Starship On-orbit refueling - Options and Discussion

[wannamoonbase]

Any thoughts on what these details are up near the nose?  Looks to be above the top dome. Remember how big this thing is, could those be an airlock and cargo bay door?  Notice the indicator lights.

The only thing that seems to be missing are docking ports for a couple dragons. Maybe they’re on the other side.
That entire nose could be a hab. No need for a header tank there.

Stowage bays for solar panels and radiators?

[DanClemmensen]

Any thoughts on what these details are up near the nose?  Looks to be above the top dome. Remember how big this thing is, could those be an airlock and cargo bay door?  Notice the indicator lights.

The only thing that seems to be missing are docking ports for a couple dragons. Maybe they’re on the other side.
That entire nose could be a hab. No need for a header tank there.

Depot is optimized for tankage. It should be tankage all the way up to the tip of the nose, so that's not necessarily "above the top dome". However, Depot also needs to carry all that funny docking and propellant transfer hardware somewhere, together with whatever active system that deploys it (Canadarm?, robots?, elaborate linkage system?) so maybe that's what is launched behind that door.

If I were designing Depot (while sitting in my armchair out here somewhere on the Internet after 3 minutes of thought) I would put the messy bits behind a disposable fairing at the nose. Depot will never re-enter, so no need for an aerodynamic shape.

[spider_best9]

Depot is optimized for tankage. It should be tankage all the way up to the tip of the nose, so that's not necessarily "above the top dome". However, Depot also needs to carry all that funny docking and propellant transfer hardware somewhere, together with whatever active system that deploys it (Canadarm?, robots?, elaborate linkage system?) so maybe that's what is launched behind that door.

If I were designing Depot (while sitting in my armchair out here somewhere on the Internet after 3 minutes of thought) I would put the messy bits behind a disposable fairing at the nose. Depot will never re-enter, so no need for an aerodynamic shape.

A depot needs power for months long operation. Those could be for solar panels deployment.

[Norm38]

Do pressurized tanks like tapering at the end?  I’m not sure they’d want to move away from a simple cylinder and dome. The depot tanks can go higher, maybe take the lower Artemis deck, up to the taper. Plenty of volume in the nose for depot operations. Maybe only 2-3 crew at a time, and not permanent.

Maybe one docking port in the nose per Orion.

Keeping the nose of a depot the same as Artemis has advantages. They can test things out, test life support before  Artemis III.

And while future depots may be able to self configure, at the beginning I’m betting some hands on help will be needed and welcome.

[TheRadicalModerate]

Any thoughts on what these details are up near the nose?  Looks to be above the top dome. Remember how big this thing is, could those be an airlock and cargo bay door?  Notice the indicator lights.

The only thing that seems to be missing are docking ports for a couple dragons. Maybe they’re on the other side.
That entire nose could be a hab. No need for a header tank there.

Note that they're common to both the depot and the LSS, which are presumably the same structures with different... accessories.  It's possible that all Starships have them, and they're the access patches for the Pez dispenser in Starlink versions.  Or they could just be servicing hatches for ground support crew.

[TheRadicalModerate]

Do pressurized tanks like tapering at the end?  I’m not sure they’d want to move away from a simple cylinder and dome. The depot tanks can go higher, maybe take the lower Artemis deck, up to the taper. Plenty of volume in the nose for depot operations. Maybe only 2-3 crew at a time, and not permanent.

v2 has three extra ring segments of tankage, but it's only one ring segment taller, so 3.6m have come out of the cylindrical portion of the payload bay, making it 4.2m tall.  If the domes are rejiggered to consume that extra space, that's about 240t of extra prop, which would be 1740t for a depot or tanker based on v2.

Remember, however, that the depot probably won't maintain subcooled prop.  So that reduces its refueling capacity to about 1640t of boiling prop.

I would expect tankers to roll to v3 first, followed by a new depot.  If a v3 depot consumed all of the cylindrical payload bay (which I believe is back to 8m high), then that's about 460t more of subcooled prop than a regular v3, for a total of 2760t.  In terms of boiling prop, that's about 2580t.

No clue if 240t for a v2 tanker could be launched, or if it provided much of a net gain in prop delivered to orbit.  Same issue for a v3 tanker with 460t of extra prop.  They obviously both have terrible T/W at launch, but maybe the extra prop can make up for it after staging.

[Twark_Main]

Do pressurized tanks like tapering at the end?  I’m not sure they’d want to move away from a simple cylinder and dome. The depot tanks can go higher, maybe take the lower Artemis deck, up to the taper. Plenty of volume in the nose for depot operations. Maybe only 2-3 crew at a time, and not permanent.

Maybe one docking port in the nose per Orion.

Keeping the nose of a depot the same as Artemis has advantages. They can test things out, test life support before  Artemis III.

And while future depots may be able to self configure, at the beginning I’m betting some hands on help will be needed and welcome.

Ironic to fret over inefficiencies in a non-optimal dome shape....  and then immediately propose an entire (unnecessary as far as I can tell) crew habitat, using that same non-optimal dome as its pressure vessel. 

"I have an annoying splinter in my eye. Could I swap it for this plank instead?" 

[Norm38]

Do pressurized tanks like tapering at the end?…

Ironic to fret over inefficiencies in a non-optimal dome shape....  and then immediately propose an entire (unnecessary as far as I can tell) crew habitat, using that same non-optimal dome as its pressure vessel. 

"I have an annoying splinter in my eye. Could I swap it for this plank instead?" 

Wasn’t fretting. A fuel tank is 6-8 bar, the nose is 0-1. Not the same thing.
Artemis III will need to use that volume and dock Orion at the nose. A depot may well need crew support occasionally. The two objectives align.

[Eka]

Do pressurized tanks like tapering at the end?…

Ironic to fret over inefficiencies in a non-optimal dome shape....  and then immediately propose an entire (unnecessary as far as I can tell) crew habitat, using that same non-optimal dome as its pressure vessel. 

"I have an annoying splinter in my eye. Could I swap it for this plank instead?" 

Wasn’t fretting. A fuel tank is 6-8 bar, the nose is 0-1. Not the same thing.
Artemis III will need to use that volume and dock Orion at the nose. A depot may well need crew support occasionally. The two objectives align.

A depot and the propellants in it are cheap compared to the cost of a manned flight. Build it with many redundancies. If all the ones for transferring propellants fail, put up a new one, and reenter the dead one over Point Nemo. A depot that has lost its maneuvering abilities can have a Starship visit it, and offload the propellants. Then use them to put the depot into an orbit that will reenter over Point Nemo. Then the Starship can change its own orbit to reenter on its own back at a starbase.

[TheRadicalModerate]

A depot and the propellants in it are cheap compared to the cost of a manned flight. Build it with many redundancies. If all the ones for transferring propellants fail, put up a new one, and reenter the dead one over Point Nemo. A depot that has lost its maneuvering abilities can have a Starship visit it, and offload the propellants. Then use them to put the depot into an orbit that will reenter over Point Nemo. Then the Starship can change its own orbit to reenter on its own back at a starbase.

A depot has an interesting risk profile, in that the later tanker/depot interactions have a higher cost of failure than the earlier ones.  The first tanker that fills a depot risks itself, the depot, and one launch.  The last tanker risks the depot, the tanker, and all of the launches that added prop before it.  Since failures are also more likely with use, it probably pays to spend more time making the system robust than you otherwise might.

That may be a pretty good argument for putting the docking hardware on the tanker and LSS side, instead of the depot.  That way, you've got an inspected set of hardware for each launch.

[Twark_Main]

A depot and the propellants in it are cheap compared to the cost of a manned flight. Build it with many redundancies. If all the ones for transferring propellants fail, put up a new one, and reenter the dead one over Point Nemo. A depot that has lost its maneuvering abilities can have a Starship visit it, and offload the propellants. Then use them to put the depot into an orbit that will reenter over Point Nemo. Then the Starship can change its own orbit to reenter on its own back at a starbase.

A depot has an interesting risk profile, in that the later tanker/depot interactions have a higher cost of failure than the earlier ones.  The first tanker that fills a depot risks itself, the depot, and one launch.  The last tanker risks the depot, the tanker, and all of the launches that added prop before it.  Since failures are also more likely with use, it probably pays to spend more time making the system robust than you otherwise might.

That may be a pretty good argument for putting the docking hardware on the tanker and LSS side, instead of the depot.  That way, you've got an inspected set of hardware for each launch.

Welcome to my thinking four days ago. 

[OTV Booster]

A depot and the propellants in it are cheap compared to the cost of a manned flight. Build it with many redundancies. If all the ones for transferring propellants fail, put up a new one, and reenter the dead one over Point Nemo. A depot that has lost its maneuvering abilities can have a Starship visit it, and offload the propellants. Then use them to put the depot into an orbit that will reenter over Point Nemo. Then the Starship can change its own orbit to reenter on its own back at a starbase.

A depot has an interesting risk profile, in that the later tanker/depot interactions have a higher cost of failure than the earlier ones.  The first tanker that fills a depot risks itself, the depot, and one launch.  The last tanker risks the depot, the tanker, and all of the launches that added prop before it.  Since failures are also more likely with use, it probably pays to spend more time making the system robust than you otherwise might.

That may be a pretty good argument for putting the docking hardware on the tanker and LSS side, instead of the depot.  That way, you've got an inspected set of hardware for each launch.

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

[TheRadicalModerate]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

[OTV Booster]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

[Eka]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

More tankers, and simultaneous launches relieves the need for purpose built depot Starships. On the other hand it puts a much much harder load on GSE to keep up with the launch rate. A two tower site with proper GSE could launch 6 Starships/tankers in a day into the same refueling window. Treat it like a tree for the pairing off for transfers. When a pair gets into orbit, they immediately start transferring to one of them. This is because they have the least to maneuver before starting fuel transfer. The first two pairs launched maneuver during fuel transfer to meet when both their transfers are done. They then transfer both double loads into a quad tanker load. yada yada yada... The empties make maneuvers to arrange for landing when slots are available.

Landing Note: If grid landing pads are available, two tankers can land just after two take off. They literally need to be going in for landing at the time the launching two are in launch preps. The boosters will be caught by the towers soon after the tankers land. Ground support now has around 6 hours to detank, and move off the spent tankers, and restack the boosters and stack the next tankers. It needs to be a well coordinated ballet, but it should be possible. With more automated Starship handlers, one could likely manage up to 4 simultaneous refueling windows in space at one time. That would be for fueling 4 Starships for higher orbits or Moon or Mars journeys. That would require launches every two hours. Would be very hard to have any humans in the launch and landing pad areas at any point.

I wonder what the minimum survival distance from a launch for a Teslabot would be? Teleoperated Teslabots could do lots of the launchpad work. They could hide in soundproofed vaults during launches and landings, and quickly come out while things are still unsafe for humans.

Note on windows and fuel transfers: I wonder how much out of phase landing could happen due to needing acceleration to make transfers happen? Constant acceleration or deceleration is needed while fuel transfers are under way. This would allow one to push forward or back the timing of tanker landings. It makes the launch ballet harder to calculate, but that's what computers are for.

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

[DanClemmensen]

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

The pad passes under the orbital plane twice a day, once on the ascending side and once on the descending side. Unless you can launch to both you get one launch opportunity per day. Unless you can land on both, you get one landing opportunity per day. If the orbit is polar, the two opportunities are 12 hours apart. As the orbit's inclination decreases, the two become closer until the inclination reaches the latitude of the pad, when there is only one opportunity per day.

Tankers are cheap. If your launch campaign takes N tanker flights, the easiest approach is to just build N tankers and launch once a day, and then use another N days to land them. 14 tankers: launch for two weeks and return for two weeks, or use two pads and launch for a week and land for a week. Four pads (two in Texas, two in Florida)  will take four days.

[OTV Booster]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

More tankers, and simultaneous launches relieves the need for purpose built depot Starships. On the other hand it puts a much much harder load on GSE to keep up with the launch rate. A two tower site with proper GSE could launch 6 Starships/tankers in a day into the same refueling window. Treat it like a tree for the pairing off for transfers. When a pair gets into orbit, they immediately start transferring to one of them. This is because they have the least to maneuver before starting fuel transfer. The first two pairs launched maneuver during fuel transfer to meet when both their transfers are done. They then transfer both double loads into a quad tanker load. yada yada yada... The empties make maneuvers to arrange for landing when slots are available.

Landing Note: If grid landing pads are available, two tankers can land just after two take off. They literally need to be going in for landing at the time the launching two are in launch preps. The boosters will be caught by the towers soon after the tankers land. Ground support now has around 6 hours to detank, and move off the spent tankers, and restack the boosters and stack the next tankers. It needs to be a well coordinated ballet, but it should be possible. With more automated Starship handlers, one could likely manage up to 4 simultaneous refueling windows in space at one time. That would be for fueling 4 Starships for higher orbits or Moon or Mars journeys. That would require launches every two hours. Would be very hard to have any humans in the launch and landing pad areas at any point.

I wonder what the minimum survival distance from a launch for a Teslabot would be? Teleoperated Teslabots could do lots of the launchpad work. They could hide in soundproofed vaults during launches and landings, and quickly come out while things are still unsafe for humans.

Note on windows and fuel transfers: I wonder how much out of phase landing could happen due to needing acceleration to make transfers happen? Constant acceleration or deceleration is needed while fuel transfers are under way. This would allow one to push forward or back the timing of tanker landings. It makes the launch ballet harder to calculate, but that's what computers are for.

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

The timing you suggest for landing immediately on the heels of a launch means a last minute scrub splashes a tanker. I don't think there are any orbital shenanigans that would delay a landing by more than a few minutes and a 30 minute scrub means a 12 hour wait for the next slot.

There's been detailed discussion of the value of a dedicated tanker build. Most center around specialized hardware; a depot once and done or everything on all tankers with recurring mass penalties. Mating arms/struts, active and/or passive cooling and PV power to name a few.


A biggie is the GSE quick disconnect plate. It needs a redesign for androgyny, a gender bender adapter or a different gender on the depot. The problem is the GSE quick disconnect is one gender with both ships and tankers having the other. The depot has to have the same gender as the GSE or it will not be able to mate with ship or tanker - but then it can't mate with the GSE QD plate.

IMO the simplest solution is to put a gender bender on the GSE when launching a depot and keep the mass on the ground. They are QUICK disconnects so it shouldn't be all that much of a bother for relatively rare depot launch, but an androgynous connector plate can't be ruled out.

Refueling campaigns will be rare occurrences until the moon or mars get heavy traffic and it's tempting to try to simplify and do without a special depot variant. Every time we look at it the tankers need one kludge after another or it won't work. Every tanker ends up being three quarters of a depot with heavy mass penalties. A dedicated depot has all the mass penalties but it only launches once and saves a bit because it has no fins or heatshield.

[Eka]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

More tankers, and simultaneous launches relieves the need for purpose built depot Starships. On the other hand it puts a much much harder load on GSE to keep up with the launch rate. A two tower site with proper GSE could launch 6 Starships/tankers in a day into the same refueling window. Treat it like a tree for the pairing off for transfers. When a pair gets into orbit, they immediately start transferring to one of them. This is because they have the least to maneuver before starting fuel transfer. The first two pairs launched maneuver during fuel transfer to meet when both their transfers are done. They then transfer both double loads into a quad tanker load. yada yada yada... The empties make maneuvers to arrange for landing when slots are available.

Landing Note: If grid landing pads are available, two tankers can land just after two take off. They literally need to be going in for landing at the time the launching two are in launch preps. The boosters will be caught by the towers soon after the tankers land. Ground support now has around 6 hours to detank, and move off the spent tankers, and restack the boosters and stack the next tankers. It needs to be a well coordinated ballet, but it should be possible. With more automated Starship handlers, one could likely manage up to 4 simultaneous refueling windows in space at one time. That would be for fueling 4 Starships for higher orbits or Moon or Mars journeys. That would require launches every two hours. Would be very hard to have any humans in the launch and landing pad areas at any point.

I wonder what the minimum survival distance from a launch for a Teslabot would be? Teleoperated Teslabots could do lots of the launchpad work. They could hide in soundproofed vaults during launches and landings, and quickly come out while things are still unsafe for humans.

Note on windows and fuel transfers: I wonder how much out of phase landing could happen due to needing acceleration to make transfers happen? Constant acceleration or deceleration is needed while fuel transfers are under way. This would allow one to push forward or back the timing of tanker landings. It makes the launch ballet harder to calculate, but that's what computers are for.

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

The timing you suggest for landing immediately on the heels of a launch means a last minute scrub splashes a tanker. I don't think there are any orbital shenanigans that would delay a landing by more than a few minutes and a 30 minute scrub means a 12 hour wait for the next slot.

There's been detailed discussion of the value of a dedicated tanker build. Most center around specialized hardware; a depot once and done or everything on all tankers with recurring mass penalties. Mating arms/struts, active and/or passive cooling and PV power to name a few.


A biggie is the GSE quick disconnect plate. It needs a redesign for androgyny, a gender bender adapter or a different gender on the depot. The problem is the GSE quick disconnect is one gender with both ships and tankers having the other. The depot has to have the same gender as the GSE or it will not be able to mate with ship or tanker - but then it can't mate with the GSE QD plate.

IMO the simplest solution is to put a gender bender on the GSE when launching a depot and keep the mass on the ground. They are QUICK disconnects so it shouldn't be all that much of a bother for relatively rare depot launch, but an androgynous connector plate can't be ruled out.

Refueling campaigns will be rare occurrences until the moon or mars get heavy traffic and it's tempting to try to simplify and do without a special depot variant. Every time we look at it the tankers need one kludge after another or it won't work. Every tanker ends up being three quarters of a depot with heavy mass penalties. A dedicated depot has all the mass penalties but it only launches once and saves a bit because it has no fins or heatshield.

I was landing the Tankers on a grid landing pad like talked about in this post of mine. That means the towers are only used for catching Boosters, and Starships returning with loads. A tanker has say 200t of ship, and 1740t of propellants, and the Booster likely is pushing it to orbit at over 2G just before hot staging. That means the skirt is likely plenty strong enough to be landed on when it returns empty. One just needs a place to get rid of the exhaust to and a bit of ground based dampening. That is why the water cooled, sprung, and dampened grid over a flame trench. I did suggest the skirt be extended down some to provide greater safety for the engine bells. So simultaneous takeoffs and landing is possible.

Orbital shenanigans: Because fuel transfer requires the tanks to be settled, and that requires constant positive or negative acceleration. This changes how long the orbits take. I was wanting the tankers to land just after takeoff so the human exclusion time is minimized. The most delay I'd think would be worthwhile is to have the tankers landing a few minutes after the booster lands.

A huge issue is before full automation of ship moves, humans need safe windows to work in, and get the pad cleared again. Imagine a person having a medical emergency during pad clear. If there isn't enough time to safely move that person off site, that would scratch all landing tankers. With the grid landing pads, two sets of them could be far enough away from the towers in opposite direction to allow landings with a non cleared launch pad. BC would have the ability to space out this far, but KSC could. Place a landing pad on each side of each end of the big runway. That gives 4 landing pads widely spaced form each other. Other layouts in other places could be used.

Genders... There are genderless quick connects. I used them on my spray rig I had for the farm.

Gender benders: Any Tanker or Starship could be sent up with an attached gender bender. This is because it is on the leeward side during reentry. So which ships are sent up with gender bender can be planned out so there is always one for each transfer. Gender benders can also be rigged with a hinged structure to flip it out of the way if needed. The last Tanker to fill a Starship could also take the gender changer back that was on the Starship. I could see sending a few gender benders to Mars just in case they are needed. Mars is a better launching off place for exploring the asteroid belt. I can see them wanting to top off a Starship before exploring.

I like to know the ways to achieve high rates so I don't box the system out of using the best approaches. Yes, stuff learned in the future can change what is best. That is life. One can only work with what one knows and invents now. If thousands of ships are going to transit every transfer window, one needs design so high throughput systems can be done from end to end. While one can stage many filled Starships in orbit waiting for the transfer window, they will need topping off before leaving due to boil off. At Mars, the landing pads will also get very busy thus needing to be designed for high throughput. Another thing is slight differences in transfer times can possibly lead to landing traffic jams. The need for more landing pads will likely be encountered at some point.

My brain is telling me I've thought too long on this. Having ME/CFS sucks!

[Eka]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

More tankers, and simultaneous launches relieves the need for purpose built depot Starships. On the other hand it puts a much much harder load on GSE to keep up with the launch rate. A two tower site with proper GSE could launch 6 Starships/tankers in a day into the same refueling window. Treat it like a tree for the pairing off for transfers. When a pair gets into orbit, they immediately start transferring to one of them. This is because they have the least to maneuver before starting fuel transfer. The first two pairs launched maneuver during fuel transfer to meet when both their transfers are done. They then transfer both double loads into a quad tanker load. yada yada yada... The empties make maneuvers to arrange for landing when slots are available.

Landing Note: If grid landing pads are available, two tankers can land just after two take off. They literally need to be going in for landing at the time the launching two are in launch preps. The boosters will be caught by the towers soon after the tankers land. Ground support now has around 6 hours to detank, and move off the spent tankers, and restack the boosters and stack the next tankers. It needs to be a well coordinated ballet, but it should be possible. With more automated Starship handlers, one could likely manage up to 4 simultaneous refueling windows in space at one time. That would be for fueling 4 Starships for higher orbits or Moon or Mars journeys. That would require launches every two hours. Would be very hard to have any humans in the launch and landing pad areas at any point.

I wonder what the minimum survival distance from a launch for a Teslabot would be? Teleoperated Teslabots could do lots of the launchpad work. They could hide in soundproofed vaults during launches and landings, and quickly come out while things are still unsafe for humans.

Note on windows and fuel transfers: I wonder how much out of phase landing could happen due to needing acceleration to make transfers happen? Constant acceleration or deceleration is needed while fuel transfers are under way. This would allow one to push forward or back the timing of tanker landings. It makes the launch ballet harder to calculate, but that's what computers are for.

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

The timing you suggest for landing immediately on the heels of a launch means a last minute scrub splashes a tanker. I don't think there are any orbital shenanigans that would delay a landing by more than a few minutes and a 30 minute scrub means a 12 hour wait for the next slot.

There's been detailed discussion of the value of a dedicated tanker build. Most center around specialized hardware; a depot once and done or everything on all tankers with recurring mass penalties. Mating arms/struts, active and/or passive cooling and PV power to name a few.


A biggie is the GSE quick disconnect plate. It needs a redesign for androgyny, a gender bender adapter or a different gender on the depot. The problem is the GSE quick disconnect is one gender with both ships and tankers having the other. The depot has to have the same gender as the GSE or it will not be able to mate with ship or tanker - but then it can't mate with the GSE QD plate.

IMO the simplest solution is to put a gender bender on the GSE when launching a depot and keep the mass on the ground. They are QUICK disconnects so it shouldn't be all that much of a bother for relatively rare depot launch, but an androgynous connector plate can't be ruled out.

Refueling campaigns will be rare occurrences until the moon or mars get heavy traffic and it's tempting to try to simplify and do without a special depot variant. Every time we look at it the tankers need one kludge after another or it won't work. Every tanker ends up being three quarters of a depot with heavy mass penalties. A dedicated depot has all the mass penalties but it only launches once and saves a bit because it has no fins or heatshield.

I was landing the Tankers on a grid landing pad like talked about in this post of mine. That means the towers are only used for catching Boosters, and Starships returning with loads. A tanker has say 200t of ship, and 1740t of propellants, and the Booster likely is pushing it to orbit at over 2G just before hot staging. That means the skirt is likely plenty strong enough to be landed on when it returns empty. One just needs a place to get rid of the exhaust to and a bit of ground based dampening. That is why the water cooled, sprung, and dampened grid over a flame trench. I did suggest the skirt be extended down some to provide greater safety for the engine bells. So simultaneous takeoffs and landing is possible.

Orbital shenanigans: Because fuel transfer requires the tanks to be settled, and that requires constant positive or negative acceleration. This changes how long the orbits take. I was wanting the tankers to land just after takeoff so the human exclusion time is minimized. The most delay I'd think would be worthwhile is to have the tankers landing a few minutes after the booster lands.

A huge issue is before full automation of ship moves, humans need safe windows to work in, and get the pad cleared again. Imagine a person having a medical emergency during pad clear. If there isn't enough time to safely move that person off site, that would scratch all landing tankers. With the grid landing pads, two sets of them could be far enough away from the towers in opposite direction to allow landings with a non cleared launch pad. BC would have the ability to space out this far, but KSC could. Place a landing pad on each side of each end of the big runway. That gives 4 landing pads widely spaced form each other. Other layouts in other places could be used.

Genders... There are genderless quick connects. I used them on my spray rig I had for the farm.

Gender benders: Any Tanker or Starship could be sent up with an attached gender bender. This is because it is on the leeward side during reentry. So which ships are sent up with gender bender can be planned out so there is always one for each transfer. Gender benders can also be rigged with a hinged structure to flip it out of the way if needed. The last Tanker to fill a Starship could also take the gender changer back that was on the Starship. I could see sending a few gender benders to Mars just in case they are needed. Mars is a better launching off place for exploring the asteroid belt. I can see them wanting to top off a Starship before exploring.

I like to know the ways to achieve high rates so I don't box the system out of using the best approaches. Yes, stuff learned in the future can change what is best. That is life. One can only work with what one knows and invents now. If thousands of ships are going to transit every transfer window, one needs design so high throughput systems can be done from end to end. While one can stage many filled Starships in orbit waiting for the transfer window, they will need topping off before leaving due to boil off. At Mars, the landing pads will also get very busy thus needing to be designed for high throughput. Another thing is slight differences in transfer times can possibly lead to landing traffic jams. The need for more landing pads will likely be encountered at some point.

My brain is telling me I've thought too long on this. Having ME/CFS sucks!

Adding more.
Tankers  = Starship Tankers for lofting propellants to orbit.
Starship = Starship meant to go to Moon, Mars, or elsewhere.
Booster is self explanatory.

More Orbital Shenanigans: What happens if we fill two Tankers at very low LEO, say 200km. Yes, I'm having them at a very low orbit. They will only be up there for days at most so air drag is basically a non issue. Then have them go to a much higher altitude orbit, or highly elliptical one, before fueling the Starship. We save the fuel cost of lofting most of the Tankers dry mass to higher orbits. We still incur the fuel costs of lofting all that fuel to the higher orbit, but not the costs of getting many Tankers there. Just two plus the Starship. This allows the Starship to leave Earth with its final burn fully fueled from a higher energy orbit.

Obviously there could be many variants of Tanker numbers and Starship to get different energies before going into the transfer orbit. This should allow higher energies thus faster transfer orbits to Mars without stranding a tanker in a long orbit to get back to Earth.

And what I'm really getting to: What it really does is call into question is using depots in fixed orbits. I always called them into question because they would often be in the wrong orbit. Even the big huge waves of ships going to Mars would all have slightly different orbits. I guess if NASA wants to spend money on them. I guess it may pay to have one in orbit around the Moon, or LEO if trips from KSC to the moon are routine.

I've been thinking some tankers should have two sets of header tanks. The second set is for a burn for returning from a very high energy orbits to LEO for lining up for landing.

[OTV Booster]

I double thunk on this one. A tanker/depot loss cost in program time far outweighs the financial cost of a few launches of propellant. A late loss cost more in $$ but early it late, the big cost is to program flow.

Depends on the reason for the loss, the cadence, and the boil-off rate.  If the cause is easy to find in the instrumentation, it may be a software fix or something that can be turned around and re-installed in a day or two.  Or it could be something that grounds the program for months.

Unless they have a spare depot even an easy fix will take a while. A hot spare sounds like a good idea. Hardware rich...yada yada.

More tankers, and simultaneous launches relieves the need for purpose built depot Starships. On the other hand it puts a much much harder load on GSE to keep up with the launch rate. A two tower site with proper GSE could launch 6 Starships/tankers in a day into the same refueling window. Treat it like a tree for the pairing off for transfers. When a pair gets into orbit, they immediately start transferring to one of them. This is because they have the least to maneuver before starting fuel transfer. The first two pairs launched maneuver during fuel transfer to meet when both their transfers are done. They then transfer both double loads into a quad tanker load. yada yada yada... The empties make maneuvers to arrange for landing when slots are available.

Landing Note: If grid landing pads are available, two tankers can land just after two take off. They literally need to be going in for landing at the time the launching two are in launch preps. The boosters will be caught by the towers soon after the tankers land. Ground support now has around 6 hours to detank, and move off the spent tankers, and restack the boosters and stack the next tankers. It needs to be a well coordinated ballet, but it should be possible. With more automated Starship handlers, one could likely manage up to 4 simultaneous refueling windows in space at one time. That would be for fueling 4 Starships for higher orbits or Moon or Mars journeys. That would require launches every two hours. Would be very hard to have any humans in the launch and landing pad areas at any point.

I wonder what the minimum survival distance from a launch for a Teslabot would be? Teleoperated Teslabots could do lots of the launchpad work. They could hide in soundproofed vaults during launches and landings, and quickly come out while things are still unsafe for humans.

Note on windows and fuel transfers: I wonder how much out of phase landing could happen due to needing acceleration to make transfers happen? Constant acceleration or deceleration is needed while fuel transfers are under way. This would allow one to push forward or back the timing of tanker landings. It makes the launch ballet harder to calculate, but that's what computers are for.

Note I'm assuming 8 hours between times when the orbit goes over the launch pad again.

The timing you suggest for landing immediately on the heels of a launch means a last minute scrub splashes a tanker. I don't think there are any orbital shenanigans that would delay a landing by more than a few minutes and a 30 minute scrub means a 12 hour wait for the next slot.

There's been detailed discussion of the value of a dedicated tanker build. Most center around specialized hardware; a depot once and done or everything on all tankers with recurring mass penalties. Mating arms/struts, active and/or passive cooling and PV power to name a few.


A biggie is the GSE quick disconnect plate. It needs a redesign for androgyny, a gender bender adapter or a different gender on the depot. The problem is the GSE quick disconnect is one gender with both ships and tankers having the other. The depot has to have the same gender as the GSE or it will not be able to mate with ship or tanker - but then it can't mate with the GSE QD plate.

IMO the simplest solution is to put a gender bender on the GSE when launching a depot and keep the mass on the ground. They are QUICK disconnects so it shouldn't be all that much of a bother for relatively rare depot launch, but an androgynous connector plate can't be ruled out.

Refueling campaigns will be rare occurrences until the moon or mars get heavy traffic and it's tempting to try to simplify and do without a special depot variant. Every time we look at it the tankers need one kludge after another or it won't work. Every tanker ends up being three quarters of a depot with heavy mass penalties. A dedicated depot has all the mass penalties but it only launches once and saves a bit because it has no fins or heatshield.

I was landing the Tankers on a grid landing pad like talked about in this post of mine. That means the towers are only used for catching Boosters, and Starships returning with loads. A tanker has say 200t of ship, and 1740t of propellants, and the Booster likely is pushing it to orbit at over 2G just before hot staging. That means the skirt is likely plenty strong enough to be landed on when it returns empty. One just needs a place to get rid of the exhaust to and a bit of ground based dampening. That is why the water cooled, sprung, and dampened grid over a flame trench. I did suggest the skirt be extended down some to provide greater safety for the engine bells. So simultaneous takeoffs and landing is possible.

Orbital shenanigans: Because fuel transfer requires the tanks to be settled, and that requires constant positive or negative acceleration. This changes how long the orbits take. I was wanting the tankers to land just after takeoff so the human exclusion time is minimized. The most delay I'd think would be worthwhile is to have the tankers landing a few minutes after the booster lands.

A huge issue is before full automation of ship moves, humans need safe windows to work in, and get the pad cleared again. Imagine a person having a medical emergency during pad clear. If there isn't enough time to safely move that person off site, that would scratch all landing tankers. With the grid landing pads, two sets of them could be far enough away from the towers in opposite direction to allow landings with a non cleared launch pad. BC would have the ability to space out this far, but KSC could. Place a landing pad on each side of each end of the big runway. That gives 4 landing pads widely spaced form each other. Other layouts in other places could be used.

Genders... There are genderless quick connects. I used them on my spray rig I had for the farm.

Gender benders: Any Tanker or Starship could be sent up with an attached gender bender. This is because it is on the leeward side during reentry. So which ships are sent up with gender bender can be planned out so there is always one for each transfer. Gender benders can also be rigged with a hinged structure to flip it out of the way if needed. The last Tanker to fill a Starship could also take the gender changer back that was on the Starship. I could see sending a few gender benders to Mars just in case they are needed. Mars is a better launching off place for exploring the asteroid belt. I can see them wanting to top off a Starship before exploring.

I like to know the ways to achieve high rates so I don't box the system out of using the best approaches. Yes, stuff learned in the future can change what is best. That is life. One can only work with what one knows and invents now. If thousands of ships are going to transit every transfer window, one needs design so high throughput systems can be done from end to end. While one can stage many filled Starships in orbit waiting for the transfer window, they will need topping off before leaving due to boil off. At Mars, the landing pads will also get very busy thus needing to be designed for high throughput. Another thing is slight differences in transfer times can possibly lead to landing traffic jams. The need for more landing pads will likely be encountered at some point.

My brain is telling me I've thought too long on this. Having ME/CFS sucks!

There's over 300 posts in that topic. Now that they're pointed out I'll read 'em but for now, can you give a quick explanation of what a grid landing pad is? Thanks in advance.

I totally agree about separating launch and landing ops. It helps with the hypothetical one hour booster turnaround but it jams up other parts of ops. If they're turning launches around in a hour it seems reasonable to have a pipeline from the landing pad supplying a ship an hour with maybe two or three ships in the pipeline. But that's for tomorrow.


For today they need to get the second OLM up and working. Then the OLMs at the cape, which IIUC, are on hold. If the CAPEs on hold and the second BC OLM is finished is when we should keep an eye out for a BC landing pad.

You reference strengthening and maybe extending the engine skirt. I guess this is something specific to the grid landing pad. Without knowing the details it makes me instinctively recoil. The engine skirt is designed to take a steady load, not a jerky load. Sounds as if it might be a large mass penalty to stiffen it for jerks (or even for nice guys ).

The thing about moving gender benders around is the classic 'ah, well boss, we don't have one of them around right now."  GBM (Gender Bender Management) adds another administrative headache that makes ops more complex. Simple is good.

The way a tanker takes props from another tanker then aggregates it into another tanker means most of the propellent gets transferred twice or more. Chilldown losses each time. When a transfer takes place both tankers maneuver a heatshield and fins plus any connecting gear and other depot kit needed. A depot has what it needs and no more. The ops sound to complicated and the hardware too massive.

There's no telling what refueling will be like by the time high numbers of ships are going to mars. In the short term, first and maybe second missions will probably be one or two dozen ships making the transfer. Even that is too far out for firm numbers but let's do a mind experiment.

Assume four depots 6 hours apart and in the same orbit. They can be topped up and ready to support four ships leaving at the earliest opening of the window. They do another four ships in the middle of the window and another four ships at the close of the window. That's 12 ships total. Then they're available to support launches the next synod.

My take is that tankers doubling as depots fix little, make ops more difficult and probably end up needing more of a final top off. Straighten me out if I'm missing it.