Good points... I've been following a number of discussions over the past several years on in-space fuel depots on nasaspaceflight.com/forums, by any number of knowledgeable people. It's a fascinating topic in and of itself. It's also one of considerable debate.

Basically, the debate centers around two things- boiloff and propellant transfer. To this point in time, all in-space fuel transfers have been hypergolic propellants-- that is to say, room-temperature storable propellants. Most of this has been done by the Russians (their later Salyut stations and Mir could take on fuel from the Progress freighter or even some Soyuz craft IIRC and tank it for use on it's station reboost engines, similar to how it's done on the ISS). Cryogenic fuel transfer, that is to say, transfer of LOX or LH2, has never been done in space. Now, the debate comes from different camps. Folks having worked on Centaur and other cryogenic stages in space, who have dealt with propellant handling and boiloff concerns on these stages, say that it's a problem that can be dealt with and overcome, no problem. Others (mainly within NASA, who's never done it, hasn't dealt with an in-space cryogenic stage since S-IVB (shuttle Centaur not withstanding) and who have a vested interest in seeing the BFRE (biggest F-ing rocket EVER!) say it can't be done or it'll take a decade and a billion or two... I tend to think the "yes we can" folks are closer to the truth... There are a number of "test kits" that are capable of being installed on Centaur to use it's residual propellants for transfer tests and storage tests. Centaur has already proven that propellant settling in the tanks for transfer can be accomplished with as little as 0.001 G's of acceleration, using the vented propellant gases from boiloff for settling by venting it through a thruster nozzle(s). Boiloff can be managed quite effectively through sunshields and sun-synchronous orbits that keep the tanks pointed away from the sun and earth (which radiates considerable heat) for short-medium durations, and for long term storage by active refrigeration of the cryogenic fuels. I've personally argued for the use of a single-propellant depot, and I'd make LOX the depot-stored propellant, for two reasons-- 1) LOX is a cryogen, but not a "deep" cryogen like LH2-- LH2 is NOTORIOUSLY leak prone and hard to handle, and at just a few degrees above absolute zero, VERY hard to cool and keep cool to prevent boiloff and excess pressure buildup and quantity losses. Oxygen has a higher liquid temperature, is a larger molecule that's easier to contain without leaks, and is easier to cool to prevent boiloff. 2) LOX accounts for a large majority of the weight of the propellants on a space mission, due to it's considerably larger density than LH2. If an EDS stage launched from Earth with just enough LOX to insert into orbit, but with enough LH2 to complete the entire mission already tanked up, it would only be about 20-25% heavier (give or take) than an EDS that had only enough propellant to achieve orbit with empty tanks, and then get ALL it's fuel (LH2 and LO2) from an orbital depot. This only requires the hydrogen tank be sized for the mission needs, and the LOX tank can be smaller, holding just enough to achieve orbit, and then take on the additional heavy LOX on the depot. While this adds to stage mass, because of the larger LH2 tank sized to hold all mission propellant from liftoff to completion, therefore reducing the mass fractions, it also greatly simplifies depot operations. It only requires ONE propellant handling/conditioning system NOT TWO, ONE set of connections, NOT TWO, it means your tanker craft can launch with a single storage tank, or have a single enlarged stage propellant tank to carry the propellant cargo to the depot, NOT TWO, and be equipped with only one connection to tranfer the propellant to the depot, NOT TWO. It means you're handling a 'medium cryogen' instead of a medium cryogen and a "deep cryogen", and yet you're still transferring most of the weight of propellants off the mission stage, making it capable of hauling more payload (a LOT more!) for a given stage size, since it would only be tanked about half-full of LOX at liftoff; the weight of the additional LOX could be directly converted to extra payload. Using the S-IVB as an example-- it carried 192,000 lbs or so of LOX. Say it could have docked with a LOX depot and transferred all the LOX needed for TLI from the depot. About half the propellants in the S-IVB were used for ascent, the other half for TLI. If the stage was 'short fuelled' on the pad, say with only 100,000 lbs of LOX, that means that the other 90,000 lbs could have been EXTRA PAYLOAD (less the propellant transfer equipment for tanking up with LOX at the depot). That's NEARLY DOUBLE the payload of the S-IVB! If you optimized the stage for this arrangement, by making the LOX tank smaller, the weight saved could ALSO be added to the payload! The hydrogen tank would remain the same size, only the LOX tank would be shrunk. The full hydrogen tank only carried about 40,000 lbs, so why complicate the works by building a two-propellant depot to tank up an S-IVB with 100,000 lbs of LOX and 20,000 lbs of LH2?? The LH2 really complicates the setup, and yet handling it really adds very little to payload capability. Now, the counterpoint to that is, IF you have a bi-propellant depot, you could COMPLETELY refill the S-IVB with BOTH propellants, and launch double the weight to TLI or more... but it hardly makes sense to do it that way. You'd have to fly the equivalent of a second vehicle to deliver that amount of propellant to the depot anyway, so why not simply launch the cargo in two loads instead of one (unless it was something indivisible, like a huge space station segment, a complete L2 space station, lunar base, nuclear power reactor, etc.) So long as you're getting your propellant from Earth's surface, this approach makes the most sense. If you start talking about propellants from the lunar surface, then obviously you'll need a bi-propellant depot, or you'll have to size your ascent/descent vehicle tanks to carry all their LH2 from the lunar surface to orbit/deep space (L2) and enough to land empty of LH2 again-- at that point it's probably better to go on and develop a full bi-propellant depot and size your tanks appropriately, and benefit from the added performance of optized mass fractions. But for Earth-supplied propellants, the LOX-only depot is an easier 'entry level' system to get experience and perfect the technology and gives almost as much benefit as the bipropellant depot. Another idea I had would be to launch the propellants as ordinary water, then pump it over to the depot as liquid water, and have the depot equipped with a solar-powered electrolysis plant to crack the water into GH2 and GO2, then convert them into LH2 and LO2 for supplying later arriving stages. This has the benefit of requiring only very small (relatively speaking) cargo tanker tanks sized for liquid water (which is quite dense compared to LH2 and LO2), easy to transfer as a non-pressurized, room-temp storable liquid, and of course is only one fluid containing BOTH propellants. BUT while greatly simplifying tanker operations, it WOULD add a THIRD liquid handling system to the depot, and complicate depot operations with the solar-powered electrolysis plant.

More to come... OL JR :)

Cont'd from above...

Now, there are other propellants that could be used in a depot setup besides hypergolics (which we KNOW will work-- just use bladder tanks like is currently done on Progress/ISS and automatic connect/disconnects) and LH2/LO2 (which needs a lot of work to prove and establish capability). One could use kerosene/LOX in a depot. Kerosene would be about as easy to handle in space as water, though slightly less dense, it is room-temp storable and would present no handling difficulties not experienced by any other liquid. The main requirement for using kerosene in space is a highly efficient RP-1 engine to burn it in... approaching the theoretical limits on kerosene ISP. LOX, being cryogenic, cannot be used with bladder tanks, but the technology to handle cryogenic propellant handling and transfer in space could be perfected using LOX and then applied later to handling LH2. Then there's the mild cryogens, fuels like butane, propane, ethane, methane, etc... LP gas (butane and propane) have been recommended as possible fuels for a depot. Being mild cryogens, their pressure requirements to remain liquified and prevent boiloff are quite manageable, so they'd handle more like non-cryogens that cryogenic fuels, only needing basic thermal shielding to prevent boiloff, and their ISP is comparable or better than kerosene/LOX. Methane has been proven as a fuel, though not applied in a flight vehicle yet (as it was to be on the CEV in the early days of Cx) and experiments have been done with butane, propane, and ethane that have proven the concepts that they could be used with appropriate engine mods.

SO, there are any number of ways to do a depot based architecture. Problem is, tankers. If you have to launch another rocket to deliver the fuel, it'd better be a DARN CHEAP ONE! Otherwise there may be no advantage over simply launching the mission with multiple modules that EOR.

Of course when you're talking about transfer to lunar orbit, L1 or L2, etc... if you want MAXIMUM payload capacity and time isn't an issue (unmanned lander/cargo transfer) then it's hard to beat SEP, solar electric propulsion. The ISP is VERY high, but the electricity requirements mean you have to have HUGE solar arrays and the thrust is quite low-- but highly efficient transfers with VERY little propellant on the order of 180-270 days are possible, which is fine for unmanned vehicles if they're engineered for such a long dormant period. The main problem is (like your wheel space station in 1,000 mile orbit) is the transit through the Van Allen Radiation Belts, which tends to play havoc with electrical equipment, long term. The radiation degrades solar cells and microprocessors and such, so they would have to be hardened against it, especially for reusability (which would be mandated due to the up-front costs of such a system). This WOULD provide a 'constant' capability of unmanned cargo transfer between L1/L2 or HLO and LEO. It would be useless for manned missions (you wouldn't want to spend weeks inside the Van Allen Belts-- better to camp out at Fukujima) but it would alleviate the need of having to launch all the mission hardware and cargo via 72 hour TLI, or other "high energy" trajectories (there are lots of interesting trajectory discussions on nasaspaceflight.com/forums as well). If VASIMIR works out, that's another possibility, and if NASA ever gets SERIOUS about deep-space exploration, sooner or later a nuclear engine for in-space applications is going to be needed. When you have that, depots will be a minor part of the equation. They'll only need to handle one propellant for tanking up nuclear or SEP or VASIMIR propelled 'space tugs'. We should also be looking at ballutes or other non-propulsive means of transfering from TEI trajectories back to LEO... aerobraking to eliminate the need for propulsive braking is an excellent idea. Chemically braking back into LEO is prohibitively expensive (requires the same amount of fuel to brake back into LEO as it did to achieve TLI with the same mass).

What it boils down to (ha ha pun intended) is that if you can handle LOX in a depot situation, you can set up any number of highly efficient operations without LH2, but ultimately LH2 will have the best performance possible-- but may not be necessary if less is 'good enough'...

Interesting discussion! OL JR :)

Excellent points, all. I hadn't thought of the partially-filled oxidizer tank option. The original Apollo EOR mission plan called for launching the S-IVB TLI stage filled with LH2, but with *no* LOX in its oxidizer tank (the S-IC and S-II stages would have injected this S-IVB into orbit, as they later did with Skylab's converted S-IVB "dry" orbital workshop). Reducing the TLI stage's LOX tank size and mass (when empty as well as full) would be of significant help in increasing the payload mass.

In-orbit cryogenic refueling, like ion propulsion (for *primary* spacecraft propulsion rather than just stationkeeping and attitude control) and rotating spacecraft and space stations to generate artificial gravity, are among those enabling technologies for true spacefaring operations that NASA either never deigns to test or else takes its own sweet time to test (four decades in the case of ion propulsion!). A simple LOX *and* LH2 in-orbit refueling test could be conducted using two separately-launched Centaur stages:

The tanker Centaur stage would have the propellant transfer outlets installed at the aft ends of its LOX and LH2 tanks. The TLI Centaur's propellant-receiving inlets would be located at the forward ends of its LOX and LH2 tanks. After the tanker Centaur docked nose-to-nose with the TLI Centaur (using Russian automated docking hardware and ranging antenna systems to minimize costs), the docked stages would be rotated slowly in order to pack the tanker Centaur's liquid propellants down onto the outlets, with pressurized (boiled-off) gases above the liquids.

Applying dark paint or panels to one side of the tanker Centaur (and then orienting that side toward the Sun as the docked stages rotated) would ensure that the ullage gas temperatures (and thus their pressures) inside its LOX and LH2 tanks would be higher than those inside the TLI Centaur's tanks, which should allow propellant transfer between the two stages without any need for transfer pumps. As in an aerosol spray can, the pressurized gaseous oxygen and gaseous hydrogen above their respective liquid forms in their tanks would force the LOX and LH2 out of the tanker Centaur's propellant tanks and into the TLI Centaur's tanks. After the TLI Centaur was refueled, the two stages would undock and the TLI Centaur could be restarted to hit a lunar or solar orbit trajectory, perhaps carrying one or more scientific "hitch-hiker" payloads to take advantage of the mission opportunity.

Sounds like a plan! There is a more gradual "stepped" plan that's supposedly 'in place' that NASA is supposedly funding, to create several "goalposts" that would be crossed in order to prove the technology. It would start with settling experiments (which has already been done on Centaur) and then some limited tranfer experiments (which Centaur plans to do with an add-on "kit" hitch-hiking on a GEO satellite mission, and being put into action after the satellite is dropped off in geosynch orbit and the Centaur does it's 'disposal burn' to get it 'out of the neighborhood'... the plan would be to settle and transfer some of the residual propellants to small tanks in the "kit". The next step in the NASA plan would be automatic couple/uncouple devices and cryogen pumping, long term storage issues, insulation, and structures. That would be followed by a subscale orbital demonstration similar to what you described (but not actually going anywhere). Once again, the long path around the bend instead of the straight line, but it's a start... IF it ever actually gets funded...

Later! OL JR :)

*SIGH* NASA seems to have gone back to their "timid" stance (such as originally planning 6 - 10 manned Mercury-Redstone suborbital shots before trying an orbital mission, and planning the first EVA [for Gemini 5, not Gemini 4] to consist of one astronaut merely opening his hatch and standing up!) that required milestone-setting Soviet space spectaculars to goad them into being more daring.

Heh, if Elon Musk wanted to prove in-orbit LOX and LH2 refueling quickly (and felt a need to do so), he could buy a couple of Long March CZ-3 launches fairly cheaply in order to conduct a proof-of-concept refueling mission. Maybe something like that would light a fire under NASA to be more aggressive in their R & D work.