Dual-Fluid Single-Launch Propellant Depots

[Note: Welcome AIAA Daily Launch readers!  Feel free to check out other posts on propellant depots, as well as the Altius Space Machines blog where I talk about what my current company is up to.]

I was going to write a short post today about a variant on the Dual-Fluid, Single-Launch Propellant Depot idea, when I realized I had never actually gotten around to explaining the idea here on the blog.  While this idea has been now discussed in a few of our AIAA papers ULA and I have published on the topic, I figured it was still worthwhile to do a brief blog post for those who haven’t had the time to read any of these papers.

Back in early 2009, Frank Zegler (of ULA) and I both independently came up with the concept for a LOX/LH2 propellant depot of decent propellant capacity that could be launched on a single Atlas V.  The basic idea, illustrated in the picture below (lifted from Bernard’s presentation from last month that is the third link above) is pretty simple:

Dual Fluid Propellant Depot (Credit: Bernard Kutter, ULA 2010)

Dual Fluid Propellant Depot (Credit: Bernard Kutter, ULA 2010)

  • You have three parts, the Centaur upper stage that is used to launch the depot, a central section that holds the depot controls, docking interfaces, propellant transfer interfaces, etc, and a depot LH2 tank that is manufactured using the same tooling and processes as the Centaur tanks.
  • When you get to orbit, you transfer all the excess LH2 from the Centaur to the depot LH2 tank, you then vent the Centaur LH2 tank down to vacuum, and permanently seal off the connection between the Centaur LH2 tank and the depot LH2 tank.
  • You then close off the Centaur LH2 tank, and transfer the remaining LOX from the Centaur into it.  It now becomes the LOX tank for the depot.
  • In order to reduce propellant boiloff to reasonable levels, ULA has typically suggested either the use of MLI and/or deployable sunshields to cut down on the heat flux into the tanks.
  • Boiled-off hydrogen is used as a “heat sponge”, to intercept heat flowing into first the LH2 tank than the LOX tank.  Eventually the now much warmer GH2 is run through a rocket nozzle to provide settling force and station reboost.  It turns out that the boiloff rate achievable with good passive system design is lower than the amount of propellant you need to use anyway for stationkeeping/reboost, so for depots in LEO or L1/L2, you get the benefits of a ZBO system with separate reboost capabilities without the difficulty of building a ZBO system.

The end result is that with a Centaur diameter depot LH2 tank, you can store around 30 tonnes of propellant.  It’s a bit on the small side, but enough to fully refuel a Centaur or Delta-IV upper stage in orbit.  It has both propellant.  It requires no orbital assembly, no EVAs, no new tank tooling, etc.   The depot tank and depot center section together weigh less than 2 tonnes, so you can actually use the Centaur to deliver a depot like this to anywhere in cislunar space, and with good passive shielding (sun shields etc) to Mars orbit or to a NEO you want to visit.

There are ways of going bigger than 30 tonnes that have been discussed, the two primary ways suggested have been having the depot LH2 tank built into the 5m diameter fairing (instead of keeping it Centaur diameter).  That gets you up to ~60-70 tonnes of propellant.  Going with ACES style tanks can get you up into the 110-120 tonne capacity, but require the development of the ACES stage.  Even 30 tonnes of propellant, if you have depots in both LEO and at L1 or L2 is getting into the range that you can do very interesting things.

Anyhow, just wanted to introduce people to the idea if they hadn’t heard of it before.  Read the papers for more details.

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.
This entry was posted in Propellant Depots. Bookmark the permalink.

48 Responses to Dual-Fluid Single-Launch Propellant Depots

  1. Coastal Ron says:

    Unfortunately the people that want HLV’s don’t want fuel depots to be proven out, because then the need for HLV’s, in the near term, go away.

  2. Actually HLVs are extremely compatible with space depots since it would enable NASA to deploy huge depots at LEO and at the Lagrange points. This would enable reusable ACES shuttles to easily transport humans from LEO to the Lagrange points or to lunar orbit and back.

    If NASA was smart, they’d replace the hypergolic Service Module concept for the next crew exploratory vehicle with an ACES 41 Service Module.

  3. Jon, have you ever considered that by throwing in your lot with the LH2 crowd you’re cutting out the non-government market? If you look at the history of cryogenic development, or even the recent difficulties of national efforts in other countries, it’s hard to imagine cryogenic propellant depot resupply ever being a free market with actual competition. LH2 handling remains black magic.

    On the other hand, LOX/hydrocarbon is mostly public technique.. you could almost call it engineering. Amateurs do it. The idea that upstart competitors can enter the market with cheap (maybe even unreliable) launchers and deliver commodity propellants is a lot easier to believe without all the government secrecy.

  4. Marcel,
    You *don’t* need HLVs to deploy impressively large depots to LEO and L1.

    But the rest of your points are actually reasonable.


  5. Trent,
    No I don’t think I’m throwing in my lot with government. Most LH2 experience is actually in the private sector. I happen to know for instance a small General Dynamics spinoff down in So. Cal that has all the facilities you need for learning how to use LH2, testing LH2 systems, etc. I think there are several potential commercial markets that could be enabled by LOX/LH2 depots.


  6. Note I said “enabled”, which kind of entails that they could only come into being if the depot came first.


  7. A_M_Swallow says:

    30 tonne of propellant as cargo is good. The LEM only carried about 10 tonnes.

  8. A_M_Swallow says:

    Batteries can be recharged but they can also be replaced. Could a ‘Dual Fluid Propellant Depot’ be used as the pre-filled fuel tanks of a lander or transfer vehicle?

    A depot would presumably hold several of these cartridges and have space arms to fit them. There is still plenty of time to change the design of BLEO spacecraft to use them.

  9. Jon, Boeing/Lockheed are not exactly nimble non-government players. Seriously, the only LH2 work I’ve seen outside dinospace is the pump XCOR did, and they did it under contract for ULA. If you see propellant depots as a way to open a free market and encourage actual change in the launch industry – LH2 and all the secrecy that surrounds it sounds like a poison pill.

    And while I appreciate an “I disagree” over nothing, something more substantial would be better if you have time 😉

  10. Jonathan Goff Jonathan Goff says:

    There’s nothing secret about handling LH2. It’s handled industrial every day, mostly by people who probably only have high school diplomas. Just like LOX. The company I had in mind was Innovative Engineering Solutions down in San Diego. I was talking with Mark Wollen from down there just yesterday. Most of the LH2 testing they do is probably for ULA (they’re also one of the leads on CRYOTE), but I didn’t actually ask him.

    Check out the uses list towards the end of this page:

    Yes, LH2 is a pain in the backside to work with, worse than most other common alt.space fuels/oxidizers.

    Yes LH2’s properties mean that the alt.space crowd probably isn’t going to have much use for it (unless they’re doing contracting work for the bigger aerospace companies).

    But no, working with LH2 is not some deep dark secret. It’s used all the time in industry, and if you have a need to learn how to use it safely, you can probably find someone who can teach you pretty quickly.

    For in-space large-scale transportation, I think it’s a much better choice than Kerosene or even Methane.


  11. Jon, making LH2 rockets is considered a national asset. There’s extra laws on top of the standard ITAR nonsense. You know this.. we’ve discussed it previously. I’m not sure why you’re dodging the question. It’s generally accepted that SpaceX will take 10 years to develop their LH2 capability. I’ve spoken with SpaceX engineers who think that is an underestimation.

    So.. let’s try this again. Who is going to supply LH2 to a propellant depot? ULA and….? who? International partners is a euphemism for national space programs of other countries. This is certainly an improvement over heavy lift….

  12. Ed Minchau says:

    I could see using a 5m tank for LH2, but the smaller, Centaur sized tanks might be more suitable for denser propellants like methane or RP1 or Xenon or Hydrazine. However, if several such tankers rendezvous in orbit, then the leftover LH2 (or RP-1) and LOX can be transferred to the appropriate holding tank. The standard plumbing interface is a key enabler. I like the idea of mooting boil-off by using it for reboost.

  13. Simple curiosity here: what is the reasoning behind making the Centaur LH2 into the Centaur & Depot LOX tank? Size thus thermal? Are the benefits really big enough to be worth it?

    (Hadn’t noticed that part before, I thought I had read all the papers but it’s been a while).

    “Theoretically” If a LOX/LH2 depot works then everyone ought to be convinced a LOX/anything depot will work as well (disclaimer: in a hypothetical sane world …right, so much for that then).

  14. In case it wasn’t clear the last bit was on how a LOX/LH2 depot at least in theory could possibly have a nice benefit as proof of concept despite Trent most likely being right.

  15. Trent,
    Jon, making LH2 rockets is considered a national asset. There’s extra laws on top of the standard ITAR nonsense. You know this.. we’ve discussed it previously.

    Can you cite these extra laws? Because I sure haven’t heard of them. Legally, there’s nothing stopping me from deciding to build a LOX/LH2 engine and test stand. What’s stopping me and most of the alt.space players is that LOX/LH2 is challenging and expensive enough that we haven’t had the need/financial justification to pursue it. I repeat that there are now “extra laws” or national secrets involved here. Any company in the US that wants to do LOX/LH2 rockets or LH2 tanks can do so whenever they want. There are companies that provide consulting on how to do this, and you don’t need a super sekrit clearance or anything to talk with them. Just folding-dead-presidents.

    I’m not sure why you’re dodging the question.

    I’m not dodging the question, I’m trying to tell you that you’re wrong. 🙂 There’s a difference.

    It’s generally accepted that SpaceX will take 10 years to develop their LH2 capability. I’ve spoken with SpaceX engineers who think that is an underestimation.

    It’s “generally accepted” by whom? I didn’t ask Tom Mueller about timelines for Raptor last time I was down there, but my understanding that the main thing holding them back was a compelling business case to do so. Remember, that many of the top guys in SpaceX’s propulsion team (Tom included) have built and fired LOX/LH2 engines before.

    Sure there’s a bunch of subtleties, but if SpaceX had the money to build a LOX/LH2 stage, I’d be really surprised at this point if it took them more than 3-4 years to do so…

    …and more to the point, a delivery tanker is nowhere near as complicated as a full LOX/LH2 stage with LOX/LH2 propulsion.

    So.. let’s try this again. Who is going to supply LH2 to a propellant depot? ULA and….? who?

    SpaceX. Notice that whenever Dallas Bienhoff (of Boeing) writes about depots, he always talks about using SpaceX as the launch provider. He doesn’t seem to think this is a big problem. That might be a hint. Admittedly he was talking about using a Boeing-designed reusable tanker flying on top of a SpaceX vehicle, but point remains–the barriers preventing SpaceX from becoming able to launch LOX/LH2 exist mostly in your head, not in reality.

    The main reasons I focus on LH2 for depots are that:

    1-That’s where the largest demand is likely to be. Of the liquid upper stages in the US right now, only SpaceX’s uses Kerosene, so I’d have only one customer if I went that route. But both ULA vehicles use LH2, all future government US’s use LH2, and even SpaceX is looking at moving to LH2.

    2-Hypergols don’t really need full-blown depots. All the technology for them exists, there just hasn’t been the market justification for them yet, and the demand for them is in enough different inclinations and orbital planes that you’ll still be talking really small “depots”. I’m hoping that will change, but it’s not there yet.

    3-Even when hypergolic demand does finally get there, it’s likely to be tons less than that for LOX/LH2. Stationkeeping just doesn’t consume anywhere near as much energy as large orbit transfers does.

    4-For commercial exploration beyond LEO, I think the extra hassles of using LOX/LH2 instead of LOX/Kero or LOX/Methane are going to be outweighed by the much lower amount of mass that needs to be shipped around just to enable transportation. To use a naval analogy, coal might be cheaper than oil, but since you need so much of it, there’s a reason why most boats started switching over to diesel when it became available.

    5-When ISRU becomes available LOX/LH2 is the best positioned to take advantage of it. If only LOX is available, LOX/LH2 has a much higher O/F ratio than any other normal propellant combination, so you can provide more of the propellant locally. If water is available you get both. Sure they have found some hydrocarbon traces in the lunar poles, but I think LOX/LH2 is still going to have a leg up.

    6-For a depot in a position where station-keeping is required, a LOX/LH2 depot with good passive thermal control is going to take the same or less total mass loss than a LOX/Kero depot. Even if the LOX/Kero depot had zero-boiloff on the LOX, you still need to do stationkeeping and reboost, and warm GH2 has a similar Isp to LOX/Kero.

    The list could go on. My point is just that while LH2 may be hard for amateurs to work with, and while there isn’t much economic reason to want to mess with it for the suborbital stuff most of the industry is focused on, that doesn’t mean it doesn’t make sense for in-space work. I’ve noticed that most of the griping about LH2 comes from people who’ve never used it. There are plenty of legitimate gripes about LH2, but most of the people who work with it on a regular basis have learned to deal.


  16. It might also be worth asking Jeff Greason what he thinks about LOX/LH2 for depots.


  17. Anonymous says:


    Is the future market for Argon propellant transfer larger than the market for LH2?

    It appears that DARPA, AFRL, and NASA are all starting to embrace solar electric propulsion using Argon as the fuel for their high delta-V missions (including LEO-to-GEO, Lagrange Point, and Mars Tranfers). There are some missions that are already funded using this approach.

  18. A_M_Swallow says:

    There may be a need to review the actual product the LEO depot sells. Looking at the figures in this thread a Centaur with a RL-10A4-2, ISP 451s and 30 tonne of propellant can send a payload of about 20 tonne to EML-1. That is more than twice what a Dragon capsule weighs.

    So the main product of the LEO depot could be inspace Centaurs.

    There would have to be a few differences to the Centaur basically man rating and a docking connector. The connector could be the Common Docking connector used in the ISS or possibly LIDS. The depot’s tug and robotic arm may have to dock Centaur and cargo.

    The EML-1 Depot would still need to sell propellant to landers, unless the engines are replaced each time (very expensive).

  19. A_M_Swallow says:

    A lot of the problems with hydrogen go away if its use is hidden from the payload people. If they can leave the hydrogen handling to the launch pad and depot operators only specialist firms needs the skills.

    Handling liquid hydrogen on Earth is a well known skill. Filling the tanker LV’s tanks is the job of the launch pad people. Transferring to and from the depot’s tanks is the job of the depot’s operators. Designing the tanks is a specialist skill, users just have to buy them.

  20. Tom D says:

    I have to agree with Habitat Hermit, once a LOX/LH2 propellant depot has been successfully demonstrated, doing LOX/Methane (if that is more practical) should be easy. I find it very frustrating that NASA has been dragging their feet over depots. With depots it looks like we could start sending people to NEOs and possibly even Mars within the next 10 years with no new launch vehicles!

  21. A_M_Swallow says:

    The Common Extensible Cryogenic Engine, or CECE, is a hydrogen burning engine under development for landing on the Moon.

  22. Googaw says:

    Doing the hardest and biggest thing first is quite the backwards way to do things. Clearly H2/O2 is the kind of depot you work on if you’re angling for a NASA (sub-)contract to do one of the various astronaut-based economic fantasies (excuse me, “exploration”. Never mind that the real exploration is being done by telescopes and robots). If you want to do something actually useful for real commerce or defense, the storables they already use are the obvious way to go: much smaller scale and much more mature technology means a much smaller and less risky capital expenditure. So the choice of H2/O2 by Zegler et. al. sends a very clear message about whose problems they are and are not trying to solve. And they aren’t the problems of real commerce or defense, or even of real exploration.

    As for commercial or DoD depots being “too small”, yes I imagine they are too small for groups used to big fat NASA astronaut project contracts. But the first “gas stations” on earth were jerry cans. Small is where reality starts outside of NASA contractor economic fantasies. And BTW, a very large proportion of commercial and DoD satellites are in a single orbit and plane, namely GEO.

    What Jon is actually doing with his actual company in the contracts he’s announced is far more useful, IMHO.

  23. Googaw says:

    BTW, Jon mentioned ISRU as a justification to go with H2/O2. That’s based on speculation on the concentrations of various volatile components (H2O, CH4, NH3, etc.) in the lunar polar ices and how much it would cost to manufacture the various alternative propellants from these ices. Before we commit to justifying any such choice based on ISRU, we should actually send some little drilling robots to the lunar poles to actually see what is there and in what concentrations, and then do comparative studies of the costs of making the various alternative propellants.

    What we do now know, hopefully, is some economics. The economics of ISRU are such that, if it makes economic sense at all, after the capital investments are paid off it’s going to drive down the cost of propellants on orbit quite substantially. But if the cost of propellants are cheap but the storage (tank etc.) remains expensive, that radically changes the economic assumptions we bring to the rocket equation. Any storage scheme that increases storage costs (weight of tanks, thermal shields, etc. launched) to save propellant mass makes less and less sense as propellant costs decrease relative to the tank et. al. costs. As the energy imparted to the propellant (either when made or when used) becomes more expensive relative to the propellant mass, high specific impulse becomes less of an advantage and indeed, with very cheap propellant a _lower_ specific impulse can actually make more economic sense (Anthony Zuppero has some good econometrics on this).

    The bottom line is that when we switch to ISRU-based propellants the economic assumptions we bring to the rocket equation change radically, and not in favor of H2/O2 which requires heavier and costlier tank, thermal shield, and associated equipment launched from earth than other propellants in order to gain advantages that decrease the cheaper propellant mass gets. Depending on what we find at the lunar poles, this may mean that H2/O2 is an extremely poor choice for ISRU, and it’s quite premature to use ISRU as a justification for that choice.

  24. Jon, I know you’re not ignorant of the technology transfer paperwork required to do LH2 work.

    Admittedly he was talking about using a Boeing-designed reusable tanker flying on top of a SpaceX vehicle

    Bingo! We have a winner. Catching on now?

    but point remains–the barriers preventing SpaceX from becoming able to launch LOX/LH2 exist mostly in your head, not in reality.

    How’s that cognitive dissonance going? My argument here in a nutshell: the if-you-build-it-suppliers-will-come argument for propellant depots doesn’t hold up for LH2 because the barrier-to-entry is too damn high. The only way competitors can enter the market is with assistance from the established players, who will ensure that the new entrants remain under their thumb.

    Now consider the alternative.. as Mr Googaw harps on about, the commercial customers for propellant don’t care about LH2, they care about hydrazine and other storables. Who can fly this stuff? Why, everyone. Unlike Mr Googaw, let me throw some numbers at you.

    both assuming pmf=0.1
    LH2 (isp 435) required to put 10t into LTO = ~12,015 kg.
    Storable (isp 312) required to put 10t into LTO = ~21,397 kg.

    For the sake of argument, let’s grossly round this up and say a storable propellant stage would have to be twice the mass of an LH2 stage. So what? This means you need twice as many flights to fill up the depot, yes. That’s fine, because we can take as long as we like to do it now and don’t have to worry as much about boiloff. (even with reboost use for LH2 boiloff it’s still an issue).

    Twice as many flights means twice the price right? Well, no. A lack of volume is precisely what makes launch so expensive. Double the launch rate and you’ll more than half the price – that’s what the koolaid says. But more importantly, the increased ease of handling storable propellants over LH2 means that more competition will be available, and that will drive the prices even lower.

    In short, more launches means more competition which means lower prices – and LH2 seems like the exact opposite way to get that. LH2 belongs to the performance blinded von Braun disciples.

  25. Paul says:

    I assume that common point (at the top of the Centaur, bottom of the central section) of the piping, is the docking/fuelling point? Does that mean this is a “dumb” depot, needing “smart” tugs?

    If you’ve flown even one LH/LOx depot, I can’t imagine you’ll have much difficulty convincing backers that you can handle cryo-nobles if there’s ever a market.

    “Before we commit to justifying any such choice based on ISRU, we should actually send some little drilling robots to the lunar poles to actually see what is there”

    The first half of that sentence in redundant. 🙂

    “The bottom line is that when we switch to ISRU-based propellants the economic assumptions we bring to the rocket equation change radically, and not in favor of H2/O2”

    Jon might correct me, but I’m assuming that this depot a) will not be the only depot ever launched, just possibly the first, and b) will not fly for very long (years, not decades.) By the time ISRU-propellant is a commercial proposition, depots might be on their fourth generation and you can build whatever type you like.

    But if LH/LOx is easier for the first non-commercial ISRU propellant, it would help you justify ISRU to your backers if the depot/fuel-transfer technology was not just possible, but flying. (Once you trick talk them into funding you, you can switch to whatever the economics justify.)

    Trent, re:ITAR and handling LH2,
    I get the feeling you think Jon is treating this proposal as the One-True-Depot-Idea-Which-Replaces-All-Others. You’re not going to fly anything unless there’s a market. LH/LOx means NASA/DoD (or subbing for ULA, or sub-subbing for SpaceX subbing for ULA). If so, that solves both your ITAR/regs problem and the LH2 handling experience required.

    IMO, if the first depot is purely commercial NewSpace, it will fly Hydrazine. But, also IMO, the first depot will probably not be purely commercial, nor purely NewSpace. So if Jon’s not watching LH/LOx, he’s not in the game.

    And once the first depot flies, it’s there for the world to see. Developing a second depot flying LH/Kero or Hydrazine or whatever is going to be vastly easier to justify, and cheaper, than the first. And if the first is LH/LOx, no one can say “Sure it’s theoretically useful, but the first depot didn’t solve the [something] issue.” LH/LOx is the hard problem. Once you’ve flown that, every other application is easier.

  26. Brad says:

    “It turns out that the boiloff rate achievable with good passive system design is lower than the amount of propellant you need to use anyway for stationkeeping/reboost, so for depots in LEO or L1/L2, you get the benefits of a ZBO system with separate reboost capabilities without the difficulty of building a ZBO system.”

    That’s wonderful. That convinces me of the practicality of the depot portion of a liquid hydrogen/LOX propellant transfer architecture. I’m not convinced however about all the other elements.

    Let’s be clear. The only practical purpose for storing and transferring in space such large amounts of propellant is for the support of manned space-flight operations with destinations beyond LEO. That being the case the question arises: is hydrogen/LOX the best choice for the first depot project? The practicality of liquid hydrogen vs alternative propellants needs to be evaluated by comparing entire proposed architectures.

    Clearly hydrogen is better for departure stages, but a good case could also be made for a hypergolic propellant depot so that manned landers and/or manned Earth return vehicles could be dry launched into LEO. Odds are both types of propellant depots would be useful, but the question is which type would be easier and more practical to do first?

    One question about a hydrogen depot that immediately pops into mind is how large does the depot supply spacecraft need to be for whole depot system to be practical? How small could you go before boil-off and/or handling issues make a hydrogen supply spacecraft impractical?

  27. Trent,
    You only have to do export control paperwork if you’re doing an export. And even then, I don’t think that exporting LOX/LH2 work is any harder than say exporting a satellite component. As far as I’ve seen there’s no extra layer of legal complexity.

    As far as SpaceX and other players not being able to make an LH2 tanker for delivery…that’s an assertion, not a fact. I’m pretty confident that with the LH2 experience that many of their lead engineers have, they could do that. It’s a lot easier to make an LH2 transfer tank than it is to make a full LOX/LH2 stage and LOX/LH2 engine.


  28. Googaw,

    Re: polar volatiles. All competing theories for polar volatiles suggest a higher abundance of Hydrogen than Methane or other fuels. The LCROSS data supports this strongly. And as I said before, LOX/LH2 has an O/F ratio of 5.5-6.0, which means that ~85% of the propellant is LOX, which is definitely available. Whereas for LOX Methane or LOX Kero, it’s a smaller percentage, and you need more of it due to the lower Isp…I think that as far as lunar ISRU is concerned, LOX/LH2 still looks pretty favorable.

    Re: hypergolic depots. While it’s true that most satellites that could be refueled use hypergols, there are very few concentrations of satellites that would justify an actual *depot* per se. GEO is probably the only one. In most other cases, you’d be better off just launching a refueling tug, and refueling it directly. Depots are most useful when the size of the supplier and the size of the demander are significantly different. I’m not saying that hypergolic refueling isn’t important, just that it doesn’t provide enough volume demand to be the end-all-be-all in my book. More importantly, the hypergolic “depots” are really mostly market problems, not technical ones. There are one or two technologies we’re doing at ASM that should help, but we’ve already got what it takes to make those happen…just nobody has been able to close a business case on doing so yet.

    re: LOX/LH2 only being useful for economic fantasies…this just isn’t the case. Even without manned exploration, propellant used for leaving LEO (either for GEO satellite deliveries, deep space probes, etc) is a major market. If you can take a satellite (NASA Science, DoD, or commercial) that would’ve taken a DIV-H to launch, and instead allow it to be put up on an Atlas V 501, with $50M worth of prop on orbit, you’ve saved a lot of money. Stopping at a depot which probably is at a moderate inclination provides a small delta-V penalty, but if the propellant is cheaper than a big LV it still may be worth it. As for deep-space probes, depots can help with those too, by making it easier to launch them, with simpler vehicles and a refueling step in space (LEO and/or L1/L2). And lastly regarding manned space “economic fantasies”, if NASA or ESA, or someone else is willing to buy propellant for stuff you think is economically irrelevant, does it really make their money any less green?


  29. Paul,
    re: your first question, yes the rendezvous/docking/transfer area would be the middle section between the two tanks. I’m working on a technology (that I hope I’ll be able to talk about publicly on the blog soon) that might enable you to use a relatively dumb tanker without needing a tug. But worst case, I’d rather go with a tug stationed at the depot than try to do smart tankers (unless they were fully reusable–but even then I’d prefer dumber tankers to enable a larger market for them).


  30. Brad,
    Re your last question, about how small can an LH2 tanker go without it becoming too much of a problem. My goal is to make it feasible to delivery propellant in chunks as small as 100-500lb, assuming there are small RLVs or other ways of launching it that small.

    Key ideas:
    1-minimal penetrations on the prop tank
    2-good passive insulation
    3-subcool the LH2 as far as you can before launch
    4-offload “smarts” to tugs or the station itself


  31. A_M_Swallow says:

    With boiloff hydrogen depots are going to need accurate meters since the quantity launched and the quantity delivered are different.

  32. AM,
    Yes, not just with LH2. That’s a concern with any cryogenic propellant. The term of art varies a bit, but “mass-gauging” is the term I see most often mentioned in the papers. There are several approaches to solving the problem, but once you’ve settled the propellant it becomes easier.


  33. BTW guys, I found out that this blog post got mentioned on the AIAA Daily Launch. If possible, let’s try to keep questions and comments to those directly related to this technical concept.


  34. Aerostadt says:

    What are the LH2 and LO2 evaporation loss rates? Are they acceptable over weeks, months, years?

  35. Jonathan Goff Jonathan Goff says:


    Good question. I don’t have the specific numbers on me right now, but my understanding was that you use the LH2 boiloff to cool the LOX enough that you completely suppress its boiloff, and then use the warmer GH2 for stationkeeping. The total amount of LH2 boiloff per year was low enough that if you were actually using the depot it was in the noise, but would be sizeable if you only used the depot occasionally. IIRC, the total percentage loss was in the mid single-digit percents, but mostly LH2…Frank Zegler’s AIAA SPACE 2009 exploration paper (http://ulalaunch.com/site/docs/publications/AffordableExplorationArchitecture2009.pdf) had more details, but I’d have to dig them up.


  36. Jonathan Goff Jonathan Goff says:

    Actually, I’ll have to dig a bit further for specific numbers.

  37. Specific Imp says:

    Actually there are a lot of good reasons to use a 28.5 inclined depot beyond crewed operations. Basically with a depot and an upper stage capable of flight operations durations measured in several days you can lift pretty much every conceivable GSO payload on the smallest possible rocket. You boost to LEO on a nice, cheap, super-reliable and low-environments Atlas 401 or Falcon 9 and rendezvous with the depot. These are nice and safe but do have the highest cost per pound of mass lifted. But you can get 20t to LEO with those simple rockets- that’s a lot of xenon or you can go with bigger solar panels, bigger antennas etc. Regardless, you max out the performance it does have since the residuals are important.

    Depot propellants are brought up on high-rate, cheaper per pound but less reliable things like Atlas 551’s or big Falcon HLV’s. Or an RLV or something made in Botswana for that matter. You buy the additional propellants you need at market rate or exercise your propellant delivery options that you invested in years before. Take on a couple tons to supplement your residuals and then perform a direct inject and inclination removal to GSO with the same ascent stage and you can then perform the soon-to-be-mandatory deorbit task with the big margins you have. Perhaps deed the vehicle over to some hangmen who arrange to have another derelict payload or two at GSO removed with the same upper stage after you are done with it. They take over control and use your residuals/margins to do real work and then finally end the mission with upper stage decommissioning and a paycheck from the international launch tax fund.

    With existing vehicles and one dinky depot you can emplace huge GSO payloads without invoking monstrous launchers. Same goes for science payloads to Mars or the outer planets which would never fly on anything other than the smallest rocket. Nice when dealing with radioactive stuff too to not have big solids around. Plus the presence of depot technology implies the practical delivery of cryogenic propellants to Mars orbit and hence a fully powered, precision-targeted descent without the necessity of having to bring ballast, heat shields the size of merry-go-rounds or contraption-intensive sky-cranes.

    Of course if you decide to do the L2 depot you can move monstrous things to Pluto if you are so inclined. L2 enables practical lunar surface operations including propellant synthesis and delivery at a cost which will eventually mean that the LEO depot gets its propellants from the moon. You just have to wait about 40 years or so after surface operations commencement.

  38. Paul says:

    Given that the Centaur is not discarded but has been made an integral part of the depot, is there really any cost saving in not creating a depot/upper-stage combo as a single purpose built item?

    Intuition says that adapting something that already exists is cheaper than starting from scratch, but in practice it’s often harder, therefore more-expensive/less-robust.

    Since you are already creating the H2 tank + control-section + dock/etc, and you’re already modifying the Centaur’s plumbing to integrate into the depot, wouldn’t it be more logical to create the entire unit from scratch?

  39. Paul,

    I kind of doubt it. You’d have to change a whole lot more to do it that way, at least I think it would. A depot done this way is probably a $200-300M project including the launch. A depot done that way could easily end up in the $1B+ range.


  40. steven rappolee says:

    Benard Kutter and myself where kicking around the idea of LHE storage in a fuel depot architecture,since that time I have concluded storage is not economical, but perhaps a L2 fuel tanker could service a space telescope with LHE , and then transfer its fuel to a depot, or act as a depot it self.

    image is from B Kutter from a private communication with S Rappolee
    Fuel depots it seems lend them self’s to transport and storage of Liqiud noble elements for ion propulsion,but the question I have is can Ion engines find a home on an existing ULA upper stage/in space stage along with the RL-10 engine or as a pure Ion stage?
    and a hybrid Ion/chemical Centaur stage,
    A liquid Xenon ion powered centaur scared the stuff out of me, this is very dense and heavy fuel, it would never work as a tug or a planetary transfer stage, so that’s when the liquid Argon thought struck me as making more since ( the VASIMER folks like it)early papers out of UM suggest it, as did a decades old paper to a mars society conference.
    So I am stuck on this hybrid LH,LO2,LAr Centaur Tanker/ in space stage
    chemical energy gets us out of the van Allen belts,the ion engines spiral us out to L2,transfer fuel or pick up more,dock with a science mission, perform a chemical burn for a L2 to earth fly by and then a ion powered coast to an outer planet.Perhaps our tanker has performed some commercial servicing missions be fore leaving earth.Removing the stage along with a decommissioned satellite from cislunar space could also be a final mission.
    Hydrogen boil off cools the liquid Argon

  41. steven rappolee says:

    couple more thoughts,
    the ion engine thrust would impinge on the RL-10 motor nozzle (?)
    so you might need out board ion engines, just not to much so as to exceed the payload fairing.
    Or the ARCTUS or CRYOTE transports the liquid Argon or Helium with boil off from the attached Centaur as coolant and later versions of ARCTUS or CRYOTE evolve into our Ion powered in space transfer stage( third stage?)** Centaur still makes the trip out to a outer planet, still makes the insertion orbit burn but then separates ( does a final LCROSS type mission perhaps?)
    NASA and some private company’s want to build satellite servicing vehicles,perhaps our L2 tanker/tug/in space transfer vehicle could stay some what “dumb” by letting this vehicle do the thinking.

    *B Kutter referd me to a 2005 paper that suggests long term cryogenic stages as planetary orbiters, I suggest that, that space craft bus should be a hybrid LHe, LH, LO2, LAr hybrid 🙂

  42. Brad says:

    Re: #30

    That is about 90% smaller than the smallest answer I expected. I’m very curious what the loss rate is, and how it might vary with a change of mass scale.

    It is this aspect, the tanker, that I had been most skeptical of for a practical liquid hydrogen propellant depot. It is one thing to succeed in minimizing long term hydrogen boil-off from a large depot in space. But I would think a much more difficult job to reduce to acceptable levels hydrogen losses from small tankers.

    If a 100 pound hydrogen tanker is practical, I then wonder if that small a mass lies within the theoretical microsatellite payload capability of the XCOR Lynx II rocketplane?

  43. Brad,
    To be honest, 100lb is probably pushing it pretty far–you could probably do it for early proof-of-concept demonstration, but 300-500lb is likely closer to an economic sweet spot.

    The big trick with small tankers (in addition to normal “minimize penetrations and put good insulation on it” type advice) is to subcool the hydrogen. LH2 actually has a pretty amazing heat capacity…looks like about 9.7 kJ/kg*K. So, say you chill it down to just over its freezing point (14.2K) and you’re ok with letting it warm up to just it’s normal boiling point (call it 20.4K). That gives you 6.2K worth of “heat sponge”. That’s ~60kJ/kg, or about 2.7MJ for the total propellant load. The trick is just making sure you can keep the total heat load into the tank between when you fill it and when you can drop it off at the depot below this amount (and providing sufficient ullage that the nearly frozen LH2 can expand into as it warms). Tricky for a 100lb tank, but not impossible, I think.

    But realistically, I’d only use the 100lb tank size for demonstration purposes, and only if that happened to be a conveniently cheaper size.


  44. Pingback: Selenian Boondocks » Blog Archive » Random Thoughts: DCSS-Derived Dual-Fluid Depot

  45. Paul says:

    steven rappolee,
    “the ion engine thrust would impinge on the RL-10 motor nozzle (?) so you might need out board ion engines”

    I may be misreading your post, but provided the two engines (RL-10 & ion) are never simultaneously fired, then couldn’t the ion drive be at the opposite end of the craft? (On the centreline, but 180 degrees to the RL-10.) You’d need to turn over when you switch between drives, but that doesn’t seem especially difficult.

  46. steven rappolee says:

    that might work for a fuel depot supply mission, but for the satellite servicing mission or the planetary transfer mission I am not sure where the payload attachment point would be if you did this

  47. Pingback: Selenian Boondocks » Blog Archive » Tooting Someone Else’s Horn: Quest Product Development Corp’s Advanced MLI Technologies

  48. Pingback: Summary of Some ULA Papers from AIAA SPACE 2015 | Selenian Boondocks

Leave a Reply

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