Starship On-orbit refueling - Options and Discussion

[Jim]

the boil off would be used for fuel cells or turbo generator.

[InterestedEngineer]

Using 15kW as the worst case, but at 24hr/day. (Ie, 15kJ/second.) And a heat-of-vaporisation figure for methane of 511 kJ/kg, that means boil-off-only cooling would cost 2.5 tonnes of methane per day.

If your refuelling mission cycle takes 30 days from empty, through accumulation, to final target-vehicle transfer and empty again, you lose 75 tonnes of methane. I don't think it's reasonable to assume a faster cycle for a first generation depot.

Heat of vaporization:  511kJ/kg
Heat of combustion:  56MJ/kg

Even at 10% efficiency cooling by burning methane is 10x more efficient than evaporating it.

[Nevyn72]

Using 15kW as the worst case, but at 24hr/day. (Ie, 15kJ/second.) And a heat-of-vaporisation figure for methane of 511 kJ/kg, that means boil-off-only cooling would cost 2.5 tonnes of methane per day.

If your refuelling mission cycle takes 30 days from empty, through accumulation, to final target-vehicle transfer and empty again, you lose 75 tonnes of methane. I don't think it's reasonable to assume a faster cycle for a first generation depot.

Heat of vaporization:  511kJ/kg
Heat of combustion:  56MJ/kg

Even at 10% efficiency cooling by burning methane is 10x more efficient than evaporating it.

Excellent, then as Jim says, burn the boil-off and get both!

[Nomadd]

 If they go with a depot and sunshade, has anybody here figured how much an issue earthshine would be?

[TheRadicalModerate]

It's going to be complicated because the angle of the ship to Earth isn't constant, but rotates 360 degrees each orbit if the nose points to the sun.

According to the view factor plot I included above, when h = H/R (H is the orbital radius, R is the Earth's radius) approaches 1, the angle becomes less relevant and everything approaches 0.45.  In a 250km VLEO, h=1.04.  This seems non-intuitive to me, but I trust whoever did that plot a lot more than I trust my intuition on this subject.

That would reduce the problem to figuring out a reasonable average albedo and specular flux from an entire Earth orbit, then using that as the basis for an approximation of the heating.

[TheRadicalModerate]

Using 15kW as the worst case, but at 24hr/day. (Ie, 15kJ/second.) And a heat-of-vaporisation figure for methane of 511 kJ/kg, that means boil-off-only cooling would cost 2.5 tonnes of methane per day.

If your refuelling mission cycle takes 30 days from empty, through accumulation, to final target-vehicle transfer and empty again, you lose 75 tonnes of methane. I don't think it's reasonable to assume a faster cycle for a first generation depot.

Heat of vaporization:  511kJ/kg
Heat of combustion:  56MJ/kg

Even at 10% efficiency cooling by burning methane is 10x more efficient than evaporating it.

I'd think that you'd let it evaporate, then burn it.  That way you get both. 

But if you're working off of evaporated stuff, then you probably need to factor in battery charging efficiency as well.  I'd think that you need to run a cryocooler for a while to get it chilled down and operating at peak (albeit lousy) efficiency.  You can't do that if you're only running your APU every time the tanks need to vent, so you'd have to dump the energy into the battery until you have enough for an efficient cryocooler run.

[Greg Hullender]

I looked online for into on methane-burning fuel cells, and everything I see suggests that they're a) experimental and b) run at 600° C or more. And I think you're still going to need a sizeable radiator.

[InterestedEngineer]

I looked online for into on methane-burning fuel cells, and everything I see suggests that they're a) experimental and b) run at 600° C or more. And I think you're still going to need a sizeable radiator.

Methane turbines are around us and generate a large part of our power, power backup for homes, etc.  Far better than 10% efficiency.

I'm curious if one could find one that would work in space and use oxygen instead of air.

[BT52]

You mean using Raptor powerhead as generator? Doable. I m sure unnecessarily tho. I would like to keep propellant passive as much as possible. Well at least how long this prop will be in orbit? 1-2days max? So any advanced sys is unnecessarily complication. I would even argue they would make slightly thicker SS skin. If was necessarily to raise operational pressure and skin stresses.

[sebk]

Using 15kW as the worst case, but at 24hr/day. (Ie, 15kJ/second.) And a heat-of-vaporisation figure for methane of 511 kJ/kg, that means boil-off-only cooling would cost 2.5 tonnes of methane per day.

If your refuelling mission cycle takes 30 days from empty, through accumulation, to final target-vehicle transfer and empty again, you lose 75 tonnes of methane. I don't think it's reasonable to assume a faster cycle for a first generation depot.

Heat of vaporization:  511kJ/kg
Heat of combustion:  56MJ/kg

Even at 10% efficiency cooling by burning methane is 10x more efficient than evaporating it.

You must include the mass of oxygen which is ~4x bigger. Also you should include oxygen heat of vaporization..

[Paul451]

So, stainless towards the earth it is.

There's no reason to have bare stainless steel. Even if you intend to re-enter the depot-ships, you can paint the dorsal side with something reflective.

It's all reasonable except the paint gets toasted during EDL. Painting's easy. Prep is the killer. That ship will never relaunch as a depot without a lot of prep.

Cleaning stainless steel and reapplying some paint doesn't seem like it could possibly cost more than building a new depot. Given the difference it makes in thermal input, whether that energy is removed via boiloff or cryo-cooling, it's worth the extra hassle even on recovered/reused depot-ships.

[OTV Booster]

Using 15kW as the worst case, but at 24hr/day. (Ie, 15kJ/second.) And a heat-of-vaporisation figure for methane of 511 kJ/kg, that means boil-off-only cooling would cost 2.5 tonnes of methane per day.

If your refuelling mission cycle takes 30 days from empty, through accumulation, to final target-vehicle transfer and empty again, you lose 75 tonnes of methane. I don't think it's reasonable to assume a faster cycle for a first generation depot.

Heat of vaporization:  511kJ/kg
Heat of combustion:  56MJ/kg

Even at 10% efficiency cooling by burning methane is 10x more efficient than evaporating it.

I'd think that you'd let it evaporate, then burn it.  That way you get both. 

But if you're working off of evaporated stuff, then you probably need to factor in battery charging efficiency as well.  I'd think that you need to run a cryocooler for a while to get it chilled down and operating at peak (albeit lousy) efficiency.  You can't do that if you're only running your APU every time the tanks need to vent, so you'd have to dump the energy into the battery until you have enough for an efficient cryocooler run.

Or, run a smaller cooler at a higher duty cycle. Unknowns: do cryo coolers gain efficiency with size? Is it acceptable to allow random subcooling throughout the duty cycle?

[OTV Booster]

I looked online for into on methane-burning fuel cells, and everything I see suggests that they're a) experimental and b) run at 600° C or more. And I think you're still going to need a sizeable radiator.

Methane turbines are around us and generate a large part of our power, power backup for homes, etc.  Far better than 10% efficiency.

I'm curious if one could find one that would work in space and use oxygen instead of air.

Turbine engines do not scale down well. Bigger = more efficient.


An idea I've put out is for a constant rpm constant load two stroke. Liquid propellant used for cooling and combustion.  Inexpensive, stone simple to maintain. When run off of liquid propellant the compression stroke has, in effect, already been done. Should be crazy efficient at small scale.


For those with minimal experience with piston ICE's, with constant rpm and load the complexity of 'breathing' and mixing are simplified and efficiency can approach the theoretical Carnot limits. Get rid of the compression stroke, add in the perfectly characterized cooling requirements, and it's better yet.


The major trade to look at is tapping unvaporized propellant vs increased efficiency. I noodled feeding it with boiloff but the cooling gets tricky and the compression stroke shows up again. Maybe gas fed can work, maybe not.


Being a simple easily maintained system, it's good for Mars use too.

[OTV Booster]

So, stainless towards the earth it is.

There's no reason to have bare stainless steel. Even if you intend to re-enter the depot-ships, you can paint the dorsal side with something reflective.

It's all reasonable except the paint gets toasted during EDL. Painting's easy. Prep is the killer. That ship will never relaunch as a depot without a lot of prep.

Cleaning stainless steel and reapplying some paint doesn't seem like it could possibly cost more than building a new depot. Given the difference it makes in thermal input, whether that energy is removed via boiloff or cryo-cooling, it's worth the extra hassle even on recovered/reused depot-ships.

I'm in the middle of deck rail and trim painting. If I let you put a sharp stick in my eye would you do it for me?


More seriously, if we're talking about a tanker with some add-ons, why bother? Let it show its heritage and let us wonks be happy. Swap in a new, or as yet unpainted tanker for the next accumulator. If it's a Depot Heavy, do a refurb.


Edit to add: what I've been reading about solar white paint mentions thickness from 3-10mm. It's more of a coating than paint. Might be a pain to remove and re-prep.


Edit 2: The paint substrate material and surface texture are an inherent part of the thermal performance. A wire wheel will most likely remove the coating and not hurt the stainless but it would alter the surface texture. Maybe to the good, maybe not.


In small scale tests investigating Y2O3, an aluminum substrate was found to add good rejection properties. Stainless had the poorest rejection of the tested substrates. My knowledge of aluminum coating is limited to vapor deposition onto glass in a good vacuum (3x10^-5mm of mercury). I'm not aware of any electroplating or chemical deposit techniques but I am widely renowned for being ignorant.    I wonder how chrome works. A little work on the fins and 1960-70's car vibes would rule.


Repreping the surfaces looks a bit more complicated than preping my deck trim.

[TheRadicalModerate]

Edit to add: what I've been reading about solar white paint mentions thickness from 3-10mm. It's more of a coating than paint. Might be a pain to remove and re-prep.

Some of the solar white work uses a silvered backing to reflect as much IR as possible.  That would be a huge pain to reapply on a reusable tanker or depot.  I'm not sure how much that contributes to the the overall reflectance.  Probably doesn't have much to do with the emissivity.

[Greg Hullender]

Edit to add: what I've been reading about solar white paint mentions thickness from 3-10mm. It's more of a coating than paint. Might be a pain to remove and re-prep.

Some of the solar white work uses a silvered backing to reflect as much IR as possible.  That would be a huge pain to reapply on a reusable tanker or depot.  I'm not sure how much that contributes to the the overall reflectance.  Probably doesn't have much to do with the emissivity.

Fron the papers cited earlier, it looks like you have your choice of paint or ceramic tiles.

The tiles have a silver backing and reflect 99.8% of solar radiation while still being over 90% emissive from 8 microns on down. Perfect for a dedicated depot (or maybe a sun shade), but too fragile for reentry.

The paint can go on any sort of reflective backing, but it's not as reflective. Applied to unpolished stainless steel, they measured it at 90% reflective with ten coats (5 mm total thickness). Maybe worth it just to reduce the size/cost of a cooler. Or to eliminate the need for one entirely, depending on the timeframe. Probably dirt cheap, so no big issue if it doesn't survive reentry.

[OTV Booster]

Edit to add: what I've been reading about solar white paint mentions thickness from 3-10mm. It's more of a coating than paint. Might be a pain to remove and re-prep.

Some of the solar white work uses a silvered backing to reflect as much IR as possible.  That would be a huge pain to reapply on a reusable tanker or depot.  I'm not sure how much that contributes to the the overall reflectance.  Probably doesn't have much to do with the emissivity.

Fron the papers cited earlier, it looks like you have your choice of paint or ceramic tiles.

The tiles have a silver backing and reflect 99.8% of solar radiation while still being over 90% emissive from 8 microns on down. Perfect for a dedicated depot (or maybe a sun shade), but too fragile for reentry.

The paint can go on any sort of reflective backing, but it's not as reflective. Applied to unpolished stainless steel, they measured it at 90% reflective with ten coats (5 mm total thickness). Maybe worth it just to reduce the size/cost of a cooler. Or to eliminate the need for one entirely, depending on the timeframe. Probably dirt cheap, so no big issue if it doesn't survive reentry.

The tiles mentioned were not really intended or explored for EDL. I wonder if anybody's been looking at different fillers. If the heat shield tiles had a Y2O3 glaze they might be able to thin them for lower mass.


Edit to add: Y2O3 is hydrophobic. That's a big thing for the tiles. Not a total solution but a step in the right direction.

[Greg Hullender]

The tiles mentioned were not really intended or explored for EDL. I wonder if anybody's been looking at different fillers. If the heat shield tiles had a Y2O3 glaze they might be able to thin them for lower mass.


Edit to add: Y2O3 is hydrophobic. That's a big thing for the tiles. Not a total solution but a step in the right direction.

I think the trouble is that the heat-shield tiles need to be emissive in the range from about 1 to 8 microns. That rules out coating them with anything like Y2O3. Or anything else that reflects most solar energy, for that matter.

[OTV Booster]

The problems with refolding and stowing the deployables, and the impact of EDL on solar white leads me (again) to the conclusion that while testing may use near plain vanilla tankers as accumulators, an operational depot will most likely never do EDL.

As pointed out by edzeba, the only specific near future use is Artemus. One in LEO or a bit higher, and one in HEO or lunar orbit. The high one is easily a throwaway. The low one could conceivably return if brute force 'launch more tankers' can successfully counter boil off. This last depends on real world understanding of boil off and how long a transfer campaign would take.

If the not yet available numbers reflecting real world experience show that more tankers can't keep up or drags out the campaign too long, boil off mitigation will be done. Everything we've looked at makes successful EDL look chancy. The low ship would then be a throwaway.

This is not necessarily a bad thing. Why cry over a wasted SS? How many have we seen scrapped before a fire was lit under them? How many more will we see? The first Artemus accumulator/depot will be crude compared to what follows.

The follow-on's will explore ways to either design for EDL or for further use in later missions. My gut says the latter. I also expect that a lot of the ones earmarked for further use will quickly become obsolescent and be splashed in - probably before ever being reused. It's what SX does, and it seems to work.

One tanker depot on LEO will align for lunar insertion every 14 days. A second, in a plane 90 degrees to the first opens up an opportunity every 7 days. Correcting the earth exit to insert to any lunar orbital plane that crosses the earth-moon line is relatively trivial if done around L1. An Earth or sun synchronous family is harder. There's a lot of operational flexibility here.

[OTV Booster]

The tiles mentioned were not really intended or explored for EDL. I wonder if anybody's been looking at different fillers. If the heat shield tiles had a Y2O3 glaze they might be able to thin them for lower mass.


Edit to add: Y2O3 is hydrophobic. That's a big thing for the tiles. Not a total solution but a step in the right direction.

I think the trouble is that the heat-shield tiles need to be emissive in the range from about 1 to 8 microns. That rules out coating them with anything like Y2O3. Or anything else that reflects most solar energy, for that matter.

Maybe I'm misunderstanding. See pg3, bottom chart showing a piece of the Y2O3 transmission spectrum. I'd cut n paste but it's not cooperating. [size=78%]https://iopscience.iop.org/article/10.1088/1757-899X/1240/1/012001/pdf[/size]


The chart shows low transmission from about 2 to 10 microns. I've assumed this to mean it readily adsorbs these wavelengths which should also mean it readily emits at these wavelengths. Wrong?


I've tried looking up an emission spectra and hit paywalls and yttria mixed with other things. Also, why 2-10 microns? Isn't this a bit past the far UV? I thought the ideal emission range was in the near IR.