Starship On-orbit refueling - Options and Discussion

[Redclaws]

There is a "clash of visions" here. From my humble POV.

Launching lots of big tanks and assembling them to create a depot is kind of the old NASA way of doing things: this is how von Braun Mars 1969 would have done it. And this is how they do it in Stephen Baxter "Voyage".
...
With expendable Saturn V !  And there it is justified, because the "tanker rocket" has to be rebuild each time, and so the infrastructure allows to "wait".

Seems Elon wants to do it a very different way: no orbital infrastructure. So one partially fueled Starship - bound for Mars later - would be refueled "in serie" and pretty fast by ten launches; ten more Starships each with 150 mt of propellants. The Starships would go into orbit, offload their propellants, return, and start all over again.
Because the vehicle is reusable and with fast turnaround, there is no point for a massive orbital infrastructure to store the fuel: the Mars-bound-Starship IS the depot all by itself.

I think Libra nails it.  Musk has - to my knowledge - never suggested starship would involve any form of orbital depot.  He’s repeatedly described what Libra just did; launch rapidly and accumulate fuel in the departing ship.  The only variant of this I see being terribly likely is filling *one* ship on orbit to fill a departure ship, so humans don’t have to wait around/experience 10 (or whatever) refuelings while onboard.  This requires almost no specific technology or engineering at all - you’re already doing refueling and long duration hold of cryogenics (though not in LEO and maybe not in the main tanks).

[steveleach]

I think Libra nails it.  Musk has - to my knowledge - never suggested starship would involve any form of orbital depot.

Ironically, Musk tweeted this yesterday...

An orbital propellant depot optimized for cryogenic storage probably makes sense long-term

[Ludus]

Much of the disagreement seems to stem from whether the application is a large synodic fleet going to Mars together or steady state operations in earth orbit. For a big fleet, that necessarily is waiting around for a long time anyway, it doesn’t make sense to require a parallel fleet of accumulation tankers.

For day to day steady operations, it doesn’t make sense to make a priority mission wait around and dock 6-10 times with tankers when you could assign one tanker to accumulate and simplify the logistical planning. For steady operations it’s a major benefit to be able to just keep launching tankers 24/7 all the time rather than starting only when a mission has been launched and is waiting. It makes risk mitigation sense to only dock once with higher value Starship with cargo aboard.

[joek]

I think Libra nails it.  Musk has - to my knowledge - never suggested starship would involve any form of orbital depot.

Ironically, Musk tweeted this yesterday...

An orbital propellant depot optimized for cryogenic storage probably makes sense long-term

One reason everyone was stepping carefully around the word "depot" (this goes back a ways, circa 2019 and earlier):

Senator Shelby called NASA and said if he hears one more word about propellant depots he’s going to cancel the space technology program.

Thus IMO why Musk studiously avoided "depot"; instead "tankers".  Things have loosened up a bit since.  Why the change?  Longer conversation and OT for this thread (Shelby, SLS, Artemis, SpaceX-Starship, depots, ...  makes for a potentially toxic discussion).

[Redclaws]

I think Libra nails it.  Musk has - to my knowledge - never suggested starship would involve any form of orbital depot.

Ironically, Musk tweeted this yesterday...

An orbital propellant depot optimized for cryogenic storage probably makes sense long-term

Hah, bit of egg on my face then.  Interesting.  I think it's an interesting discussion anyway, but I still don't think it's likely as an early phase.

[Redclaws]

I think Libra nails it.  Musk has - to my knowledge - never suggested starship would involve any form of orbital depot.

Ironically, Musk tweeted this yesterday...

An orbital propellant depot optimized for cryogenic storage probably makes sense long-term

One reason everyone was stepping carefully around the word "depot" (this goes back a ways, circa 2019 and earlier):

Senator Shelby called NASA and said if he hears one more word about propellant depots he’s going to cancel the space technology program.

Thus IMO why Musk studiously avoided "depot"; instead "tankers".  Things have loosened up a bit since.  Why the change?  Longer conversation and OT for this thread (Shelby, SLS, Artemis, SpaceX-Starship, depots, ...  makes for a potentially toxic discussion).

There's obviously a lot of ugly context around "depots" for NASA, etc, but Musk didn't just "say" tankers - He showed (consistently across the various ITS, BFR, SS/SH presentations) a bunch of ships launching to fill one ship.

But, I suppose if you put one ship up there and fill it, then it fills the departing ship, well, if you do that a few times with one of them...  It does start to look pretty depot-ish.  I just think it's highly likely that the early stuff will be just be a standard or *lightly* modified starship.

[alastairmayer]

One word.
Inclination change is expensive.
Every moon/mars opportinity will require different inclinations.

That's four words (twelve, if you count the next sentence).

But it's most expensive in LEO. The higher (further) you're going, the easier it gets. Worst case, add another tanker load of fuel.

[mikelepage]

I don't think this will work. They need to have bilateral symmetry. Starships will approach each out back to back and inverted so the heatshields of each will be on different sides

Why inverted? Back to back is enough to flip connection points, with, say, a male connector on the left and a female connector on the right.

Exactly. I must admit I've been confused by the need for inversion every time I've seen it depicted, because it seems to me your heat management processes will be served better by having both heat shields point in the same direction, and your electricity generation helped by having your solar panels pointing in the same direction. Is inversion necessary just to put the CH4 and O2 lines on opposite sides of the craft? If so it seems like quite a tradeoff. Maybe I'm missing something?

Also, I tend to find many of the 2D depictions in this thread somewhat hard to interpret, so hoping my putting my 3D models into this will help visualisation. I'm still working on my next "Starship Design speculation" video, and we now know from SN8-11 that the orientation of the raptors is two sea-level raptors on the "ventral" side and two vac raptors on the "dorsal" side.

Thank you to whoever it was who said that the prop lines would make most sense if they ran along the inside edge of the skirt (that's one reason I've made the landing leg design as it is - more on that in the video). Also I prefer the idea from this thread  that each tank would have two connectors that would be male and female - one routed to the "top" of the tank and one to the "bottom" (or what would be top and bottom if the Starship was standing with the pointy end up). Here I've depicted the CH4 lines in red and the O2 lines in blue.

I've done a second version with separate gas connection pipes (8-pipe version) that would connect to the tanks separately from the liquid fill lines? although I'm not sure whether the ability to handle the gas and liquid propellants separately would be worth the extra mass and docking complexity. 

[BZHSpace]

It is correct that Superheavy/Starship will be fuel by the bottom of the booster ? So maybe to stack Starship and Superheavy they will need an inner ring to join the CH4/LOX pipe that could be use also for Starship/Starship orbital refueling operation ?

[steveleach]

It is correct that Superheavy/Starship will be fuel by the bottom of the booster ? So maybe to stack Starship and Superheavy they will need an inner ring to join the CH4/LOX pipe that could be use also for Starship/Starship orbital refueling operation ?

Using the same plumbing for fuelling the stack and orbital refuelling is exactly their plan, and the reason why they are fuelling from the bottom.

I'm not sure what you mean by "inner ring" though.

[BZHSpace]

It is correct that Superheavy/Starship will be fuel by the bottom of the booster ? So maybe to stack Starship and Superheavy they will need an inner ring to join the CH4/LOX pipe that could be use also for Starship/Starship orbital refueling operation ?

Using the same plumbing for fuelling the stack and orbital refuelling is exactly their plan, and the reason why they are fuelling from the bottom.

I'm not sure what you mean by "inner ring" though.

Inner ring means : ring welded inside the skirt as current landing legs supports.

[libra]

Hah, bit of egg on my face then.

Egg on mine, too 

[Rossco]

Rather than butt2butt refilling why not side to side?

My thinking is you could either have two tankers fuelling a single SS at the same time or maybe leaving one/two attached for the entire trip to mars to act as a fuel tank for the main SS - doing this could require less engines on the tankers, thus saving weight?
Once in orbit at Mars they disconnect ready to be collected by a returning SS?

[AC in NC]

Rather than butt2butt refilling why not side to side?

My thinking is you could either have two tankers fuelling a single SS at the same time or maybe leaving one/two attached for the entire trip to mars to act as a fuel tank for the main SS - doing this could require less engines on the tankers, thus saving weight?
Once in orbit at Mars they disconnect ready to be collected by a returning SS?

- Because there are no GSE attachments on the side
- Because there's no need to have two tankers refueling at the same time
- Because leaving them attached requires a redesign to the thrust structure
- No, less engines wouldn't be the result
- Collected how?

No.  Just no.  There's no problem being solved here and rockets aren't Legos^TM.

[joek]

Rather than butt2butt refilling why not side to side?

My thinking is you could either have two tankers fuelling a single SS at the same time or maybe leaving one/two attached for the entire trip to mars to act as a fuel tank for the main SS - doing this could require less engines on the tankers, thus saving weight?
Once in orbit at Mars they disconnect ready to be collected by a returning SS?

- Because there are no GSE attachments on the side
- Because there's no need to have two tankers refueling at the same time
- Because leaving them attached requires a redesign to the thrust structure
- No, less engines wouldn't be the result
- Collected how?

No.  Just no.  There's no problem being solved here and rockets aren't Legos^TM.

To expand on that... "Once in orbit at Mars..." Not going to happen until Mars orbit aerocapture is proven; a significant challenge.  For the foreseeable future, it's going to be direct Mars entry per current SpaceX (as best we can tell) conops.

[Twark_Main]

Trying to shed a little light on the subject of Starship refueling ladders, I made up a spreadsheet to calculate optimal elliptical refueling orbits. I call it my Big Starship Refueling Spreadsheet of Doom^TM.

The idea here is to start with N ships (1 "mission" Starship and N-1 tankers) in LEO. The whole fleet raises into an elliptical orbit, and using on-orbit refueling one tanker empties into the remaining fleet, filling all the rest of the ships completely (the math trick here is that these two constraints, put together, allow you to solve for the optimum elliptical orbit apogee). Repeat with N-1 tankers until only the "mission" Starship remains.

This avoids the need to develop new hardware (nuclear rockets, stretched tankers, stripped tankers, etc), which is a huge advantage. It also prevents the problem where a depot gets stuck in the "wrong orbit."

The only disadvantage for manned missions is that it requires multiple passes through the radiation belt. This can be minimized with trajectory planning, or simply by discharging the radiation belts temporarily.

Using the spreadsheet we can check the math on Casey Handmer's excellent blog post on lunar Starship. We can calculate that it's possible to perform the same two-way cargo transport of 192 tonnes using only 29 flights (down from 31), raising the payload per launch from 6.2 tonnes per launch to 6.6 tonnes per launch. That's a huge win.

If anyone is interested in putting forth a serious proposal for an orbital depot, this is the "business as usual" plan that they've gotta beat.

Corrections, suggestions, and feedback welcome.

Edit:

Instructions
On the spreadsheet red highlighted cells indicate invalid values. such as
  • insufficient delta-v in LEO to complete the mission
  • an elliptical orbit gets too high and escapes into solar orbit
  • insufficient aggregate tank volume in LEO for that number of refueling flights

Typical usage is to set your mission- and vehicle-related numbers, then look at the first block ("LEO Direct") and find the first row without any red cells. Proceed left-to-right across blocks until finding a valid row.

[superg05]

can a StarShip Superheavy booters IN RAPTOR HEAVY STYLE CONFIGURATION get a fully fuelled starship to orbit? if so it can skip all the refueling of multiple tankers

[AC in NC]

can a StarShip Superheavy booters IN RAPTOR HEAVY STYLE CONFIGURATION get a fully fuelled starship to orbit? if so it can skip all the refueling of multiple tankers

No.  And if it could you'd be throwing away one or three SH Boosters.

[Thrustpuzzle]

The idea here is to start with N ships (1 "mission" Starship and N-1 tankers) in LEO. The whole fleet raises into an elliptical orbit, and using on-orbit refueling one tanker empties into the remaining fleet, filling all the rest of the ships completely (the math trick here is that these two constraints, put together, allow you to solve for the optimum elliptical orbit apogee). Repeat with N-1 tankers until only the "mission" Starship remains.

These tools are always fun to play with! Thanks for the model.

Wouldn't the strategy be optimized by first consolidating all fresh tanker launches (each holding 100t)  in LEO to form fully-fueled tankers holding 1200t, then laddering those up with the (similarly fully fueled) mission starship  into higher orbits? Every time a non-full tanker raises its orbit to PO1, its wasting propellant lifting the tanker's dry mass. You want to amortize that dry mass by having as large a fuel loading as possible.

It's most likely I'm misunderstanding the spreadsheet and this is already the model, but it looks like the 100t tankers are doing their first fuel transfers at higher orbit PO1 without an initial LEO consolidation.

[Twark_Main]

The idea here is to start with N ships (1 "mission" Starship and N-1 tankers) in LEO. The whole fleet raises into an elliptical orbit, and using on-orbit refueling one tanker empties into the remaining fleet, filling all the rest of the ships completely (the math trick here is that these two constraints, put together, allow you to solve for the optimum elliptical orbit apogee). Repeat with N-1 tankers until only the "mission" Starship remains.

These tools are always fun to play with! Thanks for the model.

Wouldn't the strategy be optimized by first consolidating all fresh tanker launches (each holding 100t)  in LEO to form fully-fueled tankers holding 1200t, then laddering those up with the (similarly fully fueled) mission starship  into higher orbits?

Yes! That's exactly what my model assumes. Sorry for not making that clear.


This spreadsheet lets you run the numbers for partially-fueled tankers too (say, two fully-fueled and one half-fueled Starship), depending on exactly how much delta-v your mission requires. But you're exactly right, the mathematical "optimum" (most delta-v for the least launch mass) is to fully fuel all vehicles in LEO, aka PO0.

Problem is, how do you deal with the "leftover fuel" on that last refueling flight? It's rather unlikely that your propellant mass will work out exactly an integer number of 100 t refuelings.

Typically you'll need to tweak the payload mass to achieve exactly that optimum you're describing. Scroll down until you find the blue cell in the "Prop" column. Blue in that column indicates that the last refueling tanker launched into LEO has extra (wasted) fuel left over. Then use the Goal Seek... tool to set the blue cell to the value of the total tank capacity by changing the outbound payload mass (cell B7). Eg if you're using one Tanker and one Mission Starship, you'd Goal Seek to set the blue cell to 1200+1200 = 2400 by changing B7. If you're using three Tankers and one Mission Starship, you'd Goal Seek to set the blue cell to 4800.