Starship On-orbit refueling - Options and Discussion

[AbuSimbel]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

Woah I think you just found the thing lol
The thing being SpaceX/Elon’s signature stroke of economic genius through which they can develop new capabilities while still competing in an actual market and not using the government as a cash cow.

Goes to show how big things are often written off as an overkill when the real problem is a lack of vision to leverage them

[edzieba]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That depends on the orbit of the depot and the target orbit of the payload. Regardless of payload mass you need to cart ~100 tonnes of Starship dry mass (plus a few thousand tonnes of prop) around for every burn, so plane changes would become extremely expensive (regardless of whether you moved Starship to match the depot or vice versa). You could well end up in a situation where you can only offload net a few hundred kg of propellant from a chuck-a-cubesat-out-the-side mission after you account for the propellant used to move between the depot orbit and the deployment orbit (and/or move the depot from its previous orbit to a matching one).
I could see a Transporter-esque service where you can fly your payload for cheap if and only if you are happy being dropped off into an orbit a low energy burn away from a depot orbit, but not a general load-prop-on-every-launch setup.

[DanClemmensen]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That depends on the orbit of the depot and the target orbit of the payload. Regardless of payload mass you need to cart ~100 tonnes of Starship dry mass (plus a few thousand tonnes of prop) around for every burn, so plane changes would become extremely expensive (regardless of whether you moved Starship to match the depot or vice versa). You could well end up in a situation where you can only offload net a few hundred kg of propellant from a chuck-a-cubesat-out-the-side mission after you account for the propellant used to move between the depot orbit and the deployment orbit (and/or move the depot from its previous orbit to a matching one).
I could see a Transporter-esque service where you can fly your payload for cheap if and only if you are happy being dropped off into an orbit a low energy burn away from a depot orbit, but not a general load-prop-on-every-launch setup.

I have not been following the transporter business. Is there a small set of customer-preferred planes? If so, put a depot in each plane. Now the problem is how to move the depots to where they will be needed after they are filled.

[edzieba]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That depends on the orbit of the depot and the target orbit of the payload. Regardless of payload mass you need to cart ~100 tonnes of Starship dry mass (plus a few thousand tonnes of prop) around for every burn, so plane changes would become extremely expensive (regardless of whether you moved Starship to match the depot or vice versa). You could well end up in a situation where you can only offload net a few hundred kg of propellant from a chuck-a-cubesat-out-the-side mission after you account for the propellant used to move between the depot orbit and the deployment orbit (and/or move the depot from its previous orbit to a matching one).
I could see a Transporter-esque service where you can fly your payload for cheap if and only if you are happy being dropped off into an orbit a low energy burn away from a depot orbit, but not a general load-prop-on-every-launch setup.

I have not been following the transporter business. Is there a small set of customer-preferred planes? If so, put a depot in each plane. Now the problem is how to move the depots to where they will be needed after they are filled.

Transporter launches fly to dropoff orbits near Starlink deployment orbits, plus whatever orbits can be reached with one or two additional burns (usually only for headline customers paying extra) e.g. an apogee raise and circularisation.

[volker2020]

I found another part of this article more interesting:
SpaceX plans to keep its low-Earth orbit propellant depots topped off with fuel for missions other than Artemis,

I see that as confirmation, that SpaceX will have many propellant depots in different planes, and will always have a easily accessible depot in reach, regardless when they start.

[InterestedEngineer]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That's slick.

It's a good idea, but alas, SpaceX only makes about 20-24 non-SpaceX non-NASA launches per year, and at 2 / month that's maybe enough to top off from venting related to cooling but not enough to stock a depot.

https://en.wikipedia.org/wiki/List_of_Falcon_9_and_Falcon_Heavy_launches#2022_2

[Robotbeat]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That's slick.

It's a good idea, but alas, SpaceX only makes about 20-24 non-SpaceX non-NASA launches per year, and at 2 / month that's maybe enough to top off from venting related to cooling but not enough to stock a depot.

https://en.wikipedia.org/wiki/List_of_Falcon_9_and_Falcon_Heavy_launches#2022_2

Who says they can’t be SpaceX or NASA launches? Starlink may be volume constrained as Starship performance is dialed in, plus they may want to limit the number of Starlink satellites per launch because of plane constraints/etc.

And 20-24 launches per year is CURRENT rate, and that number has increased massively over time and no reason to think it won’t grow in the future. 20-24 is also a pretty large number. Extra propellant from them alone may be enough for an Artemis mission.

[Robotbeat]

The most obvious candidate for launches with extra propellant is smallsat rideshares. The destination is LEO (granted, often at higher inclinations than you might like), the total payload is very small, and to be competitive it has to occur regularly even if demand is relatively low and uneven. These customers are very cost conscious (or why not get a dedicated launch?), and there should be many opportunities.

Smallsat rideshare on tanker flights is a pretty obvious idea.

[InterestedEngineer]

Pretty interesting. So they’d be able to use really lightweight missions that would normally launch on Falcon 9 (or even Falcon 1), offer it for basically the marginal cost of launch, then the “profit” would be propellant they can load into a depot for other missions.

That's slick.

It's a good idea, but alas, SpaceX only makes about 20-24 non-SpaceX non-NASA launches per year, and at 2 / month that's maybe enough to top off from venting related to cooling but not enough to stock a depot.

https://en.wikipedia.org/wiki/List_of_Falcon_9_and_Falcon_Heavy_launches#2022_2

Who says they can’t be SpaceX or NASA launches? Starlink may be volume constrained as Starship performance is dialed in, plus they may want to limit the number of Starlink satellites per launch because of plane constraints/etc.

And 20-24 launches per year is CURRENT rate, and that number has increased massively over time and no reason to think it won’t grow in the future. 20-24 is also a pretty large number. Extra propellant from them alone may be enough for an Artemis mission.

24 launches *per year*.  I doubt unused prop is going to sit in orbit for an average of 4-5 months, which at 2/month is how long it will take to fill a depot with 20t payload flights that happen to have 120t of fuel with them.   What's the prop boiling rate again?   I suspect a depot turnover time will need to be less than a month.

The non-NASA non-SpaceX flights haven't been growing.  The existing space market is dominated by very expensive low volume satellites, and Falcon-9 has already lowered their launch costs to rounding error (the insurance probably costs more than the expendable 2nd stage for example)

Falcon-9 is picking up contracts because of schedule considerations, not cost, if I read the ESA wins correctly.

[InterestedEngineer]

Have we reached a consensus on what a depot turnover time might look like?

IOTW, how long can a depot sit in LEO with 1200t of fuel before it needs to be topped off with a 150t refueling due to fuel evaporation and its venting?

My guess is 1-3 weeks, not months.   But this thread is very long, so I might have missed something.

[OTV Booster]

Have we reached a consensus on what a depot turnover time might look like?

IOTW, how long can a depot sit in LEO with 1200t of fuel before it needs to be topped off with a 150t refueling due to fuel evaporation and its venting?

My guess is 1-3 weeks, not months.   But this thread is very long, so I might have missed something.

IIRC, real world boil off rates are one of the boxes yet to be checked. And what mitigation might be used is an unknown. There would also be unknown losses from RCS. It's hard to get a consensus when all of the important numbers are unknown. But then, I'm constantly impressed by the number of people who have strong opinions on things they have no understanding of - and equate opinions with knowledge. Thankfully not so much here at NSF.


Side note: I've suggested CMG's or reaction wheels for the depot in the past with no traction. In theory active cooling can achieve zero boiloff but there would still be RCS losses.

[Robotbeat]

Cryo cooled space telescopes (some in LEO) have liquid helium (sometimes even superfluid, all the way to 2 Kelvin) that lasts for a year to years. The boiloff rate is just gonna depend on what SpaceX decides is acceptable. There’s no reason they couldn’t make it last for months between topping off, even in LEO.

[Robotbeat]

Boiloff gases can and almost certainly will be used to assist stationkeeping.

[edzieba]

Cryo cooled space telescopes (some in LEO) have liquid helium (sometimes even superfluid, all the way to 2 Kelvin) that lasts for a year to years. The boiloff rate is just gonna depend on what SpaceX decides is acceptable. There’s no reason they couldn’t make it last for months between topping off, even in LEO.

Those telescopes are also designed and built around the Helium dewar, and have to deal with a lot of work in eliminating thermal bridges to approach that performance.

[Robotbeat]

Cryo cooled space telescopes (some in LEO) have liquid helium (sometimes even superfluid, all the way to 2 Kelvin) that lasts for a year to years. The boiloff rate is just gonna depend on what SpaceX decides is acceptable. There’s no reason they couldn’t make it last for months between topping off, even in LEO.

Those telescopes are also designed and built around the Helium dewar, and have to deal with a lot of work in eliminating thermal bridges to approach that performance.

Of course. It takes aggressive engineering to get a very low boiloff rate. This is why I said the boiloff rate will be set to whatever SpaceX can tolerate.

[OTV Booster]

Boiloff gases can and almost certainly will be used to assist stationkeeping.

Yup. Boosting orbit a tad takes propellant in one form or another. Controlling attitude is another thing. I've no idea how much would be lost to this. If it's significant then CMGs or reaction wheels start to make sense. The higher the loss and the longer the expected depot lifetime, the more sense.

[volker2020]

Cryo cooled space telescopes (some in LEO) have liquid helium (sometimes even superfluid, all the way to 2 Kelvin) that lasts for a year to years. The boiloff rate is just gonna depend on what SpaceX decides is acceptable. There’s no reason they couldn’t make it last for months between topping off, even in LEO.

Those telescopes are also designed and built around the Helium dewar, and have to deal with a lot of work in eliminating thermal bridges to approach that performance.

Of course. It takes aggressive engineering to get a very low boiloff rate. This is why I said the boiloff rate will be set to whatever SpaceX can tolerate.

Yes, it should be self evident that SpaceX can add as much isolation as they feel fit, they could add reflectors that shield the ship from heat radiation and or the heat radiation from earth, worst case they could even install a cooler (strongly doubt they would need it). There is no reason, why a tanker need to be full, so basically they could use 100t for extra equipment if they wanted. The real question is, what level of cooling makes economical sense.

[Nomadd]

 I hadn't seen this before. Is it in one of the other threads?
 The depot looks big enough for two starship loads.
 *Teslarati*

[Twark_Main]

what level of cooling makes economical sense.

One (incomplete) datapoint: cryocooler mass.

This PDF gives a scaling law on page 14, which simplifies to:

mass = 0.187195 kg/W * (cooling power) + 21.1033 kg

I believe this does not include the radiator, but it's hard to tell.

[Twark_Main]

I hadn't seen this before. Is it in one of the other threads?
 The depot looks big enough for two starship loads.
 *Teslarati*

https://forum.nasaspaceflight.com/index.php?topic=50157.msg2406262#msg2406262

Depot seems to be roughly 61.5 meters tall.