Equipment for Moon Mining Operations are Being Developed

The technology needed for mining water ice on the moon and converting it into fuel is pretty straight forward. Various groups are already making the actual needed hardware. Paragon Space Development and Giner are already making key pieces of what is needed. If we are making large amounts of fuel on the moon then we are massively lowering the cost of all missions in space. The cost of anything from higher earth orbit and beyond becomes several times cheaper.

After the D-day invasion, the Allies made a temporary port. We need to move beyond thinking science missions to working on logistics and supply chains.

Processing Water into Fuel

After water is extracted from the Permanently Shadowed Regions (PSR) of the Moon, it is processed to purify and electrolyze the water into hydrogen (H2) and oxygen (O2). Paragon Space Development Corporation (Paragon) and its partner Giner, Inc. (Giner) are developing the ISRU-derived water purification and Hydrogen Oxygen Production (IHOP) subsystem through a recently awarded NASA Next Space Technologies for Exploration Partnerships-2 (NextSTEP)-2 Broad Agency Announcement (BAA) contract. Paragon’s Ionomer-membrane Water Processing (IWP) technology is optimized to perform primary water purification for this ISRU application. The purified water receives final polishing and is then electrolyzed using a Giner static feed water electrolyzer to produce H2 and O2 propellant.

The Systems Analysis of the Moon Mine Has Been Performed by Metzger

The best estimate of lunar PSR (Permanently Shadowed Region) water ice content comes from the LCROSS mission at 5.5wt% ice, while other estimates put the ice content at 10wt% and the most pessimistic estimates put it at 1wt%. This suggests that the thermal mining process should be situated at a location with more than 4wt% ice. Recent re-analysis of mission data found areas with over 30% ice. At 30% ice, the power needed to sublimate icy regolith material is 33% less than the power needed at 4% ice and is only 10% greater than the power needed to sublimate pure ice.

There should be plenty of spots where we can mine lunar ice at the polar regions using about 370 kilowatts of power to produce 2450 tons per year. This would make 1640 tons of propellant per year.

At 30% ice, we can mine an area of about 2.5 acres a year to get 1640 tons of propellant per year.

43 thoughts on “Equipment for Moon Mining Operations are Being Developed”

  1. The other reason why SpaceX chose methalox is that it would be really, really hard to manage the boiloff of LH2 for 4 or 5 months, even in header tanks, even with a hefty cryocooler (and even heftier radiators for it). But that’s not really an issue in a just-in-time cis-lunar architecture.

    Reply
  2. Your earlier figure of 1400 m/s isn’t that different. But I really have no idea what it should be. I’m just quoting from that thread I linked to (via memory, so maybe inaccurately).

    I agree the BFS seems to ignore the Moon. I think a key design requirement was for it to be self-sufficient, and not need any other infrastructure. But furthermore I think they don’t expect anyone to use Lunar water anytime soon. And they may be right: BO is painfully slow with their R&D, NASA is betting on SLS, and there aren’t many other players on the horizon with the necessary funding, attitude, etc.

    Given SpaceX’ relatively rapid development cycle and the time it would take to properly develop the Lunar resources and the rockets/tugs/etc that would use them, SpaceX can still have plenty of time to respond with their next generation of rockets, or even the one after that.

    As for hydrolox vs methalox, my impression was that they decided hydrolox was generally a PITA.

    Reply
  3. OK, your delta-v to the surface jibes with my recollection.

    I have this feeling that the whole BFS architecture is predicated on the idea that there wouldn’t be any lunar resources for a long time. Methalox has obvious advantages in terms of payload density if you’re launching a lot of it, but hydrolox would work just as well on Mars. Seems like the launch cost was probably the deciding factor.

    If lunar water turns out to be available sooner and more abundantly than SpaceX guessed, my guess is that they’re gonna have to scramble a bit. Not that a little viable competition would be a bad thing…

    Reply
  4. Not 1 km/s slow-down, but 1 km/s total delta-v from Mars orbit to surface, if aerobraking is used. As I recall, it was actually closer to 700-800 m/s, with the rest for safety etc.

    I think this is the thread: https://www.reddit.com/r/spacex/comments/73nzgq/analyzing_the_deltav_figures_from_the_bfr/

    > requires a lot more delta-v to get the tanker back, which limits its payload

    Not so much back, but to the depot. Delta-v from EML2 to LEO is just 0.33 km/s, and after that (maybe even for that?) it can use aerobreaking. Presumably it would refuel in LEO from another depot before going to EML2, but yes, it adds to the costs. But keep in mind that it only needs to carry carbon (so it would actually be a cargo BFS, not a tanker). The hydrogen and oxygen can come from the Moon with lower delta-v.

    The possible advantage (not really sure if it is) of an EML2 depot over HEEO, is that the same depot can be used with both Mars and Moon missions. Not sure if its optimal in terms of delta-v, since as you point out, it requires an extra refueling in LEO for some missions. But seemed good to me last time I thought about it.

    Reply
  5. Ah, you’re right–I had a bug in my model.

    Only 1 km/s from aerobraking doesn’t sound right. In theory, with enough lift and proper thermal management, you should be able to kill as much velocity as you want. As long as they only need MTO and about 1400 m/s to actually land, you can do it in 12 launches (1 payload + 11 tankers to get the thing full).

    In my lunar calculations, just going to the highly elliptical earth orbit (I used LEO + 2500 m/s) made a significant difference. There’s nothing wrong with L2, but it requires a lot more delta-v to get the tanker back, which limits its payload. With HEEO, you need hardly any delta-v to get home–just a few m/s to lower the perigee enough to reenter.

    Reply
  6. By my calculations, the 150 ton BFS (whatever you want to call it) had a total delta-v of 6.4 km/s when fully fueled and with full cargo. Based on https://en.wikipedia.org/wiki/Delta-v_budget#Interplanetary , the total delta-v from LEO to LMO is 6.6 km/s, and then I read elsewhere that Mars aerobreaking and landing is estimated at ~1 km/s for BFS.

    So it could get from LEO to Mars with aerobreaking, but not with full payload. But if it starts from EML2, the total delta-v drops to 4.3 km/s, which is low enough for full cargo. And if it’s fully refueled (in EML2), it has enough delta-v to land on the Moon with full cargo and get back to either EML2 or LEO, again with full cargo, without refueling again. So I would put the methalox in EML2.

    But then again, if it’s volume limited, EML2 refueling may not be helpful for Mars. And I agree that everyone else will probably use hydrolox eventually.

    Reply
  7. I’ve done a lot of modeling of the Adelaide version of BFS for lunar missions with no surface refueling, and it’s surprisingly ugly. If you want the full 150 t on the wild surface, it’s like 15 launches total: the payload launch, 6 refills of that payload BFS in LEO, then boost to HEEO and rendezvous with yet another tanker, which has also been topped off in 7 other tanker launches. Even with almost no payload, it still takes 9 launches.

    Things are considerably better for Mars, because you can aerobrake, and of course round-trip fueling is out of the question. I got 4 launches (1 payload + 3 tanker) for a 5000 m/s MTO, which is considerably more than you need.

    Landing on an airless body with non-trivial gravity and no gas station sucks.

    Lunar surface refueling helps enormously, but then you’d have to deal with the methalox issues, which will be on SpaceX’s dime; everybody else is gonna use hydrolox.

    Reply
  8. Interesting point about the extra tank. That does add some mass over the hydrolox tanks. The hydrolox tanks are smaller, but with the water tank, the total mass ends up roughly the same.

    My 1000 ton number was from BFS’ fuel capacity. At least some missions would need it fully refueled, or close to that. BFS is the earliest days possible for in-orbit refueling.

    Granted, in LEO it could refuel from tanker BFSes, the way Elon intended. But it could be useful to refuel it in ELM2 to increase the max payload to Mars and to/from the Moon.

    Reply
  9. I’d be surprised if you needed more than 50 tonnes a month on-orbit in the early days. So that would indicate you’d need about 350 kW of power.

    But I finally did the comparison of a water tanker vs. a hydrolox tanker from the surface, and the latter isn’t nearly as bad as I thought. Using the Akin mass estimating relationships ( https://spacecraft.ssl.umd.edu/academics/791S16/791S16L08.MERsx.pdf ), I got:

    I think my mistake in my guesstimate before was that water requires an extra physical tank, while hydrolox can be held in the same tanks as the prop. (BTW, all tanks are assumed to be spherical above–no fairings required!)

    So the real question is how much power you want on the surface. While you don’t have to make a lot more prop to get hydrolox to NRHO over water, you’d still have to make the payload hydrolox, which is pretty much always roughly double. That of course requires double the power for the same unit time.

    Some day I’ll figure out the difference between $/kW in NRHO and $/kW on the lunar surface. My guesstimate is that the surface power is about 3x as expensive. But that of course depends on transportation costs.

    Reply
  10. Based on my calculations for the other thread, water splitting ~5 tons/day/MW. A couple of weeks would give ~70 tons/MW, so ~15 MW to fill up ~1000 tons (give or take some inefficiency factor, which I’m not sure how big; so maybe 20-30 MW total).

    Reply
  11. Figure 45 is nuts. Nobody’s building track on the lunar surface until ISRU metallurgy and fab is up and running. Let’s talk again about 10 years after the water operation is running, and then we can start thinking about track. Until then, it’s all mobile collection systems.

    Not to beat a dead horse, but if you’re going to ferry stuff from the surface to some kind of orbit, you need hydrolox production on the surface and on-orbit. The efficiency of your system is dominated by the structural coefficient of your lander/ascender, which means that you want the smallest payload tank possible, which is a water tank, and the smallest hydrolox tanks possible–just enough for a one-way ascent. You can refuel for the return to the surface on-orbit.

    That means you need:

    1) Water collection and a small hydrolox production system on the surface (just enough to fill the ascent tanks).

    2) A much larger, just-in-time hydrolox production system on-orbit, where the cost per watt of power is about a third of that on the surface, and where water is stored until just before it’s needed by a customer. (Realistically, that’s probably a couple of weeks, but keeping cryocooling costs down as much as possible reduces your power requirements.)

    If you’re careful with your design, the hydrolox production stuff can be the same module for the surface and on-orbit, with more modules on-orbit to provide the larger capacity.

    Reply
  12. There’s no reason why you can’t do hydrolox production both on the surface and in orbit. Indeed, the more I think about it, the more it seems likely that you design a purification payload and an electrolysis + power payload. Then you snap ’em together as needed, but you only need to do the R&D once for both locations if you’re careful.

    If I ran the circus, I’d put high-volume purification on the surface with a small number (i.e. low volume) of electrolysis modules, and no purification on-orbit, with a large number (i.e. high volume) of electrolysis modules. Then the surface system makes just enough hydrolox to get the water tanker to orbit, and the on-orbit system makes just-in-time hydrolox as needed.

    As an arm-wave, I’d guess that you can put at least double the number of modules in NRHO or an L2 Lissajous as you can on the lunar surface. So the economics militate toward doing the minimum amount of prop production on the surface. (NB: doing a tanker round trip from either end is insanely expensive. You need a depot on each end.)

    Reply
  13. You need propellant to get the water into orbit, don’t you? Therefore it makes sense to do at least some of the electrolysis at the water mine.

    In fact, the plan described above has the fuel produced on site.

    Reply
  14. You need to consider the concentration of each volatile and the cost of all the extra systems needed to process it. It’s not just MOFs, it’s all the chemical processing, storage, power, etc. Below some minimum concentration, it just doesn’t make economic sense.

    Reply
  15. I’m not so sure thermal conduction will do. Lunar soil is mostly ceramic dust (various oxides) with vacuum in between. That’s a pretty decent insulator.

    Reply
  16. It may indeed be more efficient to do both in the same place – if you were going to use all the propellant there as well. Otherwise there are arguments in favor of doing the splitting in orbit, at the fuel depot (delta-v, systems costs, etc). It’s pretty easy to add some extra solar panels to provide the extra power for the sublimation.

    (Put another way, energy (in)efficiency probably isn’t the dominant cost in this case.)

    Reply
  17. 370kW of power is over 3 times the maximum combined power of all ISS solar arrays. All that “making fuel” dreaming is a dressed-up energy conversion process – there should be electric energy production first, at record levels for a space facility, and only then it would make sense to develop energy conversion processes. Even at the best spots of Lunar poles, the insolation is not constant, so the arrays would have to be elevated above terrain to maintain generation. It is not a zero-g situation, so the arrays and their support will have to be load-bearing and also elevated – that is just far out at the current state of space construction affairs. The easiest and fastest way to making Lunar fuel starts with a nuclear generator, which is the only possible power source in Lunar environment capable of providing power at hundreds kW range. The closest thing to that now is NASA’s kilopower – a very small nuclear generator, and only an experimental prototype. In summary, that picture is a solar fantasy in a place rather far from peaks of quasi-eternal light where night lasts for two weeks.

    Reply
  18. ARRGG!! REPOSTING!!

    Table of volatiles detected on Moon, by percentages.

    H@@PS://isru.nasa.gov/IMAGES/volatilesTable2.png

    EXHAUSTING!!

    Reply
  19. *REPOST PART ONE of TWO (I think for the twelfth time?)

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices. Perhaps in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. It would be foolish to develop a system for water recovery that treats all the other volatiles as waste. –See table POSTED SEPARATELY.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.

    Reply
  20. 10 plus times trying to post a comment.

    Broke it down into separate posts. Deleted. Again… deleted. Again… deleted.. oh, and deleted previous post, too. Over and over and over.

    Reply
  21. *REPOST OF PART ONE. ~10th time

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices. Perhaps in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. See table below. It would be foolish to develop a system for water recovery that treats all the other volatiles as waste.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.

    Reply
  22. *Breaking post into two parts
    *REPOSTING, again, DELETED FIRST PART — 8th or 9th attempt in total

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices. Perhaps in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. See table below. It would be foolish to develop a system for water recovery that treats all the other volatiles as waste.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.
    part two below

    Reply
  23. PART TWO — of post above

    MOFsMOFsMOFs. MOFs.

    The hydrogen sulfide (H2S) can supply hydrogen to off-set the excess oxygen from water electrolysis, while providing sulphur for Sulphur-polymeric products: resins, plastics and textiles, as well as for agricultural use.

    There’s methane for rocket fuel, and other hydrocarbons for plastics, fabrics and resins. There’s ammonia for fertilizer, radiation shielding, (high hydrogen content) and for use as rocket fuel.

    It may be easier to design to recover water only, treating the rest as waste, but don’t. Design for what is actually needed; complete volatile recovery. You are going to need and want every molecule you can capture.
    end of part two

    Reply
  24. *FOURTH ATTEMPT TO POST — Breaking it into two parts.

    PART ONE

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices. Perhaps in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. See table below. It’d be foolish to develop a system for water recovery that treats all the other volatiles as waste.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.
    continued in next post…..

    Reply
  25. *THIRD ATTEMPT TO POST

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices. Perhaps in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. See table below. It would be foolish to develop a system for water recovery that treats all the other volatiles as waste.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.

    MOFMOFMOF. MOF.

    The hydrogen sulfide (H2S) can supply hydrogen to off-set the excess oxygen from water electrolysis, while providing sulphur for Sulphur-polymeric products: resins, plastics and textiles, as well as for agricultural use. There’s methane for rocket fuel, and other hydrocarbons for plastics, fabrics and resins. There’s ammonia for fertilizer, radiation shielding, (high hydrogen content) and as rocket fuel.

    It may be easier to design for water recovery only, treating the rest as waste, but don’t. Design for what is actually needed; complete volatile recovery. You are going to need and want every molecule you can capture.

    Reply
  26. *note to self — Copy EVERY post before you post!!

    The polar shadow-craters are not the Moons only cold-sinks. There are lava tubes scattered around the Lunar surface that are just as likely to hold volatile ices.

    Possibly in even greater concentrations and quantities due to their higher surface area, greater age and protected, more thermally stable, environs.

    It’s important to remember, too, that water ices are not the only ices present. See table below.

    It would be foolish to develop a system for water recovery that treats all the other volatiles as waste.

    Using “tuned” MOFs, metal–organic frameworks, with enhanced gas selectivity, raw vapours could be pumped into a high(ish)-pressure vessel, with gas-specific MOF “windows”. The mixed-element vapours would be separated into constituent gases and drawn off for storage in MOF “sponge” tanks.

    The hydrogen sulfide (H2S) can supply hydrogen to off-set the excess oxygen from water electrolysis, while providing sulphur for Sulphur-polymeric materials: resins, plastics and textiles, as well as for agricultural use.

    There’s methane for rocket fuel, and other hydrocarbons for plastics, fabrics and resins. There’s ammonia for fertilizer, radiation shielding, (high hydrogen content) and as rocket fuel.

    It may be easier to design for water recovery only, treating the rest as waste, but don’t. Design for what is actually needed; complete volatile recovery. You are going to need and want every molecule you can capture.

    Reply
  27. I just lost yet another of my posts.

    Is it a “some one” that is deleting my posts, or just a flaw in the system? Do some commenter’s have the power to delete? What’s up?

    Reply
  28. If nothing truly mind-blowing turns up, the mere presence of microbial life shouldn’t, in and of itself, prevent mining.

    Unless they tell us “no”.

    Edit: The pharma-gods would want to sample EVERY lil’ microbe for medical and/or industrial value, and would be willing to pay bank for access, I would think.

    Reply
  29. 1) Deeper, yes. Thermal conduction.

    2) Duh. It’s in those dark and most stable of ices that the “Galactic Order of
    Knowledge Transfer” has encoded all of the info Humanity will need to
    survive the coming Singularity, and to thrive in a new era of Galactic
    membership. 🙂

    Reply
  30. The poles are not exclusive of other points on the Moon as cold-sinks.

    The lava tubes are just as likely, perhaps more so, to hold vast amounts of ices.

    Remember too, it’s not just water ice. Below is one breakdown of some other ices that also exist in the shadows, though in lesser quantities.

    There are hydrocarbons to make plastics. Methane for fuel. Amonia for agricultural use. The hydrogen sulfide (H2S) is another source of hydrogen, allowing for waters excess oxygen to be balanced out, while providing sulphur for sulfur-based plastics, resins and agricultural use.

    Use “tuned” MOFs, metal-organic frameworks, for gas separation. Feed raw vapors into a single high(ish)-pressure vessel lined with “windows” of enhanced gas selectivity MOFs and draw the separated gases off into MOF storage tanks.

    In other words, though it’s simpler to design a way to collect just the water, treating all the rest as waste, don’t. Design for the MUCH more complex processes needed to capture every molecule you can. You will need and use them all.

    Reply
  31. First, under Lunar vacuum, water ice sublimates straight to gas. There’s no liquid state. Maybe deeper underground, but how would the heat get there?

    Second, the mining would be done in permanently shadowed areas. No light/dark cycling there – just dark all the time. The crater rim is another story, but no mining there. And probably also no ice, since it would’ve sublimated away a long time ago.

    Reply
  32. Compromise: Test for organic matter as a required portion of the mining operations. Delaying space exploration in the name of pure science is a recipe for continued anemic progress. Let’s keep this SpaceX, Blue Origin led momentum going!

    Reply
  33. *reconstituting lost post

    Before any volatile mining starts, there should be a thorough survey of the ices to determine whether or not they contain any life.

    I’m not saying there IS life, but consider this…

    1) It is an accepted subject to discuss the possibility of life having been transferred between
    the Earth and Mars.

    2) If possible, then the odds of Earth-life-bearing matter having reached the Moon is far
    greater than for it to have reached Mars.

    3) There is water on the Moon.

    4) Around those craters that hold ices in permanent shadow, there will be cycling of Sunlight
    and darkness, alternately melting and refreezing ices in a sub-surface zone. Think of a
    torus shaped zone around a crater, below the surface, where the soil temperature will rise
    above, then drop below freezing, creating moist soils.

    Life-seeds + energy + liquid water + time = ?

    The Moon is a lot closer than any other option for seeking exo-life. If nothing else, it’s a good place to develop and practice the processes.

    Reply
  34. Better to use smaller operation scale to do so. That way it could be taken from crater to crater a lot easier. If only a little water is found in one it could go to the next and the next. Or on a smaller scale more craters in an area could be mined at once.

    Reply
  35. Seems inefficient to use separate energy for the sublimation step. If the electrolysis were done in the sublimation facility, the waste heat from that step could do the heating. And given the large amount of waste heat involved, you’d probably get more effective sublimation.

    Reply
  36. Doing a quick search, I find that removing trace ammonia before electrolysis is called “polishing” the water. Apparently ammonia boosts the conductivity of the water so much that a great deal of the electrolysis current is wasted.

    It wouldn’t be shocking to find lunar water contaminated with ammonia, in as much as it originates from cometary strikes.

    By the way, really not a fan of the way the new system auto-converts urls to functioning links, even if you’re in the middle of entering html. But I suppose it’s better than having links automatically removed.

    https://onlinelibrary.wiley.com/doi/abs/10.1002/jctb.4923

    Reply
  37. I’m curious what that “water polisher” step entails. Probably activated carbon or an ion exchange resin, but you’ve already run the solution through your membrane, so whatever they’d capture would have been removed already. Or, hopefully removed, in which case that’s the best membrane that could work under the operating conditions, or it’s a sad membrane.

    Reply

Leave a Comment