I suspect this will in the end be solved in utmost unsexy, boring and reliable way by solar plus huge NiFe battery and/or hundreds of kilometers of high voltage grid, manufactured by robots from local resources or asteroid mining.
Or somewhat sexier, beamed orbital power or mirrors are less technically challenging than on earth.
Either should be less expensive than hauling nuclear reactors from earth and ensuring they get reliable cooling.
What’s wrong with operating half the day? I guess it’s space, so it’s worth a lot of effort, but I’d think if they can avoid RTGs and just use solar they would.
The night is 14 days long and is deeply (cryogenically) cold, which is really bad for lots of hardware. Military component ratings are something like -55°C[1]. Lunar night is more like -130°C.
Ideally you keep as much equipment as close to an even temperature as you can, or it might never turn on again when daylight comes.
[1]: the exact definition varies, but it's in that general area. Industrial is -40°C and commercial is 0.
humans use to build with mud, sometimes with walls meters thick. Underground or natural caves would be even better. It doesn't get any less sexy. Not sure how it works with such extremes but on earth it takes months of heating or cooling to get through a few meters of soil.
Why must the build be done all at once in one go? Humans will land at the dawn, leave before dusk, working in 14 day shifts set up the base. This will repeat for several months till the base gets self-sufficient enough.
Some stations in Antarctica were built that way only during summers, in winters they were left empty till finished and operational.
Going to need roads for that. Might as well construct rails. A circum-lunar railway! Probably want to build that at 80 degrees latitude or more, the trade off being solar angle vs material costs.
If you build rails, you'll get a free good-enough conductor for getting power from the other side, even before you think about putting anything onto those rails.
How would a lunar railway handle thermal expansion? We’re back to the same problem that the nighttime lunar surface is mighty cold, and I’d expect big gaps to form
I have no idea how plausibly, but the book 2312 has a city on one of the hot inner planets that gets squeezed along a rail using thermal expansion. Even if that doesnt work, it seems like the sort of thing you could engineer around, just a more extreme version of the engineering we already do to handle thermal expansion.
Orbital power has the huge advantage that setting it up is essentially independent of the rest of the mission - we can set up most of the work in advance, so that part doesn't count towards a weight limit and we can abort/wait if it doesn't work. Once a secure base using orbital power is set up, then NiFe should be arranged for redundancy.
I think the heat part would best be Feolite, which is mostly sintered iron oxides, but with some other bits added. It has a tremendous volumetric heat capacity, nearing that of water, but you can crank it way past 100 C without worrying about pressure.
One of my all time favorite everyday engineering things I've learned is the absolute bonkers amount of energy needed to boil water (vs nearly any other fluid).
It's so commonplace yet absurd/impressive when you really think about it.
> "manufactured by robots from local resources or asteroid mining."
I feel like this just creates the new problems of operating, powering, and maintaining the robots, plus all the difficulties associated with mining asteroids and refining the mined materials somewhere (presumably somewhere that's not Earth's surface, which means we still need to generate power somewhere off Earth for this to work).
I am still awaiting the future promised me where 3d printers replicate themselves. This is orders of magnitude easier than robots making themselves in outer space, yet we are still waiting and I assume we always will be.
Let's assume 10mm copper wire to the other side of the moon. 3000km long copper wire 10mm in diameter will have 600 ohms resistance. The amount of copper needed is 235 m³, copper at 9t/m³ weighting 2115 tons in total or 21 starship flights costing $21 milion.
But even at $100m/launch, a big fat copper wire (IIRC aluminium is better per unit mass) would still make more sense than shipping up a nuclear reactor or a huge pile of batteries, and that part of this hypothetical mission would still be about 80% cheaper than the JWST.
Aluminum is also probably better because it’s easy to source on the moon - many of the surface rocks are made of aluminum oxides. Copper, on the other hand, would need to be entirely transported from Earth.
I wonder, how hard is it to make aluminium from the oxide if you're bootstrapping and in a vacuum? I know the normal process is "melt it with added cryolites and apply current for electrolysis", but if the goal is low mass rather than good energy efficiency, can you do it a different way? Like, what happens if you focus sunlight and just melt it, does the oxygen bubble out in a vacuum, or is this just going to do vapor deposition of sapphire on all nearby surfaces?
I believe molten alumina does not separate naturally - you’d still need to electrolyze it. This can be done without cryolite although at higher temperature and correspondingly higher energy usage. Otherwise the aluminum and oxygen would remain bound while they evaporate.
The $/launch numbers don't account for the refueling flights, which will be necessary to get to and return from the lunar surface.
That would bump that number up to ~$10m/$20m or so. Still ridiculously cheap and not likely to be anywhere close to the price charged as it would absolutely sink the rest of the market, which SpaceX is not interested in doing.
There wont be much left on the other end because of volt drop in cables. But can be possible to make it into a HVDC transmission line. Will add bit of weight on the starship.
You definitely need HV. At 100kV and 10A, the lost power if you send 1MW is around 60kW: a fairly serviceable 6% or 20W/km. However, your losses at a fixed voltage scale quadratically with your power draw: you'd lose 24% of 2MW.
If you sent it at 1000V, the I²R losses to send 1000A over that cable outweigh the transmitted power by 600 to 1 and your cable is burning 200kW per kilometre. Which in a vacuum would probably just melt it in fairly short order.
Which is why the bigger HVDC links get, the higher the voltage: there's a 1MV+ system in China that sends 12GW over 3000km.
Also I'm not sure how lunar regolith will work with regards to the "earth" return path so you might well actually need two wires.
600Ω (WolframAlpha gave me 510, but let's go with the bigger number) isn't that bad, and you don't need anything expect bare metal to run HV on the moon because both the vacuum and the regolith are already insulators.
How awesome would it have been if Artemis actually carried a payload to just dump off on the moon? It's just so absurd that this feat of engineering and rocket science just did a fly by/test.
Payload to low earth orbit might be 100+ tons each. But payload to lunar surface is a small fraction of that. We could assume 10 tons. So 210 starship flights.
One can see pretty quickly that any larger constructions far from earth would really benefit from maximum use of local materials.
Local aluminum will be cheaper, and stonkin' high voltage will be needed to combat the ohmic losses. The 'atmosphere' on the moon is rather extremely dry, so corona losses shouldn't be an issue, you could probably run north of a megavolt w/o too much trouble.
Now, you could lay both conductors in parallel and only need to build the power line half-way around the moon, probably save some dough on prospecting for and constructing pylon sites that way. Alternatively, you could run a single conductor all the way around. Doing that, you could establish a lunar scale magnetic field, though it'd probably be pretty wimpy unless you ran serious kA (MA?) of current, which would mean much bigger conductors etc, but it's fun to think about. Heck, with a loop that big, you'd probably get significant induction from the solar magnetic field .. which might be something to harness, or might just be a headache for your line operators.
This seems like a chicken-and-egg problem. Wouldn't you need power in the first place (and therefore transmission lines) to even extract and refine aluminium locally?
> Doing that, you could establish a lunar scale magnetic field, though it'd probably be pretty wimpy unless you ran serious kA (MA?) of current, which would mean much bigger conductors etc, but it's fun to think about.
Or a stator in orbit? But hrm… for Mars the same idea needs only 1T to 1.5T stator but has to place xt at lagrange between Mars and Sun. So, I naively guess for the stator to be far enough away that the deflected solar wind merges after the moon could /reall/ mean that the stator would be at lagrange between Earth and Sun, which could have perhaps interesting effects on Earth.
People survive Antarctic winters so I can't see that there is a particularly big challenge. There is no atmosphere on the moon so heat loss to the sky is purely radiative whereas in Antarctica you have winds blowing that cool the habitats by conduction.
Solar panels should be able to provide the necessary energy which can be stored in piles of rock and insulation should be able to to prevent the loss.
The 'article' says that the lunar night is rife with problems but doesn't mention what they are. There is no weather after all.
Any permanent habitation should include a smelter to alleviate some of the cost of shipping metals from Earth. From it oxygen or CO2 will be "waste" byproducts, liquefacting and storing them should not be a problem. I can imagine that having large reserve of oxygen will be desired safety feature.
Hydrocarbons will likely be in short supply on the moon but if there is, say, a zinc mined, it could be "burned" in zinc-air cells.
There's also the cost of the diesel fuel (or whatever fuel you burn) to consider. With no native hydrocarbons on the moon, extracting them would be impossible. So they would need to be launched from earth, at many tens of thousands of dollars per kilogram as an ongoing cost.
The biggest challenge to the obvious solutions - either solar plus lots of batteries, or nuclear - would be to crack water into oxygen and hydrogen, and then burn them during the night. The challenge there is storing the hydrogen. But if you do that, basically what you'd be doing is using hydrogen/oxygen as an energy storage mechanism, e.g. as a battery. You'd need to use solar during the day to crack the water.
Why store hydrogen when you can use the metals? I mentioned zinc as that's most mature technology, but almost all of them can be reacted with oxygen to get back energy.
12% by mass is iron, with calcium, aluminum, magnesium at 7-8%. These could be considered ores worth mining on Earth. All four metals readily burn in oxygen, even though aluminum needs some help igniting.
The lunar surface temperature is like -150C at night. And yeah, heat loss is radiative.. to a sink at roughly -273C. And night is 2 weeks long. Just use solar heaters and insulation? Well you have to fly that stuff there first, and also fly whatever equipment and people/robots to make the system.
There are various studies to have stuff survive lunar night without a nuclear source, and it takes on the order of 10kg of stuff to keep a 1kg payload alive. And in turn that takes 100kg of spacecraft/fuel to fly it all there. That’s not including the rocket and its fuel to get to Earth orbit first.
Nuclear appears to be the only realistic contender. But the mission architecture of Artemis goes around the problem by having only short stays on the lunar surface.
Even with nuclear, it can be far safer: small reactors need less containment. There is no risk of fallout without an atmosphere, and no ground water to contaminate.
Still, relying too much on it and failing (melting down) will put in danger the whole human space program. For low-power unmanned applications the RTG remains unbeatable.
What does the risk profile look like for launching nuclear fuel through our atmosphere? It seems intuitively concerning, but I don't know much about nuclear.
It isn't too big of an issue, rockets have to meet certain requirements to fly nuclear material. The flight profile is required to be over water, so intact material just gets diluted or sinks. There are some requirements about how the flight termination system works (eg it shouldn't spread payload debris over a large area) and general reliability considerations identical to those for human spaceflight.
Essentially, flying nuclear material is treated with the same level of basic care as flying humans.
Once the material is in space the regulations aren't as thorough, I imagine the most that NASA considers is for the launch to be into a direct injection orbit or at a sufficiently high parking orbit that there isn't an immediate risk of uncontrolled reentry in case of some failure.
Basically zero. Unused nuclear fuel is not very radioactive. You can hold it in your hand with a glove. Once you split atoms it becomes more hazardous, but space nuclear doesn't start up until after it's been successfully lifted.
- "There is no risk of fallout without an atmosphere"
I'd speculate that it would spontaneously propagate along the lunar surface through electrostatic forces. The fallout particles would be highly charged (self-ionizing), very small, and in a perfect vacuum -- a recipe for some seriously weird dust physics.
Space is very irradiated, and direct sunlight is radiation itself. Anything you add there will be a drop in the ocean. Atmosphere is not sufficient to resolve space radiation, you need a magnetosphere.
Rtgs are quite common already indeed and will likely be a part of the solution. But it's a stretch to call nuclear the only contender. Cables are a perfectly valid way to move power around from areas that do get solar exposure. And there are plenty of ways to create batteries or store energy. It might even be possible to use resources on the moon to build those (e.g. a heat battery would be doable).
The key limitation is going to be transporting stuff from earth. Solar panels have a key advantage here: they are pretty light and easy to deploy. There's no wind or weather that will dust them over. And per launch, you can move some significant amount of power generation.
> There is no risk of fallout without an atmosphere
Fallout is mostly irradiated material spread by an explosion, so as long as you have a surface and some amount of debris from an accident that doesn’t reach escape velocity you can have fallout.
What about a large reflector dish grabbing sunlight at some distance from the Moon in a geosynchronous orbit, bouncing energy to a receiving station at the new Artemis site?
Luna-synchronous is a long way from the moon. So far away, in fact, that you're it.
L1 would be easier, but that's still 58 megametres from the moon. Possible, of course, but I'd be surprised if that was really better than a big wire on the lunar surface.
Starship and Superheavy ("can and kicker") will not realistically cost any less than $20M per launch, plus whatever what you are launching costs to make, stow, and deploy. If it takes 100 tons to orbit, that is $200/kg minimum.
Getting a can to the lunar surface takes launching a bunch of fuel runs, so that much several times over, say $1000+/kg all told.
Bringing cans back from the moon would be counterproductive, except as needed to bring crew home. Maybe you unmount used vacuum engines, cut their bells off, and bring home the fiddly bits. Somebody should find a use for the cast-off cans, eventually, and the cut-off bells. Maybe swing the cans on the ends of a wire for artificial gravity so your bones don't dissolve; though getting in and out would be tricky. You could store energy in their kinetic motion, resolving the nighttime power problem at the expense of variable artificial gravity inside.
Nuclear is probably most practical for long term generalization, for now I think the approach all crewed landers have proposed to use is to place panels as high up as they can to maximize sunlight duration.
At the poles this could enable some permanent power generation by getting the panels high enough to access permanent sunlight.
Putting the base at the South Pole of the moon is the current plan. But missions will want to go into craters, which will lead them to have to deal with the long nights
The very large temperature difference between day and night means you can bank heat in the daytime, and cold in the night, and generate power from the difference.
You need a pair of very large bags of regolith to store the heat and cold in, and a gas to percolate through them from radiators exposed alternately to sun and black sky.
Or you can just put up very big reflectors in orbit, lighting up your solar panels. Maybe they pump laser tubes pointed at your solar panels.
Nukes would be a big nuisance, needing constant maintenance, unless you just run a naked pile at incandescent temperature and catch the light coming off photovoltaically.
> Nukes would be a big nuisance, needing constant maintenance, unless you just run a naked pile at incandescent temperature and catch the light coming off photovoltaically.
If we had piles of Pu-240 we could use RTGs, though they wouldn't be very efficient considering the amount of mass we'd have to send.
Running a naked reactor doesn't sound so bad considering there's no atmosphere or water sources to poison -- just stay away from the reactor. Running a steam turbine shouldn't be too hard, but it probably can't be serviced -- if it breaks, you replace it.
Sealed free piston Stirling engines might be even lower maintenance and can provide a number of kilowatts pretty simply. Couple with a fast reactor or a combined fuel moderator solution alas TRIGA should work.
To make your power go further, we'd want some impressive insulation in our structures. Aerogel comes to mind, but it's low density and fragility means shipping it by rocket from Earth wouldn't work. However it may be possible to make it locally - the moon has lots of aluminum oxide, but we'd have to ship the water & alcohol needed for synthesis. Removing the water and alcohol after the aerogel structure gets created is done with freeze-drying, something that is easy on the moon, and the water + alcohol can probably be recovered for reuse in the next batch.
Downside to aerogel? It makes for a lousy micro-meteoroid and radiation blocker, unlike something like rockwool that masses more.
The moon is a near vacuum, so aerogel is the wrong approach. The main form of heat loss is radiative, and the way to reduce radiative loss is with many layers of shiny things. This works like the superinsulation between layers of a Dewar flask. It could be literally a layered structure or a filling of shiny fluffy stuff inside something else. This will dramatically outperform mineral wool or really anything practical on Earth. All this insulation goes on the outside, vacuum side of whatever pressure vessel people live in.
To the extent the lunar atmosphere is dense enough to cause meaningful conductive loss, one could pump a better vacuum inside the insulation. The pressure difference between this high vacuum and the lunar atmosphere would be low, so the mechanical strength needed for whatever container holds it would be corresponding low. (This seems very unlikely to be a problem at all. The moon’s atmosphere has an extremely low pressure. Offgassing from the insulation seems like a bigger concern.
Just hang your room from a wire (or 3) and wrap the thing in aluminized (or gold) mylar reflector. You've got an almost perfect thermos. Pretty quickly, you'll have to dump waste heat. Now cooling is pretty easy since just need a radiator in shadow. You could run a full on heat pump if you wanted to collect and store power, but this isn't like earth where there's an atmosphere with wind and rain. Or a place where there isn't always a significant solid angle of 3k space to radiate to, most of the sky is always night.
It's just weird to me that they wouldn't go to the interesting and water/gas rich craters perpetually at the terminator. You've always got power there... just start digging.
Nuke the moon! Right now the topography of the poles necessitates quite tall towers to get into perpetual sunlight; but a bunch of judiciously placed nukes could easily take away most of the shadow-bringers (crater rims) and reduce the tower height necessary to get up into perpetual light to a small fraction of what it would be now.
Just remember: nuke first, then colonize. So, sooner rather than later.
They practiced on home soil and irradiated many US citizens [1] while determining that big explosions aren't all that great, these days we use precision blasting patterns for earth breakup and moving [2].
109 comments
[ 2.4 ms ] story [ 163 ms ] threadOr somewhat sexier, beamed orbital power or mirrors are less technically challenging than on earth.
Either should be less expensive than hauling nuclear reactors from earth and ensuring they get reliable cooling.
Ideally you keep as much equipment as close to an even temperature as you can, or it might never turn on again when daylight comes.
[1]: the exact definition varies, but it's in that general area. Industrial is -40°C and commercial is 0.
Some stations in Antarctica were built that way only during summers, in winters they were left empty till finished and operational.
[1] https://youtube.com/watch?v=4PUIxEWmsvI
https://en.m.wikipedia.org/wiki/2312_(novel)
It's so commonplace yet absurd/impressive when you really think about it.
I feel like this just creates the new problems of operating, powering, and maintaining the robots, plus all the difficulties associated with mining asteroids and refining the mined materials somewhere (presumably somewhere that's not Earth's surface, which means we still need to generate power somewhere off Earth for this to work).
Would heat pumps work on the moon?
Where can I book a Starship flight for $1M?
But even at $100m/launch, a big fat copper wire (IIRC aluminium is better per unit mass) would still make more sense than shipping up a nuclear reactor or a huge pile of batteries, and that part of this hypothetical mission would still be about 80% cheaper than the JWST.
Starship is supposed to be cheaper, the moon is definitely more expensive.
I wonder, how hard is it to make aluminium from the oxide if you're bootstrapping and in a vacuum? I know the normal process is "melt it with added cryolites and apply current for electrolysis", but if the goal is low mass rather than good energy efficiency, can you do it a different way? Like, what happens if you focus sunlight and just melt it, does the oxygen bubble out in a vacuum, or is this just going to do vapor deposition of sapphire on all nearby surfaces?
That would bump that number up to ~$10m/$20m or so. Still ridiculously cheap and not likely to be anywhere close to the price charged as it would absolutely sink the rest of the market, which SpaceX is not interested in doing.
If you sent it at 1000V, the I²R losses to send 1000A over that cable outweigh the transmitted power by 600 to 1 and your cable is burning 200kW per kilometre. Which in a vacuum would probably just melt it in fairly short order.
Which is why the bigger HVDC links get, the higher the voltage: there's a 1MV+ system in China that sends 12GW over 3000km.
Also I'm not sure how lunar regolith will work with regards to the "earth" return path so you might well actually need two wires.
One can see pretty quickly that any larger constructions far from earth would really benefit from maximum use of local materials.
Now, you could lay both conductors in parallel and only need to build the power line half-way around the moon, probably save some dough on prospecting for and constructing pylon sites that way. Alternatively, you could run a single conductor all the way around. Doing that, you could establish a lunar scale magnetic field, though it'd probably be pretty wimpy unless you ran serious kA (MA?) of current, which would mean much bigger conductors etc, but it's fun to think about. Heck, with a loop that big, you'd probably get significant induction from the solar magnetic field .. which might be something to harness, or might just be a headache for your line operators.
Or a stator in orbit? But hrm… for Mars the same idea needs only 1T to 1.5T stator but has to place xt at lagrange between Mars and Sun. So, I naively guess for the stator to be far enough away that the deflected solar wind merges after the moon could /reall/ mean that the stator would be at lagrange between Earth and Sun, which could have perhaps interesting effects on Earth.
Solar panels should be able to provide the necessary energy which can be stored in piles of rock and insulation should be able to to prevent the loss.
The 'article' says that the lunar night is rife with problems but doesn't mention what they are. There is no weather after all.
Hydrocarbons will likely be in short supply on the moon but if there is, say, a zinc mined, it could be "burned" in zinc-air cells.
The biggest challenge to the obvious solutions - either solar plus lots of batteries, or nuclear - would be to crack water into oxygen and hydrogen, and then burn them during the night. The challenge there is storing the hydrogen. But if you do that, basically what you'd be doing is using hydrogen/oxygen as an energy storage mechanism, e.g. as a battery. You'd need to use solar during the day to crack the water.
12% by mass is iron, with calcium, aluminum, magnesium at 7-8%. These could be considered ores worth mining on Earth. All four metals readily burn in oxygen, even though aluminum needs some help igniting.
https://en.wikipedia.org/wiki/Army_Nuclear_Power_Program
There are various studies to have stuff survive lunar night without a nuclear source, and it takes on the order of 10kg of stuff to keep a 1kg payload alive. And in turn that takes 100kg of spacecraft/fuel to fly it all there. That’s not including the rocket and its fuel to get to Earth orbit first.
Even with nuclear, it can be far safer: small reactors need less containment. There is no risk of fallout without an atmosphere, and no ground water to contaminate.
Still, relying too much on it and failing (melting down) will put in danger the whole human space program. For low-power unmanned applications the RTG remains unbeatable.
Essentially, flying nuclear material is treated with the same level of basic care as flying humans.
Once the material is in space the regulations aren't as thorough, I imagine the most that NASA considers is for the launch to be into a direct injection orbit or at a sufficiently high parking orbit that there isn't an immediate risk of uncontrolled reentry in case of some failure.
https://space.stackexchange.com/questions/17518/what-does-it...
I'd speculate that it would spontaneously propagate along the lunar surface through electrostatic forces. The fallout particles would be highly charged (self-ionizing), very small, and in a perfect vacuum -- a recipe for some seriously weird dust physics.
You take neutron radiation (well, its consequences), but that is pretty much non-existent when compared to nuclear waste.
The key limitation is going to be transporting stuff from earth. Solar panels have a key advantage here: they are pretty light and easy to deploy. There's no wind or weather that will dust them over. And per launch, you can move some significant amount of power generation.
The absence of atmosphere would make it viable to beam electrical energy down from orbit. Seems cheaper.
Studies have shown that in lunar vacuum, fine dust in the regolith can travel halfway around the globe and even end up in lunar orbit and beyond with just a little kick. https://www.theverge.com/2019/7/17/18663203/apollo-11-annive...
Fallout is mostly irradiated material spread by an explosion, so as long as you have a surface and some amount of debris from an accident that doesn’t reach escape velocity you can have fallout.
L1 would be easier, but that's still 58 megametres from the moon. Possible, of course, but I'd be surprised if that was really better than a big wire on the lunar surface.
(cue Everybody Wants To Rule the World)
https://www.youtube.com/watch?v=SRMDcC0QvFQ
Getting a can to the lunar surface takes launching a bunch of fuel runs, so that much several times over, say $1000+/kg all told.
Bringing cans back from the moon would be counterproductive, except as needed to bring crew home. Maybe you unmount used vacuum engines, cut their bells off, and bring home the fiddly bits. Somebody should find a use for the cast-off cans, eventually, and the cut-off bells. Maybe swing the cans on the ends of a wire for artificial gravity so your bones don't dissolve; though getting in and out would be tricky. You could store energy in their kinetic motion, resolving the nighttime power problem at the expense of variable artificial gravity inside.
At the poles this could enable some permanent power generation by getting the panels high enough to access permanent sunlight.
You need a pair of very large bags of regolith to store the heat and cold in, and a gas to percolate through them from radiators exposed alternately to sun and black sky.
Or you can just put up very big reflectors in orbit, lighting up your solar panels. Maybe they pump laser tubes pointed at your solar panels.
Nukes would be a big nuisance, needing constant maintenance, unless you just run a naked pile at incandescent temperature and catch the light coming off photovoltaically.
If we had piles of Pu-240 we could use RTGs, though they wouldn't be very efficient considering the amount of mass we'd have to send.
Running a naked reactor doesn't sound so bad considering there's no atmosphere or water sources to poison -- just stay away from the reactor. Running a steam turbine shouldn't be too hard, but it probably can't be serviced -- if it breaks, you replace it.
Downside to aerogel? It makes for a lousy micro-meteoroid and radiation blocker, unlike something like rockwool that masses more.
To the extent the lunar atmosphere is dense enough to cause meaningful conductive loss, one could pump a better vacuum inside the insulation. The pressure difference between this high vacuum and the lunar atmosphere would be low, so the mechanical strength needed for whatever container holds it would be corresponding low. (This seems very unlikely to be a problem at all. The moon’s atmosphere has an extremely low pressure. Offgassing from the insulation seems like a bigger concern.
It's just weird to me that they wouldn't go to the interesting and water/gas rich craters perpetually at the terminator. You've always got power there... just start digging.
Just remember: nuke first, then colonize. So, sooner rather than later.
[1] https://en.wikipedia.org/wiki/Sedan_(nuclear_test)
[2] https://www.youtube.com/watch?v=nSWURwNC8dc