Artemis III and refuel in LEO

[jfri]

I have encountered claim that 17 fuel refill missions is needed to get the lunar lander from LEO to the moon

This is obviously wrong. From what I have seen it's about one refuel. But in what way has he got it wrong ?
And exactly is this going to be done during the Artemis III mission ?

[VSECOTSPE]

So the Lunar Starship lander will refuel once from a Starship depot in Earth orbit before heading to the Moon. But that depot requires somewhere between high single digit (according to Musk/SpaceX) and high teen (according to a recent NASA presentation) Starship tanker launches/visits to fuel up.

No one knows the exact number of tanker launches/visits required because it depends on the performance of a lot of things that are still in development:  the exact performance of Starship tankers to orbit (how much propellant can go up per launch), the exact performance of the Starship depot (how much stored propellant does it lose per unit of time), and the launch cadence of Starship tankers (how long does the Starship depot have to wait between Starship tanker visits).  Even those in the know — even SpaceX and NASA leadership — have only estimates and ranges on these figures with which to make educated guesses at this time.

The guy in that YouTube video is making an overly simplistic argument about complexity and reliability.  He is right that a lunar lander that requires refueling is more complex _operationally_ than one that does not.  But there’s also systems complexity.  With only one type of engine, a stainless steel structure, gobs of mass margin, etc., Starship is a relatively simple system compared to, say, SLS, which uses three different engines each launch (and arguably four or five different engines over its life-cycle), requires advanced welding techniques, has little margin to spare if any significant payload rides with Orion, etc.  There’s also more to reliability than operational and systems complexity.  Experience and redundancy also play huge roles.  A system that can launch a lot, like Starship is intended to, provides a lot more workforce experience and a lot more opportunities to recover from accidents than a system that launches only once every year or two, like Orion/SLS.  If someone wanted to do a real engineering analysis, instead of speak to a few slides with some folksy wisdom, the trade offs would be much more complex and nuanced than what is depicted in that video (which really shows no trade offs at all).

The video also ignores a few other hard realities that have nothing to do with complexity and reliability.

One, if we’re ever going send humans to Mars, or frankly, do more at the Moon than Apollo did, we need heavy launch vehicles (HLVs) with high launch rates.  NASA’s own Mars Design Reference Mission 5.0 (NASA’s last comprehensive study of a human Mars mission architecture) requires at least 7-8 HLV launches, and somewhere in the teens depending on transit propulsion choices, to assemble and fuel the Mars mission stack in Earth orbit over the course of about a year and a half.  At best, SLS launches once every year or so — it’s utterly incapable of doing the Mars job.  Even if we ignore Mars, the best case Orion/SLS launch rate will be about half that of Apollo, and likely worse when Murphy inevitably rears his ugly head.  We may not need an HLV that has as high a launch rate as Starship plans for.  But we sure as heck need something better than what we’ve got in SLS.

Two, mass cryogenic propellant storage, transfer, and management is on the critical path to Mars and pretty much everywhere else we’d like to send astronauts.  And it’s probably on the critical path to substantive lunar activities like in situ propellant production from ice or regolith as well.  You don’t take a vacation in the family car by hitching an 18-wheel tank truck filled with gas to that car.  You gas up along the way.  (Or if you’re a explorer or settler, you graze your horse or oxen along the way; you don’t carry a barn’s worth of feed.)  Outside of very limited, Apollo-type missions, the same is even more true of human space missions because of the triple tyranny of Earth’s high gravity, thick atmosphere, and the rocket equation.  Again, most of the HLV launches in NASA’s own Mars DRM 5.0 were for fueling the mission stack in Earth orbit.  We have to learn and perfect that capability, or we’re stuck with sub-Apollo missions ad infinitum.

Third, budgets and costs matter.  Orion/SLS has consumed most of NASA’s human space exploration budget for the last decade and a half and will continue to do so for the foreseeable future just to get a few astronauts into lunar orbit once every year or two.  That leaves little for the other, arguably as critical if not more important, elements of a lunar architecture.  When NASA finally went to procure a lander, Lunar Starship was the only proposal that NASA could afford with the paltry less than a couple billion dollars that was left over after Orion/SLS had been consuming (and is still projected to consume) $4-5 billion-plus annually.   It took a couple more years and much lobbying before NASA could afford to contract for a second lander.

To close, there’s just a lot of unreality ignored reality in that video.  If a refuelable industry lander is too operationally complex, then where does the money come from to afford a non-refuelable one?  The last time NASA tried such an architecture under Constellation a couple decades ago, the cost estimators were spitting out $20-25B estimates to develop the Altair lander alone.  If a refuelable industry lander is too operationally complex, then where do we get the second SLS launch to put a big Altair-like lander up in parallel with an Orion capsule when SLS is struggling just to get to one or two launches per year?  If a refuelable industry lander is too operationally complex now, then when, where, and with what money are we going to prove out these refueling technologies and techniques that are one the critical path to Mars and substantive lunar activities, anyway?  Do we really think a launch vehicle that is as complex, expensive, and rarely flown as SLS is the path to those future architectures?  (Or heck, even a good way to redo Apollo compared to the simpler Saturn V?)

Engineering analyses and life in general are a lot more complex and nuanced than what a keynote or YouTube presenter and their handful of slides would have you believe.  There is no single, simple, good solution, only trade offs between more and less worse options.  You have to place your bets and takes your chances.  But whatever you do, don’t buy the snake oil from luncheon speaker and YouTube hucksters.

[sdsds]

Caveat: I didn't watch the video.

Beyond that, I first want to express appreciation for your commentary, both here and in general on the forum: the views are consistently well-informed and well articulated.

Nit-picking at one part of your commentary, particularly relevant to the "17 tanker missions" claim:

[...] if we’re ever going send humans to Mars, or frankly, do more at the Moon than Apollo did, we need heavy launch vehicles (HLVs) with high launch rates.

Yes, and I particularly appreciate your use of the term HLV. In that category I would put Falcon Heavy, DIV-H, SLS Block 1 and perhaps a few others in the past or, like New Glenn, in the near future.

Precisely because of on-orbit propellant transfer, we don't absolutely need super-heavy launch vehicles. What we need are launch systems that either through high individual flight capacity and/or through high flight rates provide large amounts of mass delivered to orbit inexpensively.

How to get there is still an open question. For optimal efficiency rockets and rocket engines "want" to be big. For optimal manufacturability there's a constraint on vehicle and engine size. If the metric were "tons delivered to orbit per week" it's still unclear what size or type vehicle can do that at the lowest cost..

[jfri]

So the Lunar Starship lander will refuel once from a Starship depot in Earth orbit before heading to the Moon. But that depot requires somewhere between high single digit (according to Musk/SpaceX) and high teen (according to a recent NASA presentation) Starship tanker launches/visits to fuel up.

And my question is how many Moon going Starship can be refueled by the filled Starship depot ?
Also from what I understand is what we are seing in the presentation below is the Starship capsule plus the second rocket stage
https://www.spacex.com/vehicles/starship/
Will they somehow fill the rocket stage with new fuel ?
Also this is stated as unmanned tasks. Is that not very complicated and risky ?

[eeergo]

This forum section is about SLS.

There's a Moon Missions section two places down which has threads dealing with HLS Option A (i.e. Moon Starship / "Moonship").

In the Starship section down in the SpaceX subforum there's also some threads dealing with concepts for deport architectures for this spacecraft.

Mods please move accordingly.

[VSECOTSPE]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

The Starship Depot is essentially a gas station.  It can refuel as many Lunar Starship missions as its service life allows.  (I don’t how many that is.)

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

In practice, this probably won’t happen.  Gas stations don’t normally wait until their underground tanks are nearly empty before refilling and neither will a Starship Depot.

Also from what I understand is what we are seing in the presentation below is the Starship capsule plus the second rocket stage

A Starship launch consists of two stages.  The first stage is called Superheavy and it is intended to return to the launch site, like a Falcon 9 first stage can do.

The second stage is the Starship proper.  It does not have a separate capsule (or third stage, transit stage, etc.).  The entire Starship second stage goes to orbit, and if it needs to go farther, it refuels at a Starship Depot.  A Lunar Starship lander is essentially a Starship second stage that refuels at a Starship Depot and is modified to go farther, land on, and take off from the lunar surface.

Will they somehow fill the rocket stage with new fuel ?

Yes.  We do smaller versions of this today.  The stationkeeping capability of the International Space Station is refueled via Progress visits.  You can think of Starship Depot as a much bigger, cryogenic version of that operation.

Also this is stated as unmanned tasks. Is that not very complicated and risky ?

Most human space flight tasks are run from the ground and/or by computers.  For example, rendezvous and docking at the ISS can be completely automated (current Dragon capsules) or involve berthing, capture, and docking via the astronaut-operated Canadarm (original Dragon capsules).

The devil is always in the details.  US automated rendezvous and docking systems have very few (any?) anomalies, whereas the Russian KURS systems had quite a few.

[VSECOTSPE]

Yes, and I particularly appreciate your use of the term HLV. In that category I would put Falcon Heavy, DIV-H, SLS Block 1 and perhaps a few others in the past or, like New Glenn, in the near future.

Precisely because of on-orbit propellant transfer, we don't absolutely need super-heavy launch vehicles. What we need are launch systems that either through high individual flight capacity and/or through high flight rates provide large amounts of mass delivered to orbit inexpensively.

How to get there is still an open question. For optimal efficiency rockets and rocket engines "want" to be big. For optimal manufacturability there's a constraint on vehicle and engine size. If the metric were "tons delivered to orbit per week" it's still unclear what size or type vehicle can do that at the lowest cost..

I agree.  At the small end of the spectrum, in theory a depot could even be refueled with a steady stream of something like this:

https://www.thespacereview.com/article/544/1

https://en.m.wikipedia.org/wiki/Aquarius_(rocket)

https://newspaceeconomy.ca/2023/12/06/aquarius-the-ultra-low-cost-space-launch-that-never-was/?amp=1

https://www.thespaceshow.com/loral-aquarius

[Robotbeat]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

The Starship Depot is essentially a gas station.  It can refuel as many Lunar Starship missions as its service life allows.  (I don’t how many that is.)

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

…

So actually I don’t think the depot can do 2 lunar starship fills, at least not for a round trip.

If we look at the LEO-NHRO-Surface-NHRO-disposal delta-v, it’s actually about 9.6km/s. A filled out Starship (ie with 100tonnes of furnishings/cargo/etc) can do “only” 6.9km/s. Where does that 2.7km/s come from?
-lower cargo/furnishings for HLS than 100t
-slower transit (doesn’t help much since you have cryogen boiloff to keep in mind)
-reduced delta-v for disposal burn than the assumed 0.45km/s.

Even with all that, there’s still likely a 2-2.5km/s gap to make up.

So the solution is the refueling happens in a GTO-like orbit, not LEO. Alternatively, the Depot acts like a stage and boosts HLS to a GTO-like orbit.

With such a strategy, the 2500t of capacity is enough to refuel HLS *and* put it in a near-GTO-like elliptical orbit. But not to do so twice. (On the other hand, you could still do 2 LEO refuelings of one-way cargo HLSes.)

[sdsds]

-reduced delta-v for disposal burn than the assumed 0.45km/s.

First, thanks for sharing the NASA slide comparing LLO and NHRO delta-v requirements! Where's that from?
It implicitly makes the assumption that "all delta-v is equal." While true for Starship-like systems that rely solely on methalox in a single-stage vehicle, it could be deceiving for systems that use a combination of e.g. hydrolox and storable propellants for different phases of the mission and discard some stages along the way.

Regarding the disposal burn from NHRO, are you estimating that delta-v to escape, or to DRO?

[Robotbeat]

-reduced delta-v for disposal burn than the assumed 0.45km/s.

First, thanks for sharing the NASA slide comparing LLO and NHRO delta-v requirements! Where's that from?
It implicitly makes the assumption that "all delta-v is equal." While true for Starship-like systems that rely solely on methalox in a single-stage vehicle, it could be deceiving for systems that use a combination of e.g. hydrolox and storable propellants for different phases of the mission and discard some stages along the way.

Regarding the disposal burn from NHRO, are you estimating that delta-v to escape, or to DRO?

I’m just assuming the NHRO to Earth burn in that slide is equivalent to a worst case disposal delta-v requirement.

The slide shows the decomposition of the delta-v, so you can of course pick a different Isp for each leg if you like.

[sdsds]

-reduced delta-v for disposal burn than the assumed 0.45km/s.

First, thanks for sharing the NASA slide comparing LLO and NHRO delta-v requirements! Where's that from?
It implicitly makes the assumption that "all delta-v is equal." While true for Starship-like systems that rely solely on methalox in a single-stage vehicle, it could be deceiving for systems that use a combination of e.g. hydrolox and storable propellants for different phases of the mission and discard some stages along the way.

Regarding the disposal burn from NHRO, are you estimating that delta-v to escape, or to DRO?

I’m just assuming the NHRO to Earth burn in that slide is equivalent to a worst case disposal delta-v requirement.

The slide shows the decomposition of the delta-v, so you can of course pick a different Isp for each leg if you like.

Ah, yes, disposal via Earth atmospheric re-entry is almost certainly a good "worst case" estimate.

And (for the benefit of other readers) answering the question about the origin of the presentation slide: Jason Crusan!
https://www.nasa.gov/wp-content/uploads/2016/05/20181207-crusan-gateway-reduced-v4tagged.pdf

[jfri]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

Is it possible that the starship don't need to be fully tanked for going to the Moon ?
Is double amount fuel sufficient for going to Mars ? They claim that Starship can reach Mars aand even beyond that.

[VSECOTSPE]

Is it possible that the starship don't need to be fully tanked for going to the Moon ?

Lunar Starship will need some amount of refueling in Earth orbit to reach the Moon.  We don’t know yet whether that will require a complete topping of the tanks or something less than that.

Is double amount fuel sufficient for going to Mars ? They claim that Starship can reach Mars aand even beyond that.

Starship’s concept of operation at Mars requires refueling on the Martian surface in order to return to Earth.  Given a power source, carbon dioxide in the Martian atmosphere can be extracted and turned into methane and oxygen.  Starships headed to Mars will either carry a power source and processing equipment with them or land near the same from earlier missions.

As a rough rule of thumb, it takes about the same amount of energy (or more) to reach Earth orbit as it does to travel from Earth orbit to most targets of interest for human space flight in the solar system.

[Robotbeat]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

Is it possible that the starship don't need to be fully tanked for going to the Moon ?
Is double amount fuel sufficient for going to Mars ? They claim that Starship can reach Mars aand even beyond that.

Even for one-way missions to both destinations, it actually takes less propellant to send stuff to land on Mars than it does the Moon, due to Mars’ atmosphere.

Measured by propulsive delta-v, travel to Mars is “closer” than traveling to the Moon. Doubly so for round trip missions because volatiles are more readily available on Mars than the Moon (we’ve already demonstrated making oxygen and CO fuel on Mars from the Martian CO2 atmosphere), cutting the delta-v potentially in half.

The Delta-V from LEO to NHRO and to the moon’s surface and back to NHRO is about the same as the delta-v from LEO to Mars’ surface and back to Mars orbit.

So once SpaceX demonstrates Artemis 3 and 4 plus successful Earth recovery of a Starship from orbit, they basically are ready for Mars (uncrewed to start).

[jfri]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

Is it possible that the starship don't need to be fully tanked for going to the Moon ?
Is double amount fuel sufficient for going to Mars ? They claim that Starship can reach Mars aand even beyond that.

Even for one-way missions to both destinations, it actually takes less propellant to send stuff to land on Mars than it does the Moon, due to Mars’ atmosphere.

Measured by propulsive delta-v, travel to Mars is “closer” than traveling to the Moon. Doubly so for round trip missions because volatiles are more readily available on Mars than the Moon (we’ve already demonstrated making oxygen and CO fuel on Mars from the Martian CO2 atmosphere), cutting the delta-v potentially in half.

The Delta-V from LEO to NHRO and to the moon’s surface and back to NHRO is about the same as the delta-v from LEO to Mars’ surface and back to Mars orbit.

So once SpaceX demonstrates Artemis 3 and 4 plus successful Earth recovery of a Starship from orbit, they basically are ready for Mars (uncrewed to start).

Two things come to my mind.
Moon gravity is less than Mars
Mars atmosphere is very thin. Does that not limit aerobraking significantly ?

[whitelancer64]

And my question is how many Moon going Starship can be refueled by the filled Starship depot?

If you’re asking how many Lunar Starships could be refueled at a Starship Depot before the depot needs refueling, it may be two.  Don’t quote this, but I’ve seen numbers bandied about that the Starship Depot will hold ~2500t of propellant.  The wet mass of a Starship at launch is ~1200t.  So if a Lunar Starship is also refueled with 1200t of propellant after reaching orbit, a full Starship Depot could do that twice before needing refueling itself.

Is it possible that the starship don't need to be fully tanked for going to the Moon ?
Is double amount fuel sufficient for going to Mars ? They claim that Starship can reach Mars aand even beyond that.

Even for one-way missions to both destinations, it actually takes less propellant to send stuff to land on Mars than it does the Moon, due to Mars’ atmosphere.

Measured by propulsive delta-v, travel to Mars is “closer” than traveling to the Moon. Doubly so for round trip missions because volatiles are more readily available on Mars than the Moon (we’ve already demonstrated making oxygen and CO fuel on Mars from the Martian CO2 atmosphere), cutting the delta-v potentially in half.

The Delta-V from LEO to NHRO and to the moon’s surface and back to NHRO is about the same as the delta-v from LEO to Mars’ surface and back to Mars orbit.

So once SpaceX demonstrates Artemis 3 and 4 plus successful Earth recovery of a Starship from orbit, they basically are ready for Mars (uncrewed to start).

Two things come to my mind.
Moon gravity is less than Mars
Mars atmosphere is very thin. Does that not limit aerobraking significantly ?

Mars' atmosphere is thin, but it is sufficient to reduce most of Starship's velocity. After reentry Starship will only need approximately 1 km/s dV to land.

[TrevorMonty]

Precisely because of on-orbit propellant transfer, we don't absolutely need super-heavy launch vehicles. What we need are launch systems that either through high individual flight capacity and/or through high flight rates provide large amounts of mass delivered to orbit inexpensively.

Something F9R has well and truly proven.
Most of mass any BLEO mission is fuel, dry mass of vehicles isn't that great unless wanting to place skyscraper on lunar surface .
20mt to LEO is more than enough for dry mass of crew OTV, crew lander,  in space tankers and depots.

[DanClemmensen]

Precisely because of on-orbit propellant transfer, we don't absolutely need super-heavy launch vehicles. What we need are launch systems that either through high individual flight capacity and/or through high flight rates provide large amounts of mass delivered to orbit inexpensively.

But economies of scale are real. That's the reason container ships have grown ever larger even though the size of a container has not changed, or why supertankers grew to ridiculously large sized before finally running into physical limits. The measure for tanker rockets will be total cost/kg of delivered propellant.

Something F9R has well and truly proven.
Most of mass any BLEO mission is fuel, dry mass of vehicles isn't that great unless wanting to place skyscraper on lunar surface .
20mt to LEO is more than enough for dry mass of crew OTV, crew lander,  in space tankers and depots.

[VSECOTSPE]

But economies of scale are real. That's the reason container ships have grown ever larger even though the size of a container has not changed, or why supertankers grew to ridiculously large sized before finally running into physical limits. The measure for tanker rockets will be total cost/kg of delivered propellant.

Economies of scale are real, but they can take the form of physical dimensions or of quantity (or both).  Throughput is what matters in the end.  You can try to build the biggest, most capable HLV with the tech base you’ve got, but if it can only puts 70-130t into LEO once every year or two (cough...SLS...cough), you’re better off going with a less capable HLV that can put 60t to LEO 5+ times a year for 300t+ annually (cough...F9H actually did or will do this in 2023...cough).

Also, infrastructure, nature, and neighbors put limitations on physical size.  Yeah, there are some really big supertankers, but they can only visit certain ports.  If they’re honest, Airbus will tell you that the A380 was a mistake.  Australians can run several trailers behind their tractors, but you can’t do that on US roads.  Etc.

To be clear, I have no idea if Starship will prove too big or launch too infrequently or something else.  I’m just saying that in the abstract, big is not always better.  NASA human space flight has paid dearly by not heeding that advice and doing the analysis before going down blind alleys on STS size, Orion size, Ares V, and now SLS.

FWIW...

[DanClemmensen]

But economies of scale are real. That's the reason container ships have grown ever larger even though the size of a container has not changed, or why supertankers grew to ridiculously large sized before finally running into physical limits. The measure for tanker rockets will be total cost/kg of delivered propellant.

Economies of scale are real, but they can take the form of physical dimensions or of quantity (or both).  Throughput is what matters in the end.  You can try to build the biggest, most capable HLV with the tech base you’ve got, but if it can only puts 70-130t into LEO once every year or two (cough...SLS...cough), you’re better off going with a less capable HLV that can put 60t to LEO 5+ times a year for 300t+ annually (cough...F9H actually did or will do this in 2023...cough).

Also, infrastructure, nature, and neighbors put limitations on physical size.  Yeah, there are some really big supertankers, but they can only visit certain ports.  If they’re honest, Airbus will tell you that the A380 was a mistake.  Australians can run several trailers behind their tractors, but you can’t do that on US roads.  Etc.

To be clear, I have no idea if Starship will prove too big or launch too infrequently or something else.  I’m just saying that in the abstract, big is not always better.  NASA human space flight has paid dearly by not heeding that advice and doing the analysis before going down blind alleys on STS size, Orion size, Ares V, and now SLS.

FWIW...

That's the reason I said "The measure for tanker rockets will be total cost/kg of delivered propellant." All of your points get rolled into total cost/kg. As of now, we are beginning to see limits on the number of launches per year from a site. This scarcity will begin to increase the cost of using a launch slot, in which case bigger is better, aside from any intrinsic economies of scale.