Starship On-orbit refueling - Options and Discussion

[Twark_Main]

exoterran propellants

"Extraterrestrial?" 

The thermal impact of EDL is non zero. But with solar input of what, 1.2kW/m^2 for visible light only, unless spaceships were as common as cars this wouldn't amount to a bubble of flatulence in a cyclonic disturbance. Actually I don't have any numbers. If you have a BOE, (I'm not sure how to myself), it would interesting no matter what it shows.

I think the real concern isn't thermal heating, but production of NOx on reentry, which is both ozone depleting and a super greenhouse gas.

[spacenut]

Ozone is replaced by lightning.  Lightning happens all the time around the world. 

[cdebuhr]

Ozone is replaced by lightning.  Lightning happens all the time around the world.

While that is technically correct .... it's irrelevant.  We don't need ozone in the troposphere (where lightning produces it in minor amounts), and in fact surface level ozone is harmful to us meatbags.  It's stratospheric ozone, and the depletion of same, that's the subject here.

[rakaydos]

I think a lot about this because I think for various reasons Earth launch may become limited for legitimate environmental reasons (in the long term). Additionally, by off-loading stuff to Mars, you're building up the industrial capacity there. And it might even save costs eventually.

I'm not saying that well meaning individuals wouldnt mandate offworld propellant, but that is NOT the enviromentally sound approach.

The correct apprach, which elon has even aluded to, is to mandate 100% renewable sabatier-originated natural gas- not just for space launch, but for all applications where electricity doesnt beat combustion. Yes, there's a hundred years reserve of fossil natural gas. But using it  is piling more sweaters on our already overheated planet. Sabatier methane is more expensive, but is inherently carbon neutral- even if you burn it on an escape burn, the exaust will be suborbital, returning to the earthly carbon cycle.

Martian methane for earth departure is the wrong answer because the exaust will also fall to earth- but it's not from earth, it's purely addition. Technically not a fossil fuel, but a carbon import all the same, when we want to be exporting carbon as much as possible. And that's ignoring as irrelivant aerocapture atmospheric heating and the depletion of ion propellants, which are energetic enough NOT to fall in a suborbital path back to the parent body.

All correct in a 'purest' sense but in a practical sense? There are many unknowns here.


I don't think anybody is suggesting exoterran propellants for use for launch from earth. Someone, maybe you, ran some numbers showing rocket exhaust from near earth operations staying with earth. I question this. Some will stay for sure but the exhaust vectors are many and the exhaust velocity is high. It will be CO2 and water vapor in any case. Not methane. Not good, but enough to amount to anything at even the most optimistic traffic projections?

The effective exaust velocity is the ISP times the Standard Gravity, 9.80665 m/s^2
Even a raptor vac has an ISP of only 380 seconds
That's an effective velocity of 3.724 kilomiters a second.

LEO is 7.8 km/s. Earth escape velocity is 11.190 km/s. So if a raptor in LEO fired  in the direction of motion, the exact opposite of what a departing starship would do, propellant would reach escape velocity until the starship has spent 334 m/s of delta V

Noone's suggested propulsive braking at earth orbit, so all propellant spent will be traped in earth's SoI, much of it  in suborbital or surface-intercepting elliptical orbits. what is there to question?

Ignoring earth launch and focusing on exoterran propellants in near Earth use, and assuming 75% 100% is captured, how does this compare to the overall the man made CO2 excesses? Among the unknowns here are future space traffic density, actual recapture rates and future Earth based CO2 production rates. Might as well add that we know little about the impact of CO2 and H2O at extreme altitudes.

We are not talking about now. We are talking about a future in which mars has sufficent propellant production capability  to export back to earth orbit. A future, meanwhile, wherein earth has been ravanged by centuries  of industrial revolution, resulting in events that make the current california wildfires and the american south's double hurricane look like a cloudy day.

If, in the future when mars is capable of exporting propellant, earth industrial CO2 isnt Net Zero or lower, we deserve the venesian helscape it will be delivering us.

These things may come to pass. Or, assuming we finally have a permanent presence in space, technology may transcend these problems with propulsion alternatives we can only speculate about. They'll give us a whole new set of headaches.

None of which is defending the proposal of importing greenhouse gasses, no matter how small an absolute quantity, in a time of enviromental crisis. 1200 tons of CO2+H2O per ship's worth of fuel... and for what? to avoid building on earth the same fuel production facilities you have already built up on mars? Why is that a goal?

[FutureSpaceTourist]

In context of more powerful Raptor reducing number on Super Heavy:

twitter.com/ppathole/status/1311893344297996289

And ultimately this upgrade will result into cutting the number of refuelling flights to 4 instead of 8 which would be a huge improvement!

https://twitter.com/elonmusk/status/1311907493182926849

Probably 5 or 6 with an optimized tanker, although filling up the ship in orbit isn’t required for Mars, so 4 is possible

[sdsds]

Has someone presented a plausible calculation of how many refueling flights would be required to take the maximum Starship LEO payload onward to GEO?

Specifically suppose a Starship launches with the most massive payload it can carry to LEO. Does refilling from a single tanker give the Starship enough propellant to bring its payload to GEO? (The delta-v from LEO to GEO is maybe half that of Earth to LEO, but sheesh, there are logarithms involved in the calculation.... ;-) )

[kkattula]

Has someone presented a plausible calculation of how many refueling flights would be required to take the maximum Starship LEO payload onward to GEO?

Specifically suppose a Starship launches with the most massive payload it can carry to LEO. Does refilling from a single tanker give the Starship enough propellant to bring its payload to GEO? (The delta-v from LEO to GEO is maybe half that of Earth to LEO, but sheesh, there are logarithms involved in the calculation.... ;-) )

Just GEO?  GEO and return the SS to Earth?  GTO? (and then which GTO? e.g. -1500 m/s, -1800 m/s, etc.)

As a rough estimate:

Just GEO with 150t payload and 120t SS, from equatorial LEO, 370 Isp, I get 520t propellant.  Maybe 600 from higher inclination LEO. Call it 3 optimized (200t payload) tankers.

To also take enough extra propellant (approx.100t) to deorbit the empty SS and land from GEO, another 190t or a 4th optimized tanker.

If the max payload can use it's own propulsion to insert from GTO, then about 1.5 optimized tankers, with SS returning to Earth.


I also get the max payload sent to GTO on one optimized tanker load is about 100t. (Assuming 2500 m/s from LEO to GTO, you basically need 1 ton tanker propellant for each ton of SS, return prop (30t?) and payload. That's assuming the SS has it's own return prop already, plus about 50t unused prop from the lighter payload.


Edit:  Should be about 3 optimized tanker loads to send 150t payload to Mars in a minimum delta v trajectory, including landing propellant.  So I expect Elon is still planning on faster trips since he said 4.

[sdsds]

Thanks — having estimates for a variety of mission profiles really helps clarify things! I found this observations particularly ingriguing:

[...] I also get the max payload sent to GTO on one optimized tanker load is about 100t. (Assuming 2500 m/s from LEO to GTO, you basically need 1 ton tanker propellant for each ton of SS, return prop (30t?) and payload. That's assuming the SS has it's own return prop already, plus about 50t unused prop from the lighter payload.

Just one tanker transfer might put a 100t payload onto a GTO trajectory. That's maybe a 10-fold increase over what Ariane 5 can do, but fully reusable? Wow.

[ricardoholanda511]

Thanks — having estimates for a variety of mission profiles really helps clarify things! I found this observations particularly ingriguing:

[...] I also get the max payload sent to GTO on one optimized tanker load is about 100t. (Assuming 2500 m/s from LEO to GTO, you basically need 1 ton tanker propellant for each ton of SS, return prop (30t?) and payload. That's assuming the SS has it's own return prop already, plus about 50t unused prop from the lighter payload.

Just one tanker transfer might put a 100t payload onto a GTO trajectory. That's maybe a 10-fold increase over what Ariane 5 can do, but fully reusable? Wow.

I've been lurking this site for more than three years now and this time I just couldn't control my curiosity. I haven't seen any discussion on the possibility of a Martian orbital refueling in order to reduce the risks of  a hot atmospheric breaking. 

Since Phobos and Deimos look like good candidates for a LOX factory, couldn't they also be used as a tanker base to send intercepting tankers  in a long orbit  to couple with incoming starships, giving them enough fuel to reduce delta-v and avoid the dangers of a fully atmospheric breaking?

I apologise in advance if this question is stupid but I don't know how to use KSP in order to make a simulation like that.

[cdebuhr]

Thanks — having estimates for a variety of mission profiles really helps clarify things! I found this observations particularly ingriguing:

[...] I also get the max payload sent to GTO on one optimized tanker load is about 100t. (Assuming 2500 m/s from LEO to GTO, you basically need 1 ton tanker propellant for each ton of SS, return prop (30t?) and payload. That's assuming the SS has it's own return prop already, plus about 50t unused prop from the lighter payload.

Just one tanker transfer might put a 100t payload onto a GTO trajectory. That's maybe a 10-fold increase over what Ariane 5 can do, but fully reusable? Wow.

I've been lurking this site for more than three years now and this time I just couldn't control my curiosity. I haven't seen any discussion on the possibility of a Martian orbital refueling in order to reduce the risks of  a hot atmospheric breaking. 

Since Phobos and Deimos look like good candidates for a LOX factory, couldn't they also be used as a tanker base to send intercepting tankers  in a long orbit  to couple with incoming starships, giving them enough fuel to reduce delta-v and avoid the dangers of a fully atmospheric breaking?

I apologise in advance if this question is stupid but I don't know how to use KSP in order to make a simulation like that.

Welcome to the forum!

You'd need to carry lots of prop from earth for a propulsive MOI burn, otherwise you're doing aerocapture into Mars orbit anyway, which, I've been led to understand here, is actually at least as rough as EDL is going to be.  Nice to have a Phobos depot, but if you can't get into Martiam orbit, it may as well be in another galaxy.

[tbellman]

I've been lurking this site for more than three years now and this time I just couldn't control my curiosity. I haven't seen any discussion on the possibility of a Martian orbital refueling in order to reduce the risks of  a hot atmospheric breaking. 

Since Phobos and Deimos look like good candidates for a LOX factory, couldn't they also be used as a tanker base to send intercepting tankers  in a long orbit  to couple with incoming starships, giving them enough fuel to reduce delta-v and avoid the dangers of a fully atmospheric breaking?

I apologise in advance if this question is stupid but I don't know how to use KSP in order to make a simulation like that.

From Deimos to Mars' surface, entirely propulsively, is about 5.35 km/s Δv.  For a Starship with empty mass 120 tonnes, cargo of 50 tonnes, and effective exhaust velocity of 3.7 km/s (I_sp=377 s), that requires (120t+50t)*(e^{5.35km/s / 3.7km/s} - 1) ≈ 550 tonnes of propellant.

A dedicated tanker Starship, where the propellant tanks take up the entire ship and can be filled with 2000 tonnes of propellant in LEO, will have about 385 tonnes of propellant left when it reaches Deimos (Δv = 5.3 km/s from LEO to Deimos).  Assuming a "slow" (8-9 months) transfer.  So you need three such tankers for every two ships going to Mars' surface.  Assuming a tanker from Earth's surface can bring 150 tonnes of propellant to LEO, that's another 40 tanker launches to LEO.

By stopping at Deimos, you are also adding almost 1.7 km/s of Δv for the ships going to the surface before they can be refuelled at Deimos.  (Normally they would spend almost no Δv between trans-Mars injection, TMI, and landing, but now they need to spend 1.7 km/s to reach Deimos.)  They will thus need to be filled up more in LEO, or they need to go slower than the 5 months or so Elon wants.

You have thus gone from somewhere between 10 and 15 launches from Earth in order to send two ships to Mars' surface, to something like 60 launches.  Alternatively, you could have used those 45+ extra launches to send 8-10 more cargo ships, trying to land on Mars using aerobraking.  Doing that would give you extra practice in aerobraking, and with a little bit of luck making it safe.  And those ships that don't crash or burn during that testing, will have brought extra cargo to Mars, making the manned stays safer and more comfortable.

It is absolutely possible to do what you suggest, but I very much doubt that it will be worth it.

EDIT to add:

It's worth noting that stopping at Phobos or Deimos incurs a noticable delta-v cost.  If instead you refuel during the coasting phase towards Mars, the tankers save almost 1.7 km/s Δv, and can deliver 670 tonnes of propellant each, instead of 385 tonnes.  The surface-destined ships save 1.3 km/s by not stopping at Deimos.  They will now need 625 tonnes to capture and land, which is slightly less than what one tanker could deliver.  So then you would be down to 28 extra launches from Earth per two ships going to Mars' surface.  Still a fairly large overhead.

(625 tonnes is more than the 550 tonnes I got earlier, but that's because those 550 tonnes only go from Deimos to Mars' surface, while the 625 tonnes here goes from TMI to Mars' surface.  The ships that refuelled at Deimos instead needed to have ca 100 tonnes of propellant left over after TMI in order to stop at Deimos.)

[JaimeZX]

I don't think (edit: ricardoholanda511) was suggesting Starship stop at Diemos, but rather sending tankers from Deimos to intercept the inbound ship on arrival.

[tbellman]

I don't think cdebuhr was suggesting Starship stop at Diemos, but rather sending tankers from Deimos to intercept the inbound ship on arrival.

(It was ricardoholanda511 who suggested it, not cdebuhr.)

Ah, you're probably right.  Upon re-reading again, I think I misread Ricardo's proposal.  (I actually read it that way first, then upon re-reading I managed to interpret it differently...)

But in that case, I have other objections. 

The atmospheric braking will likely only be dangerous in the early days of Mars missions.  After a few tries of landing unmanned ships, SpaceX will hopefully have made that maneuver safe for Starship.  And then the Phobos or Deimos propellant factory will no longer be needed (at least not for this purpose).

But, that propellant factory would need to be built early.  Before any manned flights to Mars.  And importantly, you would need to develop equipment for mining Phobos/Deimos crust and processing it into propellant.  Equipment that would need to work in almost no gravity (less than 1/1000th of Earth gravity).  That's probably technology readiness level 3, or possibly 4, so it will take quite a lot of time, effort, and money, before it can be deployed for real and be relied upon.

You probably also first need to do some more surveying of Phobos and/or Deimos to verify if and how they can be used for making propellant.

I believe it will be significantly faster and cheaper to just send unmanned Starships to try out aerobraking at Mars until it has become safe enough for humans.

[ricardoholanda511]

Many thanks to cdebuhr, JaimeZX e tbellman for the assessment of my question.

Tbellman's second answer really went to the point and made me realise that Elon Musk's self imposed (rightly so) time and technology constraints pretty much exclude some possibilities.

What made me worry about the atmospheric braking only option was also the impossibility of delay, in case  Mars' climate conditions go awry and pose some unknown danger to the ship.

[dgmckenzie]

In the Sands of Mars by Arthur C. Clarke the liner parked off Phobos, I don't think it was Deimos.
They pulled in in and then the passengers transshipped to a Moon to Mars shuttle.

[Norm38]

Probably 5 or 6 with an optimized tanker, although filling up the ship in orbit isn’t required for Mars, so 4 is possible

What does Musk mean my this, that filling up the ship in orbit (LEO?) isn't required for Mars?  He's saying Starship doesn't need full tanks to get to Mars?  But if it did have full tanks, couldn't it make the trip faster?
Or speed is limited by aerobraking heating and coming in faster isn't possible, so only need to fill up as full as it takes to get to that max speed?

Has this been discussed already and I missed it?

[xvel]

Probably 5 or 6 with an optimized tanker, although filling up the ship in orbit isn’t required for Mars, so 4 is possible

What does Musk mean my this, that filling up the ship in orbit (LEO?) isn't required for Mars?  He's saying Starship doesn't need full tanks to get to Mars?  But if it did have full tanks, couldn't it make the trip faster?
Or speed is limited by aerobraking heating and coming in faster isn't possible, so only need to fill up as full as it takes to get to that max speed?

Has this been discussed already and I missed it?

1. Coming in faster is possible but will probably need also propulsive braking
2. Each time you refuel, the delta v increase is smaller and smaller, example:

starship dry mass 120t, 100t payload, ISP 370s

propload dV
200t   - 2346 m/s
400t   - 3759 m/s
600t   - 4774 m/s
800t   - 5566 m/s
1000t  - 6215 m/s
1200t  - 6766 m/s

you don't get a lot on the last two refuellings

[Simbulation]

I am not sure if this has been asked yet, but I could not find this talked about in this thread. Sorry if this has been already asked.

Does anybody have any idea about the technical details of actually doing In-Orbit refueling? How do you go about physically getting propellants in 0g to transfer from one vehicle to another? SpaceX renders show SS docking with a tanker from the aft end, does this make sense?

[docmordrid]

Probably 5 or 6 with an optimized tanker, although filling up the ship in orbit isn’t required for Mars, so 4 is possible

What does Musk mean my this, that filling up the ship in orbit (LEO?) isn't required for Mars?  He's saying Starship doesn't need full tanks to get to Mars?
>

That's how I took it.

[steveleach]

I am not sure if this has been asked yet, but I could not find this talked about in this thread. Sorry if this has been already asked.

Does anybody have any idea about the technical details of actually doing In-Orbit refueling? How do you go about physically getting propellants in 0g to transfer from one vehicle to another? SpaceX renders show SS docking with a tanker from the aft end, does this make sense?

As I understand it...

SS is designed to be fuelled via feed lines from SH, through its skirt. To refuel in orbit you dock the two of them skirt-to-skirt and connect the same feed lines. Then they use RCS to accelerate slightly to settle the propellant, and pump it from the tanker to the other SS.

If you look at images/video of the refuelling you'll see that the two SSs are rotated 180 degrees with respect to each other, so the LOX feed line of one matches up with the LOX vent line of the other (and the same for Methane).