Then I can’t help you, except again to point you at the leaked numbers that the WSJ got. With $5200M in expenses, you have to squeeze in 61 flights (one of them a Falcon Heavy), launching roughly 1700 of their own satellites (and these aren’t dinky cubesats), build a pair of Crew Dragons and a Cargo Dragon, continued Starship development, paid all the fixed costs of three orbital launch pads, and so on.
You can think they’re fudging the numbers on F9 launch costs if you like–I can’t prove you wrong with the available information. But I can say that if you believe that, then everything else that they demonstrably did has to fit into a smaller piece of the pie. Building 1700 satellites for a billion is already impressive–and you want to squeeze that further? Or a superheavy rocket program developing brand new engines, etc. At a certain point, you have to pick the option of least astonishment.
To be clear, I never said anything of the sort. I said it’s hard to make the numbers work unless they have cheap flights (even with the funding rounds). They lost about $600M in 2022, which presumably wasn’t a big deal due to the external funding (and in any case, they seem to have a decent amount of cash on hand).
If you tell me that you think their real launch costs are $60M each, then that means everything else they did in 2022 has to fit into $5.2B-(61*$60M) = $1.54B. It seems highly improbable that they built 1700 satellites, two crew-certified spacecraft, a cargo-certified craft, continued development on Starship, and all the other miscellaneous stuff with that.
If you tell me no, but it’s probably $40M per flight, then I’ll just shrug. That small difference isn’t worth quibbling over given the lack of information.
One suspects there is a big difference in the cost of a Starlink versus a random commercial launch. They will have long since amortised the integration costs and have a well oiled process. Given how many launches are Starlink this likely significantly sways the average cost.
Maybe something similar can be applied to Starship fueling launches. Once they have the process sorted that is. The money needed to get to that point will have a lot of uncertainty. So a lot of accounting wizardry might be needed to claim a low per launch cost.
The launch cost of $2m that Musk has floated as an ‘aspirational’ goal is variable cost only - fuel, labor, etc.
In reality, there are many other costs:
- cost of capital
- insurance
- amortized cost of vehicle construction
- Operations amortized over flights
- maintenance of towers, etc.
- Recovery of R&D cost.
- Profit/investor return
Let’s say a Starship can fly 20 times, and costs $100 million to build. That’s $5 million per flight right there.
If we assume Starship can fly for $20 million, which is a lot closer to reality than $2 million, then a fully fueled depot in orbit will cost about $150 million to $250 million. IF it takes all that fuel to send HLS to the moon and back, that’s still only a $300-$400 million mission, including an expendable HLS. That’s pennies when it comes to deep space exploration. Less than 1/5 the cost of a single SLS launch. And HLS might be reusable, cycling back and forth between Earth orbit and the lunar surface.
And that would put 100 tons on the moon. Apollo put about 15 tons on the moon.
FWIW, they’re shooting for a far higher reflight rate than that. Possibly in the thousands, though it’ll differ between the upper and lower stages. They’re already qualifying Falcon 9 for 40 reuses and that has far more limitations when it comes to degradation.
My suspicion is that their reuse limit will be gated by the reliability of landings, which are “only” about 98-99% for Falcon 9. That obviously limits the typical reuse rate to 50-100, even ignoring everything else. Though hopefully Starship will do better, given that most of the failures are related to the ocean landing, while Starship always lands back at the tower.
I was trying to be conservative. Maybe they’ll fly much more than that, or maybe it’ll turn out less. But we really don’t know yet.
So, to get back to the original question, is this the plan ?
-
Launch a Starship with a bunch of moon stuff on it to low earth orbit (LEO). Because it has a lot of moon stuff, it doesn’t have enough fuel to get to the moon
-
Launch another Starship without any payload except an extra fuel tank to LEO. Rendezvous this rocket with the first Starship and transfer the fuel to the first Starship. Second Starship returns to Earth
-
Repeat step 2 as many times as necessarily to fill the first Starship with fuel
-
First Starship flies to moon to deliver moon stuff
Starship can’t get to the Moon at all without refueling - even empty.
In space refueling sounds like a complication, but ultimately it makes everything easier. Without it, moon trips would be extremely mass limited, and you would need many kore trips to outfit an actual base. Much better to refuel in Earth orbit, then have plenty of margin for the lunar trips.
So long as we are using chemical rockets, we need to learn in-space refueling anyway if we want to do anything serious in space. Otherwise, we are bound by the tyranny of the rocket equation. So in-space refueling is a key technology, and we need to make it cheap(ish) and common and easy.
Thanks for the clarification. The depot seems like an extra step in an already complicated process, but I guess the advantage is that the depot remains in LEO for future flights?
That’s one advantage. But I think it may prove necessary to have the boiloff prevention systems. It’ll take a while to fill–months, probably–and they’ll want to reduce the losses as much as possible. It also gives the schedule much more flexibility, as the depot can loiter up there indefinitely if some other component is delayed.
Eventually, all this may not be necessary–once Starship is launching multiple times per day, there’s no boiloff problem. But it’ll take a while to hit that cadence.
It’s also good to do as many of the unmanned parts of the mission as possible before the manned parts, so if anything goes wrong with the unmanned parts, you don’t have humans up in space waiting for things to be fixed.
Yep. Ideally, all the unmanned stuff is prepped and ready to go before humans launch. That launch can always wait. As they say, it’s better to be on the ground and wish you were in space than to be in space and wish you were on the ground.
The crazy part for me is having the huge HLS fly to the Moon empty, have the astronauts fly out separately in a capsule, then dock with a station in RLHO orbit to transfer to HLS for the ride down to the surface.
The Moon mission would be cheaper and easier without SLS and Gateway in the mix at all. Just fuel HLS in LEO, load it with astronauts and fly it to the surface of the Moon. Fly back to LEO after. Astronuts transfer back to Earth on Dragon, HLS gets refueled in orbit for another Moon mission.
But then SLS would have no reason for existence, and we can’t have that.
Count your blessings that Richard Shelby is no longer in the picture and we can at least have a depot.
…So why did Senator Shelby oppose orbital fuel depots?
He was a senator from Alabama, which serves as a recipient for large amounts of NASA pork. And SLS was the biggest, fattest pig. Orbital depots reduce the need for SLS–even ignoring Artemis, they’d enable smaller commercial rockets to take over some missions. Shelby wanted to head all that off at the pass. No depots means zero chance at SLS being cancelled. The strategy worked, at least for as long as Shelby was senator–and more specifically, chair of the Senate Appropriations Committee, which gave him huge influence on these kinds of things.