Starship Artemis Contract & Lunar Starship

[Vultur]

To me, June 2026 prop transfer feels very ambitious, but given that, the other two dates feel unambitious. Why 15 months between HLS Demo and real HLS?

Testing full size ECLSS? Testing hatch plus elevator crane to lunar surface? Lots of human factors considerations to address during that time, plus training the crew!

True, but I'd think things like ECLSS could be developed in parallel : it wouldn't need to wait on the HLS Demo landing.

[JIS]

https://x.com/audrey_decker9/status/1989352112728510935

SpaceX’s new tentative schedule for HLS, per an internal document I obtained:
- Prop transfer June 2026

That seems reasonable, assuming Flight 12 successfully completes a suborbital launch in early Q1 with the V3 Starship, Flight 13 successfully tower catches in late Q1, first tanker launched into orbit in Q2 and propellant transfer Starship launches in late Q2.

Catching the ship isn't even on the critical path, they can build ships faster than that already and the transfer demo is going to require some specific hardware that will probably not be present on the first couple V3s.  As long as the V3 transition goes better than V2, this isn't even an ambitious schedule.

I agree on all your points.

They can work on ship recovery while still checking off other developments.  A prop transfer in June 2026 is the most aggressive part of the schedule I think.

I think a demo flight and Artemis 3 landing are totally reasonable.  Building and flying HLS isn't as hard as getting the prop into LEO.

Once the prop transfer is successfully sorted the pace could really accelerate.

Isn't this naive expectation? V3 will need some tinkering, refueling will need a lot of tinkering. Long loiter, deep space operation, moon landing, moon operation and ascend...

All of these tests will need a lot of flights, dozens of flights. Maybe one hundreds of flights all based on major engine iteration of Raptor 3...   

[JIS]

It certainly is not increasing speed of development.

How do you know that? I don't see any other fully reusable SHLVs getting developed faster.

https://forum.nasaspaceflight.com/index.php?topic=50806.msg2733845#msg2733845

You might want to elaborate. By the dates you provide, LC-39 took about 5 years ('62 -'67) from breaking ground to first flight & orbit, and 7 years to reach 4/yr rate ('69). Starbase Pad 1 took about 2.5 years to first flight ('21-'23), 3 years to reach the target orbit, and less than 5 years to 4/yr rate ('24).

Starship has been constrained by flight hardware availability much more than pad availability - and perhaps exclusively by flight hardware. This is evident from the fact that they were stacking development hardware on a flight pad with B4/S20.

No, you have it wrong. Apollo build all their pads they ever needed for all Lunar landings in about 4-5 years. SpaceX build their first iteration pad maybe in 2.5year, but it was completely useless for their stated goal. It got destroyed after the first launch and nearly destroyed the Starship before the liftoff. It had to be heavily modified for second launch and needs to be rebuild again after 10launches. All the while they need maybe dozens launches for a single moon mission. So 5 years after they got HLS contract and 10years after they announced starship they might have the first proper launch pad but they still need to finish more launch pads to be able to execute the moon mission.

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place. They practically put KSC infrastructure on hold soon after 2021. Still I think they might have 2 pads ready at KSC for 2027 with rapidly reusable boosters and hopefully solved refueling. So maybe 6 years after the HLS contract the infrastructure could be ready.       

[Brigantine]

Audrey Decker
@audrey_decker9
SpaceX’s new tentative schedule for HLS, per an internal document I obtained:
- Prop transfer June 2026
- Uncrewed lunar landing June 2027
- Crewed lunar landing Sept 2028

I'm wondering how V3 "Next Gen" and V4 "Future" fit into this.

Prop transfer with V3 seems the only plausible option
Uncrewed landing with a V3 HLS and a mix of legacy V3 and early experimental V4 tankers?
Nominally plan for a crewed landing with a 'V3.5' HLS and exclusively V4 depots and tankers?
(extant V3 relegated to Starlink and commercial, anything like an experimental V5 omitted from this particular mission)

Or will they stick with V3 for longer?
It may depend on how the early V3 performance matches up, and what sort of Final Tanking Orbit it corresponds to

The long pole may be 'man-rating' (for on-orbit and lunar ops) an HLS design incorporating lessons from the uncrewed landing(s)

[DanClemmensen]

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place.

Everybody in the space industry and at NASA knew that the 2024 date was a fantasy for a contract award in 2021. SpaceX could either decline to bid at all, or bid and win and then ignore the formal schedule and make their best effort at executing. They chose the latter. Note that SLS, Orion, and the space suits also missed their schedules. Do you think that either of the other two HLS bids would have made the schedule?

[JIS]

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place.

Everybody in the space industry and at NASA knew that the 2024 date was a fantasy for a contract award in 2021. SpaceX could either decline to bid at all, or bid and win and then ignore the formal schedule and make their best effort at executing. They chose the latter. Note that SLS, Orion, and the space suits also missed their schedules. Do you think that either of the other two HLS bids would have made the schedule?

Then what is the problem? That Chinese will make their flags and footprint mission first? That is so stupid. Just land some automatic probe or lander at the south pole, keep it in operation and claim that landing spot for yourself. No need to launch your own useless flags and footprint PR stunt.

SpaceX HLS lander architecture is excessively complex and non-optimised for the Moon. But why should we care? Once SpaceX succeeds creating fully and rapidly reusable Starship and will make hundreds of launches per year, it will be so easy to make one landing every two years. Who cares whether it happens within 5 or 10 years? NASA is waiting for this mission for 50 years already.     

But maybe the problem will solve itself as New Glen is doing very good progress and MK1 is nearly ready. So multiple MK1s with MK2 crew cabin could be the best shot for Flags and Footprints. 

[Robotbeat]

I think the Starship HLS architecture is simpler than the Blue Moon one. Number of launches isn’t architectural complexity but instead number of unique elements.

[DanClemmensen]

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place.

Everybody in the space industry and at NASA knew that the 2024 date was a fantasy for a contract award in 2021. SpaceX could either decline to bid at all, or bid and win and then ignore the formal schedule and make their best effort at executing. They chose the latter. Note that SLS, Orion, and the space suits also missed their schedules. Do you think that either of the other two HLS bids would have made the schedule?

Then what is the problem?

I have no problem. You stated that it is a problem.

Apparently, Duffy, et.al. thinks it's a problem that needs to be solved, apparently by using another fantasy architecture that magically works by creating a new alternative and dictating a three-year schedule.

In my opinion SpaceX has a high probability of being ready for Artemis III by the end of 2028. SLS/Orion will be ready to execute a high-risk Artemis III mission if they succeed with the high-risk Artemis II mission. If these missions fly, I really hope they succeed, or at least that the crews survive. Success is likely, it's just that IMO the probability of failure is unacceptable.

[equiserre]

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place.

Everybody in the space industry and at NASA knew that the 2024 date was a fantasy for a contract award in 2021. SpaceX could either decline to bid at all, or bid and win and then ignore the formal schedule and make their best effort at executing. They chose the latter. Note that SLS, Orion, and the space suits also missed their schedules. Do you think that either of the other two HLS bids would have made the schedule?

Then what is the problem?

I have no problem. You stated that it is a problem.

Apparently, Duffy, et.al. thinks it's a problem that needs to be solved, apparently by using another fantasy architecture that magically works by creating a new alternative and dictating a three-year schedule.

In my opinion SpaceX has a high probability of being ready for Artemis III by the end of 2028. SLS/Orion will be ready to execute a high-risk Artemis III mission if they succeed with the high-risk Artemis II mission. If these missions fly, I really hope they succeed, or at least that the crews survive. Success is likely, it's just that IMO the probability of failure is unacceptable.

That is a very important point when comparing arquitectures.
Apollo seems to be the yardstick, but what was the real pLOC of Apollo? was is 1/300 like shuttle design? or more like the shuttle actual pLOC?
And what is the requirement today for the Artemis program?

[OTV Booster]

Note the first V3 ship has docking port fixtures (well part of it), so I think it’ll be a standard part of V3 hardware.

I suspect it's not "standard". The first several Ships will probably be multifunction test prototyes and can test propellant transfer, Starlink deployment, and long-term propellant storage, as needed. Going forward after testing, Starlink Pez does not need refill. Pez that will go to Mars for Marslink will need them but not Starlink Pez.

One firm requirement for testing transfer is the ability to get a second ship on orbit while the first ship is still alive and well. My eyes will be on the launch mount next launch.

Then watch for two ships coming together about the same time. Boosters too.

I think SpaceX specifically said they intended to do a combined test, where the first Ship would go up and demonstrate that it can stay in orbit for awhile without excessive boil-off, and the second ship would then bring up more propellant and do the propellant transfer. So yes, the first ship must remain alive and well. Presumably, if the second ship is delayed for too long or if boil-off is worse than expected, they will be forced to de-orbit the first ship without completing the transfer test. So the plan is for weeks, not days.

If the first grab attempt calls for 24hours on orbit that will IMO, be enough time for a rough boiloff characterization. This would be enough to decide on max loiter time for the transfer test.


In an ideal world (stop laughing) loiter would allow time for two tests. Schedule an early test and if the first tanker doesn't make it for some reason, chances are the second will and the program is set back only (loiter time)/2.


If either tanker works out it forms the foundation for a full topoff test. It's time for the new sausage factory to start cranking out more sausages.


The big excitement here is zeroing in on total number of tanker flights needed for a mission and maybe a refinement on if/where a second transfer will happen.

[Robotbeat]

I’m more concerned about the uncrewed demo mission. I’d prefer there to be 2 of those for greater safety and in case one runs into problems.

But those are expensive.

[DanClemmensen]

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place.

Everybody in the space industry and at NASA knew that the 2024 date was a fantasy for a contract award in 2021. SpaceX could either decline to bid at all, or bid and win and then ignore the formal schedule and make their best effort at executing. They chose the latter. Note that SLS, Orion, and the space suits also missed their schedules. Do you think that either of the other two HLS bids would have made the schedule?

Then what is the problem?

I have no problem. You stated that it is a problem.

Apparently, Duffy, et.al. thinks it's a problem that needs to be solved, apparently by using another fantasy architecture that magically works by creating a new alternative and dictating a three-year schedule.

In my opinion SpaceX has a high probability of being ready for Artemis III by the end of 2028. SLS/Orion will be ready to execute a high-risk Artemis III mission if they succeed with the high-risk Artemis II mission. If these missions fly, I really hope they succeed, or at least that the crews survive. Success is likely, it's just that IMO the probability of failure is unacceptable.

That is a very important point when comparing arquitectures.
Apollo seems to be the yardstick, but what was the real pLOC of Apollo? was is 1/300 like shuttle design? or more like the shuttle actual pLOC?
And what is the requirement today for the Artemis program?

Apollo attempted seven landing missions. Six successful, one very nearly catastrophic. You decide. I'd call it about 15% pLOC.

[OTV Booster]

https://x.com/audrey_decker9/status/1989352112728510935

SpaceX’s new tentative schedule for HLS, per an internal document I obtained:
- Prop transfer June 2026

That seems reasonable, assuming Flight 12 successfully completes a suborbital launch in early Q1 with the V3 Starship, Flight 13 successfully tower catches in late Q1, first tanker launched into orbit in Q2 and propellant transfer Starship launches in late Q2.

Catching the ship isn't even on the critical path, they can build ships faster than that already and the transfer demo is going to require some specific hardware that will probably not be present on the first couple V3s.  As long as the V3 transition goes better than V2, this isn't even an ambitious schedule.

I agree on all your points.

They can work on ship recovery while still checking off other developments.  A prop transfer in June 2026 is the most aggressive part of the schedule I think.

I think a demo flight and Artemis 3 landing are totally reasonable.  Building and flying HLS isn't as hard as getting the prop into LEO.

Once the prop transfer is successfully sorted the pace could really accelerate.

Isn't this naive expectation? V3 will need some tinkering, refueling will need a lot of tinkering. Long loiter, deep space operation, moon landing, moon operation and ascend...

All of these tests will need a lot of flights, dozens of flights. Maybe one hundreds of flights all based on major engine iteration of Raptor 3...   

Dozens? Yes. A hundred? No. This is based on past performance. Early this year was the anomaly, not the norm.

[OTV Booster]

It certainly is not increasing speed of development.

How do you know that? I don't see any other fully reusable SHLVs getting developed faster.

https://forum.nasaspaceflight.com/index.php?topic=50806.msg2733845#msg2733845

You might want to elaborate. By the dates you provide, LC-39 took about 5 years ('62 -'67) from breaking ground to first flight & orbit, and 7 years to reach 4/yr rate ('69). Starbase Pad 1 took about 2.5 years to first flight ('21-'23), 3 years to reach the target orbit, and less than 5 years to 4/yr rate ('24).

Starship has been constrained by flight hardware availability much more than pad availability - and perhaps exclusively by flight hardware. This is evident from the fact that they were stacking development hardware on a flight pad with B4/S20.

No, you have it wrong. Apollo build all their pads they ever needed for all Lunar landings in about 4-5 years. SpaceX build their first iteration pad maybe in 2.5year, but it was completely useless for their stated goal. It got destroyed after the first launch and nearly destroyed the Starship before the liftoff. It had to be heavily modified for second launch and needs to be rebuild again after 10launches. All the while they need maybe dozens launches for a single moon mission. So 5 years after they got HLS contract and 10years after they announced starship they might have the first proper launch pad but they still need to finish more launch pads to be able to execute the moon mission.

But wait, I do not criticise the decision to build minimalistic launch pad first. The problem is they made a promise to launch HLS in 2024 (2025) while doing "minimal" efforts to have necessary infrastructure in place. They practically put KSC infrastructure on hold soon after 2021. Still I think they might have 2 pads ready at KSC for 2027 with rapidly reusable boosters and hopefully solved refueling. So maybe 6 years after the HLS contract the infrastructure could be ready.       

Is SpaceX behind the curve? Yup. Is this abnormal? No. Does the contract have penalties for late? IDK. If there are no penalties I'd expect that the NASA leadership at the time of contract signing expected delays and thought nothing of it.


What about the late spacesuits? What about Orion adopting an EDL kludge to keep momentum? Even with the kludge I think they're running late.


If NASA chooses to cancel the SpaceX contract, so be it. Chances are even that SX would still beat NASA to a crewed lunar landing.

[TheRadicalModerate]

I think the Starship HLS architecture is simpler than the Blue Moon one. Number of launches isn’t architectural complexity but instead number of unique elements.

Architecturally, they're almost identical.  SpaceX uses a tanker, a depot, and an HLS, all based on a common propulsion technology and launcher, with refueling occurring before the crew boards.  Blue uses a tanker (a GS2), a depot (the Cislunar Transporter), and an HLS, all based on a common propulsion technology (BE-7) and launcher, with refueling occurring before the crew boards.

It's fair to argue that ZBO for hydrolox is a lot more complicated than ZBO or near-ZBO for methalox.  But it's equally fair to argue that an architecture that requires 3-5 launches is simpler than one that requires 10-15 launches.

The big difference between SpaceX and Blue is that SpaceX is ahead of Blue by 2-4 years.

[OTV Booster]

Audrey Decker
@audrey_decker9
SpaceX’s new tentative schedule for HLS, per an internal document I obtained:
- Prop transfer June 2026
- Uncrewed lunar landing June 2027
- Crewed lunar landing Sept 2028

I'm wondering how V3 "Next Gen" and V4 "Future" fit into this.

Prop transfer with V3 seems the only plausible option
Uncrewed landing with a V3 HLS and a mix of legacy V3 and early experimental V4 tankers?
Nominally plan for a crewed landing with a 'V3.5' HLS and exclusively V4 depots and tankers?
(extant V3 relegated to Starlink and commercial, anything like an experimental V5 omitted from this particular mission)

Or will they stick with V3 for longer?
It may depend on how the early V3 performance matches up, and what sort of Final Tanking Orbit it corresponds to

The long pole may be 'man-rating' (for on-orbit and lunar ops) an HLS design incorporating lessons from the uncrewed landing(s)

ISTM it boils down to how fast and radially the versions shift and how soon they plan on landing.

If V3 works out to be dependable and 'adequate' in some sense it could make sense to stick with it for crew. If performance is less than desired, refueling is conceptually flexible. Unfortunately more tankers means higher cost.

IMO, doing Mars without doing the moon works at small scale but at the scale Elon intends, lunar resources (ie water) make a big difference. The up front costs are high but looking at it as an investment, how many synods would it take to hit breakeven?

Just to throw out some unsupported numbers let's assume each mars bound ship uses three tanker loads. This number has been thrown out there by some who crunch numbers. IMO it's way too low but let's live with it. Assume three cargo ships for every 25 humans that transit. Assume the first synod has 12 humans per ship and two occupied ships for a 1 person down rounding. Assume the number of humans doubles each synod and the packing factor in the ships increases 33% every synod until it reaches 100 per ship. Let's assume a maximum of 100 meat hauling ships. Most of these numbers are way down on the conservative side.

The numbers add up quick. At the top end there would be 100 passenger ships requiring 1200 cargo ships and 3900 tanker loads. Delta V from lunar surface to LEO is within the ballpark of earth surface to LEO. Lunar launch can start with a rail gun to save propellant.

If water, rather than refined cryogenic propellants are shipped to another location, storage is simplified and the cost of launching carbon (to make methane) from earth is reduced.

L1 might be a convenient place for the cracking plant. Carbon could come in the form of sequestered CO2. This perfectly balances out a sabatier reaction. There is insignificant earth or lunar thermal impact on finished product. Terran natural gas extraction is zero. Terran transport, purification and launch impact is zero. Musk is too smart to not understand the long term impact of lunar water on his plans for mars.

My conclusion: Musk absolutely needs lunar ISRU to make his Mars dream viable - and he knows it. He fully intends to have a strong lunar presence and he will go to the moon with or without NASA.

[OTV Booster]

The concept Of 'flags and footprints' gets throughly dissed and given the US abandonment of lunar exploration, is not unjustified. But let's down a dose of Kissinger's 'Realpolitik' and reassess in the context of current reality.

The Outer Space Treaty (OST) forswears ownership of the moon or any part of the moon. BUT, there is a big carveout. A facility, and a reasonable buffer around it, are to be respected. Definitions of a facility and its buffer are vague. Written law, at least in the western tradition, is never settled until it's tested in the courts and the edges sharpened.

At the level of National Policy, ISTM there is a fear that China will land first, set up an automated ISRU experiment, do photo ops for a few days and game the system to claim a chunk or all of the south polar region as a buffer around its automated facility. China probably fears the US has the same intentions. How successful this ploy might be is an open question. There's a lot of room for bad endings.

The truth is probably that there are factions on both sides that want to grab everything and factions that want a more reasoned and equitable approach.

It would be great if before a landing, one party or the other were to publicly announced no intention of monopolizing lunar water with a challenge to the party to announce the same.

[thespacecow]

I think the Starship HLS architecture is simpler than the Blue Moon one. Number of launches isn’t architectural complexity but instead number of unique elements.

Architecturally, they're almost identical.  SpaceX uses a tanker, a depot, and an HLS, all based on a common propulsion technology and launcher, with refueling occurring before the crew boards.  Blue uses a tanker (a GS2), a depot (the Cislunar Transporter), and an HLS, all based on a common propulsion technology (BE-7) and launcher, with refueling occurring before the crew boards.

In SpaceX's case tanker/depot/HLS are all variant of the launch vehicle upper stage, I don't think that's the case for Blue Origin. Blue's tanker share tankage with their LV upper stage, but use a different engine, I don't think the Transporter and Mk2 lander is a variant of GS2. Thus it's correct to say Blue Moon has more unique elements and more complex.

As for conops, Starship only does refueling in Earth orbit, Blue Moon requires additional refueling in NRHO, that's also more complex.

[TheRadicalModerate]

In SpaceX's case tanker/depot/HLS are all variant of the launch vehicle upper stage, I don't think that's the case for Blue Origin. Blue's tanker share tankage with their LV upper stage, but use a different engine, I don't think the Transporter and Mk2 lander is a variant of GS2. Thus it's correct to say Blue Moon has more unique elements and more complex.

I suspect that Blue is all-in on this alternate plan.  I expect the CT and BM1.5 thrust structures to be very close to identical, and of course the engines are identical.  Since they've started floating the idea of BM1s and a BM1.5 as a conops, I assume that CT is now based on BM1 inside the fairing, and CT and BM1.5's tankage will be close to common.

As for conops, Starship only does refueling in Earth orbit, Blue Moon requires additional refueling in NRHO, that's also more complex.

If we're sticking with Orion in NRHO, then Starship needs somewhat higher energy to refuel than VLEO, and even then, the margins are very tight.  If that FCC technical annex is correct, they're planning on a refueling in an HEEO final tanking orbit, which almost certainly means they're refueling twice, the first one being in VLEO.¹

That said, SpaceX is gonna go first unless something really untoward happens.  But there's a reason that Blue's second crack at HLS was so much better than their first:  they drank the refueling Kool-Aid, and it was tasty.

_________
¹The HLS can probably launch directly into a decent-sized HEEO, but there's an odd property to doing that:  If the HLS is almost completely empty, then the depot that serves it, even if it's had the domes re-jiggered to carry as much prop as possible, can't deliver a full tank of prop to the HLS in an HEEO that's high enough to make its margins a non-issue.  You can do better if you dispose of the depot, but even then it's nothing to write home about.

[Robotbeat]

I think the Starship HLS architecture is simpler than the Blue Moon one. Number of launches isn’t architectural complexity but instead number of unique elements.

Architecturally, they're almost identical.  SpaceX uses a tanker, a depot, and an HLS, all based on a common propulsion technology and launcher, with refueling occurring before the crew boards.  Blue uses a tanker (a GS2), a depot (the Cislunar Transporter), and an HLS, all based on a common propulsion technology (BE-7) and launcher, with refueling occurring before the crew boards.

It's fair to argue that ZBO for hydrolox is a lot more complicated than ZBO or near-ZBO for methalox.  But it's equally fair to argue that an architecture that requires 3-5 launches is simpler than one that requires 10-15 launches.

The big difference between SpaceX and Blue is that SpaceX is ahead of Blue by 2-4 years.

SpaceX’s elements all are based on Starship. Same propellant combination and basically same engines for all stages.

Additionally, the number of refuelings isn’t as different. Will be way more than 3-5 for Blue Moon.