criogenic storage in space

[king of nowhere]

i often hear you can't use criogenic fuels for probes because of boiloff. so you have to use hydrazine, which has much lower Isp.

nobody mentions that you can make a criogenic tank that does not lose stuff over time. you just need to have your own internal cooling system that will keep the fuel cold enough to prevent boiloff. on earth it's rarely done, because most criogenics are dirt cheap anyway (nitrogen, methane) and it's cheaper to just lose some over time. at least, i think that's the reason. i saw that a criogenic refrigerator can come as cheap as a few thousand euros.

in space, that cost is clearly not a problem. mass, on the other hand, could be an issue. but i don't have the numbers to make the calculations. those refrigerators i saw looked like they were small enough, could be a few tens of kilograms, though i could not find exact data (most people who buy a criogenic refrigerator doesn't really care how heavy it is).

but it seems to me, adding a few tens of kilograms of dry mass should be abundantly compensated by increasing Isp by 30% or more.

does anyone know the actual math of why it's not done for space probes?

 

by the way, starship wants to use methane on mars. i suppose, to keep it cool for the nine months trip, they will use this system?

[StrandedonEarth]

Refrigeration requires power. It also requires somewhere to dump waste heat. Those systems also add mass. It may be easier to and less massy to add enough insulation and sunshielding to reduce boiloff acceptable levels. 

ULA was looking into it with their ACES system, which would have burned propellant for power. It’s currently shelved, however. 

[DDE]

31 minutes ago, king of nowhere said:

in space, that cost is clearly not a problem. mass, on the other hand, could be an issue. but i don't have the numbers to make the calculations. those refrigerators i saw looked like they were small enough, could be a few tens of kilograms, though i could not find exact data (most people who buy a criogenic refrigerator doesn't really care how heavy it is).

It was first done back before the golden age of planetary exploration. The problem is lOx, not methane - a propulsion bus is going to trend towards equal temperature across both tanks, and the equilibrium would be disfavorable. Methane and ethane were still top picks, but alternative oxidizers were explored - all of them involving fluorine.

[darthgently]

48 minutes ago, king of nowhere said:

i often hear you can't use criogenic fuels for probes because of boiloff. so you have to use hydrazine, which has much lower Isp.

nobody mentions that you can make a criogenic tank that does not lose stuff over time. you just need to have your own internal cooling system that will keep the fuel cold enough to prevent boiloff. on earth it's rarely done, because most criogenics are dirt cheap anyway (nitrogen, methane) and it's cheaper to just lose some over time. at least, i think that's the reason. i saw that a criogenic refrigerator can come as cheap as a few thousand euros.

in space, that cost is clearly not a problem. mass, on the other hand, could be an issue. but i don't have the numbers to make the calculations. those refrigerators i saw looked like they were small enough, could be a few tens of kilograms, though i could not find exact data (most people who buy a criogenic refrigerator doesn't really care how heavy it is).

but it seems to me, adding a few tens of kilograms of dry mass should be abundantly compensated by increasing Isp by 30% or more.

does anyone know the actual math of why it's not done for space probes?

 

by the way, starship wants to use methane on mars. i suppose, to keep it cool for the nine months trip, they will use this system?

The heavy part would be the radiators to get rid of the heat from the cooler.  Think something on the scale of ISS's PVs.  I'm ballpark guessing here, of course

Edited May 16 by darthgently

[king of nowhere]

1 hour ago, StrandedonEarth said:

Refrigeration requires power. It also requires somewhere to dump waste heat. Those systems also add mass. It may be easier to and less massy to add enough insulation and sunshielding to reduce boiloff acceptable levels. 

ULA was looking into it with their ACES system, which would have burned propellant for power. It’s currently shelved, however. 

 

1 hour ago, darthgently said:

The heavy part would be the radiators to get rid of the heat from the cooler.  Think something on the scale of ISS's PVs.  I'm ballpark guessing here, of course

if it's possible to insulate the tanks enough to reduce boiloff to acceptable levels - and it is, the apollo missions used crio tanks for hydrogen that would have lasted years - then you need to remove very little heat from the tanks, a power output lower than a watt.  you need little to no ratiator surface for that.

this also should address the power issue; it would require extremely little power to keep cool a small, well insulated tank in space.

[StrandedonEarth]

2 hours ago, king of nowhere said:

by the way, starship wants to use methane on mars. i suppose, to keep it cool for the nine months trip, they will use this system?

They said that the header tanks will hold enough propellants for landing on Mars. Seeing as the header tanks act as a thermos, they figure boil off should be acceptable 

[king of nowhere]

6 minutes ago, StrandedonEarth said:

They said that the header tanks will hold enough propellants for landing on Mars. Seeing as the header tanks act as a thermos, they figure boil off should be acceptable 

interesting. but it leaves me even more curious as to why not include a criogenic system - as starship is big and heavy enough that the mass of one should be negligible.

 

in general, while there are some practical problems with installing criogenic cooling units, they don't seeem nowhere near dire enough to justify accepting losses of fuel - or switching to a fuel that gives 30% less deltaV.

I got the same issue with criogenic tanks on earth. i've been researching the topic because i am a teacher and i explained storage tanks to my students last month. and for all that i tried, and i found plenty of references to acceptable boiloff, i didn't find a single source even mentioning the idea of providing internal cooling. in the case of criogenic cooling on earth, clearly replenishing more material is not an issue, but i found multiple documents stating that losses of material are 0.3 to 3% of the tank content per day - which, on a large tank, means several tons per day. And I looked the cost of criogenic units, and I saw that you can buy a refrigerator that can reach -200° C for little more than 3000 $. And while I saw dozens of documents saying "losses are acceptable", I didn't see a single mention of the question "but why not avoid losses entirely?".  nowhere does it say "preventing those losses by internal cooling would be too expensive" or "would entail too many practical problems". Everything I can find suggest that preventing those losses would be rather cheap - though i am unable to find solid data on that.

I was just hoping someone could give me some answers on the actual tradeoff of accepting or preventing boiloff besides the ubiquitous "boiloff is acceptable".

[AckSed]

3 hours ago, king of nowhere said:

Does anyone know the actual math of why it's not done for space probes?

Less maths, more practicality and some inertia.

MMH/Hydrazine, and Hydrazine as mono-propellant, have the advantage of reliability and simplicity. They can take severe swings in temperature with just insulation protecting them. Building on research done for ICBMs, they can be stored for years to decades in tanks before being mixed in the engine, and not do anything exciting. Bladders and/or pistons in the tanks can pressurise the propellants in zero gravity. The engines are often pressure-fed, either combining the two fuels or feeding one over a catalyst bed, which makes them, again, simple and reliable. Specific impulse is not the main priority here.

Cryogenic propellants are trickier. Orbital propellant depots must deal with the same issues as a deep-space probe, and this paper outlines the somewhat truncated state of the art: https://www.nature.com/articles/s41526-024-00377-5.pdf

In short, keeping heat out isn't simple because keeping cryogenic liquids floating around in zero-G from boiling isn't simple either. To reliably cool a zero-G liquid, it must be touching the cooler. Materials for bladders/pistons that can stay flexible in such temperatures aren't common. Or researched. Further, most of the research was conducted on the ground.

However, Intuitive Machines' Nova-C probe was methalox and used cryogenic propellants. When it landed on the Moon in February, it broke a leg and tipped over, but through no fault of the engines or fuel. So it has been done.

If you want to use  cryogenic, methane is not a bad bet. When shaded and insulated with MLI, it can remain at a reasonable temperature for months.

...The real issue is that orbital propellant depots, and thus long-term cryogenic propellant storage, were either paid lip-service or downright ignored by certain elements of NASA. Here's an article from 2008: https://www.thespacereview.com/article/1127/1

Quote

A bigger challenge than the technology, though, might be to get NASA and others to adopt the concept of propellant depots. While NASA administrator Mike Griffin has been open to the concept, suggesting in public speeches that NASA would be willing to purchase services from commercial fuel depots, right now ESAS doesn’t depend on the concept. That, said Bienhoff, who has briefed a number of NASA officials on his proposals, is an obstacle to gaining acceptance of the concept within NASA. “They’re bound by the architecture, and they can’t spend any money on it because it’s not in the architecture.”

Edit: Then a little later, when SLS was in its infancy: https://arstechnica.com/science/2019/08/rocket-scientist-says-that-boeing-squelched-work-on-propellant-depots/

Quote

"We had released a series of papers [in 2011] showing how a depot/refueling architecture would enable a human exploration program using existing (at the time) commercial rockets," Sowers tweeted on Wednesday. "Boeing became furious and tried to get me fired. Kudos to my CEO for protecting me. But we were banned from even saying the 'd' word out loud. Sad part is that ULA did a lot of pathfinding work in that area and could have owned the refueling/depot market, enriching Boeing (and Lockheed) in the process. But it was shut down because it threatened SLS."

 

Edited May 16 by AckSed
More links

[darthgently]

59 minutes ago, king of nowhere said:

 

if it's possible to insulate the tanks enough to reduce boiloff to acceptable levels - and it is, the apollo missions used crio tanks for hydrogen that would have lasted years - then you need to remove very little heat from the tanks, a power output lower than a watt.  you need little to no ratiator surface for that.

this also should address the power issue; it would require extremely little power to keep cool a small, well insulated tank in space.

Then why is NASA scratching their collective head over it for decades?  A mystery if it is so easy

It is actually very hot in direct sunlight in space with no convection to get rid of heat.  That insulation heats up too

[PakledHostage]

There was this article about 6 months ago about a potential new technological solution to the problem of boiloff: 

https://www.nasa.gov/general/electro-luminescently-cooled-zero-boil-off-propellant-depots/

[AckSed]

28 minutes ago, PakledHostage said:

There was this article about 6 months ago about a potential new technological solution to the problem of boiloff: 

https://www.nasa.gov/general/electro-luminescently-cooled-zero-boil-off-propellant-depots/

With Gateway and Starship HLS on the way, NASA has been forced to swim, when previously they were testing the waters with the tip of their big toe and cringing. Let's hope something comes of it.

[mikegarrison]

An issue specifically with hydrogen is that it is just difficult to store hydrogen at all. Not only does liquid hydrogen need to be kept at 20-30 Kelvin, but hydrogen is of course the smallest of all atoms, and it really likes to escape straight through the walls it is being contained in. Not only that, but it loves to nestle into the crystal structure of metals, causing hydrogen embrittlement. 

[king of nowhere]

8 hours ago, darthgently said:

Then why is NASA scratching their collective head over it for decades?  A mystery if it is so easy

It is actually very hot in direct sunlight in space with no convection to get rid of heat.  That insulation heats up too

I realize there must be practical reasons if it's not done. I am inquiring about those practical reasons - because i know enough of technology to not be satisfied of just "there are problems", but i don't know enough to know what those problems are - and most hypothesis i can make on my own seem like they could easily be solved. 

So, your sarcasm is unnecessary and uncalled for. Similarly, a superficial answer like "in space it gets hot" does not help; sure, in space it gets hot, and it gets cold, and we already  have systems in place to protect delicate instruments from that, and you are telling me we can do that but we can't handle a highly insulated tank? If that's the case, i'd at least want a more detailed explanation on the why and how.

This forum is the only place i know where i can make highly technical questions and hope for people to give good answers. Many have done so; i'm going to pour over those papers as soon as i have more time.  

I wasn't expecting to get snide remarks and not-so-subtle insults, though. Nor did i expect such an attitude to actually get upvoted

[king of nowhere]

10 hours ago, AckSed said:

Cryogenic propellants are trickier. Orbital propellant depots must deal with the same issues as a deep-space probe, and this paper outlines the somewhat truncated state of the art: https://www.nature.com/articles/s41526-024-00377-5.pdf

 

while I thank everyone who contributed with hard data, this particular paper was most informative. I regret that I only have one upvote to give you for it.

so, while cryocooling wasn't the main topic, there were experiments made. the values of thermal transfer given with the use of MLI, around 1 W/m², are consistent with a small probe only needing a few W of cooling power. With a cryocooler efficiency of 5 to 10%, which is pretty reasonable considering the theoretical carnot efficiency for the temperatures involved, a good guesstimate is several tens of watts, up to 100. that's within the capacity of probes like New Horizons or Juno, but it would place a significant strain on their power generation. For starship, the power generation needed would be in the tens or hundreds of KW - could be inconvenient, since the solar panels need to be retracted and protected during aerobraking on mars.

the main issue, anyway, is indeed the way the liquid form bubbles that interfere with heat transfer. several experiments were attempted with bringing cryocoolers in space, some were successful, and it was demonstrated a prolonged storage of LOx with zero boiloff, but some failed due to heat transfer issues. sloshing is also an issue; with MMH and NTO, the problem is avoided by using internal bladders and pistons that prevent sloshing, but such solutions would not work in cryogenic conditions.

it seems the system is potentially workable, and it is being investigated. However, the specific issues are caused by the zero-g environment, and fixing them requires experiments in zero g, which are very expensive to make. confronted with limited budget, it is preferred to stick to older, less efficient, but tested and reliable solutions.

this is turning out to be a surprisingly common answer to a lot of things about space. many practical problems trequire to think a solution, launch it to space, test whether it works, tryto make some sense from the data coming from the instruments, think another solution, wait to launch that in space too... on earth, a team of engineers could fix it with a few weeks of trial and error. but needing to launch every single iteration in space - much less being unable to put your hands on it and having to rely on limited instrumental data to figure out what's going on - increases costs and time exorbitantly.

 

it still does not answer my question on why cryocooling isn't employed on earth. it seems that taking air, cooling it, extracting nitrogen through fractional distillation, and transporting it to the factory to compensate for boiloff is considered cheaper than just keeping the nitrogen cool inside the tank. this baffles me. Still, a potential answer also lies in that paper, when it suggests the main issues of venting are during fuel transfers. so it is possible that the greatest losses to boiloff are not caused by heat transfer, but by fuel transfer - which, in a factory using liquid nitrogen, happens multiple times per day. it is possible the figure for 1% daily losses include those for fuel trasfer, while heat transfer losses are a lot lower. i would still like to hear about it from an expert, but it's no longer related to the field of space travel, and it doesn't belong here.

Edited May 17 by king of nowhere

[darthgently]

2 hours ago, king of nowhere said:

I wasn't expecting to get snide remarks and not-so-subtle insults, though. Nor did i expect such an attitude to actually get upvoted

I certainly didn't intend to be snide or sarcastic, just to challenge a basic premise in a hopefully humorous way.  I'm certain I issued zero insults.  None. 

It is a good topic.  My thoughts on it were to store water and a carbon source at non-cryo temps and use solar power to make fuel as needed rather than storing it in cryo for years.  In other words use your energy to make fuel on demand instead of using it to keep it cold.  Given enough time a crossover point would be reached where just-in-time generation beats cryo storage in energy use.

3 hours ago, mikegarrison said:

An issue specifically with hydrogen is that it is just difficult to store hydrogen at all. Not only does liquid hydrogen need to be kept at 20-30 Kelvin, but hydrogen is of course the smallest of all atoms, and it really likes to escape straight through the walls it is being contained in. Not only that, but it loves to nestle into the crystal structure of metals, causing hydrogen embrittlement. 

And hydrogen embrittlement over time also

Edited May 17 by darthgently

[darthgently]

Another factor is that it isn't just direct sunlight to be dealt with in earth orbit, but I gather earthshine can add a lot of heat also so both sides of craft are affected

[AckSed]

2 hours ago, king of nowhere said:

It still does not answer my question on why cryocooling isn't employed on earth. It seems that taking air, cooling it, extracting nitrogen through fractional distillation, and transporting it to the factory to compensate for boiloff is considered cheaper than just keeping the nitrogen cool inside the tank. This baffles me. Still, a potential answer also lies in that paper, when it suggests the main issues of venting are during fuel transfers. so it is possible that the greatest losses to boiloff are not caused by heat transfer, but by fuel transfer - which, in a factory using liquid nitrogen, happens multiple times per day. It is possible the figure for 1% daily losses include those for fuel transfer, while heat transfer losses are a lot lower. i would still like to hear about it from an expert, but it's no longer related to the field of space travel, and it doesn't belong here.

I'm sorry, we saw "Space" and immediately climbed over ourselves to explain. But this is also the Science part of this forum, so this amateur librarian will take a stab.

As for actually doing it on the ground, cryocooling is used, but for things that you have to keep cold like cryopreserved people. Even then, it is actually cheaper to either accept bulk delivery and any boil-off, or (according to manufacturers of nitrogen condensers) generate it yourself on-site.

Here's a leaflet outlining how liquid nitrogen is utilised in an IVF lab. It's quite involved.

And here's someone who built his own LN2 generator from - yes - a cryocooler: https://benkrasnow.blogspot.com/2008/08/diy-liquid-nitrogen-generator.html

Edited May 17 by AckSed
Link

[king of nowhere]

1 hour ago, darthgently said:

I certainly didn't intend to be snide or sarcastic, just to challenge a basic premise in a hopefully humorous way.  I'm certain I issued zero insults.  None. 

 

 

I see. glad the misunderstanging was cleared.

Quote

It is a good topic.  My thoughts on it were to store water and a carbon source at non-cryo temps and use solar power to make fuel as needed rather than storing it in cryo for years.  In other words use your energy to make fuel on demand instead of using it to keep it cold.  Given enough time a crossover point would be reached where just-in-time generation beats cryo storage in energy use.

the main problem is that it takes a lot of energy to turn water into fuel. a probe would need to spend weeks in preparation, making fuel for a burn. and that fuel would still need to be stored.

cryo storage requires more energy in the long term, but it does require a small, continuous amount of energy. solar panels or rtgs produce a small, continuous amount of power, so they are better suited at keeping a supercooled tank rather than operating a high-powered electrolysis device for a short time.

I can see the idea working for some specific mission that require lots of relatively short burns, separated by long time intervals. but it would be a special case.

 

[darthgently]

1 hour ago, king of nowhere said:

I can see the idea working for some specific mission that require lots of relatively short burns, separated by long time intervals. but it would be a special case.

 

For most probes, is it a special case? Honestly not sure, but given most deep space probes do multiple grav assists and burns spaced years apart I wonder

1 hour ago, king of nowhere said:

I see. glad the misunderstanging was cleared.

Yeah.  Iron sharpens iron.  Velvet  not sufficient 

Edited May 17 by darthgently

[kerbiloid]

Hydrogen consists of protons and beta-particles, so it's a pure radiation, contained in a tank.
Its penetration through the tank walls is slowed down by the low temperature.

[king of nowhere]

4 hours ago, darthgently said:

For most probes, is it a special case? Honestly not sure, but given most deep space probes do multiple grav assists and burns spaced years apart I wonder

 

all burns must be relatively short, which is not always the case. if you have a big burn along the way, like for example a jupiter capture burn, then you need to accumulate lots of fuel for it, you need to start the electrolysis weeks earlier. then you need ways to store your oxygen and methane for weeks, which basically forces you to use a criogenic tank anyway