totm nov 2023 SpaceX Discussion Thread

[Ultimate Steve]

@Exoscientist on the Artemis thread, you were talking about how to best utilize Starship to carry payloads to the Lunar surface. I had originally written this for that thread but right as I was about to post it, the mods said to stick to Artemis stuff, so I'm posting here instead as no Artemis payload is as large as the numbers I'm working with.

As I'm about to post this, I just realized that, for some reason, I had jumped to cost per ton of cargo to the surface rather launch than cost per mission. If you are thinking cost per mission, you can ignore the rest of this post as I may have made a bad assumption. Under the new assumption of a single launch of an Apollo-style mission, then yes, the launch costs of a fully expendable Starship with reasonable assumptions (as described in scenario 3) would be lower than the launch costs of a fully reusable Moon-and-back Starship, but then you have to make the expendable Apollo-esque hardware cost less than what the difference would be, which would be a challenge.

Nevertheless, I spent too much time on the cost per ton analysis to throw it all away, so here is that analysis:

Quote

But a single expendable launch would only be $90 million by which because of the higher payload SH/SS could do single launch missions both to the Moon and Mars.

There are four valid mission profiles I can think of:

• Single launch expendable Starship + Super Heavy
• Single launch expendable Starship, reused Super Heavy
• Several refueling launches, 1 expendable Starship that stays on the Moon
• Several refueling launches, 1 reusable Starship that comes back from the Moon

And I am unsure which two mission profiles you are comparing. I have analyzed three (was originally two) of them below:

Spoiler

Global assumptions (mostly carried over from Exoscientist's post with a few of my own):

• Flight operations cost is 10 million dollars
• Expending a Starship costs 80 million dollars (Exoscientist's 90 million number minus operations cost)
• Expending a Super Heavy is free (only applicable to steel man the argument, obviously it isn't free, only used once)
• The baseline is 10 tanker missions to deliver 1200 tons of propellant to LEO
• Starship can hold 1200 tons of propellant
• Direct landing, no stopping at Gateway
• Negligible boiloff
• Negligible non-instantaneous burn losses
• Negligible other losses
• Depot launch costs not considered

Scenario 1 - Single Launch Expendable Everything

Optimism level: Extremely optimistic to steel man in Exoscientist's favor. Uses the most unrealistic performance numbers I could find.

Assumptions:

• 400 ton expendable payload (speculative unrealistic V3 Starship number from Elon's Twitter)
• 40 ton dry mass (speculative minimum dry mass no fairing deep space probe launching Starship number from a long time ago)
• 382s isp (Only vacuum Raptors firing in space with the highest numbers I've seen for them)
• 1 expendable launch (80 million dollars hardware, 10 million dollars fuel)
• LEO payload is interchangeable with fuel remaining in tanks

Results (Starship launches hydrolox lander to TLI):

• 150 tons of payload to TLI
• Total cost 90 million
• Massless 450s isp hydrolox lander can land 85 tons on the surface
• 1.05 million dollars per ton to the Lunar surface assuming hydrolox lander is free

Results (Starship lands directly on Lunar surface, landing engines, landing gear, and anything needed for multi day flight are massless or not needed):

• 150 tons of payload and propellant to TLI
• Total cost 90 million
• Starship delivers 56 tons to the Lunar surface
• 1.6 million dollars per ton to the Lunar surface

 

Scenario 2 - Refueled in LEO, Expendable Lander

Optimism level: Somewhat pessimistic to steel man in Exoscientist's favor. Uses numbers so pessimistic that the Artemis 3 mission profile becomes impossible.

Assumptions:

• 150 ton payload capacity (a bit high for the optimism level but we are assuming 120 tons of propellant payload already, not much of a stretch to suggest that without retaining ~12t landing prop, in total having 132 extra tons of propellant to burn, 150t could be achieved but I'll talk more about that later)
• 150 ton dry mass
• 360s isp (Both vacuum Raptor and sea level Raptor need to burn for control, both are falling significantly short of their isp targets)
• 10 refueling launches (total 100m)
• 1 expended lander (total 90m)

Results:

• 150 tons of payload to the Lunar surface (such a pessimistic starship may not be able to lift 150 tons to LEO, but it can take 150 tons from LEO to the surface when refueled)
• Total cost 190 million
• 1.27 million dollars per ton to the Lunar surface

 

Scenario 3 - Refueled in Elliptical Orbit, Reusable Lander

Optimism level: Middle of the road (an attempt to use reasonable numbers)

Assumptions:

• 120 ton dry mass for both refuelers and lander (lander would probably be slightly heavier but makes math easier)
• Average vacuum isp of 367s (both Vactor and SLaptor fire at more or less their public isps)
• Staging orbit is an elliptical orbit 800m/s closer to the Moon (more work shifted onto tankers so the ship has to do less work)
  • This number is not optimal and I pulled it out of thin air, I could write an optimizer but I don't feel like it
• Starship can get 100 tons of payload to the intermediate orbit without refueling in LEO first (a bit of a stretch but if it doesn't work it needs 1 more tanker flight max)
• Tanker starship's LEO heat shield can handle 800m/s elliptical orbit

Intermediate steps:

• Assuming 300m/s needed for de-orbit and landing (conservatively high as shuttle de-orbited with like 90m/s and Starship's terminal velocity is 80m/s, this accounts for residuals, gravity losses, and Starship burning off-axis for part of its landing burn), a 120t dry tanker would need 12 tons of reserve propellant for landing.
• If such a tanker could originally deliver 120 tons of propellant, to enter the 800m/s elliptical orbit, it would need to use about 50 tons of propellant, leaving it with only 70 tons to deliver to the ship
• This increases the number of tanker flights to 17, costing 170 million dollars
• Assuming Earth landing fuel for tanker and lander is the same even though lander doesn't have to do a de-orbit burn because I'm lazy and it's in Exoscientist's favor

Results:

• Starship masses 367 tons upon landing on the Lunar surface (120t dry mass, 12t landing propellant, 135t return propellant, 100 tons of payload)
• Starship delivers 100 tons of payload to the Lunar surface
• Starship masses 267 tons upon taking off from the Lunar surface (120t dry mass, 12t landing propellant, 135 return propellant)
• Total cost 180 million
• 1.8 million dollars per ton to the Lunar surface

 

The expendable Starship and expendable Starship+Super Heavy analyses were effectively combined into one optimistic one using the everything expendable payload numbers but the just Starship expendable cost numbers.

Even with the incredibly optimistic numbers for the expendable version and more pessimistic numbers for the semi-reusable version (normal numbers were used for the fully reusable version if you didn't read the analysis), refueled Starship always was capable of getting more cargo to the Lunar surface than single launch expendable Starship.

For cost, discounting the crazily optimistic zero cost zero mass hydrolox lander I had included for comparison, the analysis, heavily weighted in favor of expendable Starship, gave a cost per ton of 1.6 million dollars per ton to the Lunar surface for expendable Starship and 1.27 million dollars per ton to the Lunar surface for reusable tanker, expendable lander refueled Starship.

I used the most outlandish numbers we have and semi-reusable still came out cheaper, so I can only assume Exoscientist was comparing to fully reusable.

I did the third (fully reusable) analysis mostly to compare to semi-reusable. The optimism level is what I would call realistic as when I tried to use the pessimistic numbers, Starship couldn't return from the Moon at all, even to the Gateway orbit (starting from LEO at least). I found it interesting that with the cost numbers used, it is cheaper to expend the lander even with the pessimistic numbers (for starters, that 150t dry mass could be stripped down as that was originally chosen assuming it would come back but it ended up not being able to to the analysis was swapped to expendable). I didn't expect this to be the case.

Fully expendable vs fully reusable wasn't my intention when creating the third analysis, but for completeness, expendable only just barely edges out reusable in cost per ton with all of those ridiculous assumptions (1.6 vs 1.8 million per ton) and in reality, reusable would almost surely win, but having to conduct fewer launch operations might make it an attractive option (if the numbers were actually physically possible).

I'll also reiterate that the 800m/s elliptical orbit is a number I chose from thin air because it sounded alright, I'm doubtless of by at least a hundred or two m/s from the optimal, it is possible that fully reusable beats the hyper-optimistic fully expendable profile with a more intelligently chosen parking orbit because it is only off by 13% or so.

 

TLDR, with current cost numbers and my assumptions, unrealistically optimistic expendable is barely cheaper per ton than realistic reusable. Unrealistically pessimistic reusable-tanker expendable-lander beats both of them by a significant margin, also leading both in payload per landing. With these assumptions, it makes sense to use expendable Starships for cargo, but reusable Starships to transport crew back and forth (and to return samples) are not that much more expensive.

[tater]

Most dry mass figures for HLS are closer to 80-90t (no tiles, no flaps, not strengthened for Earth EDL).

[Exoscientist]

  There is speculation on space forums the reason Starship just barely makes orbit with 0 payload in IFT-3 when it should have payload capacity of 100 to 150 tons is perhaps it was only partly fueled.

But SpaceX has said it was fully fueled:

SpaceX @SpaceX
“Propellant loading complete; Starship is fully loaded with more than 4500 metric tons (10 million pounds) of propellant”
https://x.com/SpaceX/status/1649041734062714882?s=20

 Then the question remains: when payload capability is supposed to be 100-150 tons, why does a fully fueled SuperHeavy/Starship just barely make orbit(actually slightly less) carrying no payload, fully expending its propellant?

 Think of it this way, what SpaceX demonstrated with IFT-3 was a launcher with a payload to LEO capability of 0 tons even when fully fueled and fully expending its propellant. Then how can it do Artemis Starship HLS refuelings when it gets 0 tons to LEO?

  Bob Clark

Edited March 24 by Exoscientist

[Ultimate Steve]

5 minutes ago, Exoscientist said:

  There is speculation on space forums the reason Starship just barely makes orbit with 0 payload in IFT-3 when it should have payload capacity of 100 to 150 tons is perhaps it was only partly fueled.

But SpaceX has said it was fully fueled:

SpaceX @SpaceX
“Propellant loading complete; Starship is fully loaded with more than 4500 metric tons (10 million pounds) of propellant”
https://x.com/SpaceX/status/1649041734062714882?s=20

 Then the question remains: when payload capability is supposed to be 100-150 tons, why does a fully fueled SuperHeavy/Starship just barely make orbit(actually slightly less) carrying no payload, fully expending its propellant?

 Think of it this way, what SpaceX demonstrated with IFT-3 was a launcher with a payload to LEO capability of 0 tons even when fully fueled and fully expending its propellant. Then how can it do Artemis Starship HLS refuelings when it gets 0 tons to LEO?

  Bob Clark

Is this speculation credible?

We do know they had to at least have enough propellant on board for the prop transfer demonstration, the contract specified at least 10 tons and they probably had more left, as they had planned to relight an engine for a few seconds, though they cancelled that due to bad angular velocity. That's not a lot but it is a lot more than zero.

On IFT-2 they had too much propellant (likely a re-entry constraint and they didn't want to vent it in orbit for whatever reason) and venting it caused the failure. I don't remember if they dumped on ascent during IFT-3 or not, but if they did it would give the illusion of Starship barely making orbit.

Or maybe they found some other way to burn fuel (Gimbal Raptors all the way out for cosine losses, early stage separation, etc) to not have to vent stuff at all.

If this speculation is true, V2 Starship is right around the corner and the V1 ships being a little overweight probably isn't the end of the world.

[darthgently]

16 minutes ago, Exoscientist said:

  There is speculation on space forums the reason Starship just barely makes orbit with 0 payload in IFT-3 when it should have payload capacity of 100 to 150 tons is perhaps it was only partly fueled.

But SpaceX has said it was fully fueled:

SpaceX @SpaceX
“Propellant loading complete; Starship is fully loaded with more than 4500 metric tons (10 million pounds) of propellant”
https://x.com/SpaceX/status/1649041734062714882?s=20

 Then the question remains: when payload capability is supposed to be 100-150 tons, why does a fully fueled SuperHeavy/Starship just barely make orbit(actually slightly less) carrying no payload, fully expending its propellant?

 Think of it this way, what SpaceX demonstrated with IFT-3 was a launcher with a payload to LEO capability of 0 tons even when fully fueled and fully expending its propellant. Then how can it do Artemis Starship HLS refuelings when it gets 0 tons to LEO?

  Bob Clark

There are so many variables.  They could have selected a launch profile that used more fuel so as to have less to deal with for this specific set of tests.  People can speculate on the skeptical side all they want, of course, but it seems reasonable that they'd eventually take the historical record of SpaceX skeptics ending wrong and SpaceX making wonderful strides in spite of the skepticism into account at some point.  Jeerleaders are expected and necessary to an extent for iron to sharpen iron, but when it veers into the unreasonable it undermines humanity's progress and motivations must be considered

[JoeSchmuckatelli]

Yeah - the 'barely made orbit' was the plan for this flight.  

[tater]

 

[magnemoe]

5 hours ago, JoeSchmuckatelli said:

Yeah - the 'barely made orbit' was the plan for this flight.  

This, they did not want to risk it getting stranded in orbit and falling down some random place who was wise as they got problems controlling it. Splashdown point was changed, but I strongly doubt this had to do with range some other reason. Perhaps other zone was in use its an test range after all. Or launch restrictions. 

[Exoscientist]

Marcus House speculates in response to a question from one of his viewers that the reason the SH/SS just barely made orbit on IFT-3 when it had 0 payload, and how could it reach  orbit when it does have large payload, that perhaps it was only partially fueled. See at the 5:18 point here:

SpaceX's Frantic Push to Launch the Next Starship Mission is Nuts!
https://youtu.be/1HAcza0nE34

But actually SpaceX prior to the launch said on IFT-3 SH/SS was fully fueled:

SpaceX @SpaceX
“Propellant loading complete; Starship is fully loaded with more than 4500 metric tons (10 million pounds) of propellant”
https://x.com/SpaceX/status/1649041734062714882?s=20

 Bob Clark

[magnemoe]

11 minutes ago, Exoscientist said:

Marcus House speculates in response to a question from one of his viewers that the reason the SH/SS just barely made orbit on IFT-3 when it had 0 payload, and how could it reach  orbit when it does have large payload, that perhaps it was only partially fueled. See at the 5:18 point here:

SpaceX's Frantic Push to Launch the Next Starship Mission is Nuts!
https://youtu.be/1HAcza0nE34

But actually SpaceX prior to the launch said on IFT-3 SH/SS was fully fueled:

SpaceX @SpaceX
“Propellant loading complete; Starship is fully loaded with more than 4500 metric tons (10 million pounds) of propellant”
https://x.com/SpaceX/status/1649041734062714882?s=20

 Bob Clark

As other said, they did not need more I suspect they did not run superheavy on full trust, extra capacity is not needed so take it a bit gently. Was release point earlier than expected? 

[darthgently]

I wonder how much fuel was in the booster at hot stage?  Maybe more than required for boostback and landing burn.  I could see a fully fueled ascent being a very high priority while not being as concerned with extra mass during the ocean "landing"

[tater]

They're probably lying, because by lying about Starship capabilities they make huge amounts of money. Somehow. For reasons. It adds to the money their colony of underpants gnomes is stockpiling.

 

[PakledHostage]

1 hour ago, tater said:

They're probably lying, because by lying about Starship capabilities they make huge amounts of money. Somehow. For reasons. It adds to the money their colony of underpants gnomes is stockpiling 

It's 4 O'clock somewhere? Or what is this? Some jargon I don't understand?

[Exoscientist]

17 hours ago, Ultimate Steve said:

Fully expendable vs fully reusable wasn't my intention when creating the third analysis, but for completeness, expendable only just barely edges out reusable in cost per ton with all of those ridiculous assumptions (1.6 vs 1.8 million per ton) and in reality, reusable would almost surely win, but having to conduct fewer launch operations might make it an attractive option (if the numbers were actually physically possible).

I haven’t examined your calculations in detail but note both of these values would be in the range of the current least price to just to get to LEO by the reused Falcon 9. That would be a marked advance since multiple nations now are aiming for landed missions to the Moon. The cost would be in the ca. $100 million per launch range, comparable to just manned missions to LEO, rather than the multi-billions per launch of the SLS.

  Bob Clark

[tater]

4 hours ago, PakledHostage said:

It's 4 O'clock somewhere? Or what is this? Some jargon I don't understand?

 

 

Edited March 25 by tater

[magnemoe]

8 hours ago, tater said:

They're probably lying, because by lying about Starship capabilities they make huge amounts of money. Somehow. For reasons. It adds to the money their colony of underpants gnomes is stockpiling.

Few in the space industry seams to believe in Starship in the first place, but the the only hard part is second stage recovery who is not really needed to make Starship work out better than falcon 9. 

[.50calBMG]

Just because they said fully fueled doesn't mean 100% full. There's plenty of rockets that will partially fuel the tanks for certain variants or flight profiles. Fully fueled in this case could just mean "fully fueled for this mission profile"

[Nuke]

thing is capability is an x factor. i dont think anyone here or at spacex for that matter knows what this design is actually capable of at this point. that must be determined in testing.

Edited March 25 by Nuke

[sevenperforce]

1 hour ago, .50calBMG said:

Just because they said fully fueled doesn't mean 100% full. There's plenty of rockets that will partially fuel the tanks for certain variants or flight profiles. Fully fueled in this case could just mean "fully fueled for this mission profile"

Yep, that's precisely what they meant.

Superheavy takes 3,400 tonnes of propellant while Starship takes 1,200 tonnes of propellant. Fully loaded, that would be a total of 4,600 tonnes of propellant. Yet SpaceX said "fully loaded with more than 4500 metric tons of propellant"; we can therefore pretty readily deduce that "fully fuelled for this mission profile" meant 100 tonnes LESS propellant in Starship than in an orbital mission profile.

[AckSed]

On a tangent: is anyone else pursuing the mass-production of satellites and, more importantly, satellite sensors suitable for using on probes?

Starlink and its sister project Starshield seems like it's not only a turnkey solution for communication and surveillance, but a full satellite bus for any customer to place their satellite/probe on.

Project Kuiper is 'just' a Starlink competitor, but it could be turned to the same ends.

OneWeb is... lagging behind.

If you can figure out how to mass-produce a generic suite of planetary science (like narrow and wide-field cameras with multiple filters, spectrographs and radar) that fits onto a cheap satellite, you'll have a product.

Now I think about it, the real thing stopping a swarm of sats being placed around Mars, Venus or the outer planets is now not hardware, it's the lack of ground-support: receiving signals, upkeep, debugging, interpreting data. SpaceX is already doing that with over 6,000 sats and rising.

If someone could figure out how to run a planetary science sat swarm (P3S for short) on minimal personnel, such that it could be sustained with throughput over 5 years for half the current cost, then you have an attractive product.

 

[tater]

 

[tater]

 

 

[tater]

 

[Ultimate Steve]

I wasn't aware that they could safely fire the Vactors at sea level.

[tater]

23 minutes ago, Ultimate Steve said:

I wasn't aware that they could safely fire the Vactors at sea level.

They've fired them before