Starship On-orbit refueling - Options and Discussion

[DistantTemple]

Two other points. Both questionable.
Those tanks you postulate have to then reenter and either burn up or splash down, as it would not be appropriate to leave them in orbit. So assuming SX does get sixSS's off to Mars, soon (2024 or 20260) that would be 5 x 6 = 30 reentry and spashdown (as stainless will not burn up, and CF is too expensive) events. And after a few more years, 100's of dropped tanks raining down, with practical, safety, environmental, and public comment concerns.

And two they all cost to make, in $, manpower, resources, and factory space.

The refuelling system is envisaged as one suitable for hundreds and then thousands of launches. It is supposed to moving towards airline-like operations, with only fuel as expendable. And driving the cost <$10M etc.... Mass manufacturing tanks, with navigation and propulsion doesn't marry easily with that. (I guess mass of tank+fuel/oxidiser ~150t )

And a second stage is still needed, which needs to be 100% reusable, else the tanks would become a second stage themselves!!!!!

ISTM EM called this right.

[Paul451]

And now you’re discarding that item.  It has engines and avionics on it, doesn’t it?

Engines and avionics like those in satellites.

I can't picture what you are describing. How do these tanks, loaded with propellant, get into orbit?

[kkattula]

And now you’re discarding that item.  It has engines and avionics on it, doesn’t it?

Engines and avionics like those in satellites.

I can't picture what you are describing. How do these tanks, loaded with propellant, get into orbit?

Yeah. Super Heavy by itself has about zero payload to LEO, expendable.  So you need a Starship second stage to get to orbit with payload. 

Sure, you could just have the propellant for transfer sitting in a disposable tank attached to a satellite bus, that gets dropped off in orbit.

But that costs a lot more and reduces payload per launch, compared to having it sit in Starship's own tanks, and let SS do all the maneuvering. 

Also, you can't just have an orifice in the tank and poke a hose in. The two vehicles need to be firmly connected, have thrust applied to settle the fluids, and have two pipes for each propellant. One for fluid and one for gas, so pressure differential will move the fluid one way and the gas the other.

Maybe you could do all that with a flexible 4-pipe and tether connector, but no need for thrusters. There are technologies for extending and guiding snake-like connectors.  (See Tesla auto-charger). Besides which, SS needs fueling pipes on the tail anyway, so it can be fueled through SH on the pad.

[Star-Dust]

Hi,

Just wanted to start a thread discussing orbital refueling technology proposed by SX, this is (for me) a very critical technology for SS concept success in the future, very little is known about SX approah to the matter, so any insight on technologies existing or in R&D stages is welcome here to improve our understanding.

One of SpaceX's principal engineers behind the Starship project, Paul Wooster, has identified orbital refueling as one of most difficult technology challenges the company will have to overcome in order to realize its Mars ambitions.

Thanks.

[Star-Dust]

The main issues to deal with are:

1-How to prevent cryogen from boiling-off (keep the fuel cool) for long duration.

2-How to transfert the cryogen from one container to the other.

NASA with Robotic Refueling Mission 3 (RRM3) has been trying to developpe the technology needed with relative success, but the scale needed for SX plans will require a complete review of the process for scaling up which may lead to using completly different technology.

The technology is at it early stages of maturation which means that time is still needed and this means that the technology (orbital refueling) most likely (IMHO) wont be ready for SS developement timeframe.

[sebk]

From the info extracted from Elon tweets and plugging holes with informed guesses we have rough general picture of how it would be done:

- Dock the vehicles bottom-to-bottom and obtain seal of the piping
- Vent receiving end to low internal tank pressure (probably well below atmospheric[*])
- Use thrusters (probably receiving end, see below why[*][**]) to produce ullage thrust in the same it's done before re-firing of any stage in orbit
- Open valves and let the transfer begin
- Pressure difference pushes the liquids while ullage thrust (in the order of milli-gee) keeps them settled
- Keep venting the receiving end. There's one possible difficulty here, see below [***]
- If the sending end is close full (for example transfer from accumulation tanker to receiving ship) you need also to keep it pressurized at the required level.
- Once liquid fuel is done pressure would equalize quickly. Possibly bubble detection would shut the valves before this happens as this would be a bit violent. Or maybe the system is made robust enough for it not to be a problem.
- Terminate the ullage thrust
- Purge & vent the piping between main valves on both vehicles.
- Undock

Notes:
*] Venting below atmospheric ensures the remaining fuel would be at a super-cooled temperature as intended.
**] Receiving end needs to be kept at low pressure so it has obvious source of gas for ullage thrusters
***] Receiving end may see liquid foaming and condensation foring in the ullage space. Venting liquid would be wasteful and could be a source of problems as liquid venting to vacuum will partly evaporate and partly freeze and frozen liquid may plug holes. So venting piping could use "cyclone" separator and possibly some mild heating.

From the above one could notice that we need some gas generating subsystem which could pressurize the vehicle without engines firing. So some burner, pump for it and heat exchanger to vaporize the liquid.

[Star-Dust]

From the info extracted from Elon tweets and plugging holes with informed guesses we have rough general picture of how it would be done:

- Dock the vehicles bottom-to-bottom and obtain seal of the piping
- Vent receiving end to low internal tank pressure (probably well below atmospheric[*])
- Use thrusters (probably receiving end, see below why[*][**]) to produce ullage thrust in the same it's done before re-firing of any stage in orbit
- Open valves and let the transfer begin
- Pressure difference pushes the liquids while ullage thrust (in the order of milli-gee) keeps them settled
- Keep venting the receiving end. There's one possible difficulty here, see below [***]
- If the sending end is close full (for example transfer from accumulation tanker to receiving ship) you need also to keep it pressurized at the required level.
- Once liquid fuel is done pressure would equalize quickly. Possibly bubble detection would shut the valves before this happens as this would be a bit violent. Or maybe the system is made robust enough for it not to be a problem.
- Terminate the ullage thrust
- Purge & vent the piping between main valves on both vehicles.
- Undock

Notes:
*] Venting below atmospheric ensures the remaining fuel would be at a super-cooled temperature as intended.
**] Receiving end needs to be kept at low pressure so it has obvious source of gas for ullage thrusters
***] Receiving end may see liquid foaming and condensation foring in the ullage space. Venting liquid would be wasteful and could be a source of problems as liquid venting to vacuum will partly evaporate and partly freeze and frozen liquid may plug holes. So venting piping could use "cyclone" separator and possibly some mild heating.

From the above one could notice that we need some gas generating subsystem which could pressurize the vehicle without engines firing. So some burner, pump for it and heat exchanger to vaporize the liquid.

Interesting but seems very very complexe,to add to the difficulty this will have to be done 5 times to get to Mars....................can't we just attach an external tank like for space shuttle?

[Thunderscreech]

What's the benefit?  The fuel will still need to transfer out of the docked external tank into the engines, the system they envision has them doing that into existing tankage that's already present and used to get it into earth orbit (when it operates as the second stage) and the rest of the challenges aren't materially different in terms of timing for transferring fuel in during boost versus during the RCS-ullage+pressurization scenario another user shared above.

When adding new elements to a problem to solve it, it's useful to know why the added complexity is required and what unique problem it's solving, otherwise perhaps there's a better way.

In this instance, it's not clear what shipping up external tanks that will need to be plumbed solves that doing the ship-ship transfer doesn't.

[Star-Dust]

What's the benefit?  The fuel will still need to transfer out of the docked external tank into the engines, the system they envision has them doing that into existing tankage that's already present and used to get it into earth orbit (when it operates as the second stage) and the rest of the challenges aren't materially different in terms of timing for transferring fuel in during boost versus during the RCS-ullage+pressurization scenario another user shared above.

When adding new elements to a problem to solve it, it's useful to know why the added complexity is required and what unique problem it's solving, otherwise perhaps there's a better way.

In this instance, it's not clear what shipping up external tanks that will need to be plumbed solves that doing the ship-ship transfer doesn't.

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

[Thunderscreech]

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

I think I might not have written my question clearly, please allow me to try again: What makes this simpler?  You're adding parts, you still have docking operations, you still have to transfer fuel into the vehicle (the only thing this changes is when it happens; the fuel needs to come in during the burn as opposed to during the fueling operation)....

What part of this is simpler?

[Alvian@IDN]

We ofc want a thought experiment....

But at the end of the day it's up to SpaceX who actually does it. Remember that they have a very smart engineers, who knows everything they're doing = considerations of choosing this refueling method

They obviously already has NASA on supports, just to keep in mind

[Star-Dust]

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

I think I might not have written my question clearly, please allow me to try again: What makes this simpler?  You're adding parts, you still have docking operations, you still have to transfer fuel into the vehicle (the only thing this changes is when it happens; the fuel needs to come in during the burn as opposed to during the fueling operation)....

What part of this is simpler?

I didn't add parts I just proposed a rethinking of the design, I also didn't suggest a fuel transfert but a direct fuel burn from external tank in SS engines, I think the proposed idea of mating two ends of SS is also hard to imagine, because it's not possible to get the fluid flow in a reversed direction without affecting all parts like pumps and turbines and you can't get extra plumbing in a place where it's already  congested (with engines).

We ofc want a thought experiment....

But at the end of the day it's up to SpaceX who actually does it. Remember that they have a very smart engineers, who knows everything they're doing = considerations of choosing this refueling method

They obviously already has NASA on supports, just to keep in mind

This a passive way to see things.

[Nomadd]

What's the benefit?  The fuel will still need to transfer out of the docked external tank into the engines, the system they envision has them doing that into existing tankage that's already present and used to get it into earth orbit (when it operates as the second stage) and the rest of the challenges aren't materially different in terms of timing for transferring fuel in during boost versus during the RCS-ullage+pressurization scenario another user shared above.

When adding new elements to a problem to solve it, it's useful to know why the added complexity is required and what unique problem it's solving, otherwise perhaps there's a better way.

In this instance, it's not clear what shipping up external tanks that will need to be plumbed solves that doing the ship-ship transfer doesn't.

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

You'd still have to transfer the fuel to the engines, which wouldn't be any simpler than transferring it to the SS tanks, and you'd be dragging an extra 30 or 40 tons of weight around.

[Alvian@IDN]

because it's not possible to get the fluid flow in a reversed direction without affecting all parts like pumps and turbines...

What's make it not possible? It's not like they flow the propellant directly into the engine anyways

...and you can't get extra plumbing in a place where it's already  congested (with engines).

It still had quite enough space

With extra tanks, you will throw the full reusability right off the window, about as far as a billion light years away

[rakaydos]

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

I think I might not have written my question clearly, please allow me to try again: What makes this simpler?  You're adding parts, you still have docking operations, you still have to transfer fuel into the vehicle (the only thing this changes is when it happens; the fuel needs to come in during the burn as opposed to during the fueling operation)....

What part of this is simpler?

I didn't add parts I just proposed a rethinking of the design,

A rethinking that adds external tanks to the tanks used by starship to reach orbit in the first place.

I also didn't suggest a fuel transfert but a direct fuel burn from external tank in SS engines,

"Direct fuel burn" IS a fuel transfer.

I think the proposed idea of mating two ends of SS is also hard to imagine, because it's not possible to get the fluid flow in a reversed direction without affecting all parts like pumps and turbines and you can't get extra plumbing in a place where it's already  congested (with engines).

We're not talking the height-restricted SLS engine bay here- there's plenty of room for plumbing if it's needed to get the job done. The only question is how much mass it takes- and it is very easy to be less mass than a few external tanks.

Edit: You arnt thinking the fuel will be transferred from engine bell to engine bell, are you? because there are easier ways.

[Thunderscreech]

Indeed, I think spitballing ideas is fun, for this one it just seems like it adds complexity without reducing it elsewhere.  More parts, different loading, etc yet the core problem of transferring the fuel in from another tank still exists, but now it exists under heavy thrust instead of being performed during an ullage-settled transfer between two Starships.

[guckyfan]

Starship gets fueled on the pad through internal connections from Superheavy, the first stage. Which means the connecting plumbing is already there.

[Star-Dust]

What's the benefit?  The fuel will still need to transfer out of the docked external tank into the engines, the system they envision has them doing that into existing tankage that's already present and used to get it into earth orbit (when it operates as the second stage) and the rest of the challenges aren't materially different in terms of timing for transferring fuel in during boost versus during the RCS-ullage+pressurization scenario another user shared above.

When adding new elements to a problem to solve it, it's useful to know why the added complexity is required and what unique problem it's solving, otherwise perhaps there's a better way.

In this instance, it's not clear what shipping up external tanks that will need to be plumbed solves that doing the ship-ship transfer doesn't.

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

You'd still have to transfer the fuel to the engines, which wouldn't be any simpler than transferring it to the SS tanks, and you'd be dragging an extra 30 or 40 tons of weight around.

Fuel transfert to engines is easy because turbopumps and vacuuem will take care of it, and for the weight you did mention space shuttle external tank weight which is true but the proposed concept may propose multiple external tanks jettisoned when empty.

[Alvian@IDN]

What's the benefit?  The fuel will still need to transfer out of the docked external tank into the engines, the system they envision has them doing that into existing tankage that's already present and used to get it into earth orbit (when it operates as the second stage) and the rest of the challenges aren't materially different in terms of timing for transferring fuel in during boost versus during the RCS-ullage+pressurization scenario another user shared above.

When adding new elements to a problem to solve it, it's useful to know why the added complexity is required and what unique problem it's solving, otherwise perhaps there's a better way.

In this instance, it's not clear what shipping up external tanks that will need to be plumbed solves that doing the ship-ship transfer doesn't.

May be because it's simpler to do so, just imagine space shuttle docking to her external tank and going all the way to Mars, no air drag issue, the tank is light weight, no time and energy waisted in fuel transfert, the external tank could also be imagined as a payload of a starship fairing.

You'd still have to transfer the fuel to the engines, which wouldn't be any simpler than transferring it to the SS tanks, and you'd be dragging an extra 30 or 40 tons of weight around.

Fuel transfert to engines is easy because turbopumps and vacuuem will take care of it, and for the weight you did mention space shuttle external tank weight which is true but the proposed concept may propose multiple external tanks jettisoned when empty.

Hello, full & rapid reusability..

[Robotbeat]

This is something that needs to be solved, but I consider it at least an order of magnitude easier than, say, reusable orbital reentry of the upper stage. You don’t need massive pumps or whatever. Just some light ullage thrust and a pressure differential.