Space Access 2010 Propellant Depot Presentations

It’s been a rather interesting month so far, and I’ve been under a bit too much stress lately to blog much, but I wanted to put up some of the presentations from the Propellant Depot panel I was on at Space Access this year.  If I had found the time sooner I would also say something about the advanced technology panel I was on, but it’s now been long enough I can’t recall what I was going to say.

Here is the humor slide I started out with:

We Don’t Need No Steeking Propellant Depots!

My actual presentation:

RLV Friendly Depots

Bernard Kutter’s presentation for ULA:

Near Term Depots

and Dallas Bienhoff’s presentation for Boeing:

Space Transportation Impedance Matching

I haven’t been given a copy of Rand’s presentation yet.

[Edit, here it is, Rand says he’ll probably get some annotations up later]

Anyhow, a few quick random thoughts that I don’t think anyone else has really hit upon on the intarwebs:

  • One of the concepts out of Dallas’s presentation I liked was the idea of having a space transfer tug that takes landers from EML1 (L2 would also work) to some perilune trajectory, and then returns to EML1.  I’ve been toying with variants of this idea for some time.  With a Centaur-sized trasnsfer tug, fully-tanked-up in EML-1/2, you can actually bring pretty darned big landers most of the way to the lunar surface (ie leaving 1000m/s or less of delta-V for the descent), while still having enough propellant to return to lunar orbit and from there to the L-point station.  That segment is probably one of the easiest in-space segments to start doing reusable stages, since you don’t need an aerobrake, and don’t have to deal with lunar dust, just propellant transfer, and lots of engine relights.
  • In a conversation with Jeff Greason late one night at the conference, we got off onto the topic of RLVs and propellant depots.  One of Jeff’s opinions is that in order to really have an industry for some service, you need enough demand to allow for 2-3 healthy competitors.  With only one provider, you get monopolies, three is ideal.  But for RLVs you probably want a small fleet (~3 vehicles) of RLVs so that you can provide dependable service even if you either have a mishap or have to pull one of the vehicles for maintenance or repair.  Having a single vehicle may work during the development phase where you’re transitioning into operations, but once you’re in full operations, you want enough demand for 2-3 companies with probably 2-3 vehicles per year.  And for each of those vehicles, in order to get the per flight price in a really good range, you need to fly often–Jeff says 100 times per year, but I’ve heard numbers as low as 30-50 (but any way you slice it, it’s a lot of flights).  That comes out to somewhere in the 120-900 flights per year range.  The interesting thing that Jeff mentioned was that if you postulated very small RLVs to start with (say 300-500lb to LEO net payload capacity), just one lunar mission per year would be enough to provide enough demand for an entire healthy industry by itself.  Towards the lower ends of that scale, you’d only need one “soyuz around the moon” flight, or 1-3 GEO flights that used a propellant tank-up in LEO (say using a Falcon 1 with a mini-Raptor type LOX/LH2 upper stage?) to provide enough demand for at least the starting of an industry.
    • While 300-500lb to orbit sounds tiny, that’s actually a pretty reasonable size for a first-generation RLV.  The first stage doesn’t end up being that much bigger than existing or planned suborbital vehicles, doesn’t have to have much more capability either.  The upper stage ends up down in the middle of the size range for proposed suborbital vehicles.  While it has a much higher performance requirement, and much nastier reentry environment, it’s on a size that you can realistically work with a lot easier.  Also, a lot of the TPS work can be refined by flying “expendable” upper stages on these first generation commercial suborbital launchers.
    • This would definitely require the sort of RLV-friendly depot setup I described in my presentations–you’d have to have tugs that carry all the rendezvous/docking smarts, and keep the RLV-side of the propellant system as dumb as possible
    • Propellants are a much less demanding payload than people.  Not only does this keep up-front development costs down, but it also reduces the business risk if you happen to lose a vehicle occasionally.  While high flight-rate RLVs should be capable of high reliability, we’re also talking about 1st or 2nd generation systems here, where we’re still learning a lot–and learning can be painful.
  • I also liked Bernard Kutter’s graphic of the simple, single-launch, dual-fluid depot concept.  This is a simpler version of the ideas Frank Z and I came up with last year (it uses a stock Centaur-sized tank for the LH2 side of the depot), but is still quite capable–on the order of 30mT capacity is nothing to sneeze at.  With one of those in LEO and one in L2, that’s actually enough to do an ESAS-capacity lunar transportation system without Heavy Lift.
    • One of the really interesting possibilities is that if something like this demonstrator depot were chosen as a part of the money Obama has proposed for orbital refueling technology demonstration (it wouldn’t need anywhere near the full $400M-1B that Obama mentioned per technology area), if the demo system worked, it would actually be operationally useful.  Sure, you’d want to replace it and/or upgrade it down the road with lessons learned, but I’m a fan of pressing technology demos into operational service, as that’s a good way to get a lot more data out of the deal.
  • I also liked how Bernard explained a lot of the cryo storage issues.  A lot of this stuff still needs to be proven in space, but they (ULA, LM, and Boeing) have a lot more experience doing related tasks than most people realize.

I probably have some more thoughts on the matter, but I’m home at sick with a cold today, so I’ll leave it at that.

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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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.
Jonathan Goff

About Jonathan Goff

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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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 Commercial Space, Launch Vehicles, Lunar Commerce, Propellant Depots, RLV Markets. Bookmark the permalink.

39 Responses to Space Access 2010 Propellant Depot Presentations

  1. john hare says:

    Sure, you’d want to replace it and/or upgrade it down the road with lessons learned, but I’m a fan of pressing technology demos into operational service, as that’s a good way to get a lot more data out of the deal.

    Jon I know you will have to swat me for this but, Yah it worked real good for the Shuttle. 🙂

  2. Jonathan Goff Jonathan Goff says:

    I think the difference here is I’m talking about taking something that’s intentionally a subscale demonstrator and seeing if it makes sense to milk some operational experience out of it. Shuttle was always intended to be an operational vehicle, with only one full-scale prototype bird as a sort of demonstrator. But that’s a fair warning–one has to be careful of what Ben calls the “YAGNI” effect. YAGNI stands for Ya Ain’t Gonna Need It, and is whenever you try to design stuff too far ahead of where you really are at, and end up finding out that you didn’t take the right approach. Not that I have any experience with YAGNI…no siree…


  3. john hare says:

    I agree with you, but I had to throw at least one barb at that quote. Some of my construction prototypes were good enough to be the final machine. Most were the opposite of course. Some not only failed, but busted the whole concept. Which leads to the way I do demonstrators as cheap as I can think of.

    Sometimes they are not even worth return freight.:-)

  4. Mike Puckett says:

    You could always offer a deposit on the return of your ‘drop tanks’.

  5. Rand Simberg says:

    You’re not a Jewish mother, are you?

  6. A_M_Swallow says:

    The ‘Propellant depot’ article needs a diagram. If anyone has a copyright free picture of a depot the picture could usefully be uploaded.

  7. A_M_Swallow says:

    OSC is planning to give the Cygnus spacecraft an optional arm. This would allow them to be used as docking tugs.

    Page 6 of this presentation.

  8. Critics of propellant depots point out the lack of a DRM for going BEO. It almost sounds like you addressed some parts of it, with the orbit selection, and the tugs. Some numbers on the boil-off losses, how that relates to launch rate, etc. If you’re talking about 120-900 launches/year then surely the first launch will contribute less to the depot than the last launch.. by how much?

    Does kinda sound like some handwavium on the tugs though. It would be nice to see the DRM variation between the various tug possibilities.

  9. googaw says:

    the lack of a DRM for going BEO

    Not to mention the lack of a real market for going BEO at such a large scale. It’s more astronauts for the sake of astronauts, taxpayer funded as usual.

  10. Pingback: It’s been Ages « The Four Part Land

  11. Jonathan Goff Jonathan Goff says:

    Cygnus would be a really big tug for a 300-500lb to orbit RLV. If the tug starts getting too big compared to what it’s moving, the propellant costs start going up. It might work as a tug, but is really several times bigger than I would want. With something like boom rendezvous on both ends, you could get away with a much simpler tug system.


  12. Jonathan Goff Jonathan Goff says:

    We addressed boil-off issues in a depot paper I coauthored last year. Generally a depot in LEO has a boiloff rate that’s on par with the amount of propellant you would need for stationkeeping anyway. It’s still a large amount of hydrogen (not sure what the numbers would be for this smaller depot), so you really want to have a decent throughput–if you’re doing lunar or beyond mission, you want to be constantly shipping propellant up-hill from LEO either in tankers, or in outbound payload stages. For the big ACES style depot, the resulting loss rate was ~3.3% of the yearly throughput. Now a smaller depot with lower throughput is going to suffer a higher loss percent.

    As for the tug itself, I’ve done a bit of analysis on it, and for smaller tankers, smaller tugs seem to make more sense. Lot more analysis to do. The basic groundwork for tugs has been demonstrated by vehicles like Orbital Express. Personally though, I’m a fan of using some sort of boom-rendezvous system, as it greatly simplifies a lot of the tug systems, cuts down on the odds of breaking the depot accidentally with a crash, and allows for fast rendezvous/transfer, which might be important for certain BEO payloads (like GEO comsats which aren’t designed for long duration LEO loiter times).


  13. Jonathan Goff Jonathan Goff says:

    Not to mention the lack of a real market for going BEO at such a large scale. It’s more astronauts for the sake of astronauts, taxpayer funded as usual.

    Other than GEO comsats, and deep space probes, you’re right. No non-astronaut “real markets”…


  14. Heinrich Monroe says:

    Another big incentive for putting cryogenic propellant depots at libration points (eg EM L1/2) or, to some extent GEO, is that the radiation loading is essentially all from the Sun. That thermal loading is highly controllable with a well-placed lightweight multilayer shield (a la JWST). Boiloff rates can therefore be dramatically lower than in LEO.

    JWST can passively keep the observatory temperature at 40K, but that’s limited mostly by thermal parasitics. If you do it right, passive temperatures of order 15K are not unreasonable. You won’t do quite as well as that at EM libration points, but the cryogen lifetime could be pretty impressive even there. You add some small cryocoolers, and the cryogen lifetime could be essentially infinite.

  15. Jonathan Goff Jonathan Goff says:

    Not only that but long-duration sunshields are a lot easier to do when you don’t have to cope with atomic Oxygen. You don’t necessarily need cryocoolers in EML-1/2, because you still have stationkeeping requirements, so adding a bunch of complexity to eliminate venting that you would need to use anyway seems unnecessary. Now for deep-space missions, or long transfers from LEO to EML-1/2, sure. Eliminating boiloff for those missions is a great idea.


  16. Heinrich Monroe says:

    I assume you’re referring to atomic oxygen erosion of polymers in LEO. That’s an important effect, but shield technology simply doesn’t work that well in LEO. You have to shield both Earth and Sun, which are in different directions at different times. For really high performance shielding, you need to dump the heat to cold space with a vee architecture the way the JWST shield will do it. This isn’t a matter of just wrapping tanks with MLI. That doesn’t dump heat at all. Wrapped MLI just acts as an insulator. That is, a quasi-planar five-layer vee design works VASTLY better than five layers of wrapped MLI when the radiation is coming from one side only.

    Given that, it’s kind of irrelevant, but erosion is slow enough that if you could use a vee shield in LEO, you’d just dump the eroded shield every few years and put on a new one. It’s lightweight, easily deployed, and it’s not as if you have to do a complicated unwrapping/rewrapping job. That’s what you’d do for a shield at a libration point, where eventually micrometeorite impacts and possibly outgassing contamination will compromise the shield effectiveness.

    But more to the point, a cryo depot at a libration point doesn’t even need a tank that’s wrapped with MLI at all! Just a high performance shield on kept on the sunward side.

    Interesting point about using boiloff for stationkeeping propulsion, but at EM L1/2 we’re only talking tens of meters/sec per YEAR. Seems to me that making stationkeeping dependent on thermal loading and managing pressure vessels to hold gas is just a complication one doesn’t need. Stationkeeping capability for an empty tank? Gotta have it.

    Cryocoolers are no big deal. We have space qualified ones, and you just put solar panels on the sunward side of the shield to give you plenty of power. In LEO, you really wouldn’t want to rely on cryocoolers, as they’d have to be really big.

  17. Heinrich,
    Here’s a few thoughts:

    1-Definitely agree on the difference between a sunshield and just plain MLI. That said, ULA has a concept for an LEO sunshield that’s actually planned for testing sometime around 2012-2013. More details here:

    2-While it’s a challenge to make a sunshield work 100% of the time for both earth and sun shielding in LEO, you can get a shield that works all of the time for earth, and the vast majority of the time for the sun.

    3-You’re right that you could just replace the shield every five years as necessary.

    4-You’re also right that the stationkeeping requirements at EML-1/2 are pretty low, but at say 50m/s/year, with a 390s Isp from boiled off hydrogen gas, that still comes out pretty close to the boiloff rate (single digit pounds per day) you would get with a reasonable deployable sunshield.

    5-I’m not afraid of cryocoolers, just saying that if they’re not necessary for the mission, it’s nice to avoid some complexity.


  18. A_M_Swallow says:

    Cygnus would be a really big tug for a 300-500lb to orbit RLV.

    True for a payload that small. The small tug would need a chemical thruster nearer the size used for satellite station keeping or part of an RCS. I wonder if there are suitable off the shelf bus and thruster designs?

  19. googaw says:

    Jon, I was responding specifically to Trent’s comment about BEO not to your plan, sorry for the confusion. I do applaud the inclusion of a real market in your plan, as it reduces the whole Rude Goldberg nature of it substantially. With the new scaling, assuming it can still achieve the same propellant launch cost targets, it looks superior to its oversized competitor RLV proposals.

    I am quite skeptical that anybody can develop this RLV and the necessary refuelable upper stage, at such low development costs that they can be recouped from 1-3 flights Falcon-1/mini-Raptor flights to GEO per year, if that’s what you were saying. A Falcon 1/mini-Raptor combo that needs, to err on the generous side, 2,000 lb. of propellant for a flight to GEO, 3 times per year, gives us only 20 300-lb. flights per year. Not within Greason’s range and if you are charging $2,000/lb. for propellant, that’s only $12 million per year, which minus insurance and operational and overhead costs probably leaves nothing left over to amortize the development costs.

    If you can overcome ITAR (and I’m confident you can in this case) and partner with ILS or Arianespace, and again assuming (something of which I remain skeptical) that you can greatly reduce the launch cost of propellant with a small RLV, you might be able to get half the flights to GEO on one of those launchers, 5 per year at say 8,000 lb. of propellant per pop is 133 flights of your RLV per year, which is (barely) within Greason’s range. And the revenue, again assuming $2,000/lb., is $80 million per year. If Musk switches his focus to real commerce instead of NASA the Falcon 9 too might provide such a market and then you don’t have to worry so much about ITAR. And if the service proves to be a knock-out win for the customer there are nearly 30 comsats/year going from LEO to GEO and that number will probably double by 2025, so there’s plenty of room to grow.

    You have to have extraordinarily low development costs to pull this off. Since investors will be skeptical of this, I suggest early tests that demonstrate some capabilities (e.g. full-throttle tests of the actual engines, drop tests of the actual airframe if it’s a flyback, and the like), with specific development cost targets to each test-success milestone, and give the investors the option to pull the plug and get the rest of their money back at any such milestone where the estimated costs are overrun. These “options to cancel” can substantially reduce the investors’ downside risks, encouraging them to invest in an uncertain project in the first place.

    Finally, can you actually design something much smaller than existing RLV designs and retain the cargo/total mass ratio, or at least retain enough of it to get 300 lb. to orbit? Or does the cargo get eaten by components that can’t be proportionately scaled down?

    BTW, deep space probes, being completely taxpayer-funded, are not real markets. The are closer to it is some ways (e.g. they are not robots for the sake of robots) but are still political demand, not natural demand.

  20. Bob Steinke says:

    Here’s an idea for a small tug tech demo. An octo-cubesat (8kg, 20cm cube) with a small thruster and a 6-DOF arm with propellant crossfeed plumbing. It grabs a propellant tank, makes the pumbing connections, then fires its thruster with propellant from the tank it is towing. The 6-DOF arm acts as a super-gimbal that can do 3-axis attitude control with one engine.

  21. Googaw,
    When I had been suggesting 1-3 GEO flights per year, I had been thinking about the numbers I ran on refueling a Centaur (ie to do very large commsats to GEO). A single Centaur tanking would provide 150 flights, which would be about what you needed for one supplier to eke by. 3 per year would be enough for a small industry, and anything beyond that would just go towards making things a lot better.

    A mini-Raptor would need more flights to close, but your numbers are off a bit. The Falcon 1 US as-is currently has around 5-6klb of propellant and the 1e is probably closer to 8-9klb from calculations I’ve seen, not 2000lb. You’d still need a fair number of flights at that demand level to make the market case close–I’d say bare minimum 4-6 per year to provide enough market to keep two providers alive by itself. But remember, with a full tank-load in LEO of a F1e US sized LOX/LH2 stage, you could be providing as much GEO capacity as a Soyuz launch from Kourou, for a competitive price.

    As for whether government demand is demand or not…their money is just as green. So long as they’re actually buying launch services and not trying to run the whole show (ie more like how NASA buys science flights than how they do manned flights), why is that not real? Sure, it may not be the height of libertarian purity, but if it is providing a service people actually want, and they’re willing to pay for it, and it helps close the case, why not?


  22. googaw says:

    Jon, thanks for the clarification on your demand scenario. I quite like this real market approach and my main skepticism at this point is your development costs. Assuming you can scale down the size of an RLV so greatly and still get cargo to orbit, still, how do you scale down the development costs?

    As for why I stress real markets and disparage projects which are primarily trolling for NASA contracts and calling it “commerce”, of course if you’re in a company you get almost any profitable government contract you can get. But it’s only from that point-of-view that the money is “just as green.” My point of view is very different, it is of a taxpayer deciding whether such taxpayer expenditures are worth it or whether tax cuts or other kinds of expenditures would be more worthwhile. I’m hardly a libertarian purist — among other things I’m a big fan of a strong military — I just object to the ludicrously distorted implication that political demand is equivalent to private demand. They are radically different things. Political demand has no market accountability: politicians and activists can pursue all sorts of economic fantasies, nearly arbitrarily divorced from economic reality, on the taxpayer’s dime. Real markets generally don’t do that.

    Political needs are often extremely far off the economically rational path — especially in something like HSF where’s it astronauts for the sake of astronauts instead of meeting core government needs such as defense in the most effective and efficient ways. In these situations government does not lead commerce, it leads astray, often as in the case of HSF very far astray.

    Thus, I am highly skeptical of contract trollers who cry “commerce, commerce” but expect me and other taxpayers to absorb the development costs. This kind of scam has fleeced me and my fellow taxpayers far too many times already. Raise private investment money to fund the development costs, based on a business plan dominated by revenue from private customers, and that indeed actually signs up real private customers, and this will be strong evidence that it’s not just another taxpayer-fleecing scam.

  23. googaw says:

    BTW, all except the first paragraph of the above post are generic comments, they aren’t targeted at Jon’s or anybody else’s proposal specifically, and indeed probably apply quite a bit less to Jon’s most recent proposal than to most of the projects that space activists discuss.

  24. ht says:

    It was kind of shocking and even refreshing to see Robert Zubrin’s piece this spring ( ) just to have a different take on things. Surprising that he almost seemed to be coming out in favor of moon (first) mars (next) and beyond.

    He also seemed to not want to work on propellant depots right now, but to just work on a solid mission for NASA. He supported going somewhere, and mentioned Rutan’s comments about getting something done faster.

    Of course, I’ve been promoting my friends’ website in development, , along those lines. And this is getting to the political, NASA proposals side of your subject here, but what do you make of Zubrin’s comments?

    Frankly, my favorite part of this is the humor slide, very effective graphic!

  25. Googaw,
    Yeah, keeping development costs low is critical. I have ideas along these lines, but since I’m actually trying to make them work, I don’t want to tip my hand too much.

    But I definitely agree, if you’re shooting for a price point of $500/lb to orbit, off of no more than 150 flights per year (over say 5 years), that implies you really want to keep the development cost you need to amortize down in the $30-50M range. That’s…aggressive for developing an RLV. But if you can find a way to develop the RLV where some of the pieces are useful to other groups (either for paid technology demonstrations, risk maturation, suborbital missions, funded research, hardware flight demonstration, etc), you might be able to get them to pay for some of the development. If the total development is more in the $60-100M range, but half of it has already been paid for by incremental markets along the way, the part you have to amortize off of actual flight-ops is a lot smaller.


  26. googaw says:

    Jon, Virgin Galactic’s WK2/SS2 and SpaceX’s Falcon 9 both seem to be coming in at over $300 million, so yea I’d say that’s aggressive. Of course you aren’t flying people on the one hand or launching nearly as much payload on the other, but you do however have to develop more pieces, especially an on orbit fuelable upper stage (which, for simplicity, might perhaps in the first version act as its own depot). I certainly applaud your current efforts to lower development costs.

    Another possible revenue source may be the very-small-sat-to-LEO market. At this stage that is a highly uncertain matter but is worth watching. Another possible revenue source, not a non-political market but an application that seems to have quite a bit of practical use, is the responsive sat need at the DoD. Microcosm is studying a 100 kg class launcher that would be ready with 24 hours’ notice: the “the Scorpius Mini-Sprite, a three-stage, liquid oxygen-kerosene fueled rocket”:

    I suspect that an RLV that actually flies frequently is going to be less expensive for this purpose and in practice more reliably responsive than an ELV custom-designed from scratch for the purpose that normally just sits on the pad with a ground crew that normally sits around twiddling their thumbs. It probably would greatly increase the development costs and groundside operational costs to do it under contract to the DoD to their specifications, but a follow-on version of your RLV for this purpose might be counted as a possible follow-on market, analogous to an Air Force tanker version of a Boeing airliner.

    Also, if you extend the range a bit, you might be able to do some of the things they want to do with the X-37. Burn the propellant to go inspect a polar orbit satellite instead of delivering it to a depot. But again a contract to develop specifically for this purpose would probably make the whole thing far too expensive, so look on it more as another possible follow-on side-application like responsive launch.

    A nice thing about designing in response to real market needs is that quite often being useful in one area makes it much more likely to be useful in other areas.

  27. Jon, so help me out here man.. don’t make me do the math. How many launches do you need? Say you’re launching once a day on your RLV, 200kg delivered to a depot per flight, and you want an EDS of 70 tons.. how many days do you need? The answer isn’t 70,000/200=350 days .. it’s clearly more complicated than that. The first delivery contributes to the final mass a lot less than the last delivery. There’s propellant used for station keeping and the tugs, yes, and maybe the first of those is about the same as the boil-off, and that’s really interesting, but without actual numbers its just handwavium.

  28. Wes Johnson says:

    The following from SpaceOps 2010 last week: Boeing jumped on trying to promote HLLV for launching depots (AIAA-2010-2370). As a side note, I tracked the author down and tried to discuss the fact that depots preclude the use of HLLV, he seemed completely ignorant. Another side note on this one, presentation should have been called “Boeing’s confirmation of NASA’s Ares V and Shuttle Side Mount Heavy Lift Calculations”. Additionally, an interesting paper on what the science community would like as far as large space telescopes (AIAA-2010-2290), basically they want a 10m wide by 25m tall fairing and 20-40 mT to E-S L2, I think depot could provide this if there was a way to get off the ground. There was talk of a LOX-methane L1 depot (AIAA-2010-2158), this is intriguing because the LOX and methane are basically storables (i.e. won’t boil-off due to low environmental temperatures and proper shielding). Additionally, Europe is planning an On-orbit servicing mission or two as technology demonstrators (AIAA-2010-2159).

  29. Trent,
    You’re right–it does depend entirely on a ton of details. Boiloff losses are probably on the 10% level, entirely hydrogen (you’re using the hydrogen to keep the O2 from boiling off). How much the tugs use depends a lot on a ton of details as well. But a single lunar mission will easily keep a whole industry doing well, even without other markets like GEO and deep space probes.


  30. googaw says:

    an interesting paper on what the science community would like as far as large space telescopes

    You are confusing a handful of scientists at a NASA center with the scientific community.

  31. googaw says:

    Trent, there is thankfully no longer a serious government plan to send spacecraft oversized by several orders of magnitude to the moon. Interestingly, the lunar ISRU architecture I’ve been sketching over the last year has a similar scale to Jon’s proposal, a few hundred pounds to LEO per delivery from the moon, using a small reusable lunar shuttle. Serving real markets economically leads to real infrastructure that private enterprise will actually be willing to pay money for.

    Indeed there is no near-term government plan to send oversized spacecraft anywhere beyond LEO, only hypothetical talk about asteroids and Mars two or more Presidential administrations in the future. Again, good thing, as these kinds of missions while entertaining are a big waste of taxpayer’s money.

    Meanwhile real space development continues. Over twenty real commerce spacecraft go to GEO every year, serving a market with tremendous long-term growth. The GEO run has the advantage that it can be made with a reusable tug/tractor. Not having to launch new upper stages each time may be the biggest win for refueling prior to ISRU. As Jon and I have been discussing the real commerce to GEO provides more than a sufficient market assuming the development costs can be made sufficiently low. If they can’t, then it doesn’t actually lower costs and it doesn’t make sense for NASA missions either.

  32. Pete says:

    I greatly appreciate your pointing out that a RLV should be sized based on a presumed market divided by a minimum of say 300-450 flights per year (2-3 carriers each with a fleet of ~3 vehicles and each vehicle flying ~50 times a year). This is a new space paradigm that I think worth emphasizing – and exploring with regard to possible markets and RLV designs, including the extension of this paradigm beyond LEO.

  33. Dallas Bienhoff (Boeing’s Manager for In-Space and Surface Systems with the Advanced Space exploration organization) was on The Space Show on May 3rd talking about propellant depots.

    His figure for boil-off of a H2/LOX depot was 1/10th of 1% per day of the hydrogen only (it is used to cool the LOX as Jon has said a few times). With those sorts of numbers, discussion of boil-off for a payload of fuel that sits in the depot for a year isn’t even that interesting. The “half life” is something like 692 days.

  34. With 200kg delivered per flight, 800 flights/year, and that 0.1% boiloff per day, you need about 188 days to amass 75 tons of propellant. And no, I didn’t model H2 and LOX separately in that calculation 🙂

    So the big question mark remains the tugs.

  35. A_M_Swallow says:

    Tug for depot.
    (i) The Cygnus space craft with robotic arm could be used for big objects like Mars Transfer Vehicles.

    (ii) For smaller objects like propellant tanks weighing less than 1 ton could a modified version of the Orbital Express satellite ASTRO be used? The main engine being replaced by a thruster that uses fuel from the depot plus a bigger fuel tank. ASTRO has an arm, docking facilities and possible long range sensors.

  36. AM,
    There are a ton of possibilities for tugs, including adapting any of the spacecraft that are visiting ISS (Progress, Soyuz, ATV, HTV, Dragon, and Cygnus), copying/iterating on previous tug demo vehicles (like Orbital Express ASTRO), or even doing a cleansheet approach. As it is, there are probably over a half dozen companies in the US alone that have the technical chops to do a tug, the problem is the market risk is high. And most of those companies prefer to limit their risk by sticking with government or commercial contracts instead of sticking their neck out entrepreneurially.


  37. A_M_Swallow says:

    That turns into how quickly and how cheaply the tug is produced.

  38. Yep, numbers for the tugs is the question mark.. I dunno how often I need to say this: real engineers use numbers. 🙂

Leave a Reply

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