon 9 launch is $74 million. So this is actually up-charging the government a bit (I assume there is some custom work or extra cost for arranging for the appropriate security/secrecy).
These are their normal prices, its actually reasonable cheap for military launch. And they have a fantastic margin on Falcon 9, at least 50% likely more.
SpaceX claims they launch for $20 million and thier public minimum price is lower than the charge here
Now that they are publicly traded, there may be less scope for shenanigans in those numbers. If they really had such great margins on falcon 9, an improved, cost reduced falcon rocket would make a lot of sense. Instead they're betting at all on a rocket that has a daunting list of unmet project milestones before it does what it claims it will do.

That makes me think that the cost efficiency of falcon nine rests almost totally on streamlined operations, not technology. And that there might be a shell game going on regarding real costs.

Yeah so you just don't actually understand the economics.

The thing is, unless you massively increase the volume of Falcon 9, you are not going to get it that much cheaper. Its already incredibly cheap given that its made out of fancy aerospace material that needs to be shaped and welded with complex processes. If they invested a lot, could they get it a little cheaper? Sure maybe a few % ever year.

They have been redesigning the Falcon 9 over and over again to optimize its manufacture and re-usability. They are mass producing the second stage about faster and cheaper then any other rocket structure of that complexity and size. What you suggest they should do they have already done, multiple times.

But you run into limits. You could replace the Merlin engine with an engine of the same size that uses kerolox but is staged. Not much more you can do on the Merlin itself, except very marginal improvements. But if you do that you need to make the rocket longer, but Falcon is already on the upper limit of rocket of that diameter.

So you need to increase the width. So you are already doing a new staged engine, and now your doing a very different structure already. Now you also need to change the shape of your downcomer, you have different thermodynamics of reentry, you redesign the engine bay and its heat protection. Basically you are building a new rocket at that point and the payoff for such a huge effort, is marginal at best.

Their design has effectively reached a local maximum. Marginal improvements are possible and they are doing that but its not going to change fundamentally and its not going to get to their goal. Any new rocket they develop, must be more then a marginal improvement. Airbus and Boeing don't develop new planes unless the efficiency (therefore cost) improves by 20-30%.

To actually reach the next level of profit opportunity you need to achieve full re-usability. That the next major step that can unlock much cheaper price to orbit. That's where your capital investment can have actual real payoff. Because it makes much larger constellation possible.

> That makes me think that the cost efficiency of falcon nine rests almost totally on streamlined operations, not technology. And that there might be a shell game going on regarding real costs.

Yeah and how is that streamlined operation possible you think? Maybe it has to do that they have a reusable fleet of boosters that have been highly optimized to be able to land and reused quickly. Maybe it has to do with them optimizing the shit out of their launch infrastructure, like pioneering automatic flight termination and 100s of other things. That is the technology.

And there is no shell game going on, they make a lot of profit from Falcon 9 but reinvest everything in Starship, thus their rocket division overall is losing money.

You summed up the naïve approach to the cost and reusability issue:

> To actually reach the next level of profit opportunity you need to achieve full re-usability.

Upper stage reusability is hard, heavy, and expensive. At least 5X as hard as booster reusability. On top of that they have to bet that they can scale starship launches to the point where they can economically fly 15 starship refueling missions just to get one starship out of low orbit. As of now, completing an orbit at any elevation is a trust me bro thing.

Every one of Elon's gambles has five subsidiary gambles: if starship falls 30% short of expected payload capacity, it no longer takes 15 refueling missions but 25. That's just one example.

That's what you get when one of your software guys told you about agile, and you thought let's build a rocket that way.

Blue Origin requires refuelling.

Refuelling Isn't a requirement for most of the work they plan to do.

Refueling and moon mission is not primary motivation, Starlink and so on are.

You would likely have called it naive to try to develop reusability in the first place. Yes its harder. But they also have a much, much bigger team. They have an experience engine team. Simulation team and so on. And they have the need and the financial positioning.

Re-usability is the holy grail of space flight, if in the current situation SpaceX would not go after it they would be insane.

Its certainty smarter then your idea of an incrementally improved Falcon 9 upgrade.

> On top of that they have to bet that they can scale starship launches to the point where they can economically fly 15 starship refueling missions just to get one starship out of low orbit.

That just nonsense, where are you getting your information from. The 15 missions are for getting something from LEO to the moon, land and go back to Lunar orbit.

> As of now, completing an orbit at any elevation is a trust me bro thing.

No its a 'if you actually thing about it for 2s' then you can trust it thing. They have basically achieved orbital velocity and could have gone into orbit. This is beyond question.

The important and difficult thing is getting the ship to the ground and landing. And they have demonstrated precision landing after orbital speed insertion. They have even done extra hot approach to stress the heat shield even more, and have even left many tiles out to test what happens when tiles are missing.

If they actually did a full orbit or not is completely 100% totally and utterly irrelevant to anybody except a few people who hate Musk for political reasons.

> Every one of Elon's gambles has five subsidiary gambles: if starship falls 30% short of expected payload capacity, it no longer takes 15 refueling missions but 25. That's just one example.

And that only costs them a like 1 billion $ max if they can't do the lunar flights.

Even a 30% less Starship could be useful for Starlink. And worse case the can fly Starship in Falcon 9 mode and drop the upper stage and have fucking insane payload to Orbit.

> That's what you get when one of your software guys told you about agile, and you thought let's build a rocket that way.

You get the largest rocket ever built and the best engine ever built. The largest vehicle ever returned from orbit and so on.

Do you have any actual critic of their approach. Can you actually show that some other approach would have been cheaper? What are you basing that assumption on? Starship has been reasonably cheap and incredibly so given the problem its trying to solve and how much infrastructure they are building.

Why should I trust you rather then, you know the people at SpaceX who have revolutionized the space industry? Have launched and landed more rocket then anybody. Have better data then anymore. Have worked on this problem for a decade, and have changed their design a bunch of time. What expertise do you have in the development of fully reusable launch vehicles?