Starship On-orbit refueling - Options and Discussion

[tbellman]

I agree, with one caveat:  There's a class of "commodity" payloads that could be planned and slotted into any surface architecture very quickly.  Examples:

1) 150t of solar panels.
2) 150t of batteries, fuel cells, or APUs.
3) A big tank of LOX.
4) A big tank of LCH4.
5) A big tank of water.
6) 150t of the equivalent of portland cement for regolith.

NASA could probably also buy ten thousand rubber ducks and send to the Moon.  But why?

Several megawatts worth of solar panels, or 150 tonnes of water, does not do you (or more specifically NASA) any good on their own.  For them to be useful, NASA would need other payloads that can use those resources.  And those are unlikely to pop up for NASA before mid next decade.

And even if or when they do need solar panels, batteries, water or cement in non-trivial amounts on the Moon, why would they send them to the Lunar Gateway and transfer them over to the HLS lander, instead of sending a cargo ship directly from Earth to the Moon?  Remember, my complaint was about the handwringing over how to transfer large cargo to the HLS Starship on orbit.

But these are all payloads where their most important property is scale, and HLS can't scale its payloads large enough to matter.

The HLS LSS, at least with its current conops, is a lot more marginal than most people think.  An expendable CLPS or HDL LSS, on the other hand, is not.

And without the current conops, it would not be the HLS Starship...  For example, HLS Integrated Landers must be self-sufficient; they can't rely on pre-emplaced assets.  This means the landers must carry solar panels that work on the surface of the Moon, and work regardless of the lander's orientation to the sun.  That in turn effectively prohibits HLS Starship from carrying TPS tiles and thus from returning to Earth.

Similarly, the requirements to use SLS, Orion and LOP-G, forces the HLS Starship to pick up crew and supplies in NRHO instead of on Earth, and in practice forces it to be able to stay in NRHO for a year or two between each sortie down to the Moon.

A lunar architecture that really takes advantage of Starship, would send a cargo ship with solar panels or other power sources (e.g. a small nuclear reactor), to the intended landing spot, and deploy them robotically.  Then the crewed Starship would depart from Earth, and after landing on the Moon it would connect to the power sources already there.  (You want batteries on board for several hours of operation, and a small APU is probably also a good idea.)  Perhaps pre-emplaced cryo-coolers can also be connected to?  When the crew are tired of the Moon, they disconnect from the surface assets, and return all the way through EDL down to the surface of Earth.  Since that lunar Starship can carry TPS tiles and air control surfaces, it can EDL on Earth, where it can be inspected, repaired and updated, and new supplies, medium sized cargo and crew can be loaded for the next mission to the Moon.

Now I only need to find a way to connect this to refilling of propellant to make this comment not totally off-topic for this thread. 

Right, the crew Starship (at least) would need to be refilled with propellant twice; once in LEO, and once in some later orbit; my favourite is during the trans-lunar coast.

[TheRadicalModerate]

NASA could probably also buy ten thousand rubber ducks and send to the Moon.  But why?

If you know that large commodity payloads are available, then your architectural planning gets a lot easier, the lead times on development are shorter, and the whole enterprise is cheaper.  It doesn't solve the problem of NASA figuring out what they actually want to do, but it does make it easier to do it once they have figured it out.

And even if or when they do need solar panels, batteries, water or cement in non-trivial amounts on the Moon, why would they send them to the Lunar Gateway and transfer them over to the HLS lander, instead of sending a cargo ship directly from Earth to the Moon?

That's what I meant when I said you'd send them via CLPS or HDL, which don't require a detour to NRHO.  (I'm assuming that there's nothing associated with HDL that requires it be staged from the Gateway, is there?)

Remember, my complaint was about the handwringing over how to transfer large cargo to the HLS Starship on orbit.

Yes, and I agreed with you.  The only thing I wanted to point out is that if you can get big commodity payloads in place (and you can, assuming that Starship works as advertised), then other hunks of mission planning and DDT&E become a lot easier.

[Greg Hullender]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

[Robotbeat]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

Yes.

[OTV Booster]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

Even without hard evidence I find it hard to believe most people at NASA under 30 years old aren't dreamers. The bureaucracy probably tamps it down over time but it would never go away completely. Some of those people are salivating over the potential SX is offering.


Wait for SLS to die a natural fiscal death and the budget and program momentum will open up new opportunities. 


 

[Greg Hullender]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

Yes.

Anything you'd care to share?

[TheRadicalModerate]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

Yes.

Anything you'd care to share?

Here.

[georgegassaway]

At the risk of being called "off topic" regarding the recent discussions...

Has SpaceX indicated any testing whatsoever of On-Orbit refueling hardware or anything else involved with on-orbit refueling?

I do not mean 3rd party speculators, I mean SpaceX.

[Greg Hullender]

Is there any evidence that anyone at NASA is thinking about what to do with the capability to deliver large payloads to the moon?

Yes.

Anything you'd care to share?

Here.

There's not much substance to that, though. I guess it's a proof of existence--there is at least one person at NASA (I think) who's thinking about what to do with Starship's capabilities, but there's nothing concrete in there. It's more like an appeal begging someone else to think about it.

[TheRadicalModerate]

There's not much substance to that, though. I guess it's a proof of existence--there is at least one person at NASA (I think) who's thinking about what to do with Starship's capabilities, but there's nothing concrete in there. It's more like an appeal begging someone else to think about it.

Yeah, and it's a person at Ames.  Bah!  If you don't work at JSC or MSFC, you're not a steely-eyed missileperson! (I refuse to use smileys--except accidentally when I make lists with more than eight items--but if I did, there'd be one here...)

I think your assessment is largely correct.  I also suspect that the mission planners aren't quite ready to drink the Starship Kool-Aid.  Until they are, they're gonna keep using design paradigms that they know work in the absence of obscene amounts of throw weight.  Show 'em a successful refueling test, and there'll be a stampede.

[edzieba]

Designing payloads that can only be transported via Starship also means locking out Option P landers. And given that the Option P lander contract even exists in the first place, we can be that NASA suggesting such payloads would result in congress similarly pitching a fit (and a completely coincidental lack of budget being allocated to development of such payloads). That basically limits you to payloads that are of low enough mass to be transported by all prospective landers but are modular enough that you could cram more onto a Starship lander for a reasonable and useful benefit.

[TheRadicalModerate]

Designing payloads that can only be transported via Starship also means locking out Option P landers. And given that the Option P lander contract even exists in the first place, we can be that NASA suggesting such payloads would result in congress similarly pitching a fit (and a completely coincidental lack of budget being allocated to development of such payloads). That basically limits you to payloads that are of low enough mass to be transported by all prospective landers but are modular enough that you could cram more onto a Starship lander for a reasonable and useful benefit.

I'm not sure what the relationship is between Artemis, CLPS, and the HDL (the App. P-requested cargo version of whatever HLS gets picked).  But CLPS mission planners are free to pick whatever platform will get their job done.  And if Artemis planners can rely on a CLPS LSS, they can make a payload--especially a commodity payload--as big as they want.

[OTV Booster]

Designing payloads that can only be transported via Starship also means locking out Option P landers. And given that the Option P lander contract even exists in the first place, we can be that NASA suggesting such payloads would result in congress similarly pitching a fit (and a completely coincidental lack of budget being allocated to development of such payloads). That basically limits you to payloads that are of low enough mass to be transported by all prospective landers but are modular enough that you could cram more onto a Starship lander for a reasonable and useful benefit.

I'm not sure what the relationship is between Artemis, CLPS, and the HDL (the App. P-requested cargo version of whatever HLS gets picked).  But CLPS mission planners are free to pick whatever platform will get their job done.  And if Artemis planners can rely on a CLPS LSS, they can make a payload--especially a commodity payload--as big as they want.

Is there any reason that SX has to work through NASA to deliver a lunar science package? If SX were to announce intent to deliver science packages to the moon NASA would then be only one of many customers. IIUC NASA suggested that Artemus bidders have other uses for their hardware to make the program more sustainable.

[Robotbeat]

Designing payloads that can only be transported via Starship also means locking out Option P landers. And given that the Option P lander contract even exists in the first place, we can be that NASA suggesting such payloads would result in congress similarly pitching a fit (and a completely coincidental lack of budget being allocated to development of such payloads). That basically limits you to payloads that are of low enough mass to be transported by all prospective landers but are modular enough that you could cram more onto a Starship lander for a reasonable and useful benefit.

this isn’t true.

[TheRadicalModerate]

Is there any reason that SX has to work through NASA to deliver a lunar science package? If SX were to announce intent to deliver science packages to the moon NASA would then be only one of many customers. IIUC NASA suggested that Artemus bidders have other uses for their hardware to make the program more sustainable.

They can obviously host any kind of non-NASA package, irrespective of whether it's for science, commercial development, an ITAR-acceptable foreign government, etc.  It just won't be part of CLPS or SLT.

"Commercial approach" is a sub-factor of the "management approach" evaluation in the HLS source selection methodology.  It's not a requirement.

In practice, I suspect this whole thing has the tail wagging the dog.  NASA's commitment to CLPS and SLT makes an LSS more attractive to non-NASA customers, because NASA's contracts make the financial and development risks associated with using the LSS smaller.

[TheRadicalModerate]

Something that's popped up on other threads but hasn't been discussed here (at least for a while) is the possibility of SpaceX selling prop to third parties in the near term.  An obvious potential third party would be the SLD winner.

If this were to happen, numerous folks would get their panties in a wad about dissimilar redundancy.  However, if SpaceX were to publish an open interface to the depot RPOD and prop transfer mechanisms, dissimilar redundancy is possible:

1) If Starship tankers are down for some reason, a medium-heavy launcher could fling stuff into TLI, 12-15t at a time.  The prop payloads so flung would need to be able to do RPODs and transfers to the depot.  (One such launcher is obviously Falcon Heavy.)

2) If the depot itself is down, that's more serious, but still possible to work out:  If the prop payload can emulate the depot-side interfaces, it can directly fuel the target vehicles.

So, three questions:

a) How much work would SpaceX have to do to drive an open depot RPOD/transfer standard?  Since they don't even have a proprietary version that works yet, it's obviously non-trivial. 

b) Could SpaceX short-circuit the usual technical/marketing deathmatch (anybody who's ever been involved with the IETF will understand the dynamic) by offering to accept prop into the depot and pass it through at the supplier's price, with some modest handling markup?

c) Could SpaceX do deals with third-party consumers (e.g. the other SLD service provider, a variety of heavy lunar robotic probes, or even interplanetary missions that didn't mind incurring a bit of extra cost to stage from NRHO) such that they could guarantee a price?  I would expect that price to be based on the likelihood that Starship tankers meet some particular cost, plus some premium that factored in the possibility that, if Starship tankers had a problem, SpaceX would have to provide the prop 15t at a time via Falcon Heavy.

My personal soapbox:  Dynetics should have just gone for it, negotiated a price for 60-100t of prop from a depot, and supersized their SLD architecture to use it.  If SpaceX guaranteed delivery via whatever platform works, it would have satisfied dissimilar redundancy for NASA.  If NASA believed it would work, they'd kill the BO/NT bid deader than a stone.  However, I'd put the odds of this having happened at less than 10%, since it would have leaked if they'd put that into their bid.

However, apparently Dynetics said in a recent presentation that their bid would be able to refuel from Starship.  Whether it was from a Starship tanker or a payload launched on a Starship was unclear.

We usually assume that third-party prop sales, especially in cislunar, are further out than the timeframe for the discussion we have here.  I'm not so sure that's a good assumption.  This could be a case where SpaceX could make a very generous market to dramatically increase demand in the short- to medium-term.  Such a market would not only be potentially lucrative but it would also cover the cost of a lot of flight heritage for tankers and depots.

But it all hinges on being able to standardize the RPOD/transfer interfaces.

[InterestedEngineer]

Something that's popped up on other threads but hasn't been discussed here (at least for a while) is the possibility of SpaceX selling prop to third parties in the near term.  An obvious potential third party would be the SLD winner.

If this were to happen, numerous folks would get their panties in a wad about dissimilar redundancy.  However, if SpaceX were to publish an open interface to the depot RPOD and prop transfer mechanisms, dissimilar redundancy is possible:

1) If Starship tankers are down for some reason, a medium-heavy launcher could fling stuff into TLI, 12-15t at a time.  The prop payloads so flung would need to be able to do RPODs and transfers to the depot.  (One such launcher is obviously Falcon Heavy.)

2) If the depot itself is down, that's more serious, but still possible to work out:  If the prop payload can emulate the depot-side interfaces, it can directly fuel the target vehicles.

So, three questions:

a) How much work would SpaceX have to do to drive an open depot RPOD/transfer standard?  Since they don't even have a proprietary version that works yet, it's obviously non-trivial. 

b) Could SpaceX short-circuit the usual technical/marketing deathmatch (anybody who's ever been involved with the IETF will understand the dynamic) by offering to accept prop into the depot and pass it through at the supplier's price, with some modest handling markup?

c) Could SpaceX do deals with third-party consumers (e.g. the other SLD service provider, a variety of heavy lunar robotic probes, or even interplanetary missions that didn't mind incurring a bit of extra cost to stage from NRHO) such that they could guarantee a price?  I would expect that price to be based on the likelihood that Starship tankers meet some particular cost, plus some premium that factored in the possibility that, if Starship tankers had a problem, SpaceX would have to provide the prop 15t at a time via Falcon Heavy.

My personal soapbox:  Dynetics should have just gone for it, negotiated a price for 60-100t of prop from a depot, and supersized their SLD architecture to use it.  If SpaceX guaranteed delivery via whatever platform works, it would have satisfied dissimilar redundancy for NASA.  If NASA believed it would work, they'd kill the BO/NT bid deader than a stone.  However, I'd put the odds of this having happened at less than 10%, since it would have leaked if they'd put that into their bid.

However, apparently Dynetics said in a recent presentation that their bid would be able to refuel from Starship.  Whether it was from a Starship tanker or a payload launched on a Starship was unclear.

We usually assume that third-party prop sales, especially in cislunar, are further out than the timeframe for the discussion we have here.  I'm not so sure that's a good assumption.  This could be a case where SpaceX could make a very generous market to dramatically increase demand in the short- to medium-term.  Such a market would not only be potentially lucrative but it would also cover the cost of a lot of flight heritage for tankers and depots.

But it all hinges on being able to standardize the RPOD/transfer interfaces.

Perhaps, but it has taken Tesla at least 5 years to open up to non-Tesla cars.   There's both technical and competitive reasons to delay as long as possible.   

Technical meaning "we control both sides of the software and can thus iterate and break APIs".   Much harder to do if there's a fixed standard.

[Robotbeat]

The reason it took Tesla so long is also because of competitors being stubborn (or thinking more traditional solutions were more mature), like OneWeb. Like OneWeb, though, circumstances could change.

Until a few years ago, most EV Competitors to Tesla didn’t have production cars which could charge at the 100kW minimum that would be needed to make sense, let alone 200-250kW that Supercharger v3 can do.

[OTV Booster]

Something that's popped up on other threads but hasn't been discussed here (at least for a while) is the possibility of SpaceX selling prop to third parties in the near term.  An obvious potential third party would be the SLD winner.

If this were to happen, numerous folks would get their panties in a wad about dissimilar redundancy.  However, if SpaceX were to publish an open interface to the depot RPOD and prop transfer mechanisms, dissimilar redundancy is possible:

1) If Starship tankers are down for some reason, a medium-heavy launcher could fling stuff into TLI, 12-15t at a time.  The prop payloads so flung would need to be able to do RPODs and transfers to the depot.  (One such launcher is obviously Falcon Heavy.)

2) If the depot itself is down, that's more serious, but still possible to work out:  If the prop payload can emulate the depot-side interfaces, it can directly fuel the target vehicles.

So, three questions:

a) How much work would SpaceX have to do to drive an open depot RPOD/transfer standard?  Since they don't even have a proprietary version that works yet, it's obviously non-trivial. 

b) Could SpaceX short-circuit the usual technical/marketing deathmatch (anybody who's ever been involved with the IETF will understand the dynamic) by offering to accept prop into the depot and pass it through at the supplier's price, with some modest handling markup?

c) Could SpaceX do deals with third-party consumers (e.g. the other SLD service provider, a variety of heavy lunar robotic probes, or even interplanetary missions that didn't mind incurring a bit of extra cost to stage from NRHO) such that they could guarantee a price?  I would expect that price to be based on the likelihood that Starship tankers meet some particular cost, plus some premium that factored in the possibility that, if Starship tankers had a problem, SpaceX would have to provide the prop 15t at a time via Falcon Heavy.

My personal soapbox:  Dynetics should have just gone for it, negotiated a price for 60-100t of prop from a depot, and supersized their SLD architecture to use it.  If SpaceX guaranteed delivery via whatever platform works, it would have satisfied dissimilar redundancy for NASA.  If NASA believed it would work, they'd kill the BO/NT bid deader than a stone.  However, I'd put the odds of this having happened at less than 10%, since it would have leaked if they'd put that into their bid.

However, apparently Dynetics said in a recent presentation that their bid would be able to refuel from Starship.  Whether it was from a Starship tanker or a payload launched on a Starship was unclear.

We usually assume that third-party prop sales, especially in cislunar, are further out than the timeframe for the discussion we have here.  I'm not so sure that's a good assumption.  This could be a case where SpaceX could make a very generous market to dramatically increase demand in the short- to medium-term.  Such a market would not only be potentially lucrative but it would also cover the cost of a lot of flight heritage for tankers and depots.

But it all hinges on being able to standardize the RPOD/transfer interfaces.

Here are counter questions. How much prop is at issue, and how many different vehicles would need refueling services? Short term.


Nothing out there is using props at the scale LSS needs and there would most probably be only one or two different ships besides SS/LSS needing refueling. Short term.


The idea of flinging small loads and doing direct transfers if the depot is down suggests morfing to a custom package for each customer and letting the depot do what it's designed to do - support SX ops. Customers will each have unique needs and limits that might not easily adapt to the SX system even if a more generic QD suite were available. Worse, even if a customer were using methalox that doesn't mean they won't need helium or some other fluid that the SX system is not designed to handle. Short term.


SX is so mass and volume rich that they can potentially do things almost as an afterthought that other rocketeers would see as a major design and operational goal. Launch a starship and deploy tankage and OTV from the cargo bay, designed to support one specific customer and mission. Learn lessons  that would point to a universal replenishment standard. Short term.


Long Medium term, build a more efficient universal long term solution.

[TheRadicalModerate]

Here are counter questions. How much prop is at issue, and how many different vehicles would need refueling services? Short term.

If an SLD provider were to drink the Kool-Aid, quite a bit.  If not, then probably nothing for quite a while.

The idea of flinging small loads and doing direct transfers if the depot is down suggests morfing to a custom package for each customer and letting the depot do what it's designed to do - support SX ops. Customers will each have unique needs and limits that might not easily adapt to the SX system even if a more generic QD suite were available. Worse, even if a customer were using methalox that doesn't mean they won't need helium or some other fluid that the SX system is not designed to handle. Short term.

Even hydrolox customers could benefit quite handsomely from depot-provided LOX.

And yes, you are correct that this doesn't solve the consumables problem in general.  But non-propellant consumables are a fart in a windstorm compared to the stuff coming out the flamey end.  However, if the SLD provider actually got to the point where the only reason they needed to launch a servicing mission was to replenish those consumables, I'd think that would be an excellent reason for SpaceX to provide payload support in the ogive of a tanker.

I don't disagree that getting a standard that would work for everybody would be fairly difficult, but SpaceX will be the 500lb gorilla here.  And frankly, if two Starships can nuzzle up to each other to feed, then a gnat-sized vehicle could probably find a way to work things out.

All that said, after listening to the Dynetics preso, it sounds like they are indeed building their own tanker, which would be launchable as payload on a Starship.

Maybe this is a possible compromise:  build an "adapter" tank reference design, which can provide depot compatibility to transfer prop (in either direction), but ensure that the form factor is modest enough to be able to join up with any specific needs for the target vehicle.  For example, instead of Dynetics launching a tank of methalox on whatever launcher they choose, they'd launch an empty tank, which looked like their tank on one end, but looked like a Starship on the other.  That way, if the depot is available, they can launch a bunch of empty tanks on one launcher.  But if the depot is down, they can still launch full tanks.