Random Thoughts: First Pass Analysis of a White Dragon/Xeus Lunar Sortie Mission

After yesterday’s SpaceX announcement of having booked two customers for a Apollo-8 style mission around the Moon and back, using Falcon Heavy and Dragon V2, I got thinking about ways to take the next step in lunar tourism–surface missions. Now, this “White Dragon” mission around the Moon makes a lot of sense for SpaceX even though Elon doesn’t care that much about the Moon:

  1. Hardware-wise, there really might not need to be much new work beyond what they’re already doing for Falcon Heavy and Dragon V2. The flight is short enough relative to a worst-case ISS delivery that with 2 passengers instead of 7, the ECLSS is probably already adequate. The heat shield, due to minimum gauge issues with PICA-X, is way overbuilt for LEO return, and thus is likely more than adequate for lunar return.1 The Falcon Heavy should have the throw-mass to do the mission. So they can do this without likely requiring dramatic hardware development beyond what they need for normal FH and crewed Dragon V2 flights.
  2. It’s revenue–even if it costs say $170M/flight (take a ~$140M cost for a four-person ISS delivery, subtract 60M for the Falcon 9 and add 90M for the Falcon Heavy), they can still get takers at a price of ~$85M each. Heck, even if they’re selling it through Space Adventures for $100M each (with the $15M delta being Space Adventure’s cut), this is still a good deal for SpaceX. They only have to convince two people, and they get double the revenue of a Falcon Heavy flight. That’s not quite as good as selling a several flight campaign for Iridium, but it’s not too shabby, especially if Space Adventures or someone like them is bringing the passengers.
  3. There’s a good chance it could turn into a steady stream of revenue. As I’ve previously discussed here, one of the main limitations on space tourism has been availability of spare Soyuz seats, not the price. And when you add the novelty of a lunar mission, I wouldn’t be surprised if they could get a steady White Dragon mission every year or every other year at worst. Even if the price doesn’t come down much over time.
  4. It gets them experience they want for future deep-space operations. Testing reentry heatshield wear, ECLSS behavior on long missions, how their electronics hold up outside the magnetosphere, etc.

So for all those reasons, it makes sense for Elon to say yes to White Dragon missions, even if he doesn’t care at all about the Moon. If someone wants to give you money like that, and it doesn’t cost you much to service them, why the heck not?

But as soon as you start talking about lunar landing missions, the situation changes. Now you’re talking significantly more complicated missions that are going to take non-trivial amounts of new hardware, that will unlikely be directly relevant to Mars missions. Barring someone dropping a bunch of money in Elon’s lap, I’m not sure it would make sense for SpaceX to do the mission by themselves. But, it might be an intriguing joint mission for SpaceX and ULA, assuming for sake of argument that Elon would be willing to work with his closest competitor, and that Boeing and LM wouldn’t torpedo something that could be seen as a direct competitor to SLS/Orion.

So, assume away the political challenges for a second. Could you realistically make a lunar landing mission work with a combination of SpaceX and ULA hardware? After running the numbers, I’d say tentatively yes.

Here’s what the SpaceX stack would look like:

  • Falcon Heavy
  • Large LOX/LH2 tanker (~39.4mT of prop, ~7.2mT dry)
  • Dragon V2 on top

The ULA stack would look like:

  • Vulcan/ACES 546, with the ACES having a Xeus landing kit (~1mT)
  • Small short-duration two-person crew cabin (estimated ~2mT)

Falcon Heavy would launch first, placing the crew and tanker in orbit. Vulcan/ACES would then launch shortly thereafter, with ACES performing a rendezvous with the SpaceX stack, transferring ~39.4mT of prop over (basically filling the ~70mT ACES stage). The Dragon would then separate from the tanker, and connect to the ACES/Xeus stage. The ACES Xeus stage would do a TLI burn for the stack, followed by an insertion into LLO. Dragon would then be left in orbit while the astronauts are flown down to the lunar surface by the Xeus stage, hang out for a while, and then they get flown back up to LLO by the same Xeus stage. The Xeus stage would then dock with Dragon and perform the LOI burn, sending the whole stack back to earth.

Using these assumptions, it looks like the concept closes (here’s the first-pass analysis spreadsheet: WhiteDragon-XeusCalcs), but without as much margin as you would want, and without enough propellant to propulsively capture Xeus back into LEO.

Here are some ways that the concept could be improved:

  1. Be more aggressive on the lander cabin design–2mT is actually half the dry mass of a Dragon Capsule, in spite of not needing anywhere near the capabilities–no need for a trunk or a reentry shield, no need for RCS engines (ACES and Dragon V2 both have plenty of maneuverability). So you might be able to whittle that down a bit.
  2. Use chilled propellants–between the Vulcan/ACES 546 leftover propellant and the Falcon Heavy tanker propellants, there’s actually more propellant than will fit into a stock ACES stage at normal boiling point densities. Chilling the propellant would both surpress boiloff losses, and would also allow you to cram a bit more propellant into the stage. This is already something SpaceX does for Falcon 9, so it isn’t that crazy.
  3. Stage in EML2 and have the Dragon V2 perform the earth return burn leaving Xeus at EML-2. The question is how much propellant is needed for rendezvous, reentry, and landing. By my BOTE calcs, assuming a 320s Isp on the regular Dracos, you should be able to get ~620m/s of delta-V out of Dragon V2, and you’d only need ~150m/s for the earth return/powered swingby maneuver. But you’d now be talking about a much, much longer mission.
  4. Jettison the lander cabin and/or Xeus kit prior to the earth return burn.
  5. Have the lander cabin actually be a separate lander/ascent stage. Have the ACES stage not have a Xeus kit, but do an uncrasher maneuver where the crew cabin stage separates right before landing. the ACES stage returns to lunar orbit w/o the ascent stage, and reconnects with the Dragon capsule. The ascent stage lands, the crew hangs out for a bit, and then launches again back up to the waiting Dragon and ACES stages. The crew cabin/ascent stage is left in LLO before ACES takes Dragon back to Earth.
  6. Down the road adding refueling in LLO or EML2 or on the lunar surface would make the whole thing tons easier–you’d need another launcher, but that would all of the sudden give you the margin needed to do much more ambitious missions if you didn’t have to haul the prop all the way from LEO and back.

There are probably other variations on the theme. But the interesting thing is that this concept comes close to closing using Falcon Heavy, a stock Dragon V2, with the main pieces of new hardware being a propellant tanker section for Falcon Heavy2, the Xeus kit for ACES, the fuel transfer hardware, and the crew cabin.

If you assume $150M for the Vulcan/ACES 546 flight ($90M for the bare Vulcan/ACES plus six strapons at $10M/ea), $170M for the FH + Dragon V2, $20M for the tanker/transfer hardware, and $30M each for the Xeus kit and the crew cabin, you get about $400M/mission, or about $200M/person (plus markup if your brokering it through Space Adventures). Call it an even $250M/person ticket price. That’s about 1/3 of what Golden Spike was targeting…

Anyhow, rockets aren’t legos, who knows if Elon and ULA can play nice with each other, I have no idea how ULA would get its parents to go along with a scheme that enables lunar landings without needing SLS/Orion. But the concept comes really close to closing, so I thought I’d put it out there.

The following two tabs change content below.
Jonathan Goff

Jonathan Goff

President/CEO at Altius Space Machines
Jonathan Goff is a space technologist, inventor, and serial space entrepreneur who created the Selenian Boondocks blog. Jon was a co-founder of Masten Space Systems, and is the founder and CEO of Altius Space Machines, a space robotics startup in Broomfield, CO. His family includes his wife, Tiffany, and five boys: Jarom (deceased), Jonathan, James, Peter, and Andrew. Jon has a BS in Manufacturing Engineering (1999) and an MS in Mechanical Engineering (2007) from Brigham Young University, and served an LDS proselytizing mission in Olongapo, Philippines from 2000-2002.
  1. And frankly we know a ton more about lunar return reentry heating and how to simulate/test for it than we did at the time of the Apollo program, so I’m not too worried about the heat shield not holding up.
  2. This will likely look weird, since you’d probably want to leverage ACES tankage for the job, which would have the Falcon Heavy looking like a weird cross between the spacecraft version and the Dragon V2 version.
This entry was posted in Commercial Crew, Launch Vehicles, Lunar Exploration and Development, Propellant Depots, Random Thoughts, Space Exploration, Space Transportation, SpaceX, ULA, YHABFT. Bookmark the permalink.

51 Responses to Random Thoughts: First Pass Analysis of a White Dragon/Xeus Lunar Sortie Mission

  1. Darrin Taylor says:

    Nice work =)

  2. Andrew Swallow says:

    Good idea.

    The SLS can only be used by NASA so say miners would need a commercial way of getting to the Moon. Boeing cannot sell the miners a SLS but ULA can sell them a Vulcan and a Xeus. An independent Prime Contactor may be needed to handle the competing firms and avoid the cartel laws.

  3. Dave Salt says:

    That’s a nice assessment, Jon… but does the ACES/Xeus/Cabin have to be launched on a Vulcan (i.e. could it be launched on another FH)?

  4. ken anthony says:

    CJ had similar thoughts. I have a feeling that many do.

  5. Brock says:

    Bigelow has been working on a lander for years to put their BA330 on the Moon. It’s probably worth asking if they plan on having a taxi in orbit around the Moon for that.

  6. Juho says:

    Dave Salt: Falcon has a relatively weak first stage and a very brawny second stage which does a much bigger share of the total delta-v to orbit than on ULA rockets. On FH the three first stage cores would do a bigger share of the job than on F9, but my first instinct is still that ACES and Falcon would be a bad fit. I haven’t looked at the numbers in detail, though.

  7. David Little says:

    The don’t have to “convince” two people. SpaceX was approached with this idea. They’re getting contracted for a specific flight. So they don’t have to worry about that at all!

  8. Chris says:

    Any concerns about LH2/LOX propellant transfer on orbit. Seems like this would be a technical issue that needs to be tackled. SpaceX is going to have also tackle for the Methane case as well. I’m not sure this has ever been demonstrated other than case of storables.

  9. Jonathan Goff Jonathan Goff says:

    Re the “convince” line, I was talking about follow-on flights. I’m well aware that SpaceX has the first two customers lined up. My point was just that going forward it was a lot easier than say filling up a DragonLab mission where they might need to make 100 sales to fill a flight, vs two sales to fill a White Dragon flight. Does that make sense?


  10. Jonathan Goff Jonathan Goff says:


    Thanks! If it wasn’t midnight, and I didn’t have a cold, I probably would’ve tried some more ideas to see if I could beef up the margins more. But it was a fun first-pass analysis.


  11. Jonathan Goff Jonathan Goff says:

    Dave Salt,

    I’m not sure whether ULA would ever just sell SpaceX an ACES to do a flight like this. I kind of doubt it. But in theory if they said yes, sure you could do this with two Falcon Heavy missions, but then you’d have to do them back to back, account for more boiloff, etc. Doing it with two disimilar launchers makes the back-to-back launches a lot easier, IMO. Plus, I’m an anti-monoculturalist, and don’t want to play the “what if we put everything on Falcon Heavy” game… 🙂


  12. Jonathan Goff Jonathan Goff says:

    I know Bigelow has been talking about lunar stations and landers, but from what I’ve heard from people who work there, I doubt there’s actually been much real engineering thought thrown at the problem. Pretty marketing pictures? Sure. But actual detailed engineering? I wouldn’t hold your breath.


  13. Jonathan Goff Jonathan Goff says:

    I don’t think anyone to-date has demonstrated a cryogenic propellant transfer between two initially disconnected tanks through a separable connection. But as ULA likes pointing out, every time they restart Centaur or DCSS, they’re transfering cryo propellants between the tank and engine, starting in low-g settled state (down in the 10s or 100s of microgees). The main question mark is the coupler, but even that isn’t that scary of a thing–we’ve done tons of cryogenic couplers on the ground, we just need a lightweight one for refilling stuff on orbit. My startup Altius is waiting to hear back on an SBIR Phase II selection (potentially as early as next Wednesday) for just such a cryo coupler project. If we get it, we should be taking the cryo coupler through a nearly-flight-like state. It should be good for any cryogenic fluid from Methane to LH2 (possibly with slight variations on the cryo seals designs).


  14. Bob Steinke says:

    Could the Vulcan lift the ACES with its hydrogen tank full so that the Falcon would just have to transfer LOX, which it could keep in a stretched version of its regular tankage?

  15. Live on Merritt Island have heard two spaceX people now refer to it as Gray Dragon. Had previously heard Red Dragon for Mars and Blue Dragon for Earth orbit.

  16. Jonathan Goff Jonathan Goff says:


    In theory you could have the ACEX/Xeus stage loft a separate LOX/LH2 drop tank… you’d take more of a payload hit for the plumbing and structure and stuff, but it’s not an obviously crazy idea, and would make the tanker from SpaceX not be so weird shaped… That’s not a bad idea actually. You’d have to do it as a drop tank though, because at least some of the ACES propellant is used to make it to orbit, so there’s no way to fill it all the way back up unless you have a full LH2 tank worth of prop in a separate tank somewhere. But having that be on the ACES flight would make a lot of things easier than having it on the Falcon Heavy flight…


  17. Jonathan Goff Jonathan Goff says:

    Yeah, I didn’t have the formal name, so went with White Dragon for this post. But Grey Dragon works too (and probably makes more sense).


  18. Bob Steinke says:

    But White Dragon is such a better name!

    DM: Your party rounds the corner into the room. On top of an immense pile of treasure sits a GRAY DRAGON!

    Thag the Fighter: What, did he forget to do his laundry?

    DM: Thag, make a saving roll against… oh shoot, what kind of breath attack does a gray dragon have?

  19. Doug Gage says:

    I’m with you, Bob, on “White” vs “Gray”!

    This whole concept is really intriguing, and certainly not the kind of stuff hard-over SLS/Orion partisans want to have bandied about…

  20. Robert Clark says:

    Thanks for that. This might be doable using a single Falcon Heavy.
    I did a calculation using the original cargo version of the Dragon at a ca. 4 metric ton dry mass, that you could do a lunar landing mission using two Centaur upper stages for the in-space stages:


    However, the new version of the Dragon, V2, is about 50% heavier so we would need 50% greater mass for the in-space stages. This would be about 60 metric tons for these stages. This might be doable considering the F9 full thrust improves also the Falcon Heavy payload capability. It would certainly be doable using also an additional F9.

    Bob Clark

  21. Robert,
    Based on the numbers I was running, I’m very skeptical you could make this work on a single Falcon Heavy flight, even if you assumed the lighter cargo Dragon. I wasn’t seeing much margin on a dual launch option. You could in theory do it on two Falcon Heavies, possibly even with enough spare performance to not have to expend the booster/core stages. But then you once again have to do back-to-back Falcon Heavy launches within a very short timeframe. And you need to convince ULA to use their upper stages without buying a full launch…

    If you assume that Falcon Heavy is made of magic dust and is going to cost almost nothing to fly, sure I could see the change. But by my estimation, you’re probably looking at a $60M savings using Falcon Heavy vs Vulcan/ACES 546 (and that’s assuming first stage reuse), on a $400M mission. It’s a non-trivial savings, sure. But it adds a lot of complication in the process, and makes it that much less likely to work.

    I just don’t get people’s obsession with trying to force everything into the SpaceX mold. Why do we have to replace a NASA monoculture with a SpaceX monoculture?


  22. Bob,

    Thinking about your earlier comment some more, I want to run an analysis where the Falcon Heavy only launches a LOX tanker, with the Vulcan/ACES 564 carrying all of the LH2 in a tanker. That eliminates the need for SpaceX to add any LH2 handling capabilities on their pad, reduces the number of couplings that need to be made, and makes things generally work better, assuming the numbers close.


  23. Ray says:

    I know its unlikely but surely once the ITS has been built and tested they could use this for escapades around the moon or even landing on it? It would be a great test bed for the landing and entry on the moon whilst bringing some extra revenue.

    Just because the technology is being built for the trip to Mars it doesn’t mean ti can’t be used nearby.

  24. DougSpace says:

    How many countries could afford to spend that amount of money for such a mission? I bet it’s quite a few. So, it’s not just another form of tourism but national exploration in which case the market is a fair amount bigger.

    I hope that a negotiated compromise would be: FH @ Vulcan for the Moon, HLV (e.g. SLS) for Mars. So ULA’s parents go along.

    Establishing lunar surface access might not be in SpaceX’s interest but it is in ULA’s and there are significant national reasons why the US might want to go to the Moon. So I see NASA putting up the money which could motivate both ULA and SpaceX to play nice together.

    If the first such mission were to not be crewed (no Dragon and no Xeus crew cabin) but were instead to have a one-way payload delivery to a lunar pole then there would be a shot a delivering telerobotic hardware (including power system & spare parts) to harvest ice for propellant. If successful you just developed a propellant depot on the Moon which probably solves the ACES-Xeus propulsive braking to LEO issue. This comes close to my CisLunarOne.com concept.

  25. Paul451 says:

    I think I prefer “Silver Dragon” for moon missions, rather than Grey.

  26. Doug,
    The interesting thing is that if we can get it to the point where we can get the Xeus stage back into LEO, it may be possible to do the missions in a way where an international partner could launch part of the propellant for the mission, launched on a launcher from their country, in exchange for one seat, and let SpaceX/ULA/their broker selling the other seat commercially.

    Re: SLS for Mars/Big Stuff and SpaceX/ULA for the Moon… it’s possible Boeing and LM might go for that, but it’s still a non-trivially hard sell.

    re: Uncrewed flights, I should run the numbers on that. I think a ULA/SpaceX dual launch architecture could put a pretty hefty amount of payload onto the lunar surface (likely >10mT). If I get time tonight to run the updated scenario Bob was suggesting (having all the LH2 launched on Vulcan so that SpaceX doesn’t have to interface with LH2 on their vehicle), I’ll also run a cargo one-way scenario.


  27. Jonathan Goff Jonathan Goff says:


    Theoretically, yes you might be able to do a lunar surface mission with an ITS stage. The delta-V from LEO to the lunar surface, and back to an earth intercept course for aerobraking is pretty intense though (3.2+0.8+2.0+1.8+0.8=8.6km/s), so you wouldn’t be able to fly with anywhere near a full capacity. And since it’s LOX/Methane instead of LOX/LH2, you’re less likely to be able to take full advantage of lunar ISRU. And that doesn’t even get into my skepticism that it will ever fly, let alone on a timescale to compete with this Dragon/Xeus concept.

    But long-term, if ITS ever flies, yes it could do inefficient lunar missions.



  28. You might have center-of-mass issues with the FH and the extra 39t of propellant. If you assume that you can get away with a 4m diameter tank on the FH, you’re still looking at adding an extra 11.7m to the stack. Add on the 8.1m for the D2 itself and you’re looking at about a 20m stack on top of the FH second stage.

    That’s well outside the existing F9 payload center-of-mass rules. You’ve got a lot more gimballing power with the FH, so maybe that’s OK. It’s an awfully long, skinny rocket you’re launching, though.

    This seems like a fairly fundamental issue with the FH in general. Launching propellant seems like a pretty good business for it, but you have to get length/width down somehow. Sure seems like a 5m-wide Raptor 2nd stage would be handy to have.

  29. Jonathan Goff Jonathan Goff says:


    But come on, White Dragon is so much more “metal band” than Grey or Silver Dragon… It’s like begging for a gratuitous umlaut somewhere.


  30. Another issue: If you’re doing real live fuel transfer between the tank launched on the FH and the ACES/XEUS, you’ve got a pretty hairy Earth-orbit rendezvous sequence:

    1) Rendezvous.
    2) Separate D2 from tank.
    3) Separate lander crew compartment from ACES.
    4) Dock tank to ACES.
    5) Spin up ACES-tank combo around the long axis to control the ullage during transfer.
    6) Transfer fuel.
    7) Spin down.
    8) Undock/discard empty fuel tank. (You have a disposal issue here, too: what provides the de-orbit delta-v?)
    9) Re-dock the lander.
    10) Dock D2 to the lander. (Another problem: unless you want to do an EVA to transfer to the lander, you’ve got the D2 flying ass-backwards into the unknown. If you put the D2 facing front on the ACES with the lander facing backwards during TLI, then you have to undock D2, transfer crew, undock lander, and then re-dock lander with ACES–all in lunar orbit.)
    11) Off we go.

    Maybe there’s a way to leave the lander in place and run the transfer piping either around or through it, but it sounds kinda weird and unreliable. Alternatively, maybe you wind up docking the fuel tank behind the engine, but that seems like it would violate all kinds of mission safety common sense.

    Maybe the best way to avoid all of this weirdness is to bundle three narrow tanks for the FH launch and dock them as externals to the ACES stage. You can discard them before LOI. If you do it right, they’re on a free return back to Earth, where they’ll burn up pretty cleanly. Of course, that just exacerbates your center-of-mass problems at launch (see my other post).

  31. Jonathan Goff Jonathan Goff says:

    Radical Moderate,

    I have other ideas for how to do the rendezvous/transfer that should be a lot simpler than that. Though admittedly there is the question of what orientation to have the D2 during the TLI burn. Easiest operationally would be nose attached to the Xeus (ie ass-backwards into the unknown as you put it), but nose forward could also be a possibility. I’d have to learn more about Dragon’s capabilities, and how that would drive the mass of the crew cabin.


  32. Isaac Mooers says:

    Great work Jon,

    Could you rework the design not for a single mission, but say 2 missions and year for 5 years? With the goal being to optimize hardware re-use? Lets see what difference leaving XEUS and the cabin in at EML2 with the fuel depot makes to the long term 10 mission price per mission.


  33. Peterh says:

    For fluid transfer in microgravity, I’m thinking use a pump that can accept mixed flow input and separate the liquid and gas phases. Send the gas to the far end of the source tank.

  34. Peterh:

    If you spin the tank on its long axis, the liquid will collect on the side walls of the tank. Give the inner walls a bit of a slant so the inner tank radius is larger at your outflow, and you can just pressurize the tank to do the transfer. See here.

  35. Robert Clark says:

    Jon, about the margins on:


    It is true they are quite slim. However, truthfully I’m not sanguine about using the Falcon Heavy to carry a crew anyway. And especially not for the announced circumlunar flight when it will have few flights behind it, if any.

    Better would be if the F9 were used to carry the crew anyway. By this page the upgraded F9 and Falcon Heavy will be able to get 22,800 kg and 54,400 kg to LEO respectively, for a total of 77,200 kg to LEO:


    The 77,200 kg likely would provide sufficient margin for a lunar landing flight.

  36. Jonathan Goff Jonathan Goff says:


    With how little margin I was seeing even using LOX/LH2 (which tends to minimize IMLEO), I’m really skeptical that 77mT would be enough for a crewed lunar landing flight.


  37. The circumlunar flight is obviously designed to exploit the 50 year anniversary of Apollo 8. So speculation on a mission to do the same with Apollo 11 is certainly warranted. But to be ready in time for July 2019, the plan would have to be way simpler than this.

    Could Dragon 2 carry enough hypergolics to launch itself from the Lunar surface even after a powered descent?

  38. Jonathan Goff Jonathan Goff says:

    Short answer–no, not even close. Assuming the FH upper stage does TLI, you still need 0.8+2.0+1.8+0.8=5.4km/s of delta-V to enter lunar orbit, land, takeoff, and return. the current dragon has maybe 620m/s of deltaV capability, so you’re talking about 8x as much. That would require a mass ratio of ~5.59, assuming 320s for the Draco Isp. Which, if you assumed the tanks weighed nothing, but that you need the existing 620m/s of delta-V for reentry/landing, etc would require adding (5.59-1)*7800=35.8mT of additional propellant. Ie about 5x the current mass of the DragonV2. You’d end up needing several Falcon Heavy flights to lift enough propellant to send that through TLI.

    Long story short, Dragon V2 is a good in-space and reentry capsule, but is way too heavy for a lunar lander. And shipping everything around through high delta-Vs using low Isp hypergols or LOX/hydrocarbon systems is a great way to make your system cost a lot more than you save by avoiding LH2. I’d be very reticent to use LOX/LH2 for a first stage, but once you’re in space, I think the pain is more than worth it.

    One of the things I learned the hard way at Masten is that sometimes doing things the “simple” way ends up being more of a pain in the neck than picking a few hard things to make the system level problem easier.


  39. Paul451 says:


    “I just don’t get people’s obsession with trying to force everything into the SpaceX mold. Why do we have to replace a NASA monoculture with a SpaceX monoculture?”

    If it’s a NASA architecture, sure, spread the love. But if it’s SpaceX trying to offer a commercial service, then it’s hard to imagine any context in which they’d want to pair up with another launch provider.

    Also, I’m not sure what ACES (hence Vulcan) or XEUS would offer them. Surely they have access to their own upperstage, plus their landing systems knowledge from both F9/FH first stages and (by then) Dragon 2.

    ACES, being hydrolox, will give more delta-v per tonne of launch mass, but refuelling in orbit pretty much eliminates that advantage. And SpaceX might well see a benefit in practising refuelling F9/FH upper-stages as a gating exercise to prop-transfer for ITS. And if they can also in effect sell the resulting fuelled upper-stage to paying lunar customers… gravy!

    (Contradicting myself regarding the benefits of orbital refuelling. But I’m assuming one of the upper-stages is modified as a lander. The other is just an extended tanker, carrying up to ~54 tonnes of extra prop.)

    “White Dragon is so much more “metal band” “

    Whereas “Silver Dragon” is more Chinese takeaway. But Moon and silver have as much historical association as Mars and red, and Earth and blue.

  40. Paul451 says:


    “I hope that a negotiated compromise would be: FH @ Vulcan for the Moon, HLV (e.g. SLS) for Mars.”

    The problem is that SLS/Orion actually seems pretty optimised to lunar missions, and terribly suited to Mars missions. Perhaps you can use it as a big-dumb-launcher for the Mars hardware, assembled in LEO. But it’s not a LEO optimised launcher, either; and it’s grossly overpriced for just lofting bulk components, making it extremely unlikely that you can have Mars hardware development while operating SLS. You have to pick one of the other.

  41. Jonathan Goff Jonathan Goff says:


    If it’s a NASA architecture, sure, spread the love. But if it’s SpaceX trying to offer a commercial service, then it’s hard to imagine any context in which they’d want to pair up with another launch provider.

    That was the point I was trying to make–I don’t think SpaceX is likely to go after a lunar landing venture. It’s too much of a side-track from their goal of Mars missions. They’d have to develop a decent amount of customer hardware that has almost no other commercial or Mars use. They in theory could, but I just don’t see the motive or patience being there. And frankly, I don’t think they could come up with something that would work as well as a Xeus-based architecture.

    Also, I’m not sure what ACES (hence Vulcan) or XEUS would offer them. Surely they have access to their own upperstage, plus their landing systems knowledge from both F9/FH first stages and (by then) Dragon 2.

    ACES, being hydrolox, will give more delta-v per tonne of launch mass, but refuelling in orbit pretty much eliminates that advantage.

    Refueling doesn’t eliminate the benefit of using a more efficient propulsion system. The cost of launching mass to LEO is still not that cheap, even with the cheapest likely near-term semi-reusable rate SpaceX will likely attain. Even best case w/ FH, the launch costs are going to be the biggest cost of a mission like this, and going from 3-4 launches down to 2 really can make a big overall cost difference.

    More importantly, without going Hydrolox, you’ve limited how low your transportation cost can go, because you can’t take full advantage of ISRU. Combine those two factors–needing more launch mass, and not being able to leverage ISRU going forward, and SpaceX’s non-hydrolox technology base isn’t going to be as competitive.

    And SpaceX might well see a benefit in practising refuelling F9/FH upper-stages as a gating exercise to prop-transfer for ITS. And if they can also in effect sell the resulting fuelled upper-stage to paying lunar customers… gravy!

    I agree that getting practice doing cryo prop transfer would be good for SpaceX, but I don’t think that’s a reason to not work with the company that’s been pushing in-space refueling the hardest…

    (Contradicting myself regarding the benefits of orbital refuelling. But I’m assuming one of the upper-stages is modified as a lander. The other is just an extended tanker, carrying up to ~54 tonnes of extra prop.)

    I’m not sure if you can do this with just two FH launches if you don’t go hydrolox.


  42. Jonathan Goff Jonathan Goff says:

    Paul, Doug,

    Paul has a good point that SLS is most optimized for lunar missions–that’s why the SLS huggers in Congress have been trying to kill ARM and move things back to the Moon for so long. And why in spite of being a big lunar development fan, I’m kind of terrified to see what a Moon-focused NASA will look like. They’re almost certainly going to try to turn it into a make-work scheme for SLS. And more likely than not try to sole-source as much as they can to Boeing.

    Theoretically, you could do this mission on SLS without the prop transfer, with plenty of margin. Maybe even using more or less the same lander and basic hardware, just with some “leave-behind” stuff for the lunar surface… In theory, if NASA didn’t distort things too much with their traditional human rating (heh), you could have a system they could use that also developed most of the hardware needed for doing lower-cost commercial missions as well… but the past history of Commercial Crew and Cargo makes that theory less promising than I would’ve thought 10yrs ago.


  43. Robert Clark says:

    Jon, about the Daniel’s suggestion of the Dragon as a hypergolic-fueled lunar lander, it is doable with two launches of the FH. The upgraded Falcon Heavy can get 54,000 kg to LEO. So two can get 108,000 kg to LEO, and a 100,00 kg payload number has often been considered the needed value for a manned lunar lander mission.

    Using your number of 35.8 metric tons of extra hypergolic propellant, that plus the 7.4 metric ton mass of the Dragon, would be a total mass of 43.2 metric ton. Then consider the rule of thumb that a hydrolox, Centaur-like stage can get as payload to escape velocity about the same mass as its propellant load.

    So for the escape stage we would need a hydrolox stage with ca. 44 metric ton propellant load and ca. 4.4 dry mass, for a total of about 92 metric tons all up to LEO.

    Actually, as I mentioned I really don’t like the idea of the FH carrying a crew. I would prefer the crew launched to LEO by a F9. This would give us an additional 22.8 metric tons to LEO. This now would be ca. 130 metric tons to LEO, the same as the Apollo missions.

    Bob Clark

  44. Bob,

    You can’t just say “this can all be done with hypergols and Dragon” and then handwaive into existence a SpaceX hydrolox stage! If them doing a hydrolox stage is so easy, why not use it for more of the landing, where it can make a bigger positive difference? This reminds me of a comedy sketch about an a cappella band called “Instruments Not Included” that had guitars, drums, and a piano…

    Not to mention the ~43mT Dragon stack assumed you could add 35mT of propellant without having to add extra structure, which would snowball into more propellant mass. Is it possible to make Dargon into a manned lunar landing by adding more propellant? Not really–it wouldn’t look anything like Dragon by the time you’re done. The capsule is just too darned heavy.


  45. Paul451 says:


    “More importantly, without going Hydrolox, you’ve limited how low your transportation cost can go, because you can’t take full advantage of ISRU.”

    IMO, lunar ISRU is far enough in the future that the systems designed to exploit it will be fundamentally different from the systems needed to get there to do the science to develop the methods to design the hardware to actually create propellant on the moon.

    If we’re using SpaceX as a handy cheap throw-away “getting there” service, we don’t have to feel that we’ve locked into anything permanently. It’s only if we are going the SLS route, then obviously it’ll have full institutional and political momentum behind it, preventing change.

    “I’m not sure if you can do this with just two FH launches if you don’t go hydrolox.”

    I was thinking three, once the lander is in lunar orbit. Tanker (lengthened upper-stage) launched on FH to refuel the lander in LOR. Second tanker on an FH docks with the greyish-silver-white Dragon in LEO for TLI and TEI. And Dragon with crew/passengers is launched on F9.

    However, I just did the prop calc for 311s and 4.2km/s (assuming lunar SSTO). Ick. 3:1 mass ratio for Merlins. (As opposed to 1.6:1 for RL-10s.) Assuming 20 tonnes TLI throw for FH, there’s no way you could carry enough propellant in a single launch to refuel the lander. (Unless it was some kind of open-frame micro-lander.)

    If SpaceX develops the supposed Raptor-based FH upper-stage, for DoD payloads, then that would make things a little easier. (Both boosting more payload to the moon, and providing a better base for a Xeus-style DTAL lander.)

    “Theoretically, you could do this mission on SLS without the prop transfer, with plenty of margin.”

    OTOH, for the price of dual SLS launches…

  46. Jim Davis says:

    You can’t just say “this can all be done with hypergols and Dragon” and then handwaive into existence a SpaceX hydrolox stage!

    No, but neither can you handwave into existence propellant production on the moon or propellant depots wherever.

  47. Paul451 says:

    I think Jon’s saying if SpaceX developed a hydrolox stage, why wouldn’t they base a lander on that instead of Superdracos and huge amounts of hypergols?

    I mean, if you were specially developing any kind of new upper-stage/engine to loft lunar landing infrastructure, why wouldn’t you make that part of your lunar landing development? Adding nearly 40 tonnes of extra prop to Dragon means it’s not really Dragon any more, so you are already doing extensive mod-work, so you might as well start with the better engines.

  48. Robert Clark says:

    Jon, adding hypergolic tanks to the Dragon so it could serve as a lunar lander is not my preferred architecture. I would prefer going for smaller systems that could be launched at low cost, to the extent a manned lunar landing mission could be launched for comparable prices to that of launching a manned LEO mission today.

    For this purpose the Dragon at ca. 6.4 metric ton dry mass is too heavy to serve as the lander capsule. Better would be one at ca. 2 mT or smaller dry mass a la the Apollo lander crew module.

    Orbital Sciences has been investigating giving their Cygnus capsule life support capability for use as a deep space habitat. They’ve also investigated giving it a heat shield to make it reusable. A Cygnus given a heat shield and life support could work as a lightweight capsule for a lightweight lunar sortie design, launchable on a single Delta IV Heavy or Ariane 5 using already existing hydrolox stages for the in-space propulsion:

    Budget Moon Flights.

    Bob Clark

  49. Robert Clark says:

    Jeff Bezos has just announced a new heavy lift launcher that he wishes to use to effect a return to the Moon:

    Blue Origin releases details of its monster orbital rocket.
    The privately developed rocket will also be capable of 100 reuses, Jeff Bezos says.
    ERIC BERGER – 3/7/2017, 10:00 AM

    Given its 45 metric ton payload to LEO in its reusable form, it probably has a comparable payload as an expendable to that of the Falcon Heavy at ca. 54 metric tons. Then in the next few years we will have multiple launchers capable of mounting manned lunar missions, with the Falcon Heavy, SLS, and New Glenn.

    Bob Clark

  50. Nathan Wilson says:

    Ok, but if SpaceX did decide to build an ITS-precursor that was targeted at both (round-trip crewed) Lunar and (one-way/cargo) Mars missions, what would it be like?

    Assuming it uses the ITS 7000 ton first stage, with enough stages, we can build a mission without orbital rendezvous or Lunar refueling. Let’s assume they’d use methane-O2, a re-usable first stage, and expendable upper stage(s).

    A landing capsule with enough propellant for 2.6 km/s of delta-V is in the right ball-park to land on Mars using SSRP (super-sonic retro propulsion), or take-off from the Moon and land on the Earth. The burn mass ratio for this is 2.0 with axial methane-O2 engines, maybe 2.2 with Dragon-2 style canted sidewall engines.

    At first glance, the Lunar and Mars missions differ by the extra 2.2 km/s which is required to land on the Moon (from Earth Vescape) above the 0.5 km/s required for trans-Mars injection. But an expendable second stage can throw the capsule to Mars (with poor staging efficiency). Adding an expendable landing 3rd stage to the Lunar version (to be abandoned on the Lunar surface like a LEM first stage), if it stages around GTO injection velocity, will boost the staging efficiency, and make the two capsules similar in payload capacity.

    Stacking the capsule on a Lunar landing stage will mean we use tall ladders for egress, but oh well. Having hardware for Mars and Lunar missions which differ only the addition of a Lunar landing stage could be an important way for SpaceX to fund a Mars rocket, assuming NASA only wants to buy Lunar transport services. And having a Lunar lander that can deliver 40 tons or so of cargo or a return capsule would be good too.

Leave a Reply

Your email address will not be published. Required fields are marked *